marketing theory and applications · charles a. ingene, jianfeng jiang, rahul govind 107 team...

2010 AMA Winter Educators’ Conference

Marketing Theory

and Applications

EditorsMichael K. Brady, Florida State UniversityMichael D. Hartline, Florida State University

Track ChairsChris Blocker, Baylor UniversityLeff Bonney, Florida State UniversityTerry Esper, University of TennesseeDaniel J. Flint, University of TennesseeAndrea Godfrey, University of California, RiversideKevin Gwinner, Kansas State UniversityDavid M. Hardesty, University of KentuckyCharles F. Hofacker, Florida State UniversityBetsy B. Holloway, Samford UniversityJeremy Kees, Villanova UniversityBlair Kidwell, University of KentuckyDaekwan Kim, Florida State UniversitySandeep Krishnamurthy, University of Washington, BothellKelly D. Martin, Colorado State UniversityStephanie M. Noble, University of MississippiEllen Pullins, University of ToledoMolly I. Rapert, University of ArkansasMichelle Roehm, Wake Forest UniversityM. Kim Saxton, Indiana University, IndianapolisRudolf R. Sinkovics, University of ManchesterSijun Wang, California State Polytechnic University, PomonaElizabeth J. Wilson, Suffolk UniversityZhen “Jane” Zhu, Suffolk University

Volume 21

311 S. Wacker Dr. • Chicago, IL 60606

Copyright © 2010, American Marketing Association

Printed in the United States of America

Publications Director: Francesca V. CooleyCover Design: Jeanne NemcekTypesetter: Marie Steinhoff, Southeast Missouri State UniversityISSN: 1054-0806ISBN: 0-87757-340-9

All rights reserved. No part of the material protected by this copyrightnotice may be reproduced or utilized in any form or by any means,including photocopying and recording, or by any information storageor retrieval system, without the written permission of the AmericanMarketing Association.

iii

TABLE OF CONTENTS

TABLE OF CONTENTS iii

PREFACE AND ACKNOWLEDGMENTS xvi

BEST PAPERS AWARDS xvii

LIST OF REVIEWERS xviii

EXPLORING SALESPERSON ABILITIES AND CONDITIONS

Making Salespeople to Lead Themselves: Antecedents and Consequences ofSelf-Leadership Strategies

Nikolaos G. Panagopoulos, George J. Avlonitis 1

Sales Complexity: Conceptualization and a New Measure of Complexity in theSelling Function

Christian Schmitz, Philipp Rader 3

Salespersons Listening Ability as an Antecedent to Relationship SellingTanya Drollinger, Lucette B. Comer 6

CONSUMER ASPECTS OF RETAILING

Applying Flow Theory to Consumer Behavior in Retail StoresLiz C. Wang 8

Development and Validation of the Transaction Specific Attributions ofDiscrimination (TSAD) Scale

Thomas L. Baker, Tracy Meyer 10

A Research on the Curve Relationship Between Gift Promotion Depth and theSpillover Effect from Gift Promotion

Chung-Hui Tseng 12

SATISFACTION ANTECEDENTS AND OUTCOMES

Indebtedness as a Determinant of Satisfaction: An Extension of the ExpectancyDisconfirmation Model

Akinori Ono, Takahiro Chiba 19

The Role of Salesperson-Generated Satisfaction in Retailing Performance: ADecompositional and Comparative Approach

David M. Szymanski, Lisa C. Troy 21

COUNTRY EFFECTS IN GLOBAL MARKETING

Counterfeit Inferences from Price Levels and Country of Origin: The ModeratingEffect of Perceived Risk

Kashef A. Majid, Johny K. Johansson 23

Country Resources, Country Image, and Export Performance: A Panel Data AnalysisQin Sun, Audhesh K. Paswan, Margie Tieslau 25

iv

Examining Country Effects on Average Foreign Firm Performance and the ChainSize–Performance Relationship

Carol Finnegan, S. Tamer Cavusgil, Cuneyt Evirgen 27

The Effects of Home Country Culture on MNC PerformanceDaekwan Kim, Seong-Do Cho, Pan Dapotan Sinaga, Kyoungeun Kim,Gangok Jung 28

BRAND REPUTATION

Brands You Can Rely On! How Brand Credibility Affects Core Brand Functions:An Empirical Investigation in Services

Alexander Leischnig, Anja Geigenmueller, Margit Enke 30

Ethical Reputation: Fact or Fiction? A Qualitative Consumer ExplorationKatja H. Brunk 32

Family Business Branding: A Study on the Salience and Richness of Familinesson Corporate Websites

Willem Smit, Joachim Schwass, Lise Möller 34

FIRM INNOVATION: BOARD MEMBER BLOCS, EXTERNAL PROBLEMSOLVERS, AND KNOWLEDGE CREATION

Do Board Member Power Blocs Enhance Innovation?Shannon Cummins, Ravipreet S. Sohi 36

Innovation Beyond Firm Boundaries: The Routines and Resource Investments ofSuccessful External Problem Solvers

Dominik Mahr, Aric Rindfleisch, Rebecca J. Slotegraaf 37

Managerial Ties and Corporate Innovativeness: Is Knowledge Creation a MissingLink?

Chengli Shu, Shanxing Gao, Xu Jiang, Albert L. Page 39

IMPORTANT ISSUES IN SOCIAL RESPONSIBILITY: ENVIRONMENTALCONCERN, GRATITUDE, AND MATERIALISM

Promoting Sustainable Consumer Behavior: The Moderating Effect ofPsychological Distance on the Influence of Environmental Concern

Andrea Heintz Tangari, Scot Burton 41

Recognizing Those Who Participated and Encouraging Those Who Didn’t: A SocialMarketing Perspective on Expressions of Gratitude from Post-Katrina Louisiana

Randle D. Raggio, Judith Anne Garretson Folse

Status and Materialism among College-Aged WomenAndrea Loulakis, Ronald Paul Hill 45

CONSUMER BIASES

When Does Scenario Writing Improve the Quality of Planning Decisions?Kyeong Sam Min, Hal R. Arkes 53

Product Bundling or Mental Simulation: What Is More Effective in ReducingUsage Barriers?

Sabine Kuester, Jennifer Fitzgerald, Silke Hess, Judith Bucerius 54

v

The Effect of Goals on Status-Quo Bias Within a Choice FrameworkSajeev Varki, Najam Saqib 56

The Effects of Mental Construal and Perceived Risk in Response to Loss-VersusGain-Framed Health-Related Messages

Yun Lee 58

ASPECTS OF CUSTOMER RELATIONSHIPS

Bases of Brand Relationships: Redundant Constructs or Meaningful Distinctions?Christy Ashley, E. Deanne Brocato 60

Converting Transaction-Based Services to Relationship-Based Services: A CriticalRole of Information

Haisu Zhang, Chengli Shu 62

Does Personalized Recommendation Service Increase Customer Loyalty?Anyuan Shen, A. Dwayne Ball 64

Does Consumption Know No Limits? The Moderating Role of Satiation inDetermining Customer Repurchase

Glenn B. Voss, Andrea L. Godfrey, Kathleen Seiders 399*

BUILDING AND MAINTAINING RELATIONSHIPS WITH CONSUMERS

Performance Implications of Emotional Versus Cognitive Brand RelationshipQuality: An Empirical Study of Frequent Fliers in the Airline Industry

Bettina Nyffenegger, Lucia Malaer, Harley Krohmer 65

Too Late, or Just Later than Announced? Effects on Brand Trust by DelayingProduct Introductions in Competitive Situations

Steffen Herm, Jana Möller 67

Consumer Attitudes Toward Private Brands and National BrandsRonald E. Goldsmith, Leisa Reinecke Flynn, E. Craig Stacey 69

IN PURSUIT OF METHOD EXCELLENCE

A Proposed Procedure for Construct Definition in MarketingDavid A. Gilliam, Kevin E. Voss 71

A Review of Meta-Analysis in MarketingArne Floh, Jagdish Sheth 73

LEARNING, LIKELIHOOD, LOYALTY, AND LEAVING: A LOOK ATTHE “Ls” OF TODAY’S CUSTOMERS

Customer-Focused Learning in Government OrganizationsElena Kiryanova Bernard 75

Enhancing Customer Purchase Likelihood Through Market Driving StrategiesGoran Vlasic, Ajay Kohli, Gabriele Troilo 77

The Impact of Unprofitable Customer Abandonment on Current Customers’Exit, Voice, and Loyalty Intentions: An Empirical Analysis

Michael Haenlein, Andreas M. Kaplan 79

vi

MARKETING AND TECHNOLOGY I: CONSUMER BEHAVIOR ONLINE

Why Do Users of Second Life Buy Virtual Products? Antecedents of BrandPurchasing Intentions and Behaviors in the 3D Virtual Environment

Jason Gabisch 81

Effects of Online Store Atmospheric Qualities on Consumer Behavior: Holistic vs.Experimental Approaches

George J. Avlonitis, Spiros P. Gounaris, Christos D. Koritos 91

Effects of Word-of-Mouth and E-Word-of-Mouth on Innovation Use in the Presenceof Network Externalities

Tomoko Kawakami, Kazuhiro Kishiya, Mark E. Parry 93

How Brand-Specific Internet Searches Affect Purchase Behavior: An EmpiricalStudy of Interrelations in Online Advertising

Sebastian Hild, Malte Brettel, Andreas Engelen 95

CONSUMER SATISFACTION, LOYALTY, AND COMMITMENT

How Customer Satisfaction and Loyalty Programs Influence Customers’ Cash FlowVariability: Implications for Market Segmentation

Crina Tarasi, Ruth Bolton, Anders Gustafsson, Beth Walker 97

Investigating Three-Dimensional Model of Customer Commitment in the CellPhone Service Setting

Yaser F. Alabdi, Jikyeong Kang 99

The Referral Likelihood Referral of Reward Programs: The Moderating Effect ofCommunication Media and Product Involvement

Yimin Zhu 100

BRANDING AND THE CONSUMER

Asymmetry of Consumer Uncertainty and its Impact on Brand DeclineAdina Barbulescu, Ajay Kohli 103

Unintended Consequences: How Brand Extensions Make Brands More Vulnerableto Consumer Confusion

Eric D. DeRosia, Thomas R. Lee, Glenn L. Christensen 105

THE ROLE OF SERVICE EMPLOYEES

Labor Productivity in Retailing: A Modern AssessmentCharles A. Ingene, Jianfeng Jiang, Rahul Govind 107

Team Stewardship in Customer Service Teams: Antecedents and ConsequencesJeroen Schepers, Ad de Jong, Ko de Ruyter 109

The Effects of Customer Orientation and Identification on the Service Employee:Commitment and Job Burnout

D. Todd Donavan, Brad D. Carlson, Tom Dewitt 111

vii

INNOVATION: EXPLORING THE ROLE OF PRODUCT PROGRAMINNOVATIVENESS, OPEN INNOVATION, AND INNOVATION ORIENTATION

Companies’ Innovation Orientation and Product Program Innovativeness:The Moderating Role of Interfunctional Consensus

Ruth Maria Stock-Homburg, Bjoern Six 113

When Does Open Innovation Improve Performance? An Investigation of RelationalCapability and Flexibility

Sanjay R. Sisodiya, Jean L. Johnson, Yany Grégoire 115

Which Information Sources Help Companies to Gain Superior Product ProgramInnovativeness? Insights from Company and Customer Data

Ruth Stock-Homburg, Lars Herrmann 117

CONSUMER EVALUATIVE PROCESSING

He Who Dies with the Most Alternative Fuel Wins: An Alternative Use of TerrorManagement Theory

Christine Page, Krystina Smith 119

Customer Confusion and Product Line Size: Implications for Product LineManagement

Sabine Kuester, Sebastian Buys 121

How Do Consumers React to Service Mergers?Melinda Andrews, Ronald Goldsmith 123

Sleeping with the Enemy? Attributions Triggered by an “Open Architecture” SalesModel

Winfried Daun, Daniel Wentzel, Torsten Tomczak 125

INTEGRATING SUSTAINABILITY INTO MARKETING THEORY ANDPRACTIVE

A Sustainability Typology for Nonprofit Organizations: Inherent Challenges andNormative Strategies

Robert E. McDonald, Jay Weerawardena, Sreedhar Madhavaram,Gillian Sullivan Mort 127

Sustainability and Insurance: An Approach to Sustainability SegmentationMareike Bodderas 129

An Empirical Analysis on the Adoption of Alternative Fuels: Consumers’ PerceivedRisks Associated with Natural Gas Vehicles

Klaus-Peter Wiedmann, Martin Kassubek, Nadine Hennigs, Lars Pankalla 141

SERVICE PERFORMANCE OUTCOMES

Determinants of Cross-Buying: A Re-ExaminationMarkus Blut, David M. Woisetschläger 143

Humanic Interaction in Patient Experience as a Value PropositionJie Chen, Sandra S. Liu, En-Chung Chang 145

viii

Pre- and Post-Market Liberalization Customer Service Quality Perceptions of anIncumbent Monopoly: The Case of the Saudi Telecommunications Industry

Alhassan G. Abdul-Muhmin 147

Roads to Customer Performance: Investing Wisely in the Service-Profit ChainBarbara Caemmerer, Heiner Evanschitzky 149

CULTURE AND GLOBAL MARKETING

An Assessment of the Long-Term Orientation Scale: A Replication and ExtensionBased on Findings Across 10 European Countries

Louise May Hassan, Edward Man Kee Shiu, Gianfranco Walsh 151

Cross-Cultural Examination of Antecedents of Sales Manager Effectiveness:A Study of Salespeople in Six Countries

Ning Li, William H. Murphy 153

Cross-Cultural Differences in Online Brand Communities: An Exploratory Studyof Indian and American Online Brand Communities

Vivek Madupu, Delonia Minor-Cooley 155

MANAGERIAL ISSUES IN BUSINESS-TO-BUSINESS MARKETING

Why Do Managers Give Price Concessions? The Role of Organizational PricePerceptions

Christian Homburg, Dirk Totzek 157

The Influence of a Salesperson’s Dialect: Is Regional Dialect a Driver or anObstacle in Personal Selling?

Robert Mai, Stefan Hoffmann 159

Franchising: A Review of Literature from Marketing PerspectiveSelcuk Ertekin 161

MARKET ORIENTATION: SUBMIT POWER, LEADERSHIP, AND THEMARKETING ORGANIZATION

The Role of Marketing Organization in the Market Orientation-PerformanceRelationship

Lancy Mac 173

WHEN THE SELLING GETS TOUGH

Suspicion of Ulterior Motives in Sales Interactions: Examining the Case of ValueCo-Creation

Nicole Kirpalani 182

Salesperson’s Karma Orientation: A Conceptual Framework and ResearchPropositions

Ramendra Singh, Rakesh Singh 184

When Customer-Oriented Sales Efforts Lose Ground: An Experimental Study onthe “Dilution Effect”

Peter Mathias Fischer, Walter Herzog, Sven Reinecke 186

ix

PERCEIVING PRICE INFORMATION

From a Distance We All Have Enough: Temporal Perspective and the Dual Roleof Price

Christian Homburg, Torsten Bornemann 188

Price Context, Prior Knowledge, and Working Memory CapacityShan Feng, Rajneesh Suri 190

The Impact of Favoring New Customers Over Loyal Customers on Loyal Customers’Perceptions of Price Fairness

Martin Fassnacht, Jochen Mahadevan 192

CONSUMER BEHAVIOR AND SUSTAINABILITY: ATTITUDES, PROCESSING,AND SELF-CONROL

Against the Green: An Examination of Non-Green ConsumersJ. Joseph Cronin, Jr., Mark R. Gleim, Stephanie J. Lawson 194

Does Green Really Help Being Green? The Role of Color in Processing FluencyJoon Yong Seo, Debra L. Scammon 196

The Psychological Impact of Nutrition Information: The Role of Guideline DailyAmounts

Louise May Hassan, Edward Man Kee Shiu, Nina Michaelidou 198

SERVICE FAILURE AND RECOVERY ISSUES

Linking Service Failure Types to Cognitive AppraisalJiraporn Surachartkumtonkun, Paul G. Patterson, Janet R. McColl-Kennedy 200

Service Recovery Paradox: Myth or Reality?Christian Brock, Markus Blut, Heiner Evanschitzky, Clay Voorhees 202

Setting in Versus Expecting More: Moderating Effects of Relationship Length andBrand Equity on Consumer Justice Perceptions

Jeremy S. Wolter 204

THE ROLE OF TRUST AND BONDS IN INTERORGANIZATIONALRELATIONSHIPS

Concept Providing in the Wholesaling Industry: Do Structural Customer BondsHelp or Do They Harm Competitive Advantage and Relationship Quality?

Sandra Pocsay, Hanna Schramm-Klein, Joachim Zentes, Nikolaus Hohl,Julia Naskrent 205

Production of Trust in Buyer-Seller Relationships: The Contingent Roles of ProductQualities and Life-Cycle Stage

Oliver Schilke, Malte Brettel, Gunnar Wiedenfels 207

Effects of Supplier’s Market Orientation on Channel Relationship: A Cross-Cultural Study

Trang Phuc Tran 209

x

EFFECTIVE BRAND COMMUNICATION

How Humor in Advertising Works: A Test of Alternative ModelsMartin Eisend 211

SE Habla Español: Exploring Latinos’ Emotional Reactions Toward PoliticalLanguage Tailored-Ads

Sindy Chapa, Enrique P. Becerra, Stephen Halbrook 213

The Consumer’s Perception of the Emotional Authenticity of an AdvertisingSpokesperson: Theoretical Discussion and Qualitative Study

Katja Leschnikowski, Harley Krohmer 215

NEW METHODS OF DELIVERING MARKETING TOPICS

Teaching Marketing Research BackwardsAlvin C. Burns, Peter “Danny” Weathers, Ann Veeck, Ronald F. Bush 217

Selecting a Free Vendor-Hosted Online Survey Tool for Student UseMatt Elbeck 222

Marketing Students’ Attitudes Toward and Beliefs about Commercial SignageJames J. Kellarism 224

MARKETING AND TECHNOLOGY II: MANAGING INNOVATION

Brand and Technical Integration for Radical Innovation: Do Firms That IntegrateDo Better than Firms That Do Not?

Raul Rivadeneyra, Chirag Patel, Alina Sorescu, Jaideep Prabhu,Rajesh Chandy 226

Managing Satisfying Seller-Buyer Interactions During New Product Developmentin Technology-Based, Industrial Markets

Jason Q. Zhang, Gerard A. Athaide 228

The Adoption of Technology Orientation in Healthcare IndustryOlivia F. Lee, Matthew L. Meuter, Wenzhi Wang 230

The Role of Marketing Resources in Radical InnovationKyriakos Kyriakopoulos, Mathew Hughes, Paul Hughes 232

THE IMPACT OF FRAMING ON INFORMATION PRCESSING

Comma N’ Cents in Pricing: The Effects of Auditory Representation Encoding onPrice Recall and Magnitude Perceptions

Keith S. Coulter, Pilsik Choi 234

The Influence of Assortment Size on Consumers’ Information-Processing Strategy:Rationale and Design of an Empirical Approach

Svenja Schmidt 236

Purely Incidental Numeric Priming on Product Judgments: the Effects of UsingHigh Versus Low Product Model Numbers

Yun Lee 248

xi

UNDERSTANDING GLOBAL CONSUMERS

Conceptualizing Consumer AnimosityStefan Hoffmann, Robert Mai, Maria Smirnova 250

Global Youth and Their Relationships with Global and Local Brands in EmergingMarkets

Yuliya Strizhakova, Robin Coulter, Linda Price 252

Why Consumers Buy Counterfeit Software: A Seven Country StudyElfriede Penz, Barbara Stöttinger 254

Comparing Consumer Culture Plots in Television Advertising from South Africaand Nigeria

Adesegun Oyedele 256

VISUAL BRANDING TACTICS

Exciting Red and Competent Blue: Linking Color to Brand PersonalityLauren I. Labrecque, George R. Milne 258

The Impact of Color Traits on Corporate BrandingWei-Lun Chang, Xia-Liang Lin 260

When Less Is More: The Effects of Product Quantity on Consumer BehaviorStacey G. Robinson 262

EDUCATING MARKETING LEADERS FOR THE FUTURE

The Next Generation of Business Leaders: Influences on Unethical StandardsAcross Multiple Cultures

Victoria L. Crittenden, Richard Hanna, Robert A. Peterson,William F. Crittenden 263

How Does Online Word-of-Mouth Influence Students’ Course Selections andTheir Attitudes Toward Professors?

Cong Li, Xiuli Wang 265

GLOBAL MARKETING STRATEGIES

How Do Institutional Pillars Moderate the Effectiveness of Firms’ Customer-BasedMarketing Capability Across Countries?

Fernando Angulo, Daniel Bello, Diego Prior, Josep Rialp 266

Implementation of Subsidiary Marketing Strategies: An Empirical Investigationof the Actors, Modalities, Drivers, Parameters, and Performance Outcomes

Esra F. Gencturk, Destan Kandemir 268

Perceived Barriers for Internationalization of Small Firms: Perceived Lack ofResources

Yu Henry Xie, Taewon Suh 271

When Does International Marketing Standardization Matter to Firm Performance?Oliver Schilke, Martin Reimann, Jacquelyn Thomas 272

xii

IT’S A VIRTUAL WORLD

What Is the Ideal Customization Approach? The Moderating Effect of CulturalMeaning System Activation on Customization

Cong Li 274

Alternative Hedonism, Digital Virtual Consumption, and Structures of theImagination

Rebecca Jenkins, Mike Molesworth, Sue Eccles 275

The Comparisons of Real Versus Virtual Product Purchase: The Effects ofSelf-Congruency and Product Type

Chih-Ping Wang 285

The Effects of Participation in Online Brand Communities: Findings from XBOXCommunities in Brazil

Stefânia Ordovás de Almeida, Utpal Dholakia, José Afonso Mazzon,Hugo Müller Neto 293

EXPLORING THE DIMENSIONS OF CORPORATE RELATIONSHIPS ANDSOCIAL RESPONSIBILITY

Calling or Blatant Commerce? Clients’ Views on Efforts of a For-ProfitOrganization to Engage in a Societal Issue

Johan van Rekom, Frank Go, Dayenne Calter 304

To Give and Get Back: Do Cause Portfolio Characteristics Influence the Returnon Corporate Social Responsibility?

A. Meike Eilert, Stefanie Rosen 306

The Odd Couple: Managing “Ideological Distance” in the Corporate-NonprofitRelationship

Kathryn Lefroy, Dayna Simpson, Yelena Tsarenko 308

MARKETING AND TECHNOLOGY III: TRUST AND CONFIDENCE ONLINE

Consumer’s Perceived Trust in IT-Ecosystems: Conceptualizing System-Centric andUser-Centric Determinants of Trust in Digital Environments

Klaus-Peter Wiedmann, Nadine Hennigs, Marc-Oliver Reeh, Dieter Varelmann 310

Prospective Confidence in the Knowledge of the WebKishore Gopalakrishna Pillai, Constantine Katsikeas 312

Toward a Better Understanding of Trust in Web 2.0 Social NetworksSonja Grabner-Kräuter 313

WOM: REFERRALS AND RECOMMENDATIONS

Does Rewarding Referrals Affect Perceived Sender Credibility? An EmpiricalInvestigation in Cellular Telecommunication Services

Sabrina Helm, Anne Willach 322

Exploring the Effects of Referral Reward Programs on Satisfied and DissatisfiedCustomers

Ina Garnefeld, Eva Muenkhoff, Jens Hogreve, Andreas Eggert 324

xiii

ISSUES IN RETAIL CHANNELS

An Empirical Test of the Drivers of Consumer Trust in an E-Retailer and itsOutcomes Directed Toward Customer Retention

Cuiping Chen, Chuanyi Tang, Matthew O’Brien 326

Concentration of Marketing Channels and Brand-Level Retailer MarginsKenji Matsui 328

Effects of the Extant Environment on Influence Strategy Usage in the FranchiseSystem Relationship

Kyle A. Huggins, Darin W. White 330

MISCELLANEOUS ISSUES IN BRANDING: SPONSORSHIPS ANDBRANDING CAPABILITIES

Effective Brand Strategy Implementation: A Review of the Literature andFramework for Future Research

Alexander Haas, Adele J. Huber 332

Exploring the Brand Productivity GapMaik Hammerschmidt, Tomas Falk, Frank Germann, William T. Ross, Jr.,Hans H. Bauer 334

Sponsorship Deal Characteristics, Fan Perceptions, and Sponsorship Fit: AnEmpirical Examination

David M. Woisetschläger, Christof Backhaus, Manuel Michaelis 336

The Spillover of Brand Equity Within a Corporate Sponsorship PortfolioJoe Cobbs, Mark Groza 338

POWER OF QUALITATIVE DATA

Exploring the Power of Qualitative Comparative Analysis for Marketing Research:Replication and Extension of a New Product Development Study

Andrea Ordanini, Edwin Nijssen, Fred Langerak 340

Improving the Direct Estimation of Demand by Adjusting for Incorrect Price-Statements

David Blatter, Reto Hofstetter, Klaus Miller 342

Quantifying Qualitative Data: A Novel Approach to Measure Brand AssociationsKaleel Rahman, Charles S. Areni 344

MARKETING INITIATIVES: PRODUCT DESIGN, PRODUCT PLACEMENT,AND CO-MARKETING CAPABILITIES

A Conceptual Framework of How Product Design Impacts ConsumerPerceptions and Market-Based Performance

Janell D. Townsend, Mitzi M. Montoya 346

Product Placement in the Movies: From Dawn Till Dusk?Grigori Erenburg, Ekaterina V. Karniouchina, Can Uslay 348

Co-Marketing Capability and its Impact on Marketing Alliance PerformanceMarcus Schögel, Dennis Herhausen 350

xiv

SALES AND MARKETING INTEGRATION

Conceptualizing Trust in the Relationship Between Sales and Marketing and theCustomer

Kenneth Le Meunier-FitzHugh, Jasmin Baumann,Leslie Le Meunier-FitzHugh 352

Assessing the Impact of Salespeople’s Field-Based Competitive Intelligence onProduct Competitiveness and Brand Preference

Joel Le Bon, Adam Rapp 354

A Salesperson’s Cross-Functional Orientation: Antecedents and Effects on Cross-Selling Success

Oliver Malms, Christian Schmitz 356

ADVERTISING EFFECTIVENESS

Combining Visual and Textual Rhetorical Figures in AdvertisingRobert O. Fabrize 358

Designing Advertisements for Consumers Older than 50 YearsStefan Hoffmann, Susanne C. Liebermann, Uta Schwarz 360

Examining the Effectiveness of Language Choice for Different Ad Appeals inAdvertising to Bilingual Consumers

Fernando R. Jiménez, Xiang Fang, Darrell Bartholomew 361

Predicting the Effectiveness of Celebrity Endorsements Using Balance TheorySubhadip Roy, Bashar S. Gammoh, Anthony A. Koh 363

CONSUMER PERCEPTIONS OF VALUE

Descriptions of a Radical Innovation Can Impact Consumers’ Risk and ValuePerceptions

Arjun Chaudhuri 365

How Do Consumers Form Perceptions of Economic Value: An Exploratory StudyGerald Smith, Sajeev Varki 366

Shifting Consumer Preference: The Effects of Cognitive Styles and TemporalPerspective

Yun Lee 367

REGULATORY ISSUES AND CONSUMPTION

On-the-Go Consumption: Because Consumers Want to and Not Because TheyHave To

Sabine Moeller, Tobias Schaefers 369

The Role of Self Regulatory Focus on Compliance BehaviorStephanie Dellande, Prashanth U. Nyer 371

GLOBAL PRODUCT/BRAND MANAGEMENT

Antecedents of New Product Development: A Comparison Between Germany andChina

Malte Brettel, Dominik Austermann, Andreas Engelen 373

xv

Distribution Channels for Consumer Products in India: Past, Present, and FutureDirections

Debi P. Mishra 375

Product and Environmental Influences on Marketing-Mix Standardization: ACentral Eastern European Perspective

Roxana Codita, Frank-Martin Belz 377

The Effects of Sponsorship on Brand Image: A Cross-National AnalysisChristian Lucas, David M. Woisetschläger, Alexander Eiting 380

FIRM CAPABILITIES FOR RELATIONSHIP MANAGEMENT

Promoting Buyer-Seller Collaborations: The Role of Cooperative CapacityMaggie Chuoyan Dong 382

Seeking Enhanced Understanding of Multinational Suppliers’ Choice of GlobalAccounts: An Integrative Framework and Empirical Study

Tao (Tony) Gao, Linda Hui Shi 383

The Contrasting Impact of Relational and Branding Capabilities on MarketingControl Rights Decisions Within Technology Alliances

D. Eric Boyd, Brian P. Brown 385

CURRENT ISSUES IN MARKEITNG STRATEGY: OPTIONS, KNOWLEDGEGRAFTING, DISAGREEMENT, AND STRATEGIC BEHAVIOR

Real Options: A Gold Mine, or Fool’s Gold?Abhijit Guha, William Boulding, Richard Staelin 387

Strategic Disagreement Paradox in Marketing Strategy MakingKwaku Atuahene-Gima, Namwoon Kim, Deborah Andrus 389

The Implications of Knowledge Grafting for Marketing StrategySreedhar Madhavaram, Vishag Badrinarayanan, Robert E. McDonald 391

What’s in a Name? An Analysis of the Strategic Behavior of Family FirmsSaim Kashmiri, Vijay Mahajan 393

ADDENDUM

The following papers were presented at the Summer Educators’ Conference,but were not included in the proceedings.

The Effects of Negative Emotions on Avoidance Behavior in ServicescapeJung Ok Jeon, Hee Young Han 395

From Hollywood to Broadway: A Field Study of the Effectiveness of ProductPlacements

Rick T. Wilson, Brian D. Till 397

*This paper is included at the end of the Proceedings due to late submissionby the authors.

AUTHOR INDEX 400

xvi

Preface and Acknowledgments“Marketing 2010: Strategies and Solutions for a Tumultuous Economy” is the theme for the AMA 2010 Winter

Marketing Educators’ Conference. Given the trials, tribulations, and tumult that have characterized our global economy overthe past 18 months, marketing academics and practitioners must find new solutions to ensure viability in these changing times.Traditional marketing strategies may not apply in an era in which marketing practice has become intertwined with social,economic, regulatory, legal, ethical, and consumer challenges on a global scale. Marketing academics must fulfill their roleby engaging in research that will give today’s business leaders a clear, unobstructed view of the diversity of issues facingmarketing practice in today’s organizations. Access to this type of cutting-edge research is critical if today’s business leadersand marketing academics are to have a full appreciation of the key issues that affect the strategic and tactical marketingactivities of the firm.

We have many people to thank for their involvement in developing the conference. First, we thank all the scholars whochose to share their best work with the AMA audience. Your papers, presentations, and insights are the reason we gather inNew Orleans.

Second, we thank the tireless efforts of our track chairs. These are the people who worked behind the scenes to shapethe content and overall contributions of the conference program:

Consumer Behavior David M. Hardesty, University of KentuckyBlair Kidwell, University of Kentucky

Global Marketing Daekwan Kim, Florida State UniversityRudolf R. Sinkovics, University of Manchester

Corporate Affairs, Social Responsibility, Kelly D. Martin, Colorado State Universityand Sustainability Jeremy Kees, Villanova University

Brand Marketing and Communication Michelle Roehm, Wake Forest UniversityKim Saxton, Indiana University, Indianapolis

Marketing Education Elizabeth J. Wilson, Suffolk UniversityZhen “Jane” Zhu, Suffolk University

Business-to-Business Marketing Daniel J. Flint, University of TennesseeChris Blocker, Baylor University

Research Methods Betsy B. Holloway, Samford UniversitySijun Wang, California State Polytechnic Univ., Pomona

Marketing Strategy Molly I. Rapert, University of ArkansasTerry Esper, University of Tennessee

Sales and Customer Relationship Management Ellen Pullins, University of ToledoLeff Bonney, Florida State University

Services, Service Science, and Retailing Stephanie M. Noble, University of MississippiAndrea Godfrey, University of California, Riverside

Marketing and Technology Charlie Hofacker, Florida State UniversitySandeep Krishnamurthy, University of Washington, Bothell

SIGnificant Advances among Special Interests Kevin Gwinner, Kansas State University

Social Chair and Local Events Coordinator Lauren Skinner, University of Alabama, Birmingham

Third, we thank the reviewers for their time and dedication toward reviewing the hundreds of papers and sessionproposals submitted to the conference. Your efforts are responsible for the quality of the 2010 program. We also thank themembers of the “Blue Ribbon” Selection Committee, who had the unenviable task of selecting a single paper to receive theBest Conference Paper Award.

Finally, we thank the American Marketing Association staff and volunteers who have gone above and beyond to helpus throughout this endeavor. We especially thank Lynn Reyes for her wise counsel over the past 14 months. We also thankthe members of the AMA Academic Council for giving us the opportunity to serve the AMA community. We appreciate thesupport of Kathleen Seiders (Boston College) and her willingness to take our ideas and suggestions to the Council.

Michael K. Brady Michael D. HartlineFlorida State University Florida State University

xvii

Overall Best Conference Paper

“Exploring the Brand Productivity Gap”Maik Hammerschmidt, University of Mannheim

Tomas Falk, University of MannheimFrank A. Germann, Pennsylvania State University

William Ross, Pennsylvania State UniversityHans H. Bauer, University of Mannheim

Blue Ribbon Panel (Determined Best Conference Paper)Sundar S. Bharadwaj, Emory University

A. Parasuraman, University of MiamiJ. Joseph Cronin, Florida State University

Best Paper: Track Award Winners

Consumer Behavior“From a Distance We All Have Enough: TemporalPerspective and the Dual Role of Price”

Christian Homburg, University of MannheimTorsten Bornemann, University of Mannheim

Global Marketing“Implementation of Subsidiary Marketing Strategies:An Empirical Investigation of the Actors, Modalities,Drivers, Parameters and Performance Outcomes”

Esra F. Gençtürk, Ozyegin University, IstanbulDestan Kandemir, Bilkent University, Ankara

Brand Marketing and Communication“Exploring the Brand Productivity Gap”

Maik Hammerschmidt, University of MannheimTomas Falk, University of MannheimFrank A. Germann, Pennsylvania State UniversityWilliam Ross, Pennsylvania State UniversityHans H. Bauer, University of Mannheim

Corporate Affairs, Social Responsibility, andSustainability“Recognizing Those Who Participated andEncouraging Those Who Didn’t: A Social MarketingPerspective on Expressions of Gratitude from Post-Katrina Louisiana”

Randal Raggio, Louisiana State UniversityJudith Anne Garretson Folse, Louisiana State University

Marketing Education“The Next Generation of Business Leaders: Influenceson Unethical Standards Across Multiple Cultures”

Victoria Crittenden, Boston CollegeRichard Hanna, Northeastern UniversityRobert A Peterson, University of Texas at AustinWilliam F Crittenden, Northeastern University

Business-to-Business Marketing“Why Do Managers Give Price Concessions? The Roleof Organizational Price Perceptions”

Christian Homburg, University of MannheimDirk Totzek, University of Mannheim

Research Methods“Exploring the Power of Qualitative ComparativeAnalysis: Replicating a Study on Accelerating NewProduct Development”

Andrea Ordanini, Bocconi University, MilanEdwin Nijssen, Eindhoven Technical UniversityFred Langerak, Eindhoven Technical University

Marketing Strategy“What’s in a Name? An Analysis of the StrategicBehavior of Family Firms”

Saim Kashmiri, University of Texas at AustinVijay Mahajan, University of Texas at Austin

Sales and Customer Relationship Marketing“Conceptualizing Trust in the Relationship BetweenSales and Marketing and the Customer”

Kenneth Le Meunier-FitzHugh, University of EastAnglia, Norwich

Jasmin Baumann, University of East Anglia,Norwich

Leslie Caroline Le Meunier-FitzHugh, Universityof East Anglia, Norwich

Services, Service Science, and Retailing“Team Stewardship in Customer Service Teams:Antecedents and Consequences”

Jeroen Schepers, Eindhoven Technical UniversityAd De Jong, Eindhoven Technical UniversityKo De Ruyter, Maastricht University

Marketing and Technology“Brand and Technical Integration for RadicalInnovation: Do Firms That Integrate Do Better ThanFirms That Do Not?”

Raul Rivadeneyra, Judge Business SchoolJaideep C. Prabhu, Judge Business SchoolRajesh Chandy, London Business SchoolAlina Sorescu, Texas A&M UniversityChirag C. Patel, Grenoble Ecole de Management

xviii

2010 AMA Winter Educators’ ConferenceList of Reviewers

AEsi Abbam Elliot, University of

Illinois at ChicagoAndrew Abela, Catholic University

of AmericaCarmen Abril, IE Business SchoolRaj Agnihotri, Kent State

UniversityYaser Alabdi, Manchester Business

SchoolPia Albinsson, Appalachian State

UniversityJ. Craig Andrews, Marquette

UniversityFernando Angulo, Autonomous

University of BarcelonaSyed Anwar, West Texas A&M

UniversityDominik Austermann, RWTH

Aachen University

BBarry Babin, Louisiana Tech

UniversityCem Bahadir, University of South

CarolinaStacey Baker, University of

WyomingYeqing Bao, University of Ala-

bama, HuntsvilleAdina Barbulescu, University of

TennesseeKenneth Bates, University of San

DiegoHans Bauer, University of

MannheimJasmin Baumann, University of

East AngliaJochen Becker, German Graduate

School of Management and LawJeri Beggs, Illinois State UniversityFrank-Martin Belz, Technische

Universität MünchenGuido Berens, Erasmus UniversityDoris Berger, IMC University of

Applied Sciences KremsFrederik Beuk, University of

Illinois at ChicagoSundar Bharadwaj, Emory

University

Pelin Bicen, Penn State ErieMarkus Blut, TU Dortmund

UniversityMareike Bodderas, University of

St. GallenSterling Bone, Brigham Young

UniversityP. Greg Bonner, Villanova

UniversityClement Bowen, Villanova

UniversityMairead Brady, Trinity College

DublinDavid Brennan, University of St.

ThomasMalte Brettel, RWTH Aachen

UniversityEileen Bridges, Kent State

UniversityJames Brown, West Virginia

UniversityMyla Bui, Loyola Marymount

UniversityMichele Bunn, University of

MontevalloScot Burton, University of

Arkansas

CBarbara Caemmerer, University of

StrathclydeRichard Caldarola, Troy UniversityMichael Capella, Villanova

UniversityHernan Casakin, Ariel University

Center of SamariaJesse Catlin, University of

California, IrvineWei-Lun Chang, Tamkang

UniversityCuiping Chen, University of

Ontario Institute of TechnologyShu-Ching Chen, Massey

UniversitySeong-Do Cho, Chonnam National

UniversityPilsik Choi, Clark UniversitySungchul Choi, University of

Northern British ColumbiaRoxana Codita, Technische

Universität München

Nathalie Collins, University ofWestern Australia

Lucette Comer, Purdue UniversityLarry Compeau, Clarkson

UniversityJennifer Cordero, University of

California, IrvineRobert Cosenza, University of

MississippiAmy Cox, Converse CollegeStephen Craft, Birmingham-

Southern CollegeVictoria Crittenden, Boston

CollegeKevin Cumiskey, Oklahoma State

University

DSujan Dan, University of

Tennessee, KnoxvilleJennifer Dapko, University of

South FloridaDevashish Das Gupta, Indian

Institute of ManagementBrennan Davis, Baylor UniversityCharlene Davis, Trinity UniversityDonna Davis, Texas Tech

UniversityDebra Desrochers, Tulane

UniversityThomas DeWitt, University of

Hawaii at HiloMbaye Diallo, CERGAM - Paul

Cézanne UniversityWilliam Diamond, University of

Massachusetts – AmherstAndrea Dixon, Baylor UniversityTodd Donavan, Colorado State

UniversityMichael Dorsch, Clemson

UniversityJenna Drenten, University of

GeorgiaTanya Drollinger, University of

LethbridgeCourtney Droms, Valdosta State

UniversityTurkan Dursun, West Texas A&M

University

xix

EA. Meike Eilert, University of

South CarolinaPam Ellen, Georgia State

UniversitySusan Ellis, University of

MelbourneAndreas Engelen, RWTH Aachen

University

FCarolyn Findley, University of

AlabamaAdam Finn, University of AlbertaCarol Finnegan, University of

Colorado, Colorado SpringsPeter Fischer, University of St.

GallenCarlos Flavian, University of

ZaragozaJudith Anne Folse, Louisiana State

UniversityJohn Ford, Old Dominion

UniversityStephen France, University of

Wisconsin–MilwaukeeGo Frank, Erasmus UniversityChristina Froböse, WHU–Otto

Beisheim School of ManagementGary Futrell, Florida State

University

GAndrew Gallan, Case Western

Reserve UniversityBashar Gammoh, University of

ToledoTao Gao, Northeastern UniversityIna Garnefeld, University of

PaderbornDinesh Gauri, Syracuse UniversityEsra Gencturk, Ozyegin UniversityDavid Gilliam, Oklahoma State

UniversityStephanie Gillison, University of

AlabamaTom Gillpatrick, Portland State

UniversityMark Gleim, Florida State

UniversityJohn Godek, University

Washington BothellSonja Grabner-Kraeuter,

Klagenfurt University

Yany Gregoire, Washington StateUniversity

Amir Grinstein, Ben-GurionUniversity

Abhijit Guha, Wayne StateUniversity

Julie Guidry, Louisiana StateUniversity

Gregory Gundlach, University ofNorth Florida

Kevin Gwinner, Kansa StateUniversity

HAlexander Haas, University of

BernJay Handelman, Queen’s

UniversityMary Harrison, University of

AlabamaJames Harvey, George Mason

UniversityLouise Hassan, Heriot-Watt

UniversityJon Hawes, Indiana State

UniversityKelly Haws, Texas A&M

UniversityAnne Hoel, University of

Wisconsin–StoutStefan Hoffmann, Technische

Universität DresdenJens Hogreve, University of

PaderbornMary Holden, Waterford Institute

of TechnologyThomas Hollmann, North Carolina

State UniversityRob Hooker, Florida State

UniversityTarique Hossain, California State

Polytechnic UniversityYansong Hu, University of

WarwickKyle Huggins, James Madison

UniversityMathew Hughes, University of

NottinghamPaul Hughes, Loughborough

UniversityJohn Hulland, University of

PittsburghMichael Hutt, Arizona State

University

IChiharu Ishida, Illinois State

UniversityKarthik Iyer, University of

Northern Iowa

JRuey-Jer Jean, Manchester

Business SchoolDouglas Johansen, Jacksonville

UniversityGang Ok Jung, Hannam University

KAnna Kaleka, Cardiff Business

SchoolDestan Kandemir, Bilkent

UniversityJun Kang, University of

Massachusetts–AmherstEric Karson, Villanova UniversitySaim Kashmiri, University of

Texas at AustinMartin Kassubek, Leibniz

University HannoverKathleen Kelly, Colorado State

UniversityThomas Kilian, University

Koblenz–LandauTracey King, American UniversityWerner Kunz, University of

MassachusettsPiotr Kwiatek, Poznan University

of EconomicsKyriakos Kyriakopoulos, ALBA

Graduate Business School

LRussell Lacey, University of New

OrleansAlexander Lakotta, ESCP EuropeStephanie Lawson, Florida State

UniversityJoel Le Bon, ESSEC Business

School–SingaporeOlivia Lee, St. Cloud State

UniversityRuby Lee, Florida State UniversityAlexander Leischnig, Technische

Universität BergakademieFreiberg

Ning Li, George Mason University

xx

Beichen Liang, East TennesseeState University

Mark Ligas, Fairfield UniversityXin Liu, California State

Polytechnic University, PomonaChristian Lucas, TU Dortmund

University

MJun Ma, Indiana University–Purdue

UniversitySreedhar Madhavaram, Cleveland

State UniversityVivek Madupu, Missouri Western

State UniversityVijay Mahajan, University of

Texas at AustinDominik Mahr, University of

AntwerpKashef Majid, George Washington

UniversitySuzanne Makarem, Temple

UniversityNaresh Malhotra, Nanyang

Technical UniversityMichael Mallin, University of

ToledoAvinash Malshe, University of

Saint ThomasAlan Malter, University of Illinois

at ChicagoCarlo Mari, University of MoliseMelissa Markley, DePaul

UniversityIngrid Martin, California State

University, Long BeachMarlys Mason, Oklahoma State

UniversityShashi Matta, Ohio State

UniversityRobert McDonald, Texas Tech

UniversityJeannette Mena, Michigan State

UniversityNorma Mendoza, University of

Texas at El PasoStefan Michel, IMDT.C. Mims, Texas Woman’s

UniversityDelonia Minor-Cooley, Texas

Southern UniversityDebi Mishra, State University of

New York at BinghamtonSaurabh Mishra, McGill UniversitySabine Moeller, European Business

School

Gillian Sullivan Mort, La TrobeUniversity

Kyle Murray, University of AlbertaBrian Murtha, University of

KentuckyJun Myers, California State

Polytechnic University, PomonaMatthew Myers, University of

Tennessee, KnoxvilleNiklas Myhr, Chapman University

NKent Nakamoto, Virginia

Polytechnic Institute and StateUniversity

Christopher Newman, Universityof Arkansas

Yuefeng Ni, Beijing Institute ofTechnology

OMatthew O’Brien, Bradley

UniversityShintaro Okazaki, Universidad

Autónoma de MadridAndrea Ordanini, Bocconi

UniversityLinda Orr, University of AkronAdesegun Oyedele, Saint Cloud

State University

PKadayam Padmanabhan,

University of Michigan–Dearborn

Nikolaos Panagopoulos, AthensUniversity of Economics &Business

Lars Pankalla, Leibniz UniversityHannover

Gary Pansino, QuantiViewMark Parry, University of

Missouri, Kansas CityAudhesh Paswan, University of

North TexasChirag Patel, Grenoble Ecole de

ManagementPaul Pavlou, Temple UniversityIryna Pentina, University of ToledoElfriede Penz, Vienna University of

Economics and BusinessMark Peterson, University of

Wyoming

Noemi Pezderka, ManchesterBusiness School

Joanna Phillips, Western KentuckyUniversity

Kishore Pillai, University of LeedsRobert Ping, Wright State

UniversityChris Plouffe, Florida State

UniversityWesley Pollitte, University of

Southern MississippiPaulo Prado, Federal University of

ParanaCarolyn Predmore, Manhattan

CollegeDiego Prior, Autonomous

University of Barcelona

QTianjiao Qiu, California State

University, Long Beach

RRandle Raggio, Louisiana State

UniversityPushkala Raman, Texas Woman’s

UniversityAdam Rapp, Clemson UniversityEddie Rhee, Stonehill CollegeJosep Rialp, Autonomous

University of BarcelonaRichard Robinson, Marquette

UniversityMatthew Robson, University of

LeedsMichael Rodriguez, Elon

UniversityStefanie Rosen, University of

South CarolinaMark Rosenbaum, Northern Illinois

UniversityWilliam Ross, Pennsylvania State

UniversitySubroto Roy, University of New

HavenBrian Rutherford, Purdue

University

SRitesh Saini, University of Texas at

ArlingtonTakisha Salley, Saint Louis

UniversitySaeed Samiee, University of Tulsa

xxi

Juan Samper, CESAMatthew Sarkees, Pennsylvania

State UniversityDebra Scammon, University of

UtahTammy Schakett, Franklin

UniversityHanna Schramm-Klein, University

of SiegenHeather Schulz, University of

Texas at AustinJoon Yong Seo, University of UtahLaura Serviere, University of Texas

of the Permian BasinAnyuan Shen, State University of

New York at New PaltzSimon Sheng, Adelphi UniversityEdward Shiu, University of

StrathclydeJeremy Sierra, Texas State

University, San MarcosGeorge Simon, T.A. Pai

Management InstituteRamendra Singh, IIM AhmedabadRakesh Singh, XLRISatyendra Singh, University of

WinnipegNancy Sirianni, Arizona State

UniversitySanjay Sisodiya, University of

IdahoK. Sivakumar, Lehigh UniversityLauren Skinner, University of

Alabama at BirminghamMaria Smirnova, St. Petersburg

State UniversityWillem Smit, IMDGerald Smith, Boston CollegeRobert Spekman, University of

VirginiaFredrika Spencer, Duke UniversityRaji Srinivasan, University of

Texas at AustinMona Srivastava, Harvard Business

School India Research CenterDenise Steckstor, German

Graduate School of Managementand Law

Nancy Stephens, Arizona StateUniversity

Dave Stewart, University ofCalifornia, Riverside

Barbara Stoettinger, ViennaUniversity of Economics andBusiness

Amy Stokes, University ofArkansas

Andreas Strebinger, YorkUniversity

Yuliya Strizhakova, RutgersUniversity

Rodney Stump, Towson UniversityUrsula Sullivan, University of

Illinois at Urbana/ChampaignQin Sun, University of South

Florida, St. PetersburgEsther Swilley, Kansas State

University

TYihui (Elina) Tang, University of

MissouriAndrea Tangari, University of

ArkansasCrina Tarasi, Central Michigan

UniversityDavid Taylor, University of North

TexasAlke Toellner, TU Dortmund

UniversityPingsheng Tong, California State

University, SacramentoJanell Townsend, Oakland

UniversityTrang Tran, University of North

TexasAnna Turri, University of Arkansas

UNancy Upton, Northeastern

University

VGautham Vadakkepatt, Texas

A&M UniversityDeena Vaughn, University of

Maryland University CollegeVijay Viswanathan, Emory

UniversityShiri Vivek, Eastern Michigan

UniversityFlorian von Wangenheim,

Technische Universität MünchenClay Voorhees, Michigan State

UniversityKevin Voss, Oklahoma State

University

WTrent Wachner, Creighton

UniversityStephan Wagner, Swiss Federal

Institute of TechnologyTillmann Wagner, Texas Tech

UniversityMargot Wallace, Columbia

College, ChicagoGianfranco Walsh, University of

Koblenz–LandauDou Wenyu, City University of

Hong KongPete Whalen, University of DenverDarin White, Samford UniversityKlaus-Peter Wiedmann, Leibniz

University of HannoverGary Wilcox, University of Texas

at AustinWilliam Wilkie, University of

Notre DameElizabeth Wilson, Suffolk

UniversityCarl Witte, Roosevelt University

XYing Xie, Washington University

in St. LouisYu Henry Xie, College of

Charleston

YChun-Ming Yang, Ming Chuan

UniversityWael Yassin, Vision ProSengun Yeniyurt, Rutgers

UniversityFrederick Yim, University of

Akron

ZM. Abrahim Zaka, Nyenrode

Business UniversityHaisu Zhang, University of Illinois

at ChicagoJingyun Zhang, Bowling Green

State UniversityYimin Zhu, Sun Yat-sen University

American Marketing Association / Winter 2010 1

MAKING SALESPEOPLE TO LEAD THEMSELVES: ANTECEDENTSAND CONSEQUENCES OF SELF-LEADERSHIP STRATEGIES

Nikolaos G. Panagopoulos, Athens University of Economics & Business, GreeceGeorge J. Avlonitis, Athens University of Economics & Business, Greece

SUMMARY

Leading salespeople represents an activity of para-mount importance for many B2B organizations. Due tocertain idiosyncrasies, however, salespeople are oftenless readily susceptible to control and leadership interven-tions. As a consequence, firms have been seeking alterna-tive ways for controlling and leading their salespeople.One such way that is gaining momentum among practitio-ners and researchers is empowering employees to engagein self-leadership strategies (SLS). In spite their impor-tance, however, prior studies have approached leadershipas a top-down phenomenon, paying less attention towhether salespeople can and should lead themselves.Against this background, this study seeks to contribute tothe current literature by examining (a) what are the keydimensions of self-leadership strategies (SLS), (b) howindividual differences influence SLS, and (c) whetherSLS are related to performance.

We take a predisposition perspective to study theantecedents and outcomes of SLS. Specifically, we drawon the big five personality research (Goldberg 1992), coreself-evaluation trait theory (Judge et al. 1999),self-regulation theory (Carver and Scheier 1981) and self-management theory (Manz 1986) to develop a compre-hensive conceptual framework. Specifically, we reviewseveral diverse lines of literature to identify importantconstructs with documented importance to SLS research.In short, we theorize that SLS are influenced by a set ofprominent personality traits: organizational-based self-esteem, need for autonomy, internal locus of control,extraversion, conscientiousness, and self-monitoring. Inaddition, we posit that SLS influence salesperson perfor-mance both directly and indirectly through self-efficacy;thus, we hypothesize a partial mediation effect of self-efficacy on the relationship between SLS and perfor-mance. Finally, we include several covariates in ourmodel.

The framework is tested using data from 306 sales-people working for five major organizations operatinginto pharmaceuticals (1 firm), telecommunications (3firms), and financial services (1 firm). Measurement scalesare subjected to CFA to establish their validity and reli-ability. Hypothesized relationships are subsequently testedby using a covariance-based SEM approach.

Our findings indicate that SLS are influenced bycertain personality traits such as extraversion, organiza-tion-based self-esteem, and self-monitoring, whereas SLSare found to be related to both self-efficacy and salesper-son performance. Moreover, self-efficacy exerts a directeffect on performance whereas we find support for theposited partial mediation effect of self-efficacy on therelationship between SLS and performance.

Developing leaders at all levels of the sales organiza-tion has been recognized as an important research priority(Ingram et al. 2005). Our study is a first attempt towardhighlighting that self-leadership is a meaningful andimportant construct for boundary-spanners. This is espe-cially important since salespeople are usually left outalone in the field, struggling with a relentless economyand increased customer expectations. Moreover, thepresent study provides insights into what drives sales-people to engage in SLS and whether SLS influencesperformance. Specifically, by developing and testing atheoretical framework of SLS antecedents and conse-quences, our findings have important ramifications forboth theory and practice.

From a theoretical standpoint, the results show that-at least to some extent-SLS are a function of personalitytraits. In addition, SLS are highly effective since theyexplain a relatively large proportion of salesperson per-formance, even after controlling for several variables.These results increase our knowledge into whether andhow companies should invest in developing such skills intheir salespeople. In addition, we provide a validated scaleof SLS that can be employed in future studies. Specifi-cally, we provide evidence that self-leadership is a higher-order construct comprising three general dimensions/strategies: (a) behavior-focused strategies, (b) naturalreward strategies, and (c) constructive thought patternstrategies (Houghton and Neck 2002).

Our study offers a series of important conclusions forsales executives as well. In particular, SLS might bepursued in order to reduce job demands on the designatedleader and increase salespersons’ empowerment. Thismight be fundamental for addressing challenges, whichare posed by an uncertain and complex environment.Having demonstrated the importance of SLS for enhanc-ing performance, companies should motivate salespeople

2 American Marketing Association / Winter 2010

to engage in SLS. Moreover, recruiting policies need to beadjusted in that assessment should take personality factorsinto account such as those that were found to explain SLS.Sales executives might also want to consider specifictraining interventions for salespeople who are naturallyless inclined to practice SLS. Even salespeople who areless extroverted, for instance, will benefit from learning

and utilizing self-leadership strategies to seek new oppor-tunities. In fact, SLS can be learned and used effectively,even by people who are not natural self-leaders (Stewartet al. 1996). Therefore, we propose that it would be timewell spent for nascent and experienced salespeople tobecome more familiar with self-leadership concepts. Ref-erences are available upon request.

For further information contact:Nikolaos G. Panagopoulos

Athens University of Economics & Business76 Patission, Athens, 104 34

GreecePhone: +302108203461

E-Mail: [email protected]


SALES COMPLEXITY: CONCEPTUALIZATION AND A NEW MEASUREOF COMPLEXITY IN THE SELLING FUNCTION

Christian Schmitz, University of St.Gallen, SwitzerlandPhilipp Rader, University of St.Gallen, Switzerland

SUMMARY

Introductions and Definition of “Sales Complexity”

Today’s business world is characterized by dynamicchanges in customer demand and competitor action, whichsuccessful companies have to react to or even forego byadapting their own internal routines and products accord-ingly. Sales representatives operating at the edge of mar-ket and company spheres in a boundary-spanning positionare exposed to this increasing complexity. However, aholistic model which examines the phenomenon of com-plexity in the selling function has not been developed yet.

In sociology, and within this field especially in sys-tems theory, complexity is interpreted as greater freedom(Vanderstraeten 2005), and complex system as consistingof numerous dynamically and nonlinearly interactingelements (Cilliers 1998, 2000; Backlund 2002). Psychol-ogy predicts that the more complex a task becomes themore decision necessities and information processing arenecessary (Haerem and Rau 2007; Bonner 1994; Wood1986). Other researchers stress missing transparency (i.e.,the non-availability of information), dynamic changes,unclear goals, and the number and (non-linear)interconnectivity of variables as characteristics of com-plex problems (Funke 1991; Frensch and Funke 1995).Combining the essentials of psychological and sociologi-cal complexity definitions, these approaches give way toa coherent definition of sales complexity: Sales complex-ity represents a status of the salesperson’s working envi-ronment that is characterized by a significant number ofnew, intransparent and polytelic tasks, time and/or re-source constraints, and dynamic changes.

Conceptualization

Remembering the boundary spanning position of thesalesperson (Jaramillo et al. 2006), Grössler et al. (2006)suggest to categorize possible complexity drivers in acompany-external and company-internal sphere.

External Sphere of Complexity

Technological Complexity contains the time-con-suming search for information about the frequent innova-tions from outside the company (i.e., from competitorsand customers) to convince costumers of the advantages

of the own product (Ingram 2004). Competitive Complex-ity consolidates all factors related to globalization andadvancing deregulation, which give formerly distant com-petitors a chance to enter previously protected marketsand hamper to achieve high margins – as the increasingsimilarity of products does, too (Ingram 2004; Matthyssensand Vandenbempt 1998). Customer Complexity resultsfrom the customer’s demand for individualized solutions(Moorman and Rust 1999; Du et al. 2003; Pine 1993) andrequires time-intensive assessment and data managementof customer behavior and needs as well as individualsupport, which can be perceived as “complexity costs”(Blecker et al. 2005; Blecker and Abdelkafi 2006). Besides,the increasing presence of buying centers in B2B relationsdemands even more efforts from the sales representativesto achieve above-average margins (Ingram 2004).

Internal Sphere of Complexity

Role Complexity: Within the organization, visiblecontrol systems facilitate higher performance (Babakuset al. 1996; Cravens et al. 2004), whereas intransparentstructures might stimulate, “role stress” (Jackson and Tax1995) which eventually decreases organizational com-mitment and overall performance (Jaramillo et al. 2006;Low et al. 2001). Incentive System Complexity derivesfrom intransparent and incomprehensible compensationor incentive programs, which complicate a correctprioritization of conflicting demands from customers, themanagement, and predetermined sales objectives (Larsonand Resney 2004; Sharma 1997; Filipczak 1993). SalesTechnology Complexity reflects the advantages of stan-dardized, more efficient information systems, reduced byoccurring media disruptions, unreliable IT and the neces-sity to integrate and learn the handling of these tools(Parthasarathy and Sohi 1997; Rangarajan et al. 2005;Backlund 2002).

Study Findings and Conclusion

Based on a first study in which qualitative expertinterviews with 24 sales managers of industrial manufac-turing companies allowed to confirm the major complex-ity drivers identified in literature, the final questionnairewith 32 items was generated. A second quantitative studywith 231 valid questionnaires, sent to 288 salespeople ofa multinational industrial manufacturer, allowed a scalepurification with detailed item-analyses, exploratory fac-


tor analyses with principal component factoring and ob-lique rotation, confirmatory factor analyses, and an initialassessment of scale reliability and unidimensionality ofsubscales (Churchill 1979; Gerbing and Anderson 1988).

The procedure revealed a six-factor structure close tothe assumed one, for which exploratory factor and reli-ability analysis results were sufficient. A confirmatoryfactor analysis indicated solid psychometric properties ofour scales and a good overall fit of the model (χ2 /d.f. =1.23, RMSEA = .04, CFI = .97, TLI = .97, Gamma hat =.98) (cf., Gerbing and Anderson 1988; Hair et al. 1998;Bentler and Bonett 1980).

In conclusion, this research project has firstly estab-lished a holistic definition of sales complexity and sec-ondly identified six dimensions of the complexity sales-persons are confronted with.

Organizational drivers from the internal sphere of thecompany are:

• “Incentive System Complexity,”

• “Role Complexity,” and

• “Sales Technology Complexity,”

whereas the external sphere of environmental driverscontains

• “Customer Complexity,”

• “Competitive Complexity,” and

• “Technological Complexity.”

A next step is a validation and generalization with thehelp of a broader sample. Furthermore, a close examinationof the nomological validity of the model is of majorinterest. Lastly, uncovering further sub-factors of eachcomplexity driver could allow calculating a “complexityelasticity” regarding sales peoples’ performance, revenues,and profits.

REFERENCES

Babakus, E., D.W. Cravens, K. Grant, T.N. Ingram, R.W.LaForge (1996), “Investigating the Relationshipsamong Sales, Management Control, Sales TerritoryDesign, Salesperson Performance, and Sales Organi-zation Effectiveness,” International Journal ofResearch in Marketing, 13 (4), 345–63.

Backlund, A. (2002), “The Concept of Complexity inOrganizations and Information Systems,” Kybernetes,31 (1), 30–43.

Bentler, P. and D. Bonett (1980), “Significance Tests andGoodness of Fit in the Analysis of Covariance,”Psychological Bulletin, 88, 588–606.

Blecker, T., G. Friedrich, B. Kaluza, N. Abdelkafi, and G.Kreutler (2005), Information and Management Sys-tems for Product Customization. New York: Springer.

____________ and N. Abdelkafi (2006), “Complexityand Variety in Mass Customization Systems: Analy-sis and Recommendations,” Management Decision,44 (7), 908–29.

Bonner, S.E. (1994), “A Model of the Effects of AuditTask Complexity,” Accounting, Organizations, andSociety, 19 (3), 213– 34.

Churchill, G.A. (1979), “A Paradigm for DevelopingBetter Measures of Marketing Constructs,” Journalof Marketing Research, 16 (1), 64–73.

Cilliers, P. (1998), Complexity and Postmodernism: Un-derstanding Complex Systems. London: Routledge.

____________ (2000), “What Can We Learn from aTheory of Complexity?” Emergence, 2 (1), 23–33.

Cravens, D.W., F.G. Lassk, G.S. Low, G.W. Marshall,and W.C. Moncrief (2004), “Formal and InformalManagement Control Combinations in Sales Organi-zations. The Impact on Salesperson Consequences,”Journal of Business Research, 57 (3), 241–48.

Du, X., J. Jiao, and M. Tseng (2003), “Identifying Cus-tomer Need Patterns for Customization and Person-alization,” Integrated Manufacturing, 14 (5), 387–96.

Filipczak, B. (1993), “Why No One Likes Your IncentiveProgram,” Training, 30 (8), 19–25.

Frensch, P. and J. Funke (1995), Complex Problem Solv-ing: The European Perspective. Hillsdale: LawrenceErlbaum, 3–25.

Funke, J. (1991), “Solving Complex Problems: HumanIdentification and Control of Complex Systems,” inComplex Problem Solving: Principles and Mecha-nisms, R. Sternberg and P. Frensch, eds. Hillsdale:Lawrence Erlbaum, 185–222.

Gerbing, D.W. and J.C. Anderson (1988), “An UpdatedParadigm for Scale Development Incorporating Uni-dimensionality and Its Assessment,” Journal of Mar-keting Research, 25 (May), 186–92.

Grössler, A., A. Grübner, and P. Milling (2006), “Organi-zational Adaptation Processes to External Complex-ity,” International Journal of Operations & Produc-tion Management, 26 (3), 254–81.

Haerem, T. and D. Rau (2007), “The Influence of Degreeof Expertise and Objective Task Complexity on Per-ceived Task Complexity and Performance,” The Jour-nal of Applied Psychology, 92 (5), 1320–31.


Hair, J.F., R.E. Anderson, R.L. Tatham, and W.C. Black(1998), Multivariate Data Analysis. Upper SaddleRiver, NJ: Prentice Hall.

Ingram, T. (2004), “Future Themes in Sales and SalesManagement: Complexity, Collaboration, andAccountability,” Journal of Marketing Theory andPractice, 12 (4), 18–28.

Jackson, D. and S. Tax (1995), “Managing the IndustrialSalesforce Culture,” Journal of Business & Indus-trial Marketing, 10 (2), 34–47.

Jaramillo, F., J. Mulki, and P. Solomon (2006), “The Roleof Ethical Climate on Salesperson’s Role Stress, JobAttitudes, Turnover Intention, and Job Performance,”Journal of Personal Selling & Sales Management, 23(3), 271–82.

Larson, M. and R. Resney (2004), “Why ‘Sales ForceEffectiveness’ Isn’t,” Handbook of Business Strat-egy, 5 (1), 233–37.

Low, G., D. Cravens, K. Grant, and W. Moncrief (2001),“Antecedents and Consequences of Salesperson Burn-out,” European Journal of Marketing, 35 (5/6), 587–611.

Matthyssen, P. and K. Vandenbempt (1998), “CreatingCompetitive Advantage in Industrial Services,” Jour-nal of Business & Industrial Marketing, 13 (4/5),339–55.

Moorman, C. and R. Rust (1999), “The Role of Market-ing,” Journal of Marketing, 63, 180–97.

Parthasarathy, M. and R. Sohi (1997), “Salesforce Auto-mation and the Adoption of Technological Innova-tions by Salespeople: Theory and Implications,” Jour-nal of Business & Industrial Marketing, 12 (3/4),196–208.

Pine II, B.J. (1993), Mass Customization: The New Fron-tier in Business Competition. Boston: Harvard Busi-ness School Press.

Rangarajan, D., E. Jones, and W. Chin (2005), “Impact ofSales Force Automation on Technology-RelatedStress, Effort, and Technology Usage among Sales-people,” Industrial Marketing Management, 34 (4),345–54.

Sharma, A. (1997), “Customer Satisfaction-Based Incen-tive Systems: Some Managerial and SalespersonConsiderations,” Journal of Personal Selling & SalesManagement, XVII (2), 61–70.

Vanderstraeten, R. (2005), “System and Environment:Notes on the Autopoiesis of Modern Society,” Sys-tems Research and Behavioral Science, 22 (6), 471–81.

Wood, R.E. (1986), “Task Complexity: Definition of theConstruct,” Organizational Behavior and HumanDecision Processes, 37 (1), 60–82.

For further information contact:Christian Schmitz

Institute of MarketingUniversity of St. Gallen

Dufourstr. 40a9000 St. Gallen

SwitzerlandPhone: +41(0)71.224.25.01

Fax: +41(0)71.224.28 35E-Mail: [email protected]


SALESPERSONS LISTENING ABILITY AS AN ANTECEDENT TORELATIONSHIP SELLING

Tanya Drollinger, University of Lethbridge, AlbertaLucette B. Comer, Purdue University, West Lafayette

SUMMARY

It has been argued to be effective in building success-ful relationships with customers’, salespeople should havegood communications skills, be able to develop trust andmutually beneficial relationships. In the present studyActive Empathetic Listening was proposed as being ameans in which salespeople were able to build thesequalities in relationships through a superior form of listen-ing that enhances understanding the needs and prioritiesof their customers, as well as relaying the message that thesalesperson is indeed concerned for the well being of thecustomer. Active Empathetic Listening (AEL) was definedas listening practiced by salespeople in which activelistening is combined with empathy to achieve a higherform of listening (Comer and Drollinger 1999). AEL isproposed as superior method of listening used by sales-people in developing open, communicative relationshipsthat focus on a full understanding of the customer as wellas concern for their position.

In order to examine the role of AEL as an antecedentto relationship selling Palmatier et al. (2006) frameworkwas employed. Palmatier et al. (2006) developed a meta-analytic framework in which they proposed antecedents,mediators, moderators, and outcomes of relationshipmarketing. In order to focus on professional selling ratherthan a larger framework of marketing the authors focusedprimarily on aspects of the meta-analytic model thatrelated to professional selling. Further, only those conceptsin the model that were well supported in the meta-analysiswere chosen as concepts of interest for the present research.Drawing from the relationship model developed byPalmatier et al. (2006) and the theoretical framework ofAEL in the personal selling process (Comer and Drollinger1999) a model was developed. In the present study AELwas purposed as being an antecedent to a salesperson’scommunication skill, the quality of relationship betweenthe buyer and seller and ability to build trust. Further,communication skill, relationship quality, and trust wereall proposed to moderate the connection between AELand the outcome of sales performance.

Empirical Study

Salespeople were chosen as the target sample for thepresent study considering that many of the hypothesized

relationships focus on personal attitudes and beliefs. Amail survey was sent to salespeople who were randomlyselected from a commercial mailing list. One hundredseventy five (175) completed and useable questionnaireswere received, the questionnaire consisted of severalmulti-item instruments that were measured using Likert-type 7 point scales. Following Anderson and Gerbings’(1988) two step procedure in testing structural equationmodels all instruments were examined.

Results and Discussion

Overall, model fit indices show the relationshipsrepresented in the model were a good fit with a reportedGoodness of Fit (GFI) .954, CFI .953, and AGFI (.893)that indicated that the model was statistically plausible.The hypothesized relationships in the model were gener-ally supported. More specifically the relationship betweenAEL and interpersonal communication skills (H1) wassignificant with a standardized path coefficient of .443.Active Empathetic Listening was also hypothesized to bepositively related to relationship quality (H2) and receivedsupport through significance and a resulting standardizedpath coefficient of .819. Active Empathetic Listening andtrust (H3) was also positively and significantly relatedwith a coefficient of .825. Interestingly the standardizedcoefficient of the path between AEL and trust was thegreatest indicating that as AEL goes up by one standarddeviation trust goes up .825 standard deviations. Commu-nication skills were also positively and significantly relatedto performance (H4) with a path coefficient of .257.Relationship quality was found to be positively associatedbut not significantly related to performance therefore H5was not supported. Lastly, trust was found to be positivelyand significantly related to performance with a standard-ized path coefficient of .241 (H6).

When examining the hypothesized relationships underthe Relationship Marketing framework presented byPalmatier et al. (2006) it is apparent that they weregenerally supported directionally and via their signifi-cance. AEL was positively related to greater communica-tion ability, fostering relationships and high levels oftrust. When salespeople utilize AEL they enable them-selves to better understand the needs and priorities of theircustomers and in the process instill a sense of sincerity andcredibility. Managers whose sales force is engaged in


dyadic communications with buyers and are interested indeveloping long term relationships with their customersmay benefit most from this study. AEL may be seen as a

means in which salespeople can enhance all of theserelationship building skills and in turn enhance overallperformance of their sales force.

For further information contact:Tanya Drollinger

University of Lethbridge4401 University Drive West Lethbridge

Alberta, Canada T1K 3M4Phone: 403.329.2174

Phone: [email protected]

Lucette B. ComerConsumer Sciences and Retailing

Purdue UniversityWest Lafayette, IN 47907



APPLYING FLOW THEORY TO CONSUMER BEHAVIORIN RETAIL STORES

Liz C. Wang, University of Dallas, Irving

SUMMARY

A flow experience is described as an intrinsicallyoptimal state during which an individual is intenselyengaged in an activity (Csikszentmihalyi 1990). Although“flow” is a salient determinant for providing a compellingonline shopping experience, little research has be done tostudy consumers’ flow experience beyond the confines ofthe Internet. As retail stores account for 90 percent ofretail sales, it is important to know whether existing flowtheory can be extended to retail stores.

Theoretical Background and Research Hypotheses

Massimini and Carli (1988) suggested that challengeand skill are the antecedents of a flow experience. Mathwicket al. (2004) further summarized the four states of mind asfollows.

Flow state – high challenges and high skill. In orderto deal with a highly challenging task, skilled con-sumers at a flow state are highly focused, enjoy theactivity at hand and feel in control.

Boredom state – low challenges and high skills.Consumers may feel bored when they are highlyskilled to deal with an activity, but the activity at handis not sufficiently challenging.

Anxiety state – high challenges and low skills. Con-sumers may feel anxious when they don’t have thenecessary skills to cope with the task or activity theyface.

Apathy state – low challenges and low skills. Inaddition, consumers may have little desire to workwith the activity at hand.

Nevertheless, the online flow literature mainly focuseson navigational challenges and consumer Internet/com-puter skills. This study takes the first step to propose in-store shopping as a flow activity by constructing thechallenge and consumer skill associated with this activity.The marketing literature clearly suggests that a consumer’spurchase process may involve challenges due to thenature of product complexity and inherent risks(Swaminathan 2003; Keller and Staelin 1987), and thatconsumer skills may vary given the same product cat-egory (Park and Lessig 1981). Thus, a single shopping

task can be associated with various levels of challengesand skills.

Flow experience includes intrinsic enjoyment, controland concentration dimensions (Ghani and Deshpande1994). Csikszentmihalyi (1990) described an individual’sconcentration as psychic energy and one’s totalconcentration is required in order to deal with a highlychallenging task (Mathwick et al. 2004). A person feelsdominant when he is able to control the situation that he isin. (Ressell and Mehrabin 1974). An individual in a stateof flow is lacking a sense of worry about losing controlbecause all his cognitive resources are free to addresscurrent challenges, and to exercise his or her skills in anattempt to overcome these challenges (Csikszentmihalyi1990). Such experience may increase shopping excitementand pleasure (Koufaris 2002). As previous researchsuggest, these dimensions are positively related toconsumer future shopping intention (Kourfaris 2002;Wang et al. 2007). Therefore, the research hypotheses aredeveloped:

H1: During the course of store-based shopping, thelevel of concentration is the highest for consumersin the state of flow among the four states of mind.

H2: During the course of store-based shopping, thelevel of perceived control is the highest for con-sumers in the state of flow among the four states ofmind.

H3: During the course of store-based shopping, thelevel of shopping enjoyment is the highest forconsumers in the state of flow among the four statesof mind.

H4a: Concentration is positively related to future shop-ping intention.

H4b: Control is positively related to future shoppingintention.

H4c: Enjoyment is positively related to future shoppingintention.

Research Method and Analysis

A survey (n = 240) was used to test the researchhypotheses. Each respondent recalled a goal-directed


shopping task at retail stores within a month, and reportedhis or her shopping experience for that particular shop-ping activity in the self-reported survey. Approximately54.6 percent were female. The age groups in the samplevaried greatly with 33.7 percent of the sample between18–25, 28.4 percent between 26 and 35, 16.6 percentbetween 36 and 45, 18.3 percent between 46 and 60, and2.9 percent 61 or older.

The measures included shopping challenge, skill,three flow dimensions and future shopping intention,which were adapted from previous online flow research(e.g., Koufaris 2002; Mathwick et al. 2004; Wang et al.2007). A confirmatory analysis was followed to assess thevalidity. A cluster analysis was used to classify respon-dents into four states of experience according to theirperceived challenges and skills with their shopping tasks(Mathwick et al. 2004). A K-means clustering procedurewith the initial centers and the four-cluster solution wasused to obtain the final clusters. ANOVA results indicatedthat the four clusters were significantly different withrespect to challenge (F

(3, 236) = 177.41, p < 0.000) and skill

(F(3, 236)

= 201.44, p < 0.000). ANOVA analysis with pair-wise comparisons among the four states was followed totest each hypothesis. H1, H3, and H4 (a) – (c) weresupported. For H2, the mean of the perceived control

(5.17) in the flow state was statistically higher than thosein the apathy (4.41) and anxiety (4.51) states, but lowerthan that in the boredom state (5.82).

Research and Managerial Implications

This study first provides a theoretical foundation todevelop the antecedents (challenges and skills) associatedwith in-store shopping activities. Second, it providedempirical evidence to examine the impacts of flowexperience on consumer behaviors at different states ofmind in retail stores settings. The study also helps retailersbetter understand consumers’ states of shoppingexperience. For retailers selling more complex products,shopping for these products can be very challenging. Thefindings imply that retailers may switch their shoppersfrom the anxious state to the flow state by increasingconsumer skills, such as providing help with productknowledge. On the other hand, consumers who havegreater skills to handle less challenging shopping activitiestend to feel bored, but perceive a higher level of control intheir shopping experience. The study found that perceivedcontrol is positively related to future shopping intention.Hence, these retailers should at least ensure their consumersfeel in control with their shopping activities.

For further information contact:Liz C. Wang

University of Dallas1845 East Northgate Drive

Irving, TX 75062Phone: 972.721.4058

Fax: 972.721.4007E-Mail: [email protected]


DEVELOPMENT AND VALIDATION OF THE TRANSACTION SPECIFICATTRIBUTIONS OF DISCRIMINATION (TSAD) SCALE

Thomas L. Baker, Clemson UniversityTracy Meyer, University of North Carolina – Wilmington

SUMMARY

Discrimination, and attributions of discrimination,have been and will continue to be a critical issue not onlyin the United States but around the world. Discriminatoryattributions occur when perceptions of unfair treatmentare attributed to group membership rather than individualfactors (Major, Quinton, and McCoy 2002). Marketplacediscrimination, which Crockett, Grier, and Williams (2003)define as “differential treatment of customers in the mar-ketplace based on perceived group-level traits that pro-duce outcomes favourable to ‘in-groups’ and unfavorableto ‘out-groups,’” can be said to be a special case ofdiscrimination occurring market transactions. In fact, anumber of researchers (e.g., Chung and Myers 1999;Feagin and Sikes 1994; Hill 2001; Oliver and Shapiro1995) have argued that dealing with discrimination in themarketplace is an almost everyday concern for manyconsumers.

To date, most research in marketing focusing ondiscrimination and discriminatory attributions has beenconducted relative to racial or ethnic discrimination (e.g.,Baker, Meyer, and Johnson 2008; Siegelman 1997;Rosenbaum and Montoya 2006). However, there are anynumbers of other types of discrimination which can occurin the marketplace including discrimination due to age,gender, disability, and sexual orientation. And whilesome have argued that the incidence of discrimination hasdeclined over the past decades, Deitch et al. (2002)indicate that overt discrimination has been replaced bymore covert forms of discrimination which may be moreharmful due to the “attributional ambiguity” associatedwith covert discrimination (Crocker and Major 1989).Thus, it would appear that additional research relative tomarketplace discrimination would be warranted. This isparticularly true as it relates to the provision of servicessince at its core service provision is an interpersonalinteraction between two people.

However, one issue which might inhibit research inthis area is a measure of what might be termed “transactionspecific attributions of discrimination.” To date mostmeasures of discrimination that exist might best be con-sidered measures of “long-term” discrimination in thatthey assess the extent to which a person has been exposedto actions which might be attributed to discriminationover a period of time. These measures do not allowresearchers to assess the extent to which a consumer’sperception of a specific transaction is being attributed todiscrimination. We believe there is potentially a broadapplication for a measure like this, particularly when thepurpose is to study interactions between customers andemployees.

This paper reports two studies, the first using studentsand the second using a national sample, designed todevelop a measure of Transaction Specific Attributions ofDiscrimination (TSAD) which can be used to assess theextent to which a person makes attributions of discrimina-tion when faced with a negative event such as a servicefailure. Analyses provide strong support for a five-itemmeasure (see Table 1).

Given the critical importance of the interactionbetween the customer and the employee, we believe theinclusion of a measure of transaction specific attributionsto discrimination might prove to be useful in helping todiagnose those situations when the outcome from a ser-vice encounter or sales presentation is less than optimal.Our conceptualization and operationalization of the mea-sure at the transaction specific level is important given thatmost measures of discrimination asses what might betterbe labeled “long term” or “chronic” discrimination.Researchers will now have a multi-item, psychometri-cally sound measure that can be used to assess discrimina-tory attributions arising from a particular situation.


TABLE 1Summary of Scale Validation

λλλλλ valuesb,c λλλλλ valuesb,d

Factor (t-values) from (t-values) fromItem Loadingsa Study 1 Study 2

1. This service failure was due to age (racial) .83 .83e .70discrimination. (f) (d)

2. If I were older (White), this failure would not have .82 .86 .83occurred. (19.36) (9.82)

3. I believe the service failure was a purposeful act of .67 .66 .83vengeance due to my age (race). (13.30) (9.83)

4. I received poor service because I am young (Black). .92 .93 .8621.66 (11.28)

5. This subtle slight is an act of discrimination. .86 .82 .96(18.13) (11.31)

Average Variance Extracted .68 .74

Composite Reliability .95 .96

a Factor loadings from exploratory factor analysis.b Values from confirmatory factor analysis.c CFA Fit Statistics: χ2 = 6.85

(5, p = .23), CFI = .99, IFI = .99, RMSEA =.04.

d CFA Fit Statistics: χ2 = 5.98(5, p = .31)

, CFI = .99, IFI = .99, RMSEA =.03.e Completely standardized loadings.f No t-value since item was constrained to 1 to set the scale metric.

For further information contact:Thomas L. Baker

Department of Marketing234 Sirrine Hall

Clemson UniversityClemson, SC 29634

Phone: 864.656.2397E-Mail: [email protected]


A RESEARCH ON THE CURVE RELATIONSHIP BETWEEN GIFTPROMOTION DEPTH AND THE SPILLOVER EFFECT

FROM GIFT PROMOTION

Chung-Hui Tseng, Tamkang University, Taiwan

ABSTRACT

This research conducted two studies to investigatewhether the discounting perception toward the gift in giftpromotion would spillover to other products sharing thesame brand as the gift, or spillover to the same product ofanother brand. Results showed that promotion depth andspillover effect showed a curve relationship.

KEYWORDS: Spillover effect, Gift promotion, Pro-motion depth, Perceived value.

INTRODUCTION

Nowadays, more and more firms are investing fundsin promotional activities. According to a statistical reportfrom Promotion Marketing Association in 2008, morethan $350 billion of packaged goods coupons are offeredannually. In other words, the importance of promotion isobvious and ubiquitous. Among kinds of promotion strat-egies, gift promotion is an essential strategy in actualpractice. Sales from gift promotions totaled almost threehundred million in 1996, and it will keep growing at a rateof 18 percent annually (Bertrand 1998; D’Astous andLandreville 2003). Since gift promotion is perceived as aprominent strategy by marketers, it is necessary to studygift promotion more deeply.

In reviewing literature related to gift promotion, itseems that most researchers took it for granted that giftpromotion can positively influence sales. However, willthe perceived value of the gift promotion also improve asthe gift price rises? In this research, the author inferredthat when the gift price is too high, consumers wouldquestion the quality of the gift, and a negative effect mightresult. Then, the negative impression could spill over toinfluence the evaluations of products relative to the gift.The author referred this phenomenon as the spillovereffect of gift promotion in this research, and it is the mainsubject of the investigation.

The spillover effect has long been discussed in thefields of brand alliance and brand extension. Researchersfound that evaluation of brand A would spill over to theevaluation of brand B under a relationship of brandalliance (Balachander and Ghose 2003; Baumgarth 2004;Desai and Keller 2002; Janiszewski and Van Osselaer2000; Park, Jun, and Shocker 1996; Samu, Krishnan, andSmith 1999; Simonin and Ruth 1998; Vaidyanathan and

Aggarwal 2000). However, researchers seldom discussedthe spillover phenomenon in relation to gift promotion;that is, whether evaluation of the gift of brand A may spillover to influence the evaluation of the same gift productof brand B; reviewing the literature on gift promotion, itseems that only Raghubir (2004) has discussed this par-ticular issue. He proposed that discounting the evaluationof a gift would negatively influence the evaluation of thesame product of other brand with the same gift. However,his research only offered empirical findings related to thisphenomenon; it lacked discussions regarding the exactreason and theoretical foundation required to explain whythe phenomenon occurred. Therefore, this research pro-poses a theoretical foundation of to determine the spillovereffect in gift promotion, and investigate it more deeply.

In brief, this research employs two studies. Thepurpose of study one was to verify whether the spillovereffect in gift promotion did exist. If it did exist, study twowould further investigate the trend of the spillover effectfrom gift promotion as the price of the gift increased. Inother words, study two would examine whether thespillover effect in gift promotion would present a linear ora curve type.

STUDY ONE

Literature Review and Hypotheses Development

A review of the literature related to the spillovereffect, most researches focused on brand extension orbrand alliance (Balachander and Ghose 2003; Baumgarth2004; Desai and Keller 2002; Janiszewski and Van Osselaer2000; Park, Jun, and Shocker 1996; Samu, Krishnan, andSmith 1999; Simonin and Ruth 1998; Vaidyanathan andAggarwal 2000). The spillover effect in brand extensionimplies that the evaluation of a new product (cf., Adidasmobile) will spill over to affect the evaluation of the parentbrand (Adidas); Balachander and Ghose (2003) definedthis phenomenon as the forward spillover advertisingeffect. On the other hand, the evaluation of the parentbrand would also spill over and affect the evaluation of anew product, which was defined as the reciprocal spilloveradvertising effect (Balachander and Ghose 2003). Inaddition, researches on brand alliance also attested to theexistence of spillover effect (Park, Jun, and Shocker 1996;Simonin and Ruth 1998). Brand alliance means twobrands forging an alliance in order to sell products together,for example Disney’s toys and MacDonald’s hamburgers


constitute an alliance to sell their products together. Underthis situation, the evaluation of the products in the alliancewould spill over to influence the evaluation of the indi-vidual brands, MacDonald’s and Disney.

Researchers adopted economics of information tomake some explanation about the cause of the spillovereffect (Morein 1975; Balachander and Ghose 2003).According to economics of information, the parent brand,through the Halo effect, may influence related products,called umbrella-branded products. In order to simplifyinformation processing, consumers tend to make evalua-tions of umbrella-branded products on the basis of impres-sions regarding the parent brand (Wernerfelt 1988), andattesting to the appearance of the spillover effect. Con-cerning the direction of the spillover effect, past researcheshave not shown consistent results. Some researchersclaimed that the spillover effect is a positive one (Simoninand Ruth 1998; Washburn, Till, and Priluch 2000), othersstated that it was negative (Keller and Aaker 1992; Lokenand John 1993; Till and Shimp 1998).

Raghubir (2004) proposed that the spillover effectwas negative in relation to gift promotion. He found thatconsumers tended to discount the gift’s value, and thediscounting perception would carry over to the sameproduct category with the same gift; in his experimentalsetting, the promotion ads read: “Buy a bottle of NapoleonCourvoisier US$52 (or a bottle of Hennessy XO US$90),and get a pearl necklace.” After participants read this ad,they read another ad: “Majorica pearl necklace US$149,”and then they were asked to write down their perceptionregarding the true value of this “Majorica” pearl necklace.The results showed that the perceived price revealed asignificant underestimating. Therefore, negative spillovereffect had occurred. Analyzing this experimental manipu-lation, we can observe a spillover effect from the giftpromotion; however, it was not clear whether the spillovereffect could appear in other product settings or was causedby the price or brand. Therefore, this research developedtwo studies: study one examined whether the spillovereffect of gift promotion would reoccur in other promotionsettings, and study two investigated the relationshipbetween the gift’s price and the spillover effect.

Based on Raghubir’s study (2004), the spillovereffect in gift promotion appeared to exist; however,Raghubir did not offer a strong theoretical foundation toexplain this phenomenon. In fact, the spillover effect fromgift promotion may be explained by Assimilation andContract Theory (Herr, Sherman, and Fazio 1983),whereby people tend to retrieve a category from theirmind in order to process external stimuli, and then ifanother external stimulus is similar to the former one,people will classify the latter stimulus within the samecategory. Thus, the assimilation effect occurs. Conversely,

if a second external stimulus is dissimilar to the formerone, people will classify the latter stimulus within anothercategory. Thus, contrary effects occur.

In the light of the assimilation effect, the evaluationof the former stimulus may spill over to affect the latterstimulus. The gift in promotion ads could be a formerstimulus, and a product related to the gift, as a latterstimulus. Because the two stimuli are similar, the assimi-lation effect may take place. Thus, the gift evaluation mayspill over to influence the evaluation of other productsrelated to the gift. Hypothesis one is as follows:

H1: Gift promotion will cause a spillover effect on aproduct related to the gift.

Method

The purpose of study one was to verify the existenceof spillover effect in gift promotion. The author subdi-vided gift promotion into two groups: present or absent.Dependent variable was spillover effect in gift promotion.The operative definition of spillover effect from giftpromotion was the variation of perceived value of aproduct relative to that of the gift. The author designed themeasurement of the variation of perceived value of aproduct relative to that of the gift on the basis of Raghubir’smeasurement (2004). There were two steps in the mea-surement process. First, before participants read a giftpromotion ad, participants needed to write down howmuch money (X) they would like to pay for a pair of jeans.After they read a gift promotion ad, they again needed towrite down how much money (Y) they would like to payfor a pair of jeans of brand A, which is the same brand asthe gift, but a different product category from that of thegift. Thus, the author measured the variation of perceivedvalue of a product relative to the gift as (Y–X)/X. If thevalue was not equal to zero, the spillover effect occurred.

This study adopted an experiment with two groups(gift promotion: present or absent) between-subject design.The experimental setting of gift promotion involved an adwith the content: “Now, Subscribe for a year of StudioClassroom English magazine at NT$2500, Get a pair ofAdidas running shoes for Free (Market value NT$500).”After reading this gift promotion ad, participants wouldread another page with a picture of a pair of jeans. Then,they needed to write down how much money they wouldlike to pay for a pair of Adidas jeans. In order to avoiddemand artifact, participants needed to fill out two differ-ent questionnaires from different research institutes, asindicated on the top of the two questionnaires. If thespillover effect still took place under this arrangement, itwould strongly support the existence of the spillovereffect. Participants in this study were students from acollege class, chosen via convenient sampling. The author


gathered fifty valid questionnaires; half of them includeda gift promotion ad, while the other half did not, but justsome questions in the first questionnaire.

Results

Regarding the descriptive statistical results in thisstudy, the average amount of money for a pair of jeansbefore reading a gift promotion ad was NT$ 2,921.33 forthe present group and NT$ 2,533.78 for the absent group.After reading a gift promotion ad, the average amount ofmoney for a pair of Adidas jeans was NT$ 1,220.44 for thepresent group and NT$ 2,436.33 for the absent group.Thus, the variation of perceived value of a product relativeto that of the gift was -.58 (SD .26) in the present group,and -.04 (SD .02) in the absent group. The two values ofvariation showed a significant difference (t = 3.23, p <.01); hence, H1 was supported. In addition, the spillovereffect was negative.

Discussion

Based on results of study one, the author found thatthe spillover effect in gift promotion did exist. In addition,the spillover effect occurred with a product of the samebrand but different category compared to the gift. Raghubir(2004) found that the spillover effect would take placewith a product under the same category as the gift (cf., anameless versus a “Majorica” pearl necklace), while thisstudy found the spillover effect also occurred with aproduct under a different category to that of the gift (cf.,Adidas running shoes vs. Adidas jeans).

Following is an investigation into the relationshipbetween price and the spillover effect. Specifically, as theprice is one of the more important factors in consumers’decision-making, study two discussed how price influ-enced the spillover effect in relation to gift promotion. Inbrief, the purpose of study two was to examine whethergift price and spillover effect in gift promotion wouldshow a linear or curve relationship.

STUDY TWO

The purpose of Study 2 was to investigate the rela-tionship between gift promotion depth and spillover effect,as well as to test whether the spillover effect found in study1 would also appear in other brands or products related tothe gift. The findings would be helpful for generalizingthe spillover effect of gift promotion.

Literature Review and Hypotheses Development

Generally, a bundle of gift promotion includes a mainproduct and a gift, and the prices of both usually show uptogether in an ad. Thus, this study proposed that consum-

ers would tend to consider both of these prices in makingtheir evaluation, and that this information processing issimilar to a case of product bundles. Thus, this study tooka variable, gift promotion depth, consisting of the pricesof the gift as well as the main product. In other words, thisstudy took gift promotion depth as a price factor to discussits influence on spillover effect. Reviewing the concept ofpromotion depth (DelVecchio, Krishnan, and Smith 2004;DelVecchio 2005), researchers defined promotion depthas a percentage of promotion value divided by productnormal price, and usually regarded it as an essential factorwhen consumers evaluate promotion activity (DelVecchio,Krishnan, and Smith 2004; DelVecchio 2005). In terms ofpromotion depth, this study defined gift promotion depthas a percentage of the gift’s price divided by the price ofthe main product.

Since gift promotion depth reflects the type of bundleproducts, the author adopted anchoring and adjustmenttheory (Tversky and Kahneman 1974; Lopes 1982; Yadav1994) to make inference about evaluation of the gift aspast researchers had inferred evaluations of bundle prod-ucts. The anchoring and adjustment theory proposed threesteps people use to process complex information, espe-cially information concerning product bundles. First,people scan the combination of product bundles. Second,people will choose the most essential product as an anchor.Finally, people will evaluate the anchor product and thenadjust their final evaluation by other minor products.

In terms of the anchoring and adjustment theory(Tversky and Kahneman 1974; Lopes 1982; Yadav 1994),consumers will choose an anchor product in the giftproduct bundle. When gift promotion depth was low,consumers would tend to choose the main product as ananchor because the price of the main product was obvi-ously higher than the gift’s price; the gift, at this moment,was not taken as an important factor. Then, when consum-ers later evaluate the products relative to the gift, based onthe assimilation effect (Herr, Sherman, and Fazio 1983),the former evaluation of the gift will spill over lightly tothe product relative to that of the gift, as inferred in studyone. Therefore, the negative spillover effect (based onstudy one) from low gift promotion depth was of smallextent.

As the gift promotion depth was higher, but stillbelow a reasonable depth, consumers should still choosethe main product as an anchor, because the price of themain product was still higher than the gift’s price under areasonable promotion depth. At the same time, higherprice of gift can be a plus to consumers’ evaluation. Then,when consumers would later evaluate products relative tothat of the gift, based on assimilation effect (Herr, Sherman,and Fazio 1983), the former positive impression wouldspill over to products related to the gift. Thus, a negative


spillover effect from low to reasonable gift promotiondepth would be offset to some extent, so it would remainslight.

Finally, as gift promotion depth was above a reason-able depth and kept higher, the gift’s price may becomehigh enough to make consumers pay greater attention toit, whereby the gift might possibly be an anchor. Consum-ers may become suspicious regarding its value and conse-quently devaluate the gift’s value. Then, when consumerslater evaluate products relative to that of the gift, based onassimilation effect (Herr, Sherman, and Fazio 1983), theformer devaluation would spill over to products relative tothat of the gift. Hence, the negative spillover effect fromreasonable to high gift promotion depth would becomemore critical.

On the basis of the above discussion, this studyproposed that the relationship between gift promotiondepth and spillover effect would exhibit a curve type. Inaddition, the spillover effect on products related to the giftcan be present in regard to two kinds of conditions. Oneis same product category with the gift but different brandfrom the gift (called “SCDB” for short), and the other isthe same brand with the gift but different category fromthe gift (called “SBDC” for short). Therefore, this studyproposed another two hypotheses as below.

H2: The gift promotion depth and spillover effect on thesame product category as the gift but different brandfrom the gift (SCDB) would be a curve relationship.

H3: The gift promotion depth and spillover effect on thesame brand as the gift but different category from thegift (SBDC) would be a curve relationship.

Method

The purpose of study two was to examine the rela-tionship between gift promotion depth and spillover effectfrom gift promotion; the independent variable was giftpromotion depth, which the author defined as a percent-age of gift’s price divided by the price of main product.The author manipulated gift promotion depth in ten groups:10 percent, 20 percent, 30 percent, . . , 90 percent, 100percent. The two dependent variables were spillover effecton the same product category as the gift and that on thesame brand as the gift. The operative definition of spillovereffect from gift promotion was the variation of perceivedvalue of a product relative to that of the gift, and themeasurement was the same as in study one.

This study adopted an experiment with a 10-group(gift promotion depth) between – subject design. Theexperimental setting of the gift promotion, 10 percentdepth for example, was an ad with the content: “Now, Buya pair of Nike running shoes for NT$2,000, and Get a

Casio watch for Free (Market value NT$200).” In addi-tion, sunglasses were an experimental product in spillovermanipulation. Because the brand Casio does not producesunglasses, evaluations toward Casio sunglasses wouldbe almost fair due to the lack of past purchase experiences.Regarding spillover brands, this study picked four brandsbased on a consideration of national or local brands: PierreCardin (national brand), Philips (national brand), MUJI(local brand) and Working House (local brand) watches.

In the questionnaire, participants were first asked towrite down how much money they would like to spend onbuying a watch. Later, they were invited to read a giftpromotion advertisement, and then asked to write downhow much money they would like to spend for the fourbrands and one product. In order to avoid demand artifact,participants needed to fill out two different questionnaires(one on gift promotion, and the other on spillover effect)from different research institutes as indicated on the top ofthe two questionnaires.

Participants in this study were college students. Byconvenient sampling, the author invited students in acollege library to fill out the questionnaire; then theycould participate in a sweepstake activity. This studygathered four hundred valid questionnaires finally.

Results

Four experimental brands and one product all showednegative perceived spillover value (see Table 1). In addi-tion, the trend of perceived spillover value on each brandor product from 10 percent~100 percent promotion depthmanifested a curve shape (like an inverted U shape, basedon Table 1).

In order to test the non-linear relationship betweengift promotion depth and spillover effect, this study adoptedthe quadratic regression model for the analysis. Thequadratic beta value of gift promotion depth of all experi-mental brands and products were significant (Table 2).The quadratic beta value of gift promotion depth on thePierre Cardin watch was: -.46 (p < .01), on the Philipswatch: -.76 (p < .01), on the MUJI watch: -.68 (p < .01),on the Working House watch: -.84 (p < .01), and on Casiosunglasses: -.74 (p < .01). The result showed a curverelationship between gift promotion depth and spillovereffect both existed in regard to the same product categoryas the gift but with a different brand from that of the giftand on the same brand as the gift but with a differentcategory from the gift. Thus, H2 and H3 were supported.

Discussion

Based on the results of study 2, the spillover effectfrom gift promotion was extended to four relative brandsand one product category. In addition, a curve relationship


between gift promotion depth and spillover effect werealso found. In brief, the results supported the hypotheses.Based on the inferences above, if gift promotion depthwas higher than the perceived reasonable depth, the evalu-ation of the gift would start to change and result in a curvetype, as we can see according to the statistical results.

CONCLUSIONS

According to study one, the author found that thespillover effect from gift promotion did exist. Study twoaimed to verify if the spillover effect still appeared inregard to different products or brands, or if it wouldexperience some price factor fluctuation. According tothe results of the two studies, although the experimentalproducts in the two studies were different, the spillovereffect appeared in both studies. Besides, there were fourbrands in the experiment of study two, and the authorfound that the spillover effect would occur in regard tonational brands as well as local brands. Thus, we con-cluded that the spillover effect from gift promotion canoccur in relation to different products and brands indifferent regions; apparently, the spillover effect is ageneral phenomenon in gift promotion. In addition, basedon the results of this research, we found that gift promo-tion depth was a significant factor influencing the appear-ance and fluctuations of the spillover effect. Specifically,gift promotion depth was calculated by the prices of thegift and the main product; thus, price also plays animportant role in the spillover effect of gift promotion.

TABLE 1Average Perceived Spillover Value vs. Each Gift’s Promotion Depth

Perceived Spillover Value

Gift Promotion Pierre Cardin Philips MUJI Working Casio SampleDepth Watch Watch Watch House Watch Sunglasses Size

10% -.14 -.40 -.34 -.45 -.39 40

20% -.09 -.39 -.29 -.40 -.38 39

30% -.08 -.31 -.25 -.3 -.25 41

40% -.05 -.26 -.20 -.24 -.24 40

50% -.09 -.30 -.26 -.26 -.24 40

60% -.14 -.31 -.26 -.30 -.25 39

70% -.15 -.42 -.29 -.36 -.25 41

80% -.23 -.48 -.34 -.41 -.44 40

90% -.25 -.48 -.41 -.49 -.44 40

100% -.34 -.53 -.49 -.58 -.55 40

Implications

This research offers some extended inference fromtwo classical theories. Specifically, this research appliedassimilation and contrast theory (in study one) and anchor-ing and adjustment theory (in study two) to infer thespillover effect in gift promotion. Anchoring and adjust-ment theory was used for adoption in product bundles, andthis research extended it to the product bundle of giftpromotion; the anchoring and adjustment theory is aclassical theory in the marketing field. Moreover, thisresearch tried to combine anchoring and adjustment theoryalong with the assimilation and contrast theory to infer therelationship between gift promotion depth and the spillovereffect from gift promotion. Application of classical theoryin this research may offer some insight for researcher.Furthermore, this research offered a new circumstanceregarding the spillover effect. In the related literature, thespillover effect was discussed in regard to brand extensionor brand alliance; the present research found that it alsoexisted in gift promotion, thereby offering a new perspec-tive of the spillover effect to researchers.

This finding also offers a practical suggestion formarketers; they need to pay more attention to gift promo-tion strategy. If the gift and the main product are toosimilar, for example sharing the same brand or the sameproduct category, a negative spillover effect would beharmful to the gift promotion activity. According toresults of study two, the relationship between gift promo-


tion depth and perceived spillover effect was non-linear.In brief, as gift promotion depth kept increasing, theperceived spillover effect increased at the beginning, thendecreased and then remained negative. This finding impliedthat higher priced gifts are not always good for promotionactivities. If the price of the gift is too high, consumersmay question the gift’s value and make a negative evalu-ation of other brands or products related to the gift. Thus,marketers should make very careful decisions about choos-ing gift price.

Limitations and Suggestions

This research was an initial investigation of thespillover effect in gift promotion, and the experimental

TABLE 2Quadratic Regression Model of Spillover Effect

Beta Value Quadratic BetaSpillover Df F of Gift Promotion Value of GiftTarget Value Depth Promotion Depth

Perceived Pierre Cardin watch 2 16.06** .31* -.46**

Spillover Value Philips watch 2 27.84** .59** -.76**

on SCDB MUJI watch 2 18.12** .55** -.68**

Working House watch 2 22.75** .73** -.84**

PerceivedSpillover Value Casio sunglasses 2 17.61** .63** -.74**

on SBDC

*p < .05 **p < .01

products, based on time and cost, were limited to severalproduct categories and brands. In order to better under-stand the phenomenon of the spillover effect in giftpromotion, future researches could employ different cat-egories and brands. Second, the product bundle of the giftpromotion in the experiment comprised a main productand a gift. Nowadays, a main product with several gifts tobe chosen is more common type in gift promotion; whetherthe spillover effect still happens in this type of gift promo-tion may be worthy of examination in future research.Also, participants in the experiment were college stu-dents; this arrangement tends to decrease the differenceswithin experimental groups. Future researchers could testthe spillover effect in other participants to enhance itsgeneralization.

ENDNOTE

The author appreciated Taiwan government’s subsidy forthis research. Subsidy from National Science Coun-cil (No:NSC 97-2410-H-032-066).

REFERENCES

Balachander, Subramanian and Sanjoy Ghose (2003),“Reciprocal Spillover Effects: A Strategic Benefit ofBrand Extensions,” Journal of Marketing, 67 (Janu-ary), 4–13.

Baumgarth, Carsten (2004), “Evaluations of Co-Brandsand Spillover Effects: Further Empirical Results,”Journal of Marketing Communications, 10 (June),115–31.

Bertrand, Kate (1998), “Premiums Prime the Market,”Advertising Age’s Business Marketing, 83 (5), s6.

D’Astous, Alain and Valerie Landreville (2003), “An

Experimental Investigation of Factors Affecting Con-sumers’ Perceptions of Sales Promotions,” Euro-pean Journal of Marketing, 37 (November/Decem-ber), 1746–61.

DelVecchio, Devon, H. Shanker Krishnan, and Daniel C.Smith (2004), “Cents or Percent? Promotion Fram-ing and Consumer Price Expectations,” Workingpaper, University of Kentucky.

____________ (2005), “Deal-Prone Consumers’ Re-sponse to Promotion: The Effects of Relative andAbsolute Promotion Value,” Psychology and Mar-keting, 22 (5), 373–91.

Desai, Kalpesh K. and Kevin L. Keller (2002), “TheEffect of Ingredient Branding on Strategies on HostBrand Extendibility,” Journal of Marketing, 66 (1),73–93.

Herr, Paul M., Steven J. Sherman, and Russell H. Fazio(1983), “On the Consequences of Priming: Assimila-tion and Contrast Effects,” Journal of Experimental


Social Psychology, 19, 323–40.Janiszewski, Chris and Stijn M.J. Van Osselaer (2000),

“A Connection Model of Brand Quality Associa-tions,” Journal of Marketing Research, 37 (3), 331–50.

Keller, Kevin L. and David A. Aaker (1992), “The Effectsof Sequential Introduction of Brand Extensions,”Journal of Marketing Research, 29, 35–50.

Loken, Barbara and Deborah R. John (1993), “DilutingBrand Beliefs: When Do Brand Extensions Have aNegative Effect,” Journal of Marketing, 57, 71–84.

Lopes, Lola L. (1982), “Toward a Procedural Theory ofJudgment,” Working Paper 17, Wisconsin HumanProcessing Program, Department of Psychology,University of Wisconsin, Madison.

Morein, Joseph A. (1975), “Shift from Brand to ProductLine Marketing,” Harvard Business Review, 53 (Sep-tember/October), 56–64.

Park, C. Whan, Sung Youl Jun, and Allan D. Shocker(1996), “Composite Branding Alliances,” Journal ofMarketing Research, 33 (4), 453–66.

Raghubir, Priya (2004), ”Free Gift with Purchase: Pro-moting or Discounting the Brand,” Journal of Con-sumer Psychology, 14 (1&2), 181–86.

Samu, Sridhar, Shanker H. Krishnan, and Robert E. Smith(1999), “Using Advertising Alliance for New Prod-

uct Introduction,” Journal of Marketing, 63 (1), 57–74.

Simonin, Bernard L. and Julie A. Ruth (1998), “Is aCompany Known by the Company Its Keeps?” Jour-nal of Marketing Research, 35 (1), 30–42.

Till, Brian D. and Terence A. Shimp (1998), “Endorsersin Advertising: The Case of Negative Celebrity In-formation,” Journal of Advertising, 27, 67–82.

Tversky, Amos and Daniel Kahneman (1974), “Judgmentunder Uncertainty: Heuristics and Biases,” Science,185 (September), 1124–31.

Vaidyanathan, Rajiv and Praveen Aggarwal (2000), “Stra-tegic Brand Alliance,” Journal of Product and BrandManagement, 9 (4), 214–26.

Washburn, Judith H., Brian D. Till, and Randi Pluch(2000), “Co-Branding: Brand Equity and Trial Ef-fect,” Journal of Consumer Marketing, 17, 591–604.

Wernerfelt, Birger (1988), “Umbrella Branding as a Sig-nal of New Product Quality: An Example of Signal-ing by Posting a Bond,” Rand Journal of Economics,19 (Autumn), 458–66.

Yadav, Manji S. (1994), “How Buyers Evaluate ProductBundles: A Model of Anchoring and Adjustment,”Journal of Consumer Research, 21 (September),342–53.

For further information contact:Chung-Hui Tseng

Department of International TradeTamkang University

151, Ying-Chuan Rd. TamsuiTaipei, Taiwan

Phone: 886.2.26215656, Ext. 2859Fax: 886.2.26209730



INDEBTEDNESS AS A DETERMINANT OF SATISFACTION:AN EXTENSION OF THE EXPECTANCY

DISCONFIRMATION MODEL

Akinori Ono, Keio University, JapanTakahiro Chiba, Keio University, Japan

SUMMARY

The question about what determines the level ofcustomer satisfaction has been answered quite often usingthe expectancy disconfirmation model of satisfaction (e.g.,Oliver 1997, Chapter 4). The model indicates that serviceperformance exceeding customers’ prior expectation hasimpacts on disconfirmation and, in turn, on satisfaction.However, do unexpected, higher quality services alwayslead to a higher level of satisfaction? Although the answeris yes according to the model, it may sometimes be no. Forexample, when contact personnel try to serve a customertoo politely, the customer may actually be less satisfiedbecause he/she feels indebted (Greenberg 1980; Greenbergand Westcott 1983) having received the “special” serviceand thinks that there is need to “repay” the contact person.This study aims to describe that higher quality services donot always result in a higher level of satisfaction; rather,they may cause customers to feel indebted to the contactperson and, thus, actually lower the level of satisfaction.

This study employs the theory of indebtedness(Greenberg 1980; Greenberg and Westcott). According tothis theory, “indebtedness” is defined as “a state of obli-gation to repay another” (Greenberg 1980, p. 4), andinfluenced by four major factors: (1) the donor’s motivefor aiding the recipient, (2) the magnitude of the rewardsand costs incurred by the recipient and donor as a result ofthe exchange, (3) the locus of causality of the donor’saction, and (4) cues emitted by comparison others(Greenberg 1980, p. 5). Moreover, Greenberg and Westcott(1983) listed three reasons why a customer who feelsindebted forms negative emotions: (1) attribution ofmanipulative intent, (2) restriction on recipient’s free-dom, (3) costs of repayment (p. 87). For one or more ofthese reasons, customers may be dissatisfied with theunexpected, “excessive” quality/quantity of service, con-trary to the implications of the expectancy disconfirmationmodel of satisfaction (Oliver 1997). Therefore, we hypoth-esize that expectation and performance have both positiveand negative effects on satisfaction as well asdisconfirmation.

We used consumer data pertaining to the shoppingexperience with contact personnel at a high-fashion bou-tique. The questionnaire was developed based on existingscales. Subjects are 235 university students, who wereasked to consider a recent purchase at the boutique. Bothconvergent and discriminant validity were confirmed.The reliability of all constructs as well as AVE (averagevariance extracted) were within acceptable range.

The structural equation model was estimated using amaximum likelihood procedure (CALIS of SAS/Stat 9.1).Absolute, incremental, and parsimony fit indices (GFI,CFI, and PNFI, respectively) were within acceptablerange. All parameter estimates have adequate signs aspreviously hypothesized and were significant at the onepercent level, with the exception of two hypotheses:negative effect of expectation on disconfirmation and onindebtedness.

In this study, we proposed indebtedness as an addi-tional determinant of satisfaction in the expectancydisconfirmation model. The study provides theoreticaland empirical support for extending the model to includethe concept of indebtedness. The indebtedness theorysuggests that when perceived performance exceeds expec-tations, customers can experience a negative emotionalresponse. The results suggest that service performancepositively affects both disconfirmation and indebtedness,which in turn affects satisfaction with opposite signs.Indebtedness can play important roles in determiningsatisfaction and, therefore, we believe that this study fillsan important void in the literature regarding the expect-ancy disconfirmation model of satisfaction in the field ofservice marketing.

Moreover, future studies should not only retest theimpact of indebtedness on satisfaction, but also considervarious consumer and/or product (service) characteristicsthat might yield different results. The incorporation ofthese recommendations in future research should result inan increased understanding of the determinants of cus-tomer satisfaction in the service sector.


For further information contact:Akinori Ono

Keio University5–4–2 Mita Minato-ku

Tokyo 108–8345Japan

Phone: 81.3.3453.4511Fax: 81.45.322.1781



THE ROLE OF SALESPERSON-GENERATED SATISFACTION INRETAILING PERFORMANCE: A DECOMPOSITIONAL AND

COMPARATIVE APPROACH

David M. Szymanski, Texas A&M University, College StationLisa C. Troy, Texas A&M University, College Station

SUMMARY

Review of the personal selling and customer satisfac-tion research streams indicates that salesperson behaviorscan be important to selling success, customer satisfactionwith the salesperson can be important to overall satisfac-tion levels, and overall satisfaction levels can positivelyimpact the financial performance of businesses, includingretailers. It is not counter-intuitive, therefore, to expectthat a proportion of customer satisfaction that is generatedfrom salesperson behaviors (i.e., salesperson-generatedcustomer satisfaction) would play a disproportionate andhighly critical role in the bottom-line performance ofretailers. However, exploring the relationship betweensalesperson-generated customer satisfaction and retailperformance in greater detail can provide importantinsights into the various dimensions of satisfaction thatmay have a greater or lesser impact on the bottom line,providing key insights for managers making decisionsregarding the allocation of scarce resources to variousretail functions such as salespeople, merchandising, etc.Yet, the role of the salesperson in generating satisfiedcustomers has not been empirically documented withrespect to its impact on objective business performance,retail or non-retail. We begin to address this gap in extantresearch by investigating the relationship between sales-person-generated customer satisfaction and retail storeperformance using cross-sectional panel data for anational specialty retailer of moderate to high-priced hardgoods. Our study addresses two key research questions:(1) how prominent a role does salesperson-generatedcustomer satisfaction play in retail store performance,especially in contrast to merchandise-generated satisfac-tion, and (2) which attributes of salesperson-generatedsatisfaction are most critical to bottom-line storeperformance?

In order to test hypotheses regarding the relativeimpact of salesperson-generated versus merchandise-generated satisfaction on retail performance as well as theimpact of more specific dimensions of both merchandise-generated satisfaction and salesperson-generated satis-faction, we draw on customer satisfaction and store per-formance data gathered from a multi-billion dollar spe-cialty retailer of medium- to high-priced hard goods.Satisfaction data was gathered post-purchase and merged

with monthly store performance figures so that the finaldatabase contained monthly panel data on 238,575 cus-tomers from 782 stores. After identifying key controlvariables and verifying validity and reliability of scales,we estimated three customer satisfaction – retail perfor-mance models with both revenues and profits as ourperformance focus and lagged measures of satisfactionand our control variables as key independent variables.

Findings from our decompositional examination ofcustomer satisfaction and retail performance (1) provideadditional support for extant customer satisfaction researchthat overall satisfaction and performance are positivelyrelated on average, (2) offer evidence that salesperson-generated customer satisfaction plays a more prominentrole in retail performance than merchandise-generatedsatisfaction on average, (3) indicate that mixed effects ofindividual attribute-level findings can mask overall mea-sures of customer satisfaction, especially in the context ofmerchandise effects, (4) show that salesperson-generatedsatisfaction attributes are stronger drivers of retail perfor-mance than merchandise effects, and (5) suggest thatperformance impacts of salesperson-generated satisfac-tion are strongest for interpersonal skills, all else equal.

Understanding that specific elements of salespersonand merchandise satisfaction differentially impact storeperformance and further identifying which individualsalesperson and merchandise satisfaction attributes canhave a strong positive or negative impact on the bottomline can enable managers to efficiently prioritize prob-lems and effectively allocate increasingly scarce resourcesto maximize revenues and profits. In addition to thesemanagerial implications, a decompositional approach suchas the one taken here contributes to the extant marketingresearch by providing evidence that specific marketingstrategy factors (in particular, those associated with thesalesperson) can directly impact performance at theattribute level. However, because this study was limited toexamining the role of salesperson-generated satisfactionon retail performance, especially compared to merchan-dise-generated satisfaction, future research could investi-gate the role of customer satisfaction with other marketingelements such as promotions, channels, e-commerce, orstore atmospherics, etc. for their differential impact onperformance.


In conclusion, much remains to be learned about themany nuances of customer satisfaction and its impact onretail performance. Additional contributions to bothresearch and managerial understanding can be made by acontinued focus on discerning and documenting the criti-

cal associations among the research paradigms in per-sonal selling, customer satisfaction, and retailing, withparticular emphasis on decomposing the key attributesthat may differentially impact the bottom line.

For further information contact:Lisa C. Troy

Department of MarketingTexas A&M University

4112–TAMUCollege Station, TX 77843–4112



COUNTERFEIT INFERENCES FROM PRICE LEVELS AND COUNTRYOF ORIGIN: THE MODERATING EFFECT OF PERCEIVED RISK

Kashef A. Majid, The George Washington University, Washington, D.C.Johny K. Johansson, Georgetown University, Washington, D.C.

SUMMARY

The global market in counterfeit products is immense.According to statistics, it is estimated to cost legitimatebusinesses over $200 billion annually (OECD 2007). Thecounterfeit markets cover a wide variety of products, fromaircraft components to designer t-shirts. In addition, theproduction of counterfeit products is not limited to the“usual suspects” in East Asia, but is estimated to takeplace in over 30 nations (European Commission 2007).Some of the named producer countries can be surprising.For example, while China is by far the largest producer ofcounterfeit DVDs, Switzerland leads the way in the pro-duction of counterfeit medicines (European Commission2007). Since consumers may use the country of origin asa cue to judge whether a product might be counterfeit, theproduction of counterfeit goods within the borders of acountry raises interesting questions about the negativeimpact on its country of origin image.

The case of toy recalls from China in 2007 providesa good example of how poorly made products and countryof origin interacts and provides an illustration of what mayhappen when counterfeit goods interact with countryperceptions. After the toy recalls, consumers boycottednot only toy products but also other, similar products,leading to a large scale avoidance of products from China(Beamish and Bapuji 2008). The main explanation for thestrong consumer response was that toys were a high riskproduct due to the possibility of harm they could cause tochildren and consumers used the country of origin toidentify toys that might be harmful (Beamish and Bapuji2008). A similar situation has the potential to occur inregards to counterfeit products. Though often perceivedas affecting mainly luxury goods such as purses or DVDs,counterfeit products can also be found in car parts andmedicines. In China the government-owned ShenzenEvening News estimated that 192,000 people died in 2001due to the absence of active ingredients in fake medicines(Naim 2005) and the World Health Organization (2007)now estimates that approximately 1 percent to 10 percentof total medicines sold globally are counterfeit.

The present paper argues that high risk counterfeitgoods can have a significant impact on the image of acountry and on product choice toward products from thatcountry. Specifically, the present paper looks at high riskcounterfeit products and argues that fear of product failurefrom a high risk counterfeit product may cause country of

origin effects to play an enhanced role in consumerattitudes and consumer choice. Perceived risk of productfailure entails the amount of risk that a consumer associ-ates with a product should a product fail to perform asexpected.

We tested several product categories and found thatdrugs represented a high risk product category. Twocountries, Switzerland and India, were chosen to testcountry of origin effects. Both are home to a large numberof counterfeit pharmaceuticals, and the countries differ inlevel of positive country of origin. In a pre-test it wasfound that participants had a significantly higher positiveimage of Switzerland versus India (M_Switzerland =5.6521 vs. M_India = 4.583, F(1, 45) = 5.288, p < 0.001).

The study design was a three (country of origin:Switzerland vs. India vs. no country) between participantsby two (price: $15 vs. $6) within participants randomizedblock design. Cell sizes ranged from 22 to 34 participantsper cell for a total of 92 participants. Our first finding wasthat participants who learned that counterfeit pharmaceu-ticals had originated from Switzerland reported signifi-cantly lower perceptions of Switzerland than participantswho were unaware of counterfeit pharmaceuticals origi-nating from the country (Mpre_Swiss = 5.652 vs.Mpost_Swiss = 4.697, F (1, 43) = 22.90, p < 0.001). Theimage effect on India was not as significant. Our secondmajor finding revealed that a suspicion of counterfeitpharmaceuticals from India caused participants to choosemore units of lower priced pharmaceuticals from Indiathan higher priced pharmaceuticals. The reverse patternappeared for pharmaceuticals from Switzerland, for asignificant interaction effect (F(3,84) = 4.45, p < 0.05).When a low image country’s products are suspected to becounterfeit, consumers see no reason to pay higher prices.

The results of the present research indicate that coun-terfeit goods can hurt both countries that have positiveimage perceptions and negative country of origin percep-tions. For countries that have high positive image percep-tions such as Switzerland, high risk counterfeit goods canlower country of origin images. For countries wherecountries of origin effects are not very high, consumersmay devalue more expensive and likely legitimate goodsdue to the presence of counterfeits in the marketplace.Both managers and policy makers would be wise tounderstand the negative impact that counterfeit goods canhave by virtue of their originating from a specific country.


REFERENCES

Beamish, Paul W. and Hari Bapuji (2008), “Toy Recallsand China: Emotion vs. Evidence,” Management andOrganizational Review, 4 (2), 197–209.

European Commission – Taxation and Customs Union(2007), “Report on Community Customs Activitieson Counterfeit and Piracy,” Customs Union 1968–

2008.Naim, Moises (2005), Illicit: How Smugglers, Traffick-

ers, and Copycats Are Hijacking the Global Economy.New York: Doubleday.

OECD – Organization for Economic Co-operation andDevelopment (2007), The Economic Impact of Coun-terfeiting and Piracy. Paris, France: Secretary Gen-eral of the OECD.

For further information contact:Kashef A. Majid

The George Washington University2201 G Street, NW

Washington, D.C. 20052Phone: 202.374.4610



COUNTRY RESOURCES, COUNTRY IMAGE, AND EXPORTPERFORMANCE: A PANEL DATA ANALYSIS

Qin Sun, University of South Florida, St. PetersburgAudhesh K. Paswan, University of North Texas, Denton

Margie Tieslau, University of North Texas, Denton

SUMMARY

The influences of an exporting country’s image onthe export strategies of multinational corporations (MNCs)have been extensively investigated in the internationalbusiness and marketing literature (Vrontis and Thrassou2007). Although national differences have strong influ-ences on MNCs’ international marketing practices (Levitt1983), the majority of extant studies focus on eitherconsumers’ perspectives on and preferences for foreignproducts (Archarya and Elliot 2001; Kaynak and Cavusgil1983), or on MNC managers’ global standardization oradaptation strategies (Cavusgil and Zou 1994; Ryans,Griffith, and White 2003; Szymanski, Bharadwaj, andVaradarajan 1993). Few studies have empirically exam-ined the impact of a country’s resources (e.g., cultures,economic conditions, policies, etc.) on its export perfor-mance, although these factors are usually regarded asimportant environmental forces that influence consum-ers’ or managers’ decisions (Yip 1989). Thus, this studyintends to explore export performance at the macro-national level instead of the micro-consumer or micro-manager level.

The resource advantage theory of Hunt and Morgan(1995, 1996) serves as the theoretical foundation of thisstudy. According to this theory, comparative advantagesof a firm’s resources will lead to its competitive advantagein the marketplace position, which will result in the firm’ssuperior financial performance. Although resourceadvantage theory focuses on firms, we argue that a nationcan be regarded as a large organization with many stake-holders. Similar sentiments have been voiced in the nationbranding literature (O’Shaughnessy and O’Shaughnessy2000; Papadopoulos and Heslop 2000, 2002; Paswanet al. 2002). Thus, like any organization, a nation canachieve superior “financial” performance (that is, increasedexports) by strategically employing its resources to achievea competitive advantage in the global market. Therefore,we identify several critical resources based on the litera-ture in international business, global marketing, and nationbranding, and strive to connect these factors to a country’sexport performance.

Research Method

This study uses a panel data model to analyze therelationship between a country’s critical national resources,

its image, and its export. Archival data were used for theeconomic factors such as economic development, indus-try type, infrastructure, exports, as well as cultural, geo-graphical and political factors. The data sources includedatabases from World Development Indicators (WDI),International Monetary Fund, The Wall Street Journal andThe Heritage Foundation. A survey was used to collectdata for the product and industry factors. A total of 24nations were analyzed in order to build, develop andvalidate the panel data model. The economic data for eachcountry were collected for the period of 1995–2006. Ouranalysis includes 24 countries: Australia, Brazil, Canada,China, Denmark, England, France, Germany, India, Italy,Japan, Mexico, Malaysia, New Zealand, Norway, Peru,Russia, Singapore, South Africa, South Korea, Spain,Thailand, Turkey, and United States.

This study first built a basic one-way random effectspanel data model without the moderator (i.e., NBI).

(1) = 26.40 – 2.60 CULTUREi + 6.20 GDPPC

it

+ 16.38 BRANDi + 0.20 INDUSTRY

i + 3.33

NetUsersit + 5.34 NATRES

i + 0.39 EFI

it

We then built on the basic model to develop the “fullmodel,” which included the moderator variable NBI andall interaction variables.

(2) = 1306.45 + 7.60 CULTUREi – 51.46

GDPPCit + 6.86 BRAND

i – 43.71INDUSTRY

i

+ 12.69 NetUsersit – 53.09 NATRES

i – 17.90 EFI

it

– 18.646 NBI + 0.93 NBI*GDPPCit

+ 0.93NBI*NATRESi + 0.31 NBI*EFI

it

– 0.16NBI*NetUsersit – 0.21 NBI* CULTURE

i

– 0.0 NBI*BRANDi + 0.94NBI* INDUSTRY

i

Discussion

The findings of this study provide insight into theinfluence of existing country image on a countries’ exportperformance. Country image, as measured by the percep-tions of consumers or people outside the country, has asignificant indirect impact on a country’s exports. Acountry’s existing nation image will interact with indi-vidual factors or resources to either positively or nega-tively influence the country’s exports. On the other hand,the poorer the image of a country, the greater will be thechange in exports that results from an improvement in the

itPX̂E

itPX̂E


country’s infrastructure. If a country develops its commu-nications infrastructure, then it could counter some of thenegative effects of a poor national image (due to thenegative moderating impact on the relationship betweeneconomic factors and export). One limitation of this studyis the selection of the countries used in our analysis. Other

factors such as per capita annual expenditure on infra-structure could be important variables that should beincluded in future studies. This preliminary studyintended to provide an impetus for research on export at amacro-national level and the related nation brandingissues. References are available upon request.

For further information contact:Qin Sun

College of BusinessUniversity of South Florida, St. Petersburg

140 7th Avenue South, COB 359St. Petersburg, FL 33701

Phone: 727.873.4087Fax: 727.873.4192



EXAMINING COUNTRY EFFECTS ON AVERAGE FOREIGN FIRMPERFORMANCE AND THE CHAIN SIZE–PERFORMANCE

RELATIONSHIP

Carol Finnegan, University of Colorado at Colorado SpringsS. Tamer Cavusgil, Georgia State University, Atlanta

Cuneyt Evirgen, Sabanci University, Turkey

SUMMARY

This study explores how industry and country effectsexplain differences in foreign affiliate performance ofretail firms. We demonstrate how country resources (i.e.,population, GDP, competitive structures) and institu-tional processes (i.e., regulatory policy) moderate market-ing processes on performance. Our retail sample consistsof wholly-owned subsidiaries, joint ventures, acquisi-tions, cooperatives, and franchise operations of 373 chainsin 32 countries. We focus our analyses on one industry toavoid problems associated with varying investment struc-tures and returns across industries. The unit of observa-tion is the retail chain by format (i.e., discounter, super-market, warehouse club, etc.). We use a hierarchical linearmodeling approach to partition the effects of foreignaffiliate decision making processes and resources fromthe effects of the host country regulatory processes andcharacteristics.

Foreign retailers can expect an average market shareof .46 percent when expanding into each foreign country.Fast growing, but moderately to highly concentrated,markets represent the best opportunities. When we lookedat business level characteristics and strategies, we findthat, on average, mean market share varies by format.Supermarkets and discount stores achieve average marketshare. Hypermarkets and warehouse clubs tend to makemore than the average market shares, while conveniencestores and drug stores attain lower than average marketshares after controlling for average host country experi-ence and chain size. By including business characteristicsin the model, we have explained approximately 54.3percent of the variation in business performance of for-eign retailers within countries.

In our country-level model, our findings suggest thatcountry factors explain more than 25 percent of the

difference in performance between foreign affiliates.Country growth rate, market concentration and averagechain size explain approximately 44 percent of thedifferences between countries in average foreign affiliatemarket share. Moreover, level of development andregulatory policy explain 26 percent of the differencesbetween countries in the relationship between chain sizeand performance.

A counter-intuitive finding is that government restric-tions strengthen the relationship between chain size andaverage market share. That is, when governments try torestrict foreign retailers in their operations, the averagerelationship between chain size and market share increases,controlling for level of development. Government restric-tions do not prevent, on average, foreign companies fromgaining more market share for larger chains. In otherwords, foreign retailers operating in countries with regu-latory restrictions on retailing attain greater market sharefor each store opened as compared foreign retailers oper-ating in countries with no restrictions on retailing.

The purpose of this study was to understand keycountry factors influence the relationship between busi-ness characteristics (chain size, experience, dominantformat, and retail formats) and performance. Understand-ing how market attractiveness, industry concentration,and average foreign chain size influence average marketshares allows policy makers to learn from experiences ofpolicy makers in comparable countries. Policy makersneed to understand the drivers of foreign retailer perfor-mance to be able to understand the implications of indus-try specific laws and regulations. This study also providesvaluable information to aid retailers in their countryselection process. We provide retailers performance infor-mation taking into account the differences between coun-tries.

For further information contact:Carol Finnegan

University of Colorado at Colorado Springs1420 Austin Bluffs ParkwayColorado Springs, CO 80918

Phone: 719.255.4319 ♦ Fax: 719.255.3494E-Mail: [email protected]


THE EFFECTS OF HOME COUNTRY CULTURE ONMNC PERFORMANCE

Daekwan Kim, Florida State University, TallahasseeSeong-Do Cho, Chonnam National University, Korea

Pan Dapotan Sinaga, UD Holy Wood IndonesiaKyoungeun Kim, Chonnam National University, Korea

Gangok Jung, Hannam University, Korea

SUMMARY

Introduction

Culture has received tremendous research attentionas a potential impediment of operations of multinationalcorporations (MNCs) in the literature (Nakata andSivakumar 1996; Winsted 1997; Steenkamp, Hofstede,and Wedel 1999; Steensma, Marino, Weaver, and Dickson2000; Nakata and Sivakumar 2001). There are multiplefacets of cultures that have been studied in the literature(Yaprak 2008). However, studies on cultures in the litera-ture mostly focus on the “host” country culture quiteexclusively (Craig and Douglas 2006; Nakata andSivakumar 2001; Nakata and Sivakumar 1996;Vanderstraeten and Matthyssens 2008). A significant roleof “home” country culture in the success of MNC’soperations in a foreign market has remained intact in theliterature.

In prior research, home country impression was stud-ied under the topic of country of origin (COO) (Bilkey andNes 1982; Han and Terpstra 1988; Lawrence, Marr, andPrendergast 1992; Ahmed and d’Astous 1996; Niss 1996).The stream of literature on COO primarily deals withconsumers’ perception on the home country, based mostlyon the direct or indirect purchase experience that theygained previously. Although COO concerns how con-sumers perceive the image of a foreign country and itseffects on purchase intention, the literature does notaddress how consumers’ exposure to the home countryculture influences the performance of the MNCs (Fongand Burton 2008; Michaelis et al. 2008). This study isdifferent from those existing COO studies in that it intro-duces the role of home country culture in inducing con-sumers purchase experience to the literature for the firsttime.

Specifically, by incorporating perceived product qual-ity and corporate image of an MNC from a foreign countryas the mediator of country image consumers gained fromtheir exposure to host country’s cultural elements such asmovies, dramas, and pop-songs, this study clarifies howthe consumers’ exposure to such incoming foreign cul-tural elements leads to their purchase experience system-

atically. That is, we argue that consumers’ positive emo-tions such as sympathy and empathy from their exposureto the incoming home country cultural elements affect theimage of the home country and, subsequently, of productsand MNCs from the home country, eventually influencingtheir purchase experience with such products and firms.Furthermore, such process that governs the transfer ofcountry image formed by emotions such as sympathy andempathy to purchase experience of consumers is likelyaffected by the level of consumer ethnocentrism.

Hypotheses

In order to investigate how the emotional responsesto the cultural elements of an MNC home country influ-ence the consumer’s purchase experience through per-ceived product quality and corporate image this studyoffers the following hypotheses.

H1: Sympathy induced by home country cultural ele-ments affects its home country image positively.

H2: Empathy induced by home country cultural elementsaffects its home country image positively.

H3: Country image of an MNC as perceived by consum-ers affects perceived product quality positively.

H4: Country image of an MNC as perceived by consum-ers affects perceived corporate ability positively.

H5: Corporate ability of MNC affects perceived qualitypositively.

H6: Perceived product quality of MNC as perceived byconsumers affects purchase experience positively.

H7: Corporate ability of MNC as perceived by consumersaffects purchase experience positively.

H8: Low ethnocentric consumers, compared to high eth-nocentric consumers, have more positive purchaseexperience toward MNC’s home country productsgiven perceived product quality.


H9: Low ethnocentric consumers, compared to high eth-nocentric consumers, have more positive purchaseexperience toward MNC’s home country productsgiven corporate ability level.

Method

We collected consumer data from college students inIndonesia. Of 602 questionnaires initially distributed, 533questionnaires were collected and subsequently used totest our proposed model. This study used four unfamiliarproducts to prevent the influence of prior knowledge.

We conducted confirmatory factor analysis (CFA)using EQS to test the measurement model. The results ofthe final CFA model indicate an excellent fit of themeasurement model with the empirical covariance withChi-square = 1,768.53 on 824 degrees of freedom,nonnormed fit index (NNFI) = .926, comparative fit index(CFI) = .932, standardized root mean-square residual(SRMR) = .043, and root mean-square error of approxi-mation (RMSEA) = .049 (Hu and Bentler 1999). The CFAresults reveal that the study constructs have an adequatelevel of convergent and discriminant validities with ac-ceptable construct reliabilities. Subsequently, the pro-posed model with all measurement items from the CFAmodel is estimated. The results of a multi-group analysisindicate that the overall fit of the model is good with Chi-Square = 2,735.2 on 1,310 degrees of freedom, CFI =0.918, SRMR= 0.073, and RMSEA = 0.058. According tothe result, seven hypotheses are supported except for H8and H9.

Discussion and Implications

This study is the first study that explored the directand indirect impacts of foreign cultural waves on consum-ers’ responses empirically. By testing the conceptual

framework that includes key variables related to a foreigncultural wave, this study explains how a foreign culturalwave gives advantages to the MNCs and products fromthe home country.

In the current study, country image functions effec-tively as a halo to transfer positive emotions from expo-sure to the home country cultural elements to perceivedquality and ability of products and MNCs from the homecountry. Through enhancing people image and economiccountry image, consumers seem to speculate the ability ofthe MNCs and quality of their products from the homecountry. This result is in line with the key argument of thecountry-of-origin literature.

In our research, not only perceived product qualitybut also corporate ability influence purchase experiencepositively. Furthermore, corporate ability positivelystrengthens perceived product quality. These results sug-gest that not only the product specific attribute evaluationlike perceived product quality but also peripheral cuessuch as country image and corporate ability influenceconsumer purchase experience. This is an important con-tribution of the current study as the literature does notidentify such peripheral cues as country image and corpo-rate ability as factors of consumer purchase (e.g., Chaudhuriand Holbrook 2001).

For managers, this study results suggest that MNCs’marketing strategies need to reflect such potential impactof their home country cultural wave more proactively.That is, product placement, as a promotional vehicle, ininternational marketing could be much more effectivethan domestic marketing. Furthermore, featuring relevantdrama or movie stars in their advertisements would beparticularly effective in such local markets where homecountry cultural waves are prevailing. References areavailable upon request.

For further information contact:Daekwan Kim

Florida State University314 Rovetta Business Annex

P.O. Box 3061110Tallahassee, FL 32306–1110

Phone: 850.644.7890Fax: 850.644.4098



BRANDS YOU CAN RELY ON! HOW BRAND CREDIBILITY AFFECTSCORE BRAND FUNCTIONS: AN EMPIRICAL

INVESTIGATION IN SERVICES

Alexander Leischnig, Technische Universität Bergakademie Freiberg, GermanyAnja Geigenmueller, Technische Universität Bergakademie Freiberg, Germany

Margit Enke, Technische Universität Bergakademie Freiberg, Germany

SUMMARY

One of the most basic characteristics of services isintangibility, referring to the fact that services cannot beseen, felt, tasted, or touched like one can sense tangibleproducts (Parasuraman et al. 1985). According to Zeithaml(1981), intangible services are characterized by experi-ence and credence attributes, providing fewer cues forconsumers to assess. Consequently, because of this stateof imperfect information (Erdem and Swait 1998), con-sumers cannot readily evaluate service quality, whichelevates consumer uncertainty and in turn, perceived riskof purchasing services. When imperfect and asymmetricinformation characterize a market, consumers and firmsmay use signals in order to convey information about theircharacteristics (Spence 1974). Addressing this problem, aneed arises for mechanisms by which service providerscan inform consumers about and convince them of thequality of their services offered. As literature reveals,brands can serve as valuable signals indicating productquality and positioning (Erdem and Swait 1998; Wernerfelt1988). However, in order to reduce perceived risk inpurchasing decisions as well as information acquisitionefforts, brands have to be credible.

Focusing on studies investigating credibility of ser-vice brands, a literature review reveals a clear lack ofresearch. Prior research on brand credibility has mainlyfocused on different product classes, but neglects ser-vices. According to Erdem and Swait (1998) brand cred-ibility comprises two facets, namely trustworthiness andexpertise. While trustworthiness can be defined as thebelief that the firm is willing to deliver on its promises,expertise refers to the belief that the firm is capable ofdelivering on its promises (Erdem and Swait 1998;Sweeney and Swait 2008). Our study intends to close theresearch gap by extending existing knowledge to theservices setting. More precise, we investigate (1) brandcredibility and its effects on core brand functions (i.e.,information efficiency and risk reduction), as well asconsumers’ repurchase intention and (2) the effects ofinformation efficiency and risk reduction on consumers’repurchase intention using service brands.

The services selected for the testing of the modelwere fast food restaurants, airlines, and retailers. Whilewe utilized the brand “McDonald’s” representing a fastfood restaurant, we selected the brand “Lufthansa” torepresent an airline. Further, we selected the brand“Starbucks” to represent a retailer. The data were col-lected via paper-and pencil surveys from graduate stu-dents as well as administrative and academic staff of amajor university. In sum, we received two hundred andforty-six completely answered questionnaires. We em-ployed structural equation modeling for model estimationusing the Amos 17.0 software program.

Results provide strong empirical support for thepositive effects of service brand credibility on core brandfunctions, namely information efficiency and risk reduc-tion. Further, data analysis shows a significant positiveeffect from brand credibility on repurchase intention.Focusing on the impact of core brand functions on con-sumers’ repurchase intention, results reveal a significanteffect of information efficiency on repurchase intention.However, we found no support for the impact of riskreduction on repurchase intention.

In summary, our research contributes to a moredetailed understanding of the phenomenon of servicebrands. In particular, our research sheds light on the roleof brand credibility in services. First, we explicate theimpact of credibility of service brands on core brandfunctions, namely risk reduction and information effi-ciency, and consumers repurchase intention. Second, weexplicate the influence of core brand functions on con-sumers’ repurchase intention. From a managerial view-point, our results highlight that service providers shouldconsider brands as valuable tools in order to conditionconsumer response. Because brand credibility has a strongimpact on information efficiency, risk reduction, andconsumers’ repurchase intention, managers are asked to(1) evaluate their brands questioning how credible theyare, (2) efficiently and effectively build credible brands,and (3) invest into credible brands. References are avail-able upon request.


For further information contact:Alexander Leischnig

Marketing DepartmentTechnische Universität Bergakademie Freiberg

Lessingstrasse 4509599 Freiberg

GermanyPhone: +49.3731.39.3922

Fax: +49.3731.39.4006E-Mail: [email protected]


ETHICAL REPUTATION: FACT OR FICTION? A QUALITATIVECONSUMER EXPLORATION

Katja H. Brunk, Université Libre de Bruxelles, Belgium

SUMMARY

Background

Over the past two decades, organizations have increas-ingly focused on corporate identity (CI), image and repu-tation – concepts recognized as major assets capable ofenhancing corporate value (Berrone, Surroca, and Tribo2007; Fukukawa, Balmer, and Gray 2007). Positivereputational capital can set a company apart from itscompetitors, and through creating vital points-of-difference, may act as a source of competitive advantage(Fombrun 1996; Fukukawa et al. 2007). However, numer-ous companies have experienced a converse scenario,when their reputations faltered because of business prac-tices which were perceived as unethical by the public.Hence, in addition to elements such as product, service, orbrand quality, an organization’s reputation is inherentlylinked to how un/ethically it is perceived to conduct itsbusiness(es), meaning moral perceptions play a majorrole in corporate reputation formation (Bendixen andAbratt 2007). In spite of ample evidence confirming theimpact of un/ethical perceptions on consumer attitudesand purchasing behavior, to this day, little is known abouthow an un/ethical perception emerges in the consumer’smind. This qualitative study explores the formation ofethical perceptions held by consumers, a major stake-holder group. Findings are based on 20 long interviews(McCracken 1988) with general consumers.

Key Findings

Naturally, the primary ways that consumers formperceptions about companies is based on either first-handexperiences or, alternatively, informational sources suchas the media, corporate communications, and interper-sonal communication. However, the absence of suchconcrete information or experience is frequent and “nor-mal” in the case of ethical evaluations, which is whyconsumers may resort to inferential processes. The studydemonstrates that consumers’ ethical beliefs can be in-ferred and identifies four types of cues that may instigateinferences in ethical perception formation: product, com-pany, category, and origin-related cues.

Product-Specific Cues. Consumers’ inferentialbeliefs may be rooted in explicit product-specific featuresand attributes. For instance, ethicality is suggested by aproduct’s name and packaging. The most prominent inter-

attribute inference is instigated by a product’s retail price,meaning a low price can arouse consumer suspicion ofunethical dealings.

Company-Specific Cues. Multiple company-specificcues exist from which consumers may derive ethicalperceptions. These relate to either the structural set-up ofa company or are of economic nature. For instance, thetype of ownership of a company can be diagnostic. Prof-itability was identified as the main economic indicator,exemplified by the observation that corporate earningsperceived as unjustifiably high can signal unethical busi-ness practices. Similarly, the size and market share of acompany as well as its degree of global reach can implyun/ethicality.

Category-Specific Cues. Inferences can also relate toa company’s industry sector or product category, inducedby certain production-related characteristics, such aslabor intensiveness or environmental properties. Anunethical perception may be evoked by a faltering reputa-tion of an entire product category or the result of anegative spillover effect from irregularities or violationsof a close competitor (Roehm and Tybout 2006). Further-more, consumers make a connection between the degreeof competition in a given industry sector and the likeli-hood of unethical business practice. A prevailing stereo-type is that a company in a monopoly-like setting is morelikely to act unethically by taking advantage of theconsumer.

Origin-Specific Cues. Consumers may associate acompany or brand with the place, i.e., a region or acountry, from which they believe it originates. This studyidentifies origin as an important source of consumers’ethical beliefs. The country bias is evoked by severalfactors, such as the level of development and education,the economic ideology, the prevailing political climate,the presence of a social welfare system, and the culture,mentality, and lifestyle of the country’s inhabitants.

Implications

The observed consumer inference processes high-light the crucial issue of reputation controllability from amanagement perspective. By nature, inferences operate ata subconscious, hence personally uncontrollable leveland often fail to reflect the (objective) reality of a situa-tion. Furthermore, consumers are not always able to


distinguish between facts and fiction and have difficultyseparating implicit (inference) from explicit (informa-tion/experience) conclusions in memory (Kardes 1988).The findings of this study therefore substantiate existingresearch suggesting that a company may have only limitedcontrol over its reputation and brand equity (Bromley2001; Keller and Lehmann 2006), as well as the brandingliterature on Doppelgänger brand images and hijacked

brands (Thompson, Rindfleisch, and Arsel 2006;Wipperfürth 2005), and implies that through previouslyunobserved inference processes, controllability is evenfurther diminished. In the liquid times of sensory andinformational overload (Bauman 2000), inferences playan increasingly vital role and perceived identities, reputa-tions and images are therefore co-produced, not only bycustomers or partners, but also by competitors.

REFERENCES

Bauman, Zygmunt (2000), Liquid Times. London: PolityPress.

Bendixen, Mike and Russell Abratt (2007), “CorporateIdentity, Ethics and Reputation in Supplier-BuyerRelationships,” Journal of Business Ethics, 76 (1),69–82.

Berrone, Pascual, Jordi Surroca, and Josep A. Tribo(2007), “Corporate Ethical Identity as a Determinantof Firm Performance: A Test of the Mediating Role ofStakeholder Satisfaction,” Journal of Business Eth-ics, 76 (1), 35–53.

Bromley, Dennis B. (2001), “Relationships Between Per-sonal and Corporate Reputation,” European Journalof Marketing, 35 (3/4), 316–34.

Fombrun, Charles J. (1996), Reputation: Realizing Valuefrom the Corporate Image. Boston, MA: HarvardBusiness School Press.

Fukukawa, Kyoko, John M.T. Balmer, and Edmund R.Gray (2007), “Mapping the Interface Between Cor-

porate Identity, Ethics and Corporate Social Respon-sibility,” Journal of Business Ethics, 76 (1), 1–5.

Kardes, Frank R. (1988), “A Nonreactive Measure ofInferential Beliefs,” Psychology & Marketing, 5 (3),273–86.

Keller, Kevin Lane and Donald R. Lehmann (2006),“Brands and Branding: Research Findings and FuturePriorities,” Marketing Science, 25 (6), 740–59.

McCracken, Grant (1988), The Long Interview. NewburyPark, CA: Sage.

Roehm, Michelle L. and Alice M. Tybout (2006), “WhenWill a Brand Scandal Spill Over, and How ShouldCompetitors Respond?” Journal of MarketingResearch, 43 (3), 366–73.

Thompson, Craig J., Aric Rindfleisch, and Zeynep Arsel(2006), “Emotional Branding and the Strategic Valueof the Doppelgänger Brand Image,” Journal of Mar-keting, 70 (1), 50–64.

Wipperfürth, Alex (2005), Brand Hijack: Marketing With-out Marketing. New York: Portfolio.

For further information contact:Katja H. Brunk

Solvay Brussels School of Economics and Management (SBS–EM)Centre Emile Bernheim

Université Libre de BruxellesAve F.D. Roosevelt 21

1050 BrusselsBelgium

Phone: +32.2.6502346E-Mail: [email protected]


FAMILY BUSINESS BRANDING: A STUDY ON THE SALIENCE ANDRICHNESS OF FAMILINESS ON CORPORATE WEBSITES

Willem Smit, IMD, SwitzerlandJoachim Schwass, IMD, Switzerland

Lise Möller, IMD, Switzerland

SUMMARY

The lack of attention given to branding issues atfamily firms is unjustified. Why? First of all, familybusinesses are omnipresent in Western economies. Sec-ond, having such a prominent presence and generating somuch employment suggest that many family businessfirms are highly competitive. On top of that thePricewaterhouseCooper Family Business survey (2007/2008) revealed that family businesses get competitivestrength particularly from their product offerings, cus-tomer relationship management and their brands. It showedthat family enterprises increasingly realize that for them tostay successful branding is and remains critical.

Theoretically like any other type of firm, familyenterprises try to leverage important resources they pos-sess. The family-based identity is one of those assets forthem. The identity and what it represents is known to bethe most important intangible asset for many companiesand is arguably a primary source for competitive advan-tage (Aaker 1991). While there is extensive literaturedevoted to branding, most of it examines branding amonglarge, well-established, mostly publicly-traded compa-nies (cf., Keller 1993). From their very founding, familyfirms are often associated with the family name. Familyconnections, legacy, and history play a vital role in theirability to mobilize resources (Aldrich and Cliff 2003), andthese associations potentially strengthen customers’ brandand authenticity perceptions (cf., Pine II and Gilmore2007). Not only theoretically, there is a beginning ofempirical evidence. Craig, Dibrell, and Davis (2008)showed that a minority of family firms actively promotetheir familiness in their communication with externalconstituents. Interestingly, actively communicating thefamily-based brand identity positively contributes to thefirm performance. As such, it is worthwhile to investigatewhether more family firms should communicate theirfamily-based brand identity and whether that lead tomaking use of this untapped potential. Before addressingif and how family firms ought to effectively manage thefamily business brand (FBB), this research deals with thefirst step into that direction. We first will explore howfamily firms brand their family businesses – i.e., howcommunicate their familiness.

Family-Based Identity or Familiness

Family business scholars conceptualize the phenom-enon “family business” as a “Dual System” with proper-ties of both the family and the business (Swartz 1989,p. 331). Business and family systems are regarded asseparate entities in terms of structure, goals and tasks. Apure typical Business System appreciates concepts suchas “rationality,” “competence,” “contract,” “future,” and“money,” whereas a pure typical Family System values“emotional,” “birthright,” “relationship,” “memories,”and “love.” The interaction of these two systems (Benson,Crego, and Drucker 1990) is the familiness and that relatesto how family firms effectively manage to harmonize,unite and strengthen the often-seemingly conflicting Busi-ness values with Family values.

Content Analysis

To examine how family firms speak of their firm toexternal stakeholders and to define a typology of familybusiness branding strategies, we collected observationsfrom the corporate websites of family businesses (publishedin the English language). To systematically analyze thetext, we used manual and computer-aided content analysisprocedures in order to generate quantitative measures ofendorsing the family-based identity of a family firm. Inthe dictionary-building process, we followed Dyer (1994)who has constructed content analysis dictionaries toidentify frequencies of high level concepts. Conceptssuch as “family” and “business” are types of abstractionsencompassing conceptually similar lexical terms of greaterspecificity. The page-level-specific sets of records werethen uploaded into SIMSTAT in order to conduct adetailed quantitative analysis of words and phrasesunderlying each category. SIMSTAT is a powerfulstatistical package with an integrated content analysismodule – WORDSTAT – which provides counts of targetwords and phrases, cross-tabulation, and statisticalanalysis.

Sample

We had two procedures in place to draw samples offamily businesses. The first procedure was to use the


companies listed in the Credit Suisse Family BusinessIndex (HOLT FBI) as a sample frame. The second sam-pling procedure was to randomly pull company namesfrom the alumni database of a business school with aspecialization in executive education to family businesses.Five hundred eighty companies are stored in the database.In order to match the number of Credit Suisse set we drewan equally large sample from this database resulting in asample size of 340 (international) companies. The CreditSuisse HOLT FBI list had 62.7 percent from Europe and34.9 percent from the USA. The other sample was pre-dominantly European (58.2%). We found 226 websites:166 from the Holt list, and 60 other family businesses.From 140 websites, we detected 103 company profilepages, and 89 history pages.

Analysis

Salience or relative prominence of either family-orbusiness-words is measured by counting the number ofwords from each dictionary. On the homepages, on aver-age less than one family-related term (mean = .61) waspresent, while the mean for business-related terms wasmore than five times higher: 3.42 words on average. Forrichness, we took as an indicator the maximum number ofwords related to family and business appearing on each ofthree different levels (homepages, company profile, andhistory). To explore the different types of family business

branding, we followed the three-stage clustering approach(see Cannon and Perreault 1999; Homburg, Jensen, andKrohmer 2008; Homburg, Workman, and Jensen 2002).The outlier identification using nearest-neighbor cluster-ing made us delete 22 cases. We then determined thenumber of clusters with hierarchical clustering algorithmthat Ward (1965) developed.

Results

Our cluster variables were amount of words, salienceand richness dimensions of homepages. Iteratively, wefound the three cluster solution to have clusters suffi-ciently large in size and with a reasonable interpretation.The first cluster (n = 59; 29%) consists of corporatewebsite homepages that are short introductions; on aver-age 63.39 words, of which 1.41 words directly relate to thebusiness lexicon. The second and largest cluster (n = 131;64%) is made up from homepages that are executivebusiness texts. Texts have a length of 248 words onaverage, of which more than seven words relate to Busi-ness, and 1.35 words are Family-related. The homepagesin the third cluster (n = 14; 7%) are longer and richerstories on both Business and Family. They have more thandouble the amount of total words than the averagehomepage (677 words). The longer texts give the oppor-tunity to talk more about both Business (almost 23.50words) and family (6.57 words).

For further information contact:Willem Smit

Research Fellow in MarketingIMD

Ch. de Bellerive 23P.O. Box 915

CH – 1001 LausanneSwitzerland

Phone: +41.21.618.02.74Fax: +41.21.618.07



DO BOARD MEMBER POWER BLOCS ENHANCE INNOVATION?

Shannon Cummins, University of Nebraska – LincolnRavipreet S. Sohi, University of Nebraska – Lincoln

SUMMARY

This paper introduces intra-industry board memberpower blocs as a mechanism of enhancing firm innovationthough enhanced cooperation between competing firms.Board member Interlocking Directorates (ILDs) betweenfirms have been discussed as anti-competitive, studied fortheir impacts on capital access, and prohibited by Con-gressional acts. While many studies and theoretical papershave reported on the importance of board members to firmstrategy and execution, there is still little consensus ontheir firm-level impacts. In this paper we introduce thenotion of power blocs – a linear quadruple of firmsconnected by shared directors – as a unit of ILD analysisto the marketing literature.

Using secondary data from twelve industries, weinvestigate the effects of intra-industry board-memberpower blocs on firm innovativeness. We also investigatethe moderating impact of industry competition and envi-

ronmental turbulence. Our findings indicate that intra-industry board member power blocs have a significantpositive impact on firm innovation. Competition enhancesthe positive effect of power blocs on innovation. How-ever, environmental turbulence lessens the positive impactof board member power blocs on innovation, suggestingthat while interlocks are beneficial, they are not alone themagic bullet for industries and firms in challenging andchanging times.

This study contributes two key findings to the litera-ture in addition to introducing the marketing literature toboard member power blocs. First, it expands currentdefinitions of industry networks to include intra-industrynetworks organized to leverage horizontal synergies withinan industry. Second, it raises questions about the currentunderstanding of board member interlocks, the reasoningbehind their prohibition, their actual prevalence in themarketplace, and the firm outcomes of such linkages.

For further information contact:Shannon Cummins

University of Nebraska – Lincoln310 CBA; 512 N. 12th St.Lincoln, NE 68588–0492

Phone: 402.472.2316Fax: 402.472.9777



INNOVATION BEYOND FIRM BOUNDARIES: THE ROUTINES ANDRESOURCE INVESTMENTS OF SUCCESSFUL EXTERNAL

PROBLEM SOLVERS

Dominik Mahr, University of Antwerp, BelgiumAric Rindfleisch, University of Wisconsin, Madison

Rebecca J. Slotegraaf, Indiana University, Bloomington

SUMMARY

Innovation is widely recognized as a key driver offirm growth and profitability by both scholars and practi-tioners (e.g., Hauser, Tellis, and Griffin 2005; IBM 2008).An increasing number of firms are adopting open innova-tion approaches that solicit the knowledge and skills ofexternal entities (Chesbrough 2003). Although such exter-nal innovation is believed to provide several advantages,such as reduced development time and increased novelty(Huston and Sakkab 2006; Nambisan and Sawhney 2007;Seybold 2006), it also poses considerable challenges. Incontrast to the relatively closed and tightly controllednature of internal innovation, external innovation requiresa considerable openness and a release of organizationalcontrol (Seybold 2006; von Hippel 2005). Moving towardopen innovation, which relies on both internal and exter-nal innovation, is a difficult paradigm shift for manyfirms, due to concerns about the potential loss of propri-etary information and doubts about the innovation abili-ties of external entities (Cook 2008). As a result, firmstend to release limited information, which produces great-er challenges for external entities who attempt to developsolutions with considerably less information about thenature, history, and background of the problem thanindividuals working within the firm (Teece 2003; vonHippel 2005). In essence, external innovation is typicallycharacterized by a high degree of decontextualizationcompared to traditional internal innovation initiatives.This poses a challenge for marketing scholars, who havetraditionally focused their inquiry upon issues related tointernal innovation activities (e.g., Ruekert and Walker1987; Cooper and Kleinschmidt 1995; Sethi 2000; Sethiand Iqbal 2008). Thus, the knowledge gleaned from thesestudies may possess limited applicability in an openinnovation context.

Given the recency of this development, research onthis topic is scant and little is known about the keys forsuccessful open innovation. Our research addresses thisneed by focusing on one important type of open innova-tion, i.e., external problem solving (EPS) during theinnovation process and examines reasons why some exter-nal problem solvers are more successful than others. Forinstance, InnoCentive provides a digital environment thatoffers the expertise of more than 160,000 potential solvers

for numerous firms, including large pharmaceutical com-panies (e.g., Eli Lilly), fast moving consumer goodscompanies (e.g., P&G), software firms (e.g., SAP), andnot-for-profit organizations (e.g., The Rockefeller Foun-dation). Other examples of EPS networks includeNineSigma, IdeaConnection, and innovation exchange.During EPS, firms (i.e., seekers) seek the help of externalexperts (i.e., solvers) to solve innovation related problemsoften through the help of these EPS networks (Hargadonand Sutton 1997; Verona, Prandelli, and Sawhney 2006),who act as both a matchmaker (by posting these problems)and a protector (by securing intellectual property rights).In spite of increasing use of EPS networks and the grow-ing numbers of registered external solvers, only a limitednumber of problems are solved and only few solvers aresuccessful (Lakhani et al. 2007). Thus, it is an importantissue for innovation brokers, seekers, and solvers alike tounderstand what makes some external problem solversmore successful than others. Our study addresses thisneed and seeks to enrich and extend the research on openinnovation by examining the reasons determining thesuccess of external problem solvers.

Specifically, our investigation focuses on the prob-lem-solving routines and resource investments of exter-nal problem solvers. We identify critical routines bydrawing from extant research on expertise (e.g., Spenceand Brucks 1997; Prietula and Simon 1989; Teece 2003)and creativity theory (e.g., Amabile 1983, 1998; Im andWorkman 2003) that highlights different paths to success.Expertise theory purports that experts are superior to non-experts in filtering relevant from irrelevant information,understanding complex and ill-structured tasks, and solv-ing problems with analytical skills, i.e., deliberate rou-tines (Ericsson, Krampe, and Tesch-Roemer 1993). Incontrast, creativity theory attributes superior performanceto developing novel ideas, approaching problems moreflexibly and imaginatively, and drawing associations withknowledge from other fields, i.e., creative routines(Oldham and Cummings 1996). Moreover, these twotheoretical traditions suggest contrasting effects regard-ing the impact of temporal and monetary resources investedin a problem solving task.

To answer our research question, we test the impactof the two contrasting problem solving routines (i.e.,


deliberate routines vs. creative routines) and resourceinvestments (i.e., time and money) via a sample of 453external problem solvers who submitted solutions toinnovation problems posted by InnoCentive. This samplecombines both survey (for our independent variables) andarchival (for our dependent variable) data.

Our findings reveal that creative and deliberate rou-tines can each enhance problem solving success but thattheir effectiveness is contingent on the resources invested.More specifically, deliberate routines are more successfulwhen external problem solvers invest substantial time butlittle money. In contrast, creative routines are more suc-

cessful when external problem solvers invest substantialmoney rather than time. Overall, our results suggest thatsuccessful external problem solvers employ either inspi-ration or perspiration and purposefully invest theirresources. Our findings challenge common practice ofinternal problem solving, where employees are encour-aged to think out of the box and to use creative techniquesin the innovation process. It appears that solvers’ deliber-ate skills may compensate for information and structurethat is present inside firms but lacking in thedecontextualised EPS setting. Our inquiry provides anumber of implications for open innovation theory andpractice. References are available upon request.

For further information contact:Dominik Mahr

Marketing DepartmentUniversity of AntwerpBE–2000 Antwerpen

BelgiumPhone: +32(3)265.50.42

Fax: +32(3)265.50.40E-Mail: [email protected]


MANAGERIAL TIES AND CORPORATE INNOVATIVENESS:IS KNOWLEDGE CREATION A MISSING LINK?

Chengli Shu, Xi’an Jiaotong University, China, and University of Illinois at ChicagoShanxing Gao, Xi’an Jiaotong University, China

Xu Jiang, Xi’an Jiaotong University, ChinaAlbert L. Page, University of Illinois at Chicago

SUMMARY

Managerial ties, which are senior managers’ personalnetworks with top managers at other firms and officials inthe government and industries, contribute substantially tofirm performance because of their strategic roles. Extantresearch tends to construct a direct mechanism for themanagerial ties–corporate innovativeness link. This paper,however, provides and investigates an indirect ties-innovativeness argument which is that organizationalknowledge creation, including knowledge exchange andknowledge combination, is a mediator. The knowledge-based view argues that the firm can be regarded as abundle of knowledge which is the basis for competitiveadvantage. From this perspective, acquiring, creating, andpossessing heterogeneous knowledge becomes the fore-most mission for effective managers to lead their firms inturbulent environments. Moreover, social networks theorysuggests that managerial ties provide opportunities formanagers to access external resources and establish trust-ful external relationships. However, the network perspec-tive also suggests that managerial ties only play a “con-duit” role by providing possibilities and opportunities to

approach external resources. How should firms turn thesepossibilities and opportunities into internal knowledgestock and further transform it into market competitive-ness? This paper aims to answer this research questionbased on and integrating social networks theory and theknowledge-based view.

We conducted this research in China because itprovides an appropriate context to address our researchquestion. We used structural equation modeling to test ourresearch hypotheses (in Figure 1) and the empirical find-ings from 270 firms indicate complex relationships amongthe studied variables. It reveals that business ties aredirectly and positively related to knowledge exchangeand knowledge combination, while political ties onlyexert weak effects on knowledge exchange. Interestingly,knowledge combination exerts positive and significantinfluences on both product innovation and process inno-vation, two widely used sub-dimensions of corporateinnovativeness, while knowledge exchange only impactsproduct innovation significantly. Differing frompast research findings that managerial ties directly influencecorporate innovativeness, we find that the influence of

FIGURE 1Conceptual Model and Research Hypotheses

Business Ties

Political Ties

Knowledge Exchange

Knowledge Combination

Process Innovation

Product Innovation

H1a: +

H1a: +

H1b: +

H1b: +

H2: +

H3a: +

H3a: +

H3b: +

H3b: +

Managerial Ties Knowledge Creation Corporate Innovativeness

H4: Mediation


managerial ties on corporate innovativeness is indirect.More importantly, we found that knowledge exchangeand knowledge combination are different constructs andthe former positively influences the latter in organiza-tional knowledge creation processes. These findingsextend our understanding regarding the ties-innovativeness

link by providing a more complete understanding of howfirms can access and internalize external resources andtransform them into market competitiveness throughknowledge creation processes. References are availableupon request.

For further information contact:Chengli Shu

Department of Managerial StudiesUniversity of Illinois at Chicago

601 S. Morgan Street, MC 243, Room 2229Chicago, IL 60607

Phone: 312.996.5529Fax: 312.996.3559



PROMOTING SUSTAINABLE CONSUMER BEHAVIOR: THEMODERATING EFFECT OF PSYCHOLOGICAL DISTANCE

ON THE INFLUENCE OF ENVIRONMENTAL CONCERN

Andrea Heintz Tangari, University of Arkansas, FayettevilleScot Burton, University of Arkansas, Fayetteville

SUMMARY

This study focuses on a public service advertisementpromoting the use of energy efficient products using twodifferent psychological distance frames (e.g., Libermanand Trope 1998; Trope and Liberman 2003). The purposeof this study is to (1) understand how temporal and socialdistance framing influence consumers’ responses to energyefficiency ad messages, (2) address if a consumer’s envi-ronmental concern influences their response to the mes-sage frames, and (3) examine if the interaction betweenthe temporal distance frame, social distance frame, and aconsumer’s environmental concern will influence atti-tudes and behavioral intentions.

Conceptual Background and Hypotheses

The distance framing effects examined in this studyare related to construal level theory (Liberman and Trope1998; Trope and Liberman 2003). The two distance framesused are temporal distance and social distance. Temporaldistance is the perceived distance between the presenttime and when a focal event occurs and social distance isthe perceived distance from one person to another personor social group.

Studies have indicated that environmental concerncan influence certain pro-environmental behaviors (Ellen,Wiener, and Cobb-Walgren 1991; Kilbourne and Picket2008). Although we predict a main effect of environmen-tal concern, some studies have suggested that environ-mental concern may play a weaker role in affectingpurchase intentions or behavior (e.g., Alwitt and Pitts1996; Smith, Haugtvedt, and Petty 1994). Given themixed findings, it is possible that environmental concernis influential in certain situations but less so in others. Wehypothesize that under certain message framing condi-tions, environmental concern will influence sustainableconsumer behavior, but under other framing conditions,the effect of environmental concern will be weakened.

Study Method and Results

This study employs an experimental design where thetemporal and social distances are manipulated within anad while a consumer’s environmental concern is mea-sured. The temporal distance frame was manipulated by

indicating how long (one month or one year) it would taketo save a significant amount of money by using energyefficient products. The social distance manipulation statedthat the majority of two different target groups (varying insocial distance from the study participants) would buyenergy efficient products. Dependent measures in thisstudy included attitude toward the ad (α = .92), purchaseintentions r = .96), perception of purchase intentions ofothers (α = .98), and light bulb choice.

We first tested the hypotheses on the three intervalscale dependent variables by performing a 2(temporaldistance frame) X 2(social distance) X 2(environmentalconcern) MANOVA. There was a positive main effect ofenvironmental concern on the dependent variables of adattitude (F(1,218) = 9.61, p < .01), purchase intentions(F(1,218) = 11.71, p < .001) and purchase intentions ofothers (F(1,218) = 17.79, p < .001).

We predicted that there would be a three-way inter-action between temporal distance frame, social distance,and environmental concern. This was generally supportedfor all three interval-scale dependent variables; purchaseintentions (F = 9.44, p < .005) and purchase intentions ofothers (F = 4.12, p < .05) were significant, while adattitude was only marginally significant (F = 3.01, p <.10). We predict that when the distance frames differed(i.e., mismatched frames), the dependent variables will behigher for those high versus those low in environmentalconcern. This was supported for purchase intentions (p’s <.02) and purchase intentions of others (p’s < .01). How-ever, when both social and temporal frames matched,there was no effect of environmental concern on purchaseintentions (p’s > .40) or purchase intentions of others(p’s > .10).

In terms of bulb type choice, when the distanceframes did not match, participant were more likely tochoose the CFL bulb if they were higher on environmentalconcern (p’s < .05). When both frames match, choice ofthe CFL bulb versus the incandescent bulb did not differsignificantly (p’s > .20) between those high and low inenvironmental concern. In combination, these resultsprovide substantial support for our predictions.


Discussion

This study examined the message framing effects oftwo different psychological distance dimensions, temporaland social distance, in conjunction with a consumer’senvironmental concern, on responses to an ad promotingenergy efficient products. There was an overall maineffect of a consumer’s environmental concern. The three-way interaction results indicate that when the messageframes do not match, we find a stronger effect of

environmental concern. When the message frames domatch, the effect of environmental concern is weakenedsuch that there are no significant differences across thedependent variables for consumers who are higher orlower in environmental concern. These findings should beof interest to marketing managers, policy makers who areinterested in promoting sustainable consumer behavior,and consumer researchers. References are available uponrequest.

For further information contact:Andrea Heintz TangariUniversity of Arkansas

1 University of ArkansasFayetteville, AR 72701Phone: 479.575.8748



RECOGNIZING THOSE WHO PARTICIPATED AND ENCOURAGINGTHOSE WHO DIDN’T: A SOCIAL MARKETING PERSPECTIVE

ON EXPRESSIONS OF GRATITUDE FROMPOST-KATRINA LOUISIANA

Randle D. Raggio, Louisiana State University, Baton RougeJudith Anne Garretson Folse, Louisiana State University, Baton Rouge

SUMMARY

In response to the outpouring of financial and volun-teer support from both public and private sources follow-ing Hurricane Katrina, the state of Louisiana expressed itsgratitude through a series of “thank you” campaigns.Louisiana’s “thank you” campaigns can be evaluatedfrom a social marketing perspective, as the goal can beseen as improving society by publicly upholding socialnorms of gratitude and encouraging future prosocial behav-iors. Here, we consider gratitude as the emotion that ariseswhen an individual (beneficiary) perceives that anotherperson (benefactor) or source (e.g., God, luck, fate) hasintentionally acted to improve the beneficiary’s well being(Fredrickson 2004). While the role of gratitude clearlyoffers theoretical and managerial implications in socialmarketing, its role in promoting prosocial behaviors isrelatively unknown.

The objectives of this research are threefold. First, weconsider the impact of expressions of gratitude on theprosocial behavioral intentions of Volunteerism and Finan-cial Donations. Next, we identify the potential mecha-nisms responsible for the effects of gratitude. More spe-cifically, we propose that both benefactors’ feelings ofbeing appreciated and their perceptions of beneficiaries’worthiness will explain gratitude effects on these prosocialbehavioral intentions. Finally, we consider these effectsover time to determine whether gratitude is enduring.

Hypotheses were tested using data collected fromtwo different online surveys conducted during two differ-ent time periods. The first data collection occurred eightmonths following Hurricane Katrina, and the second wasgathered seven months later. In both surveys, respondentswere recruited from a U.S. panel of over 50,000 potentialrespondents, all of whom were over the age of 18. TheWave 1 data collection resulted in 1,782 completed sur-veys. The Wave 2 data collection produced 2,689 com-plete surveys.

Results

In H1, we proposed the effects of seeing or hearing a

“thank you” message (Seen) would promote prosocialbehavioral intentions. In support of H

1, there is a signifi-

cant univariate main effect of Seen on both Volunteerism

(F (1, 4372) = 86.85; p < .001; Mseen =

3.71 vs. Mnot seen

=4.32) and Financial Donations (F (1, 4372) = 103.91; p <.001; M

seen =3.94 vs. M

not seen = 4.61). The univariate results

reveal that Seen does not interact with Wave (p > .05), andWave (1 vs. 2) has no significant main effect on eitherVolunteerism (F (1, 4372) = 1.58; p > .05) or FinancialDonations (F (1, 4372) = .17; p > .05).

In H2, we predict that the impact of Seen on

Volunteerism and Financial Donations will be mediatedby benefactors’ feelings of being acknowledged for theirbeneficence. Our measure of Seen serves as the predictorvariable. Our dependent measures (Volunteerism, Finan-cial Donations) serve as outcome variables. The Sobel test(Baron and Kenny 1986) confirms partial mediation ofSeen on both Volunteerism and Financial Donations.Similarly, in H

3, we predict that perceptions of beneficia-

ries’ unworthiness will mediate the impact of Seen. Unwor-thy is not a significant predictor of Volunteerism whenSeen is entered simultaneously in the regression equation,indicating no mediation. However, Unworthy partiallymediates the relationship between Seen and FinancialDonations.

Hypotheses H4a

and H4b

propose the effects of “thankyou” among those who were not associated with the stateto more clearly assess an important element to socialmarketing – does saying thank you promote prosocialbehaviors even among those who are not associated orpersonally affected by the situation. As such, we con-ducted a 2 (Wave: 1 vs. 2) x 2 (Seen: Y vs. N) MANCOVAon Volunteerism and Financial Donations, with Age, Sex,and Distance as covariates on respondents who are notfrom Louisiana, did not participate in relief/rebuildingactivities, are not related to or friends with Louisianaresidents, and are not former residents of the state. Consis-tent with the analysis used to test H1, Seen has a positiveunivariate main effect on Volunteerism (F (1,1283) =4.06; p < .05; M’s 3.07 vs. 3.39) and Financial Donations(F (1,1283) = 5.47; p < .025; M’s 3.25 vs. 3.62). Again, theunivariate main effect of Wave was nonsignificant, as wasthe interaction between Wave and Seen (p’s > .05). Theseresults are substantially identical to the results for H

1,

thereby supporting H4a

.


In H4b

, we predict that for those unassociated with thestate, Unworthy will mediate the relationship betweenSeen and the two dependent measures the same way it didfor the overall sample. The Sobel test (Baron and Kenny1986) confirms that no mediation occurs for eitherVolunteerism or Financial Donation. These results par-tially support H

4b as in both groups (overall sample in H

3

and those “not associated” in H4b

), there was no mediationof Volunteerism, which supports the hypothesis. But thereis also no mediation of Financial Donations, which we didfind in the overall sample. Thus, we find mixed support forH

4b.

General Discussion

We find evidence for two prosocial effects of expres-sions of gratitude: First, for those who participated inrelief/rebuilding efforts, expressions of gratitude have apositive impact on intentions to volunteer and donatemoney in the future. Second, we find a similar positiveimpact of expressions of gratitude on the behavioralintentions of those that did not participate in previousrelief/rebuilding activities and are not associated with the

state at all: The impact of expressions of gratitude spillover to those who are not targets of the message. Further,we find evidence that perceptions of beneficiaries andperceptions that benefactors will properly be thankedmediate the relationship between seeing or hearingexpressions of gratitude and intentions to volunteer ordonate money. Most importantly, the results hold for bothwaves of the survey, indicating that the impact of expres-sions of gratitude are as powerful within the first twoweeks of the official state campaign as they are nearlythree months after all campaigns ended, testifying to theenduring effects of such expressions. This is good newsfor those who are considering whether acknowledgingparticipation is a worthwhile component of a social mar-keting campaign as it indicates a potential positive returnfrom such efforts. These results suggest that it is importantto continue to focus on existing benefactors (“custom-ers”) past their initial involvement with expressions thatrecognize and reward their efforts. Based on our results,expressions of gratitude deserve consideration as a help-ful social marketing tool and should be considered animportant component of disaster management. Refer-ences are available upon request.

For further information contact:Randle D. Raggio

Louisiana State University3122 C Patrick F. Taylor Hall

Baton Rouge, LA 70803Phone: 225.578.2434



STATUS AND MATERIALISM AMONG COLLEGE-AGED WOMEN

Andrea Loulakis, Villanova UniversityRonald Paul Hill, Villanova University

ABSTRACT

Our investigation unpacks how clothing, associateditems, and other public possessions are used by femaleundergraduates to develop and advance their identities oncampus. Additionally, this study includes where clothingwill be worn and who will see it in an attempt to examinewhat role materialism plays in status formation.

INTRODUCTION

The importance of materialism to our understandingof consumer behavior in the developed world cannot beoverstated (Wang and Wallendorf 2006). This conceptreflects the high level(s) of importance buyers/ownersplace on products acquired over time, with an emphasis onvalue to social standing or status (Nguyen 2003). Thus,consumers with more materialistic lifestyles see productsas tools for social networking and drivers of status ratherthan merely objects reflective of their individual needsand desires. Materialism is defined as “the importancepeople attach to owning . . . possessions” (O’Cass 2004).It also may be viewed as a coping mechanism for consum-ers to deal with feelings of uncertainty (Fitzmaurice2006). As a consequence, materialistic individuals areoften less satisfied and self-actualized, more anxious, andmore dependent on external items (Christopher andSchlenker 2004). They place these goods at the center oftheir lives, judge others based on their possessions, andstrive to achieve happiness through product ownership/utilization (Wang and Wallendorf 2006).

Materialism aids our understanding of consumption,self-identity, quality and satisfaction of life, and generalwell-being (Zinkhan 1994). It has become a guidingprinciple in the lives of countless individuals. Consumersgrounded with materialistic ideals place possessions at thecenter of their lives and may have trouble being happy dueto ever-escalating needs (Wang and Wallendorf 2006).Such individuals claim that possessions symbolize theirlikes and dislikes, but research instead correlates theirbelongings with fear of rejection and negative evaluationsfrom their peers (Christopher and Schlenker 2004). Theunderlying value of a possession depends, in part, on thelevel of involvement, and materialistic persons tend to bemore involved with items that bolster their societal roles(O’Cass 2004). Possessions can be used to form conclu-sions about self and others, and O’Cass (2004, p. 869)claims “life’s meaning, achievement, and satisfaction is

judged in terms of what possessions have or have not beenacquired.”

Materialistic individuals tend to distinguish betweenpublic and private meanings of objects, and they are likelyto hold public items in the highest regard (Wang andWallendorf 2006). Consumers may prefer goods thatsignal their desired self-images, recognizing that someproducts define who one is or is not (Nguyen 2003).Further, the outward display of certain possessions en-ables individuals to identify or distance themselves froma particular social group (Thompson and Haytko 1997).For example, clothing worn in public may reveal how onewants to be perceived by people important to them, andthe items one chooses to wear may change dramaticallydepending on who he or she will come in contact. Clearly,branded items, worn by those of all age groups, aredesigned to convey status, prestige, affiliation, and wealth(Clark 2007; Nguyen 2003). Adolescents, however, aremost susceptible to become involved with such identitymarkers, as they make clothing selections that allow themto try on hoped-for personas. As adolescents enter thecollege setting, they are bombarded with new consumerchoices and develop a budding desire to embody a particu-lar persona. Therefore, the purpose of this research is toexamine the meaning of possessions, emphasizing cloth-ing and other public goods, to college-aged women asthey seek social standing among their peers. An interdis-ciplinary literature review is next, followed by our explor-atory method and results. Implications for the study ofmaterialism close the paper.

RELEVANT LITERATURE

Kamptner (1995) discusses the importance of pos-sessions to consumers as they mature from early child-hood and into adolescence. Her findings show that prod-uct meanings deepen over a lifetime as individuals seekexternal successes and develop the confidence whichenables them to be slowly less dependent on comfort-security belongings. Interestingly, females experienceearlier and greater levels of attachment to possessionswith important social implications than similar-aged males.The popular press reports eleven year olds have alreadydeveloped complex associations with branded goods thatdrive creation of self-identities as members of a con-sumer-driven culture surrounded by status-seeking images (Beckstrom 2008). Adolescents also havestrong ties to their belongings because they are in a time


period where they must separate themselves from theirparents both physically and emotionally and it is comfort-ing to look toward possessions for happiness and memo-ries. Belk (1988) states that such dependence on belong-ings decreases with age, though material emphasis willalways be evident. Clarke (2007) argues that adolescentscreate a unique sense of what is important due to the factthat adolescents finally have control over the objects theypurchase, consume, and display.

As an example, teenagers, for the most part, begin tochoose their wardrobes at this stage in personal growth:

Clothing is very important in the world of adoles-cence. As a transitional object, it can serve as anexternal link to, or confirmation of, an adolescentdeveloping internal self. It provides a way for teens toexperiment with their burgeoning confidence andexpress their struggles to cope with their inner emo-tional lives as they emerge from childhood intoadulthood (Clark 2007, p. 52).

One potential outcome is compulsive buying definedas “chronic, repetitive activity that becomes a primaryresponse to negative events or feelings” (Saraneva andSääksjärvi 2008, p. 76). For young consumers, this addic-tion is fueled by insecurities and used to escape reality,which often results in a rollercoaster of emotions thatincludes various stages of disappointment. Such individu-als become consumed by the constant influx of advertise-ments and innovative product selections, rather thanadhering to personal needs and preferences.

At the same time, young consumers are less con-cerned about indebtedness and simultaneouslyhave entitled views of the necessities of life (Penman andMcNeill 2008). There is an ever-present list of justifica-tions for their materialism, including the need to befashionable, to enjoy life, to fit in, and to be entertained.For instance, college-aged students believe they should beable to reward themselves for good grades or other ordi-nary/expected achievements (Penman and McNeill 2008).Such beliefs and/or expectations often are blamed on thewestern “culture of consumption,” but it is equally likelythat the underlying cause is identity issues, especiallyamong adolescent women. They tend to be influenced byothers with whom they shop and may fear purchasingitems of which friends may disapprove, resulting in angstabout acquisition and conformity.

In order to better understand the role of possessionsin the lives of consumers, various categories with differ-ent meanings have been advanced. For example, Richins(1994a, 1994b) divides possessions into four respectivepurposes: utilitarian value, enjoyment, representation ofinterpersonal ties, and self-identity and expression. Thefour subgroups suggest commodities are valued because

of their practical importance or their ability to facilitatesocial relationships. Christopher and Schlenker (2004)agree with the possibility of multiple types of possessions,but are more intrigued by the idea that some represent aninternal self while others an external self, which is similarto Richins (1994b) public/private meanings. Further, theirresearch aims to draw upon the connection between socialidentity and the fear of negative evaluation, suggestingthat more materialistic individuals are more likely to useexternal/public goods to define themselves (Kamptner1995). Material possessions have the power to mask theindividual who owns and displays them. Thompson andHaytko (1997) note that tangible objects can often serveas a means of “identity management,” allowing the ownerto portray them in a way which projects how they wish tobe viewed. Adolescents specifically look toward shapingdifferent identities for themselves as they learn who theyare and how they wish to be perceived. It is important forthem to acquire products that “soothe the anxieties ofseparation and provide the comfort of their own develop-ing personality” (Clark 2007, p. 47) rather than hidingtraits of individuality.

Possessions may serve as symbols that can mediateimportant social relationships (Wang and Wallendorf2006). Specifically, adolescents behave in accordancewith social expectations/norms because they feel as thoughthey are continually being watched and judged (Nguyen2003). They also believe their possessions are imbuedwith symbolic messages and serve as an indicator of socialstatus. Clothing and accessories are an interesting cat-egory of goods because of its ability to meet the require-ments of a host of previously-mentioned subgroups. Cloth-ing is functional and intrinsic, as people must wear cloth-ing on a daily basis to survive and navigate their dailyroutines. Elements of fashion can be used to align indi-viduals with certain cultural values (Murray 2002). Cer-tain outfits also may be used to demonstrate standing andprestige such as expensive and designer suits. For many,buying clothing is enjoyable, provides happiness, or re-lieves stress. Pieces of one’s clothing can even holdmemories of the past (e.g., old sports uniforms or souve-nirs of a family vacation). As Clark (2007, p. 52) states,adolescents use their wardrobe to exude individuality andto “express how they want to be perceived, who theyassociate with or identify with, and a plethora of otherdimensions of their developing selves.” Thompson andHaytko (1997) state that dressing in a particular way is ameans of deterring others from making snap judgmentsabout personality or other internal qualities. Ultimately,“clothing is contingent upon a belief that others will noticeand care about one’s appearance” (Thompson and Haytko1997, p. 22).

This review demonstrates that much is known aboutmaterialism and the accumulation of possessions, butthere remain numerous areas for new research. For instance,


while college-aged students receive regular attention fromscholars, females are rarely singled out and the focus oftenis the relationship between materialism and psychologicalwell-being rather than whether and how products defineyoung consumers. Therefore, our study follows O’Cass’s(2004) lead and unpacks how clothing, associated items,and other public possessions are used by female under-graduates to develop and advance their identities whenthey arrive on campus. Additionally, this investigationincludes where clothing will be worn and who will see it.While Richins (2004b) does examine public and privatemeanings of possessions as suggested by our inquiry, shedoes not analyze which types of possessions fit into whatcategories or whether possessions are seen as having bothpublic and private meanings. As a result, our project seeksto fill this gap using an undergraduate female populationwho are in a place where they are finally able to makemany consumption decisions.

METHOD AND RESULTS

Methodology and Analysis

This study takes into account the considerations andfindings of other scholars in their attempts to definematerialism, conspicuous consumption, and possession-related meanings. Further, it parallels many of the priorstudies’ approaches and questioning techniques. How-ever, this study aims to fill the gaps in previous researchand identify which types of possessions people own forthemselves, and what things people buy for the sake ofothers. Richins (2004b) identifies public and private mean-ings, but in the university setting, much of what one ownsbecomes public because many live with others and it isdifficult to choose which possessions one permits othersto see. However, college students are able to determine, toa certain degree, which items they wish to bring to schooland which ones they prefer to leave at home. College is acoming-of-age tale for many. Away from their parents,students may decide what they wear, who they associatewith, and what aspects of their lives to project to others. Inthe face of this independence, do undergraduate womenseek unique identities, or do they conform to others? Dothese women value that which sets them apart, or thatwhich makes them part of an idealized group? What typesof possessions best represent their owners, and whichones represent what the owners wish to be? This studyattempts to answer such questions and further exploresrelationships with products.

Two types of data collection occurred with this study –possession inventories and in-depth interviews. Posses-sion inventories involved informants looking throughtheir rooms and accounting for all clothing, accessories,and other products that they own. Quantitative measure-ments are important in this phase, as informants were

asked to count each possession and identify who pur-chased it and why it traveled with them to school. Infor-mants were asked to fill in a table of possessions undersupervision of the first author. The second form ofresearch data aims to uncover which possessions mostdistinctly define the individual. Informants partook in anhour-long interview, staged in their bedrooms so theywere surrounded by their possessions while they spokeabout them. The interview was conducted in a veryrelaxed manner; informants usually sat on their beds or attheir desks. Such an environment placed informants atease when expressing their beliefs and attitudes.

The various subgroups of possessions were definedby the interviewer at the outset, and included:

• Clothing – Anything which one has at school withthem, including shoes and other accessories, used toaccent dress.

• Beauty Aids – Products and objects which allegedlyenhance one’s beauty. This category has electronictools (e.g., blow dryers), makeup, lotions, perfume,and other products whose purpose is to make anindividual look or feel better about their appearance.

• Signs and Symbols – Mementos, photos, emblemsand embellishments in one’s room. This categoryincludes amounts and types of decorations, and anyelement of the room which is not necessary to ordi-nary living or excelling in school.

• Functional Possessions – Any object which one usesas a tool for fulfilling their role as a student. Func-tional possessions are practical and range from staplesand notebooks to calculators and printers. Theseitems are useful in daily life and help attain a particu-lar goal.

The interviews were broken into five parts. The firstportions were reflective of the above four subgroups ofpossessions. Informants were asked to respond to ques-tions about their roles as women at this private university,and then more specifically regarding their feelings andactions toward them. The final portion asked informantsto draw conclusions about self identities in relation to theirpossessions. Informants were asked which possessionsmost defined them, which items do not serve as indicatorsof who they are, and what a stranger might say about themif they were to see this collection of goods. The lastquestion allowed informants to think about the meaningsbehind possessions, and the extent to which individualsare judged by belongings. Finally, informants were askedto look at their cumulative experiences at the university,which ranged from several months to three years. Theywere then asked to compare personal styles from high


school to current styles, and to account for which posses-sions had been purchased after making the transition tocollege.

The purpose of the possession inventories was toallow the informants to begin to think about the amount ofthings they have with them at school. For the most part,they laughed and were seemingly embarrassed by theexcess of goods. This result served as a good basis formaking comparisons across individuals. For example,each coed had between seven to fourteen pairs of jeans ordesigner jeans. Most women also had designer handbags,designer bedding, and large quantities of expensive shoesand accessories. The rest of the clothing yielded similarnumbers among the informants. All but two had an iPodand all ten had cell phones; most had printers, digitalcameras, calculators, televisions, and DVD players aswell as other electronics. Only one informant wasrequired by parents to purchase all of her clothing, toilet-ries, and room decorations, while four bought at leastsome of their clothing. Analysis of interviews followedHill and Gaines (2007) to develop a thematic outline offindings. For descriptive purposes, informants are notedas Ashley, Jodi, Christine, Beth, Laura, Julie, Leslie,Emily, Whitney, and Maria, and they ranged from fresh-men to juniors. Hometown, academic major, living andfamily situations, and levels of financial independencevaried.

Relevant Findings

A host of relevant findings that inform our examina-tion of the meaning of possessions to college-agedfemales are presented as they were revealed in the datacollection and analyses. For example, in the major areaclothing, results center on the volume of clothing amassed,the public persona presented by their usage, and howclothing is used to judge others. For example, informantswere asked to examine their unique possession invento-ries and reflect upon each type of clothing. When askedabout whether the quantity of clothing she owned wasadequate or common, Julie stated, “I definitely don’t needall of what I have . . . but I’m not as bad as nearly ¾ of thegirls at this school.” It is clear that at this particularuniversity, even those who desire to have excess note thatit is not a normal environment. Yet even though individu-als recognize that their college setting is not comparableto the majority of the world, they buy into the consumertrends. Christine added,

I think it depends who I compare myself too. Somepeople definitely have nicer clothes than I do. Somepeople I do not see in the same outfit more than once.Some people, in my classes maybe, may wear thesame shirt once a week to a class. I find that I haveenough clothes to not have to do my laundry for 3–4weeks, and I still don’t wear it all. I’d like to say I’m

about average, but maybe I’d say I have too muchstuff at school.

According to Christine, whether one owns too muchclothing depends on who they compare themselves to, nottheir own necessities. She also notes that the individualsshe sees replicating outfits are often those she has classeswith, not necessarily the same students she relates to orconsiders friends. Ultimately, the adequate quantity levelis still driven by peer perception and what others have;girls at this university want to measure up against othergirls. The two most common reasons for having “morethan necessary” are laziness and personal image. Notsurprisingly, women do not enjoy laundry duties and feelit necessary to own sufficient quantities so cleaning couldhappen every three or four weeks.

Though many would view owning a full month’sworth of clothing as excess, it is the ordinary at thisinstitution. However, informants also mentioned that theydo not like to re-wear any element of their wardrobes toomany times to project a sense of style and affluence. Emilystated,

I would say I have the perfect amount for me. I don’tusually wear things more than once, so I’ll wear itthen remove it from my closet and bring it home towear. I keep replacing. Out with the old and in withthe new. To go out at night, I don’t really repeatoutfits.

Emily touches on another identified “social rule” atthis school: do not repeat outfits when you go out at night,especially if pictures will be taken of you. Maria adds toEmily’s response by explaining that she and her friendsprior to college did not share the same mentality ofwearing clothing only once, but her presence at collegechanged her outlook.

If you go out with a dress, it’s a one-time deal. Onceyou wear that dress and it’s in a picture, you don’twear it anymore. At [this school] there is definitely anidealistic image. You have to almost dress to im-press . . . [this school] is very stylistic. I think we havean imaged-based campus that is hard to break through.I think that the majority of the students really careabout their image and how they’re perceived by theirpeers.

Along with the idea of clothing as a means of main-taining image, the informants were asked to discuss theimportance of brands. Julie responded, “I think that de-signer jeans, especially at this school, are extremelyimportant for a girl to have because everyone knows thebrands. Everyone comments on people’s jeans. I nevernoticed them in high school. Now it’s Sevens and TrueReligion.” Nine out of ten informants agreed that they


wear primarily designer jeans, which can cost up to $300per pair. Five of nine were not familiar with most brandsin high school, yet now consider them essential parts oftheir wardrobes.

Beth presented the only alternative view and claimed,

All of my jeans are from Target. I just pick the rightfit for me. In high school people didn’t talk about theirjeans based on the brand and my friends who go to bigstate schools don’t seem to be like that. I still buywhatever I like . . . I have a lot of different stuff, notname brand things. I’ve always had my own style. Ilike to be different. I feel like everyone is just tryingto be like everyone else.

Beth is unique to the majority of the student body, inthat she does not spend unnecessary amounts of money onjeans. However, she also mentioned in her interview thatsince coming to college, she learned to recognize thepatterns and logos of various designers, and now can spota certain style from across the room. Beth is the youngestof the informants interviewed, and has only been at schoolfor a semester; more exposure to the common culture atthis university may persuade her to buy into the trend, too.

Finally, informants were asked about judgments thatcome with clothing usage, and whether their clothing istruly an indicator of their personal style. Ashley claimed,

My clothing is a representation of my personality.Every once in a while you get an idea from someoneelse but I try to have my own looks. You see some-thing on a celebrity or you see something on a normalperson. It’s all inspiration. I think [I dress] for myself.I consider clothes an art form. If I could sew I wouldlove to make them myself. I don’t dress to impressother people; I just find that it is a creative way toexpress yourself. I think my clothing represents mebecause I would say most of my clothes are splitbetween a different look, a unique look, and a com-mon look. I’m creative but I’m not so different.

Ashley takes pride in her clothing as a reflection ofherself and believes it is an art form. However, even insuch cases when one enjoys self-expression, scholarsargue that they instead embody “the perceived uniquenessof wanting to be unique” (Thompson and Haytko 1997,p. 22).

Julie admitted to completely changing her style andthe places she shops after entering college. Nonetheless,when asked how heavily her peers influence her ward-robe, she claimed, “I don’t think I’d stop wearing some-thing because a friend didn’t like it if I liked it, but I thinkif everyone didn’t like . . . if a ton of people are telling you‘that’s not cool,’ you’re not going to wear it.” Clearly Julie

takes into account the opinions of those around her whenchoosing which items in her wardrobe are acceptable towear in public. She also admits that girls look at others’clothing and dictate what is and is not fashionable. Emilyexplained that clothing is a means for judging on campus,“Just day to day people look at you, and I look at otherpeople. I’ll judge what their wearing, whether I like it ornot, whether it’s my style or not, and whether it looks goodon them or not.” If clothing is a measure of self expression,girls would not look to others to obtain approval, or makehasty remarks about those whose styles they do not like.Further, if clothing is an individual choice, it should notmatter how much one owns, how often they wear it, orwho they wear it around. However, at this universityclothing is not for self expression or a personal necessity.Girls use clothing as a means of fitting in to the socialscene. “Specific fashion styles stand for a larger socialidentity” (Thompson and Haytko 1997, p. 23), and girls atthis university know that one of the most effective ways ofentering a social group or obtaining an idealized image isthrough fashion choices.

Informants also were asked to discuss the comple-mentary category of beauty aids. In this vein, Ashleynoted that, “I could say hair and makeup are for others, justbecause I want to look presentable. Clothes are a creativeway to express myself. Hair and makeup I feel is just aboutlooking presentable for others.” Jodi added, “I think nomatter who you are . . . people wear makeup more forothers. Informants admitted to wearing makeup on a dailybasis, adding that they do not usually do their makeup ifthey will be alone all day. Makeup is worn when girls arearound others, or may be around others soon. Further,Maria explained how her habits in regard to makeup havechanged:

I think a lot of people wake up in the morning, makesure their hair is done and make sure they are puttogether, [including] their outfit. I think it has a lot todo with the image they are showing other people andtheir peers accepting their image. I don’t think it’s toomuch [makeup], I just think it’s always on. Mymakeup has definitely changed since high school.Here at [this school] I definitely make a consciouseffort to say, “Okay . . . I have the eyeliner on, thelashes are curled, the mascara’s on, the blush is on,the foundation is on.”

Another consistent beauty aid involves tanning. Onlyone of the ten informants went tanning regularly in highschool but most go now. Jodi claimed that, “I probablywould not go tanning as much if not everyone wenttanning. I definitely like the way I look better when I gotanning, even though I know it’s bad for me.” Further,when asked if tanning was about personal preferences ora means of benchmarking against others, Emily added,“It’s definitely influenced by my friends because they


look tan and I look pasty white in pictures. I never likedgoing tanning really until I got here [to college].” Again,attention is on the idea of photos being taken, a pressurewhich has increased dramatically as a result of Facebookand other social media. Online photos enable girls toalways be seen and judged by their peers. Emily summedup the importance of these two categories as follows:“Everywhere you go people are judging you . . . [and] youare constantly thinking about other people.”

Additionally, informants were asked to look aroundtheir rooms and identify which decorations they feltreflected who they were or wished to become. Responsesdiffered, as rooms varied widely in their decorations.Items included pictures of people at school because col-lege friends are “the ones I’m making memories withnow” (Emily). All ten rooms had at least one picture offamily members and at least one photo of friends fromhigh school. Whitney claimed,

Freshman year I brought so many pictures of my highschool friends because it was my first year of collegeand I wanted to be reminded of home. I think [Idecorate] for me; I brought them for myself. Butobviously I wanted people at college to know that Ihad an awesome group of friends at home. I think a lotof my stuff has to do with my past. If you look at mypictures, all of it is either my high school friends ormy family. There’s a few current pictures, but defi-nitely more of my past.

Unlike the last set of possessions, coeds believed thatthey decorate their room for themselves; Jodi said, “I likethe way it looks. I have to sit here every day and look at it.”Emily claimed, “You’re away from home and want tomake it homey. Having curtains and a seat cushion andflowers kind of represents my room at home.” Theseresponses seem to reflect that one’s room at collegeprovides an important transition between past and presentlives. However, like many public possessions such asclothing, one’s room can also be under “surveillance.”Julie acknowledged,

I think you decorate your room so people come in andsay, “oh this is cute” and they get a sense of who youare as a person. But if you’re just by yourself, youknow who you are so you wouldn’t need to . . . I thinkpeople would be lying if they said that they don’twant to impress people. I think that you want yourroom to be super cute.

Julie’s assertion shows that even one’s bedroom, themost private place one has on a college campus, may bedecorated for the purpose of impressing peers. Addition-ally, a bedroom is a place which one must invite others to,suggesting that students’ lives are under scrutiny even bythose with whom they feel most comfortable.

After a discussion of functional possessions, respon-dents were asked to consider the following scenario: If astranger came in your room and looked through yourthings, what assumptions would be made about you?What generalizations would they make about your life,values, and you as a person? What do your possessionsalone say about who you are? Though the responsesvaried, most informants thought their possessions wouldmake people think that they were part of the societal elite.Responses included, “Probably that I was stuck up. . . .Almost everything in my closet does have a label on it.”“Probably that I was a brat. I think that the girls at thisschool have a lot of things. I wouldn’t consider myselfsomeone that was spoiled, but I’m comparing myself togirls on this campus.” “I would think that they wouldassume that I have a lot of money because of the amountof clothing and shoes and things that I have in my room.”Only Beth and Christine argued differently for a moreaverage rating. Further, the interviewees were not embar-rassed by the thought of a stranger viewing them asaffluent. Many went on to say that if the stranger went tothe same school that she did, they would assume the roomto be average, but that a stranger from outside of campuswould presume otherwise. Students seemed to prefer theidea of being viewed as overly wealthy and fashionablerather than have others believe that they were unstylish.Girls at this university believe that status outweighs all.

Finally, respondents were asked to think about whatothers would say about their storehouse of possessions,and how they might publicly define an individual. Theresponses once again varied, but two perspectives domi-nated. For example, Leslie suggested her possessionsstood alone as potential indicators of personhood: “I thinkmy possessions do, pretty accurately, say who I am. So Ithink my possessions tell a general picture of me. Theydon’t tell my exact interests or exactly what’s most impor-tant to me but I think it gives a good beginning.” Ashleyoffers a revision: “I don’t have one specific thing at schoolthat defines me. I feel like possessions I bring to school area combination of who I am.” On the other hand, mostinformants argued: “I do not think your possessionsdefine who you are. I think the people you surroundyourself with and your personality in general define whoyou are” (Whitney). In the final analysis, these womenturned away from their previous reactions toward tradi-tional markers of identity such as family, friends, faith,and reputation.

DISCUSSION AND IMPLICATIONS

O’Cass (2004) suggests that female consumers suchas our informants clearly are materialistic and more inter-ested in the symbolic nature of possessions than theirfunctional qualities. For example, most of their discus-sions involved public meanings of goods such as cosmet-ics and other beauty aids, with the only real exception the


wearing of comfortable clothing when alone. The activemanipulation of style was described regularly in conver-sations, as informants portrayed how use of items likeclothing can be easily changed and/or combined withother items to meet social goals and objectives on campus.This finding is consistent with the belief that brandedgoods are important signifiers to boys and girls as early aseleven (Beckstrom 2008), and by the time one reachesadolescence, clothing can convey both status and groupaffiliation (Nguyen 2003). Thus, clothing and other pub-lic possessions act as transition objects allowing youngpeople to develop and confirm self-identities while simul-taneously coping with interpersonal insecurities (Clarke2007). As this research indicates, tension is not in the self-expression but in the anxiety about outside judgments.

The responses regarding possessions used to deco-rate rooms also support our discussion about public goodsto the extent to which informants believed their roomswhere a showcase to friends. This result is consistent withprevious research by Kamptner (1995), who noted thatyoung women place the highest levels of importance onsocial meanings of belongings such as photographs andmemorabilia. Richins (1994b) suggests the bedroom is aplace filled with possessions which have private mean-ings other than status or show. For instance, informantsshould be able to decide who they want to invite into suchprivate quarters, and presumably only one’s close friendsare ever allowed. If this belief holds true, the fact that someof our informants felt pressured to have dwelling spacesthat pass external scrutiny (e.g., a “cute room”) shows themost private possessions receive some amount of publicconsideration. In these circumstances, Wang andWallendorf (2006) might argue that our informants areinsecure and more materialistic than other more stable andmature consumers.

Throughout the interviews, informants claimed pos-sessions reflected styles, personalities, and self-prefer-ences. If so, the projection that an observer of theirstorehouse of goods would view them as financiallyspoiled or snobby seems counterintuitive. This reaction istelling given informants were also certain that their faithand family are both strong indicators of who they are. Theseeming paradox may be explained, in part, by relativeaffluence and strong sense of traditional values of thecommunity, with one-fourth of the current freshman classhaving parents who make in excess of $250,000 per yearand the vast majority sharing the same religious back-ground. Beth reveals that, “[This school] is unlike mostschools because everyone seems to come from the sametype of background, whereas at other schools, it seemslike there’s more of a variety of people that come from allover, different types of families.” Jodi agrees but goes onestep further:

I think [students are] molded once they get here. Ididn’t come [here] and look around campus and seethat there were girls with good style. I came [here]because I knew that it was a well-known schoolacademically. … But then once I got here and I startedbecoming friends with girls, and older girls; that’swhen I started becoming more aware of what every-one was wearing.

Of course, care should be taken in generalizing theseresults. College-aged students are more prone than othergroups of individuals to use their possessions as a meansof creating impressions by others. Similarly, college-agedwomen are in a time in their lives when they are able to tryon new personas to determine who they will become overtime. Consumption of possessions is an easy way toexperiment with different identities. Though this processmay seem to result in superficial ways of modeling anadult perspective of self, college may be the first timethese young women sought self-definitions without theinfrastructure and stability provided by parents, relatives,and their original communities. Whether the type ofinstitution (religious-affiliated, private, east coast, etc.)plays any additional important role is the subject ofanother project that looks across universities. However,the patterns of development, socialization, and experi-mentation likely occur regardless, especially when thecollege is primarily residential.

Expansion of this research project could involveexamination of individuals from very different economicsituations. Our study involved students with a ratherhomogeneous life-perspective that may be representativeof affluent female consumers. With this possibility inmind, what variation did occur may be related to theirrelative access to possessions rather than need or desire.In addition to study of different socioeconomic classes,tracking of individuals as they acculturate into a univer-sity community over time could be useful, especially thelooking at what belongings come with them, what itemsare accumulated early on to later in their academic careers,and how their meanings evolve as students mature. Clarke(2007) argues that individuals become more unique asthey age, which may yield results from freshman thatconcentrate attention on more common and public posses-sions that create a new identity as a college student thanupperclassmen and women who are focusing on the nexttransition into work life. In fact, a longitudinal study of thevarious transitions women pass through from grade schoolto empty nesters or its single equivalent may provide aninteresting perspective on the dynamic role of differentcategories of possessions as women mature into andthrough their existences.


REFERENCES

Beckstrom, Maja (2008), “Tweens Want Hip Stuff, butSelf-Esteem is the Real Need: U Study Finds Boost-ing Kids’ Confidence Deflates Materialism on theCusp of Adolescence,” McClatchy – Tribune Busi-ness News, 26 (November).

Belk, Russell W. (1988), “Possessions and the ExtendedSelf,” Journal of Consumer Research, 15 (Septem-ber), 139–68.

Christopher, Andrew and Barry R. Schlenker (2004),“Materialism and Affect: The Role of Self-Presenta-tional Concerns,” Journal of Social and ClinicalPsychology, 23 (April), 260–72.

Clarke, Whitney Goodrich (2007), “A Psychology ofMaterialism: Contributions of Development andCulture to our Material Longing,” Ph.D. dissertation,Alliant International University, San Francisco Bay,United States – California. Retrieved January 2,2009, from Dissertations & Theses: Full Text Data-base, (Publication No. AAT 3282273).

Fitzmaurice, Julia and Charles Comgys (2006), “Materi-alism and Social Consumption,” Journal of Market-ing Theory and Practice, 14 (Fall), 287–99.

Hill, Ronald P. and Jeannie Gaines (2007), “The Con-sumer Culture of Poverty: Behavioral Research Find-ings and Their Implications in an Ethnographic Con-text,” The Journal of American Culture, 30 (March),81–95.

Kamptner, Laura (1995), “Treasured Possessions andTheir Meanings in Adolescent Males and Females,”Adolescence, 30 (Summer), 301–18.

Murray, Jeff B. (2002), “The Politics of Consumption: ARe-Inquiry on Thompson and Haykto’s (1997)‘Speaking of Fashion,’” Journal of Consumer Re-search, 29 (December), 427–40.

Nguyen, Lan T. (2003), “Growing up in a Material World:

An Investigation of the Development of Materialismin Children and Adolescents,” Ph.D. dissertation,University of Minnesota, United States – Minnesota.Retrieved January 2, 2009, from ABI/INFORM Glo-bal database, (Publication No. AAT 3087778).

O’Cass, Aron (2004), “Fashion Clothing Consumption:Antecedents and Consequences of Fashion ClothingInvolvement,” European Journal of Marketing, 38(March), 869–82.

Penman, Sarah and Lisa S. McNeill (2008), “SpendingTheir Way to Adulthood: Consumption Outside theNest,” Young Consumers, 9 (November), 155–69.

Richins, Marsha L. (1994a), “Special Possessions and theExpression of Material Values,” Journal of Con-sumer Research, 21 (December), 522–33.

____________ (1994b), “Valuing Things: The Publicand Private Meanings of Possessions,” Journal ofConsumer Research, 21 (December), 504–21.

____________ (2004), “The Material Values Scale: Mea-surement Properties and Development of a ShortForm,” Journal of Consumer Research, 31 (June),209–19.

Saraneva, Anna and Maria Sääksjärvi (2008), “YoungCompulsive Buyers and the Emotional Roller-Coasterin Shopping,” Young Consumers, 9 (January), 75–89.

Thompson, Craig and Diana Haytko (1997), “Speaking ofFashion: Consumers’ Uses of Fashion Discoursesand the Appropriation of Countervailing CulturalMeanings,” Journal of Consumer Research, 24 (June),15–42.

Wang, Jeff and Melanie Wallendorf (2006), “Materialism,Status Signaling, and Product Satisfaction,” Academyof Marketing Science Journal, 34 (Fall), 494–505.

Zinkhan, George M. (1994), “Advertising, Materialism,and Quality of Life,” Journal of Advertising, 23(June), 1–4.

For further information contact:Ron Hill

Villanova School of Business800 Lancaster AvenueVillanova, PA 19085Phone: 610.519.3256



WHEN DOES SCENARIO WRITING IMPROVE THE QUALITY OFPLANNING DECISIONS?

Kyeong Sam Min, University of New OrleansHal R. Arkes, Ohio State University, Columbus

SUMMARY

People tend to be optimistic in their planning deci-sions. For example, many consumers purchase a productthat offers a mail-in rebate reward but they often fail toredeem it because they discover later that redemptionrequires a large amount of time and effort. Marketingmanagers may think that they can market a new productvery soon even though unforeseen production delays hadfrequently made them postpone the product launch in thepast. Such a tendency to generate optimistic predictions isdefined as a “planning fallacy” (Buehler, Griffin, andRoss 1994; Kahneman and Tversky 1979). While thisfallacy has been documented in various planning con-texts, we still know very little about how to reduce it. Thepurpose of this paper is to investigate the factors thatinfluence and the processes that underlie an individual’soptimistic prediction bias.

According to the ease of generation literature (Schwarz2004), when considering an event, a person who hasdifficulty generating focal thoughts about its occurrenceor who has no difficulty generating focal thoughts aboutan alternative event’s occurrence should lower thatperson’s evaluation of the event. For example, Sanna andSchwarz (2004) observed a decrease in the optimisticprediction bias when their participants generated eithermany, rather than only a few, reasons that might lead themto successfully complete an upcoming event, or only afew, rather than many, reasons that might lead them tounsuccessfully complete the project.

Such ease of generation effects have been mostlydocumented using a thought-listing task that asks peopleto list various reasons why a particular outcome is gener-ated. However, we demonstrate similar effects using ascenario generation task that asks consumers to create

multiple sequential steps to reach a particular outcome. Aspredicted by the ease of generation theory, we hypoth-esized that, difficulty in generating an optimistic scenariowould be more likely to decrease consumers’ optimisticprediction bias than ease in generating such a scenario. Incontrast, we predicted that consumers might think thatthings would go as poorly as they possibly can when theyfound it easy to write a pessimistic scenario as opposed todifficult to write such a scenario.

However, recent studies showed that the traditionalease of generation effect remained the same when indi-viduals interpreted that ease of generation would be goodbut that it was reversed when individuals perceived thatease of generation would be bad (e.g., Briñol, Petty, andTormala 2006). Building upon Briñol et al.’s (2006)findings, we also hypothesized that the meaning of sub-jective ease would qualify the ease of generation effectsuch that experiencing difficulty in generating an optimis-tic scenario would reduce individuals’ optimistic predic-tion biases if they thought that ease of generation wasgood, that is, leads to accurate estimates. Additionally, wepredicted that such a planning difficulty effect would bereversed if consumers thought that difficulty in generatingsuch a scenario was good.

The results of real-world planning decisions acrosstwo experiments provided consistent support for ourpredictions mentioned above. Specifically, we extendedour understanding of the optimistic prediction bias byidentifying debiasing strategies that would help to reducean individual’s planner’s optimistic bias. The findings ofthis research are expected to help marketers, consumers,and policy makers improve their decision quality. Wepresent the limitations of this research, theoretical andmanagerial implications, and suggestions for futureresearch. References are available upon request.

For further information contact:Kyeong Sam Min

Department of MarketingCollege of Business Administration

University of New Orleans2000 Lakeshore Dr.

New Orleans, LA 70148Phone: 504.280.6195



PRODUCT BUNDLING OR MENTAL SIMULATION: WHAT IS MOREEFFECTIVE IN REDUCING USAGE BARRIERS?

Sabine Kuester, University of Mannheim, GermanyJennifer Fitzgerald, University of Mannheim, Germany

Silke Hess, University of Mannheim, GermanyJudith Bucerius, University of Mannheim, Germany

SUMMARY

The continuous development and introduction ofnew products is vital for companies’ long-term success.High expenditures on innovative projects require a swiftreturn on investment that can be achieved by a rapiddiffusion of the innovation in the market (Montaguti,Kuester, and Robertson 2002). Individuals’ adoptionbehavior is a serious problem as customers’ resistance toinnovation is stated as a major cause of new productfailure (Garcia, Bardhi, and Friedrich 2007; Ram 1989).Usage barriers represent the most common reason forcustomers’ resistance to innovations (Ram and Sheth1989). Usage barriers are factors within the innovationdecision process which inhibit innovation adoption (Brownand Venkatesh 2003). Therefore, it is of major importanceto identity marketing instruments that can reduce usagebarriers and facilitate adoption.

We identified three types of usage barriers whichhave an effect on innovation adoption: usage difference(Hoeffler 2003), required behavioral change (Robertson1971; Veryzer 1998) and difficulty of use (Veryzer 1998).To overcome usage barriers and to utilize a productsuccessfully, customers need to change their behavior andhabits and learn how to make use of the innovation. Thisis especially true for radically new products or if custom-ers’ experience in the respective product category islimited (Hoeffler 2003; Veryzer 1998). Since the percep-tion of usage barriers depends on contextual factors suchas product newness and expertise, innovative companiesneed to know how they can reduce usage barriers in eachsituation. Our research aims to clarify when the productstrategy and product bundling is appropriate and intro-duce the communication strategy of mental simulation asa more efficient method of reducing usage barriers.

Product bundling, “the sale of two or more separateproducts or services in one package” (Stremersch andTellis 2002, p. 57), allows customers to make inferencesabout the unknown product based on information aboutthe familiar product (Popkowski Leszcyc, Pracejus, andShen 2008). Product bundling facilitates the evaluationand understanding of innovations by providing a contextin which to evaluate the new product thereby reducingcustomers’ usage barriers. Mental simulation, the mentalrepresentation of an event or series of events (Taylor and

Schneider 1989), is an effective instrument to reduceperceived usage barriers (Dahl and Hoeffler 2004; Hoeffler2003). We propose that the relative effectiveness of theseinstruments on reducing customers’ usage barriersdepends on context factors, specifically product newnessand user expertise.

Based on a review of extant literature, we identifiedthree types of usage barriers; usage difference, requiredbehavioral changes and difficulty of use. Usage differ-ence exists because a customer notices that the use of theinnovation is different from the use of conventional prod-ucts. This perceived difference affects the customer’sevaluation of the new product (Zhou and Nakamoto2007).

Behavioral changes make up the second type of usagebarriers. Once the incompatibility of the innovation toexisting usage patterns is perceived, the customer realizesthat a change in behavior is necessary in order to incorpo-rate the innovation. Customers have been shown to refuseadoption if the amount of necessary adjustment and ac-commodation is perceived as too high (Veryzer 1998).

Difficulty of use forms the third type of usage barri-ers. As part of the decision process, customers assess thedegree of difficulty to use an innovation. Difficulty of useis influenced by the perceived complexity of the product(Mukherjee and Hoyer 2001). An experienced user wouldperceive an innovation to be less complex than a novice.

Due to their category knowledge, experts compre-hend innovative products better than novices (Moreau,Lehmann, and Markman 2001). This competent evalua-tion allows experts to more easily overcome usage barri-ers. Our results establish mental simulation as an effectivetool to reduce usage barriers for novices. Reduced usagebarriers will then decrease innovation resistance andenhance adoption intention.

Similar to customers’ expertise, perceived newnessof an innovation influences the evaluation and compre-hension of new products (Rogers 2003; Veryzer 1998).To analyze the impact of different levels of newness, wedifferentiate between incrementally new products (INP)and radically new products (RNP).


Participants evaluated an RNP and an INP in sce-narios simulating product bundling and mental simula-tion. We found that the effect of these instruments de-pends on the newness of the innovation – incrementallynew (INP) vs. really new (RNP) – and level of customerexpertise. For experts, mental simulation was the onlyinstrument capable of reducing usage barriers to RNP.Both instruments were effective in reducing usage barri-ers for novices in INP and RNP. This confirms that levelof expertise is a significant factor in the adoption process

and new product marketing must be tailored to customerexpertise.

It is critical for managers to consider the fact that thesuccess of an innovation depends on customer-relatedfactors in order to make use of the proper method ofreducing customers’ usage barriers and increase the abil-ity of an innovation to succeed. References are availableupon request.

For further information contact:Jennifer Fitzgerald

University of MannheimL9, 1–2, Room 4.1068131 Mannheim

GermanyPhone: +49(0)621.181.2389



THE EFFECT OF GOALS ON STATUS-QUO BIAS WITHINA CHOICE FRAMEWORK

Sajeev Varki, University of South Florida, TampaNajam Saqib, Qatar University, Qatar

SUMMARY

Status-quo bias, the observed preference for one’scurrent choice, assumes importance in a marketing con-text since new brands have to overcome the thresholdimposed by status-quo bias. In a recent paper, Chernev(2004) has extended the literature on status-quo bias byshowing that goal orientation, or the manner in which oneapproaches goals, impacts the extent of status-quo bias.He shows that prevention oriented individuals, i.e., indi-viduals who focus on duties and obligations (as opposedto hopes and aspirations in promotion focus) exhibit agreater preference for status-quo.

In our paper, we extend Chernev’s (2004) work byexploring how consumer goals per se interact with goalorientation in their impact on status-quo bias. Within thecontext of the goal framework postulated by Bagozzi andDholakia (1999), goal orientation deals with how onewants to achieve a goal (the goal-striving component ofBagozzi and Dholakia’s framework), whereas the goalsper se relate to what one wants to achieve (the goal settingcomponent of Bagozzi and Dholakia’s framework). How-ever, the literature stream that examines the effect of goalson consumer decision making has strangely been silent onthe effect of goals on status-quo bias. This question is ofdirect relevance to marketers who would like to knowwhether goal activation via appropriate communicationthemes could result in a consumer overcoming the naturalthreshold imposed by status-quo bias. This questionassumes added relevance given Chartrand et al.’s (2008)finding that goal activation can even happen uncon-sciously and still have a strong influence on choice.However, whether such activation is sufficient to over-come the threshold imposed by status-quo bias remainsunanswered.

The specific questions of interest that we examine inthis paper as they pertain to the role of goals in choice areas follows. Do goals influence status quo bias and thuschoice? Do goals moderate the effect discovered byChernev (2004), namely, that status-quo bias is strongeramong consumers with a prevention-focus orientation?Or more specifically, does the type of goal influence theeffect of goal on status-quo bias?

Two studies are conducted. In Study 1, a 2 (goal/no-goal) by 2 (promotion/prevention focus) between-subjectdesign is adopted. The goal primed was the thrift goal ofsaving money, and it was induced by getting subjects todecide on two alternative ways to reduce credit card debtfor a fellow student. The experimental scenario involvedthe choice of two apartment rentals that vary by rent anddistance to work. Based on data from 190 student subjects,we find that goals per se exert as strong an influence asgoal orientation (cf., Chernev 2004) on status-quo bias.However, the interaction between goals and goal orienta-tion is not significant, indicating that promotion andprevention focus individuals are equally likely to switchout of status-quo when a thrift goal was involved. Toexplore whether this non-significant interaction was theresult of the nature of the goal, a second study wasconducted where the nature of the goal was varied. InStudy 2, a 2 (hedonic/utilitarian goal) by 2 (promotion/prevention focus) between-subject was adopted. Thehedonic goal was primed by getting subjects to decide onvacation choices for a fellow student. The utilitarian goalremained that of thrift (money saving) as in Study 1. Basedon data from 278 student subjects, we replicate the earlierresults in that when utilitarian goals are involved (e.g.,thrift goal of saving money), both promotion and preven-tion focus individuals switch out of status-quo in equalnumbers. It is only when hedonic goals are present thatprevention focus individuals exhibiting a greater prefer-ence for status-quo. Also, we find that hedonic goals(across self regulatory focus conditions) have a strongerattenuating effect than utilitarian goals on status-quo bias.

Our research results complement the work of Chernev(2004) in several ways. First, we show that non-consciousgoals can influence status-quo bias just as much as thegoal orientation or regulatory focus of the individuals.Second, the effect of prevention focus on the enhancedpreference for status-quo may differ depending upon thetype of goal involved. Our results suggest that whenminimal goals in the form of the utilitarian goal of savingmoney are present, there is no difference among preven-tion and promotion focused individuals in terms of theirpreference for status-quo. Third, we show that the forfei-ture of hedonic goals weighs more heavily than theforfeiture of utilitarian goals, and this effect is amplifiedeven more when individuals are in promotion focus.


REFERENCES

Bagozzi, Richard P. and Utpal M. Dholakia (1999), “GoalSetting and Goal Striving in Consumer Behavior,”Journal of Marketing, 63, 19–32.

Chartrand, T., J. Huber, B. Shiv, and Robin J. Tanner

(2008), “Effects of Non-Conscious Goal Priming onConsumer Choice Behavior,” Journal of ConsumerResearch, 35 (August), 189–201.

Chernev, Alexander (2004), “Goal Orientation and Con-sumer Preference for the Status Quo,” Journal ofConsumer Research, 31 (December), 557–65.

For further information contact:Sajeev Varki

MarketingCollege of Business Administration

University of South FloridaTampa, FL 33620

Phone: 813.974.6227Fax: 813.974.6175



THE EFFECTS OF MENTAL CONSTRUAL AND PERCEIVED RISKIN RESPONSE TO LOSS- VERSUS GAIN-FRAMED

HEALTH-RELATED MESSAGES

Yun Lee, University of Iowa, Iowa City

SUMMARY

Many people continue to engage in negative healthbehaviors, even when they know the risks (Fisher andMisovich 1990; Berger and Rand 2008). It is estimatedthat 40 percent to 70 percent of all premature deaths andup to 66 percent of all disabilities could be prevented bycontrolling a few health risk factors (i.e., diet, exercise,and alcohol abuse) (Signorielli 1993). The importance ofeffective health communications to reduce national healthproblems (e.g., AIDS, heart disease, drug abuse, anddriving and drinking), therefore, has risen over the lastdecade (Block and Keller 1995). It is also increasinglyimportant to understand how to increase the effectivenessof health messages to maximize their impact on people’shealth-related thoughts and behaviors (Block and Keller1995; Rothman et al. 2006). Optimizing the influence ofhealth messages has long been of interest to public healthexperts and communication researchers (Salovey andWegener 2003) and substantial efforts have been devotedto shape people’s perceptions on health issues (Rothmanand Salovey 1997).

Persuasive risk communication should involve pro-viding effective health messages relevant to the behav-ioral issue at hand in a manner that intended impact can bemaximized on people’s thoughts and mind, and thusencourage people to adjust their behavior and to stimulatefavorable outcomes in adopting suggested particularbehaviors (Rothman et al. 2006). To increase the effec-tiveness of a health appeal, health relevant messages haveoften incorporated message framing method in terms ofthe expected benefits (gains) or costs (losses) associatedwith a particular behavior (Rothman and Salovey 1997).Message framing has provided a theoretical ground to thedevelopment of effective health messages (Rothman et al.2006). An interesting pattern of findings on messageframing and health promotion is that loss-framed mes-sages motivates more in-depth processing, and whenparticipants process health information in-depth, loss-framed messages are more persuasive than gain-framedmessages (Block and Keller 1995; Maheswarna andMeyers-Levy 1990), whereas gain-framed messages aremore persuasive than loss-frame ones when there is littleemphasis on detailed processing (Maheswaran andMeyers-Levy 1990).

These findings suggest that loss-framed messageswill be more motivational when they are incorporatedwith concrete and detailed level of health information,whereas gain-framed messages will be more persuasivewhen they are incorporated with abstract and general levelof health information. To test this prediction, this studyemploys construal level theory. According to construallevel theory, people mentally construe objects or eventsthat are psychologically near in terms of low-level, detailed,and how-laden features, whereas at a distance they con-strue the same objects or events in terms of high-level,abstract, and why-laden characteristics (Förster, Fried-man, and Liberman 2004; Trope and Liberman 2003). Asfor linking message framing with temporal distance, in aconsumption context, a recent study demonstrated thatpeople prefer loss-framed ad messages, when a purchaseis temporally about too made, whereas when a purchase istemporally distant, gain-framed product messages becomemore appealing than loss-framed ones (Mogilner et al.2008).

Drawing on this theory, the current study proposesthat when low mental construal level of health informa-tion is loss-framed than when it is gain-framed, and whenhigh mental construal level of health information is gain-framed than when it is loss-framed, favorable attitudesand greater intentions to follow the recommended behav-ior will be led. However, we predict that the effects ofmental construal and message framing on attitudes andbehavioral intentions will be observed only when healthrisk information is highly relevant to participants, but notwhen it is less relevant to them, because people are more(vs. less) likely to integrate information into a weightedoverall evaluation when their perceived risk is high (vs.low) (Block and Keller 1995; Gleicher and Petty 1992;Rothman et al. 1993).

The results of two experiments demonstrate the influ-ence of high versus low mental construal level informa-tion which is either loss- versus gain-framed on attitudestoward health-related messages and behavioral inten-tions. The results showed significant three-way interac-tion effects (i.e., experiment 1: F(1,127) = 5.605, p < .02;experiment 2: F(1,152) = 11.277, p = .001). Follow-upcontrasts showed that when low mental construal levelinformation was loss-framed, and when high mental


construal level information was gain-framed, attitudestoward the health messages were more favorable (experi-ment 1) and behavioral intentions to follow the recom-mended behavior was greater (experiment 2). As expected,the joint effects of message framing and mental construalwere observed only when the health information pre-sented was highly relevant to the participants. When theissue of health risk information was less relevant to them,however, the effect of matching the mental level of healthinformation with different types of message framing wasdisappearing.

Across two experiments, the current study demon-strated that mental construal is an important antecedent ofreactions to loss- versus gain-framed health-related mes-sages. We believe that this research is one of the startingpoints in developing a framework for understanding infor-mation processing implications for persuasive and effec-tive health relevant risk communication, where mentalconstrual, perceived risk, and message framing are pri-mary drivers. References are available upon request.

For further information contact:Yun Lee

Department of MarketingUniversity of Iowa

21 East Market St. S252 PBBIowa City, IA 52242Phone: 319.335.0967



BASES OF BRAND RELATIONSHIPS: REDUNDANT CONSTRUCTS ORMEANINGFUL DISTINCTIONS?

Christy Ashley, East Carolina University, GreenvilleE. Deanne Brocato, Iowa State University, Ames

SUMMARY

Differences exist in the literature regarding the mea-surement and description of relational constructs. Forexample, emotional bonds between consumers and brands,including firms and service providers, have been alter-nately described as emotional attachment, affective com-mitment, relatedness, identification, and brand love. Areresearchers measuring different constructs, or simplyusing different names to describe the same construct?

To respond to this question, this paper presents atypology of selected constructs used to describe emotion-laden relationships with service providers, including:identification, relatedness, emotional attachment, andbrand love. In order to address the lack of clarity aroundthe constructs, which are frequently used interchangeablyor measured using the same items, we define and proposerelationships between the constructs. The goal of theproposed reconceptualization is to create a more fine-grained understanding of relationships between the con-structs. If the constructs overlap, future research can usefewer constructs to create more parsimonious models. Ifthe constructs do not overlap, researchers should stopusing constructs interchangeably and further the researchusing common definitions.

Each construct is defined and categorized within atypology to help elucidate the role of each construct inconsumer-brand relationships. We separate the relatedconstructs into two groups: constructs that describecharacteristics that exist within and across relationshipclasses (identification and relatedness) and constructs thatdescribe forms or classes of relationships (emotionalattachment and brand love). Further, we proposerelationships between the constructs and importantoutcomes of relationships, including commitment, loyalty,and brand support.

The individual hypotheses include:

H1a–d: Identification is positively related to Related-ness, Emotional Attachment, Brand Love, andCommitment.

H2a–c: Relatedness is positively related to EmotionalAttachment, Brand Love, and Commitment.

H3a–b: Emotional Attachment is positively related toBrand Love and Commitment.

H4a–c: Brand Love is positively related to Commitment,Brand Support and Loyalty.

H5a–b: Commitment is positively related to Brand Sup-port and Loyalty.

We test the proposed model using 356 studentresponses to a survey about Target, a mass merchandisediscount store. We demonstrate that each construct, whenproperly conceptualized, has a unique contribution toconsumer relationships with service providers. The stron-gest relationship between the constructs occurs betweenthe identification and relatedness constructs, which areclassified as constructs that could be measured across alltypes of consumer relationships with retailers. Identifica-tion is conceptualized as a cognitive evaluation based onshared values. Relatedness is conceptualized as theretailer’s ability to make the consumer feel welcome,close or connected to the retailer. The strong relationshipsbetween commitment and loyalty and commitment andsupport were anticipated.

In spite of our ability to discriminate between theconstructs, the findings also suggest that the constructs westudied are highly interrelated. When one construct, likeidentification, is elevated, cognitive and behavioralprocesses may elevate other constructs, like relatedness oremotional attachment. As a result, researchers may betempted to treat these constructs or items that measure theconstructs as interchangeable with one another. We hopethat our findings inspire researchers in the consumerrelationship area to make their assumptions explicit aboutthe relationships between these constructs and to put thoseassumptions to empirical tests. Consequently, we viewour findings as a starting point, not an ending point. Welook forward to greater consistency and clarity, which iscritical to advance our understanding of consumerrelationships.


For further information contact:Christy Ashley

East Carolina UniversityMail Stop 503

Greenville, NC 27858Phone: 252.328.6099


E. Deanne BrocatoIowa State University

3256 Gerdin Business BuildingAmes, IA 50011

Phone: 515.294.9634Fax: 515.294.2534



CONVERTING TRANSACTION-BASED SERVICES TO RELATIONSHIP-BASED SERVICES: A CRITICAL ROLE OF INFORMATION

Haisu Zhang, University of Illinois at ChicagoChengli Shu, University of Illinois at Chicago

SUMMARY

Extant literature has shed light on the role of servicerelationships in marketing (e.g., Bitner 1995; Gummesson1998). Although not every service situation is perceivedas relationship-based (Bitner 1995; Berry 1995), due togreat mutual benefits customers and service providers stilldevote growing efforts to establishing relationships witheach other (Parasuraman, Berry, and Zeithaml 1991). Inline with prior studies (e.g., Dwyer, Schurr, and Oh 1987;Gutek 1995), the relationship-based service (RBS) isconceptualized as one where the customer, based onprevious agreements, interacts with the service providerin an ongoing process, while the transaction-based service(TBS) refers to one where the customer interacts with theservice provider with no relational foundations (e.g.,shorter duration and sharper ending).

Information processing is suggested to facilitate rela-tionships between customers and service providers(Parasuraman, Zeithaml, and Berry 1985). According tothe commitment-trust theory (Morgan and Hunt 1994),information fosters the establishment of a relationship

through positive effects of communication on trust. Inaddition, prior studies suggest that information process-ing aids consumers’ evaluation of services (Zeithaml1981) and service providers’ innovations (Blazevic andLievens 2008). In this paper, we maintain that the conver-sion from TBS to RBS depends on similarity betweenacquired information and customers’ knowledge. In linewith extant studies, in our research a customer’s informa-tion acquisition consists of three stages: pre-service, ser-vice encounter (i.e., during service), and post-service.Consistent with Solomon et al. (1985), we refer to theservice encounter as interactions between customers andproviders, and it reflects the “during-service” stage ofinformation acquisition.

Borrowing literature from both services marketingand consumer behavior areas, our research contributes tothe literature by examining (a) the role of customers andtheir information/knowledge in services, (b) the differ-ences in information processing between RBS and TBS,and (c) the conversion of TBS to RBS. Figure 1 reflects theframework of our research. References are available uponrequest.

FIGURE 1Information Processing Services

Pre-ServiceInformation Acquisition

Information Acquisitionin the Service Encounter

Post ServiceInformation Acquisition

Relationship-BasedService (RBS)

Transaction-BasedService (RBS)

CustomerKnowledge

(Time 1)

CustomerKnowledge

(Time 2)

CustomerKnowledge

(Time 3)

CustomerKnowledge

(Time 4)

Conversion

TrustSatisfactionAnticipation

Pre-ServiceInformation Acquisition

Information Acquisitionin the Service Encounter

Post-ServiceInformation Acquisition


For further information contact:Haisu Zhang

Department of Managerial StudiesUniversity of Illinois at Chicago

601 S. Morgan Street, MC 243, Room 2229Chicago, IL 60607

Phone: 312.996.2680Fax: 312.996.3559



DOES PERSONALIZED RECOMMENDATION SERVICE INCREASECUSTOMER LOYALTY?

Anyuan Shen, State University of New York at New PaltzA. Dwayne Ball, University of Nebraska – Lincoln

SUMMARY

Service firms such as Netflix.com and Amazon.comuse recommendation systems to interact with individualcustomers, learn customer preferences, and recommendpersonalized products and services (Flynn 2006). Whenreturning customers sign in, they get recommendationstailored to their interests the service firms have learned orinferred from previous service interactions. We refer tothis individual marketing practice as personalized recom-mendation service (PRS hereafter).

Service firms believe that PRS will help them fostercustomer loyalty. However, the claimed loyalty effect hasnever been empirically tested. First, does PRS reallyincrease loyalty? The individual marketing literature hasraised serious doubts about the merit of learning relation-ships: customers often do not have stable, well-definedpreferences for marketers to learn and, consequently, mayfail to appreciate customized offers and learning relation-ships (e.g., Simonson 2005). Second, how do differentaspects of PRS differentially influence loyalty, assumingthat there is a loyalty effect? The individual marketingliterature implies that the outcome aspect, i.e., offerstailored to individual preferences, may be the majorcontributor to loyalty (e.g., Pine, Peppers, and Rogers1995). However, customers may respond not only to theoutcome (what is recommended) but also to other aspectsof recommendation services such as the process (how it isrecommended) and the intentionality (inferences of whyit is recommended). Yet, little is known about the relativeimportance of recommendation outcome compared withsuch aspects as process or intentionality of the recommen-dation services.

We theorize that PRS has three dimensions – accu-racy, process value, and benevolence. We define recom-

mendation accuracy as the extent to which a customerperceives that system recommendations closely meet his/her personal needs, interests, or preferences duringextended service interactions. Recommendation processvalue refers to the extent to which a customer perceivesthat the process of interacting with the recommendationsystem is personally desirable over and beyond the out-come benefit of accurate recommendations. Recommen-dation benevolence refers to the extent to which a cus-tomer perceives that recommendations are personalizedto represent the best interests of the customer as opposedto primarily the best interests of the service firm.

We test the validity of the three dimensions, theireffects on and their relative importance in influencingcustomer loyalty by surveying 273 customers who hadused personalized recommendation services. Data analy-sis supported our dimensionality of PRS. Our findings oftheir effects and relative importance were both affirma-tive and surprising. We found that all the dimensions hadnontrivial contributions to loyalty. We also found thatprocess value and benevolence had a comparable orbigger effect than accuracy.

Our research provided preliminary evidence for theloyalty effects of PRS, suggesting that it may be necessaryto reconsider the implications of preference constructionfor individual marketing. Our findings also indicate thatthe typical focus on accuracy may be seriously incompletefor marketers who wish to influence loyalty and othercustomer relationship variables. Other aspects such asprocess and benevolence are not merely “adjuncts” to theaccuracy of PRS; they are every bit as important to designinto the system as is accuracy.

For further information contact:Anyuan Shen

State University of New York at New Paltz1 Hawk Drive

New Paltz, NY 12561Phone: 845.257.2669



PERFORMANCE IMPLICATIONS OF EMOTIONAL VERSUSCOGNITIVE BRAND RELATIONSHIP QUALITY:

AN EMPIRICAL STUDY OF FREQUENTFLIERS IN THE AIRLINE INDUSTRY

Bettina Nyffenegger, University of Bern, SwitzerlandLucia Malaer, University of Bern, Switzerland

Harley Krohmer, University of Bern, Switzerland

SUMMARY

In today’s highly competitive business environmentstrong and long lasting consumer-brand relations play akey role for a brand’s sustainable competitive advantage(Srivastava, Fahey, and Christensen 2001). Thus, re-searchers and practitioners alike are highly interested inthe prediction and control of the maintenance of con-sumer-brand relationships. A focus of interest has been onthe quality of these relationships (so-called brand rela-tionship quality (BRQ); e.g., Fournier 1994, 1998). BRQcan be defined as a “customer-based indicator of thestrength and depth of the person-brand relationship”(Fournier 1994, p. 124). BRQ is perceived by the con-sumer and reflected in his thoughts, feelings, and behav-iors toward a brand (Fournier 1994). Prior research hasusually described and operationalized brand relationshipquality as a higher-order construct with different interre-lated relationship facets such as commitment, intimacy,and love (e.g., Aaker, Fournier, and Brasel 2004; Fournier1998).

We argue that BRQ includes a cognitive and anemotional component. The cognitive component of BRQresults from an evaluative judgment based on cognitivebeliefs and evaluations of the brand and its performance.Emotional BRQ, on the other hand, is reflected in theemotional feelings toward the brand and the personalconnection to the brand. The differentiation betweenemotion and cognition is rooted in the classical distinctionbetween feeling and knowing as two facets of humanexperience (Hilgard 1980) and has been applied in differ-ent academic fields. In attitude research, it has been statedthat attitudes consist of affective and cognitive dimen-sions (e.g., Batra and Ahtola 1990; Breckler and Wiggins1989; Crites, Fabrigar, and Petty 1994). Furthermore,social psychologists generally agree that relationshipsconsist of cognitive and emotional components (e.g.,Berscheid and Peplau 2002; Gabriel and Gardner 2004).Finally, in studies on brand relationships (e.g., Aaker,Fournier, and Brasel 2004; Fournier 1998; Park, Kim, andKim 2002) marketing researchers have already identifieddimensions of BRQ that are rather cognitive (such assatisfaction) and dimensions that are rather emotional(such as love and passion).

The explicit distinction between emotional and cog-nitive components of BRQ is of high relevance since priorresearch in psychology has argued that affect and cogni-tion/evaluation have distinct influences on human behav-ior (Millar and Tesser 1986; Wilson and Dunn 1986).Consequently, we also expect the emotional and cognitivecomponents of BRQ to have different consequences onconsumer behavior and brand performance. To the best ofour knowledge, the implications of cognitive versus emo-tional components of BRQ on brand performance havenot yet been empirically examined.

In our study, we address this research opportunity.First, we conceptually discuss and empirically examinethe emotional versus cognitive components of BRQ.Second, we analyze the brand performance implicationsof these components among real customers. Specifically,our objective is to examine the effects of emotional versuscognitive BRQ on word-of-mouth communication, con-sumers’ willingness to pay a price premium, consider-ation set size, share of wallet, and revenue per customer.Third, we investigate whether these performance out-comes vary across types of consumers by analyzing theeffect of consumer’s product involvement as a moderatorof the relationship between (cognitive and emotional)BRQ and brand performance.

Results and Implications

We tested the effects of emotional versus cognitivebrand relationship quality in the context of an airline’sfrequent flyer program. Based on a sample of 631 realcustomers of a large European airline company, the empiri-cal study shows that BRQ has different implications forbrand performance – depending on which component ofBRQ is taken into consideration (cognitive versus emo-tional BRQ) and depending on which consumers we lookat (consumers with low product involvement versus con-sumers with high product involvement). Our findingsindicate that emotional BRQ has a stronger overall impacton brand performance than cognitive BRQ. While theeffect of emotional BRQ on brand performance becomeseven stronger among low involved consumers, cognitiveBRQ becomes somewhat more relevant for brand perfor-mance among highly involved consumers.


Due to its impact on brand performance, BRQ hasimportant implications for brand management. In analyz-ing real customers of a company and their actual turnover,we were able to show that BRQ really pays off, also interms of economic performance implications. Overall, theresults show that emotional BRQ is a stronger predictor ofconsumers’ loyalty behavior (in terms of a reduced con-sideration of competitive brands, a higher willingness topay a price premium for the respective brand, and finallya higher revenue) than cognitive brand relationship qual-ity. Cognitive BRQ, however, was found to stronglyinfluence the consumers’ word-of-mouth. Thus, while theemotional relationship quality mainly increases the loy-alty behavior of existing customers, cognitive BRQ helps

to attract new customers via positive word-of-mouthcommunication of existing customers. Both the retentionof existing customers and the attraction of new customersare crucial drivers for the sustainable future of a brand(e.g., Bruhn 2003). Marketing managers should, there-fore, try to positively influence both emotional and cogni-tive BRQ of their customers. However, especially amonglow involved consumers, it may pay off to place moreemphasis on the emotional component of a consumer-brand relationship, which supports the call for emotionalbranding as a central corner stone of differentiation andsustainable competitive advantage (e.g., Atkin 2004).References are available upon request.

For further information contact:Bettina Nyffenegger

Institute of Marketing and ManagementUniversity of Bern

Engehaldenstrasse 43012 Bern

SwitzerlandPhone: 0041.31.631.4547

Fax: 0041.31.631.8032E-Mail: [email protected]


TOO LATE, OR JUST LATER THAN ANNOUNCED? EFFECTS ONBRAND TRUST BY DELAYING PRODUCT INTRODUCTIONS

IN COMPETITIVE SITUATIONS

Steffen Herm, Technische Universität Berlin, GermanyJana Möller, Freie Universität Berlin, Germany

SUMMARY

Announcements concerning future products fosterearly awareness and facilitate a quick diffusion of aninnovation. Most of the companies in computer hardware,software, and telecommunications announce their newproducts prior to launch. However, approximately 70percent of the announcing firms encounter some delay inintroducing their preannounced products (Wu et al. 2004).On the one hand, it is difficult to keep an exact schedulefor a launch, especially for innovative products that arecomplex and require intense development efforts. On theother hand, marketing managers induce some launchdelays on purpose in order to gain competitive advan-tages. An early preannouncement increases the willing-ness to wait for a new product rather than purchaseexisting offerings, and a delay may raise consumer atten-tion. However, such a strategy could cause negative word-of-mouth, negative publicity, and a loss in trust. Conse-quently, postponements could violate valuable consumer-brand relationships. Managers should consider positiveand negative outcomes of launch delays carefully in orderto utilize the advantages of preannouncements and toavoid risks for their brand.

Scholars assume that trust (reputation and credibil-ity) concerning a firm is likely to decline if companiesintroduce new products later than announced (Bayus et al.2001; Eliashberg and Robertson 1988; Kohli 1999; Sorescuet al. 2007). However, research has not yet providedempirical evidence. In the same vein, what is the impact ofdelays on trust toward more reliable competitors in themarket? In two experiments we examine effects of launchdelays on brand trust and identify risks managers mightface in the case of postponing a preannounced launch. Ourfindings contribute to resolve the question of whether ornot to preannounce a new product prior to its launch.

According to the commitment-trust theory of rela-tionship marketing, trust leads to commitment and bothare well-accepted key mediators to relational success(Morgan and Hunt 1994). However, one can also think ofa moderating effect of the more stable brand commitmenton the development of brand trust. Building on the ideathat firms can violate customers’ trust by postponing thelaunch of a new product, this paper examines if highlycommitted consumers are likely to react to violations ofbrand trust in a different manner than less committedconsumers. By analyzing the feedback effect of commit-ment on the development of brand trust, this researchextents the rich stream of literature examining the effect ofincreased trust on commitment.

Method and Results

We conducted two experiments: (A) a laboratoryexperiment with a fictitious product launch of a singlebrand and (B) a field-experiment using real productlaunches of two competing brands and their existingcustomers. The overall experimental design for both stud-ies was a 2 (delay: yes/no) x 2 (commitment: high/low)factorial design. The two studies differed in the productcategory. Both include repeated measures of brand trust asthe dependent variable. While we measured brand trusttoward the same brand in experiment (A) we measurebrand trust toward competing brands in experiment (B).

Findings of both experiments reveal that brand trustis likely to decline as a result of a delayed productintroduction. Further, we find a strong interaction be-tween delay and commitment: A high degree of brandcommitment is less likely to protect the violation of brandtrust. In particular, highly committed consumers are morelikely to feel betrayed by their devoted brand. The declin-ing trust in the committed brand is associated by risingtrust in a competitor’s brand. References are availableupon request.


For further information contact:Steffen Herm

Department of MarketingTechnische Universität Berlin

Wilmersdorfer Str. 14810585 Berlin




CONSUMER ATTITUDES TOWARD PRIVATE BRANDSAND NATIONAL BRANDS

Ronald E. Goldsmith, Florida State University, TallahasseeLeisa Reinecke Flynn, University of Southern Mississippi, Hattiesburg

E. Craig Stacey, New York University

SUMMARY

Brands today face consumers who are increasinglyretrenching as a result of the recession plus growingcompetition from global companies. In addition, market-ers of branded goods, especially consumer packagedgoods, find that competition from private label goods(store brands, own labels) represents a growing threat totheir financial health (Levy 2009; Retail Forward 2009).Although sales of private labels have grown overall fortwo decades in the U.S. (Retail Forward 2009), theirrelative gains versus national brands fluctuate with theoverall state of the economy and with consumers’ feelingsof affluence or anxiety reflected in price sensitivity (Levy2009). Sellers of private labels pursue strategies to expandsales by improving quality, lengthening product lines, andstrengthening their marketing, while national brandscounter by investing in brand image, innovating newofferings, and stressing value for quality

The present study was motivated by the questions,“what do private label buyers think and feel about thebrands they buy” and “how do their evaluations comparewith those of national brand buyers?” The purpose was toexpand our understanding of private label buyers bycomparing their attitudes with those of national brandbuyers.

A review of the relevant literature resulted in thefollowing hypotheses:

H1: Buyers of private label peanut butter brands feel thatbrands are less important or less of a concern than donational brand buyers, thus they are less brand con-sciousness.

H2: Buyers of national peanut butter brands are less likelyto say that private label peanut butter brands arerelevant or fit their values than are private labelbuyers.

H3: Private label buyers are more likely than are nationalbrand buyers to feel that private label peanut butterperforms equivalently to national peanut butter brands.

H4: Private label buyers of peanut butter rate the perfor-mance of private label peanut butter higher than donational brand buyers.

The data were collected via telephone interviewswith members of a large advertising research firm’s onlinepanel, matching U.S. demographics and limited to con-sumers 18 years of age and older. Although 21 packagedgoods categories were included in the study we chosepeanut butter, a commonly purchased category featuringsignificant private label competition. The original sampleconsisted of 5,389 panel members with each respondingto a subset of the 21 categories. The sample size for peanutbutter was 240, and the five brands included were Jif,Skippy, Peter Pan, Smuckers, and “Store’s Own Brand ofPeanut Butter.”

There were some demographic differences betweennational and private brand buyers. Men were more likelyto be national brand buyers (95%) than women (82%), andnational brand households tended to be smaller and tohave higher incomes than did private label households,consistent with results reported by Hoch (1996) and byHowell (2004).

The hypotheses tested whether the private brandbuyers would (1) care less overall about brands of peanutbutter, (2) feel that private labels performed as well asnational brands, and (3) manifest more positive attitudestoward their private labels. All hypotheses were sup-ported.

The results suggest that private label buyers aremotivated to buy and are sustained in their purchase ofprivate labels by more than just price. These buyers alsoseem to develop positive evaluations of these brands thatare similar to those of national brand buyers. It is clear thatthey feel the quality (performance) of their brands isequivalent to that of national brands, and it appears thatthey also develop emotional attachments to these brandssimilar to those of national brand buyers. It is arguable thatin the value-equals-benefits-minus-cost tradeoff that con-sumers make when they chose brands, private label buy-ers easily see equivalent benefits and lower prices yield-ing equivalent or even value superior to that of nationalbrands. This is a topic worthy of further study.


REFERENCES

Hoch, Stephen J. (1996), “How Should National BrandsThink about Private Labels?” Sloan ManagementReview, 37 (2), 89–102.

Howell, Debbie (2004), “National Brands Feel the Pres-

sure,” DSN Retailing Today, 43 (20), 22, 24–27, 57.Levy, Piet (2009), “Marketing in Aisle 5,” Marketing

News, 43 (9), 12–15.Retail, Forward (2009), “Private Brand Reach Tipping

Point,” report downloaded at [http://rfkb.retailforward.com/default.aspx].

For further information contact:Ronald E. Goldsmith

The Richard M. Baker Professor of MarketingFlorida State University

821 Academic WayP.O. Box 3061110

Tallahassee, FL 32306–1110Phone: 850.644.4091



A PROPOSED PROCEDURE FOR CONSTRUCT DEFINITIONIN MARKETING

David A. Gilliam, Oklahoma State University, StillwaterKevin E. Voss, Oklahoma State University, Stillwater

SUMMARY

Construct definition plays a critical role in the devel-opment of sound, generalizable theory in marketingresearch. Researchers should spend the effort necessary toproperly define constructs prior to advancing to measure-ment development (Churchill 1979). Doing so will put theresearch program on a firm foundation. MacKenzie (2003)says that poor construct definition is common in manu-scripts submitted to marketing journals. Summers (2001)adds that it is a major reason that manuscripts are rejected.Jarvis, MacKenzie, and Podsakoff (2003) state that poordefinitions can cause measurement model misspecification.

Scientists have long struggled with the problem ofdefinition and philosophers of science have expendedconsiderable effort in formulating guidelines for creatingdefinitions (Achinstein 1968). In marketing, researchersoften face the added difficulty of defining constructs.Latent constructs do not have a physical manifestation asthey represent ideas in the mind of the theorist (Edwardsand Bagozzi 2000). Constructs allow marketing and othersocial scientists to represent and explore unobservablephenomenon that would otherwise remain inaccessible toresearch.

The abstract nature of constructs presents seriousproblems when defining them. One problem involvesfinding terms to use in the definitions that do not requiredefinitions themselves. So called primal terms can stopthis infinite regress, but no commonly accepted group ofprimal terms exists (Hunt 1991; Hempel 1952). Somehave suggested the use of a formal language making useof only primal terms and terms defined by primal terms,but marketing research does not currently have such aformal language (Teas and Palan 1997). Others havesuggested the use of very narrow definitions for con-structs that include all of the components of the constructand the specific rater entity providing the measures; whilethis extreme specificity assists with proper definition, theresulting loss of generalizability limits cross study com-parison (Rossiter 2002).

Construct Definition

Researchers use nominal definitions for latent con-structs; these follow the rule of replacement whereby theconstruct and its definition can replace one another in a

sentence (Hempel 1952). There are acknowledgedrequirements for proper nominal definition. The defini-tion must attend to the constructs denotation, intension,extension, and connotation (Caws 1965; Teas and Palan1997). Multiple constructs for the same phenomena causeproblems with comparison across studies and accretion ofknowledge in an area and so conservation in creatingoverlapping constructs is needed (Wacker 2008). Also,defining constructs via antecedents or outcomes presentsa problem as the resulting theory can be true by definitionand defining constructs merely by examples yields defini-tions that are typically not adequately delimited (Sum-mers 2001; MacKenzie 2003).

Proposed Procedure

The following describes a simple, programmatic andbroadly applicable procedure for construct definition orredefinition.

1. Write the preliminary definition. This allows maxi-mum flexibility to preserve creativity prior to furtherchecking of the salient literature. Attention to theconstruct’s level of abstraction, what is and is notincluded in the construct’s domain, and taking care toavoid defining the construct in terms of relationshipswith other constructs dominate this step.

2. Build the nomological network. Cronbach and Meehl(1955) argue that the construct’s placement withinthe nomological network gives the construct animplicit definition. In this step we attend to buildingand evaluating this network and assess the stability,fertility, and changeability of the construct (Davis1985).

3. Consult the literature and refine the construct. Herean assessment of the contribution to the explanatoryand predictive power is made while searching forsimilar constructs in the extant literature. If the con-struct shows promise, development should continue,if not, the researcher should employ an existingconstruct.

4. Refine the definition. Procedures to fight ambiguityand vagueness include the use of specific nouns andadjectives, elimination of unnecessary words andcare in the use of conjunctions.


5. Expert judging process. At least two outside expertswithin the domain are asked to judge the constructand its contribution to the nomological network.

6. Adjust definition and iterate. The adjustments sug-gested by the earlier steps are integrated and a deci-

sion is made as to whether the researcher shoulditerate steps three to six, use an existing construct orenter the construct into the next step of scale develop-ment. References are available upon request.

For further information contact:David A. Gilliam

Oklahoma State University405C Business Building

Stillwater, OK 74078Phone: 405.744.8624



A REVIEW OF META-ANALYSIS IN MARKETING

Arne Floh, WU Wien, AustriaJagdish Sheth, Emory University, Atlanta

SUMMARY

In its 100-year history, marketing scholars haveproduced a tremendous outcome of their research activities.A keyword search for marketing in electronic databasesyields over 66,000 articles published in more than 100scholarly journals. Given the enormous amount ofmarketing knowledge available, it is very difficult for anindividual to capture all aspects of previous work. It seemsreasonable when marketers and scholars argue that thediscipline has matured to the point where it is indispensableto summarize the accumulated knowledge systematically.

Although research synthesis can be done in manyways, meta-analysis is the most used instrument in sciencein general, as well as marketing. An elaborate searchstrategy conducted by the authors yielded a set of 163meta-analyses in the field of marketing. Online databaseslike ABI/INFORM, EBSCO, and Science Direct wereused for a computerized bibliographic search. An issue-by-issue search of several marketing journals was con-ducted while the Web was searched for working papers,books, abstracts and conference proceedings. Althoughthe amount of conducted studies lags compared to otherdisciplines, the absolute number of meta-analyses in thefield of marketing is substantial.

The purpose of this study is to critically review theuse of meta-analysis in marketing. The need for researchsynthesis and the high popularity of the technique call fora comprehensive examination of all previous conductedmeta-analyses in the field of marketing. The authors usedBuzzell’s definition of a scientific inquiry for summariz-ing and evaluating previous studies. According to hisframework, two criteria are essential for assessing the useof meta-analysis in marketing: the nature and scope andthe methodological rigor of previous studies. Addition-ally, these criteria enable the authors to make statementsabout the discipline of marketing and its scientific status.

The identified studies are highly skewed in terms ofpublication source, title, and year. Currently, the JMR haspublished the most meta-analyses (14.8% of all studies).The AMA Conference Proceedings have published 10papers on research synthesis while the next cohort ofstudies are almost evenly distributed among six marketingpublications: JBR, JM, IJRM, JPPM, ML, and MS. Stud-ies in the six journals account for approximately 23percent of all meta-analyses. The remaining meta-analyses(approximately 45%) are spread over 42 different journals.

The publication date of the studies shows anunderlying trend. While 33 research syntheses wereconducted in the seventies and eighties, the number ofstudies almost doubled in the nineties to 59 studies.Between 2000 and 2008, the number of meta-analysessurpassed the prior two decades: 70 meta-analyses werepublished in the last eight years. In short, quantitativeresearch synthesis in general and meta-analysis in particularhas grown in importance as the discipline matures.

Turning to the substantive findings of this researchstudy, note that the total number of 163 meta-analyses isgrouped into 10 categories. The most dominant sub areasare promotion (which includes advertising, sales manage-ment and direct marketing (41; 25.2%)) and researchmethodology (which includes meta-analyses about therigor of the technique itself and studies about responserate and reliability (36; 22.1%)). Categories such as prod-uct management (18; 11.0%), international marketing(16; 9.8%), marketing strategy (13; 8.0%), pricing (12;7.4%), and consumer behavior (11; 6.7 %) are moderatelyresearched areas. Only partially covered by meta-analysisare channel management (6; 3.7%), and business-to-business marketing (5; 3.1%). Five papers, which couldnot be grouped logically with one of the nine abovementioned categories, were categorized under “miscella-neous.” They account for 3.1 percent of all studies.

In addition to the substantive characteristics, themethodological rigor of previously published meta-analyses was also analyzed. For this purpose, all paperswere reviewed for their analytic procedures:

Coding is a crucial preliminary step in conducting ameta-analysis. Therefore, it is highly recommended thatmultiple coders are used and the categories are thoroughlydefined in advance. Coding reliability is a measure ofconsistency between two or more coders. Categoriesdiffer substantially in terms of reported coding reliabilities.Relative frequencies range from 62.5 (marketing strat-egy) to 0 (pricing) percent.

Adjustments due to effect size are even more impor-tant in conducting a meta-analysis. Two approaches areapplied frequently in marketing meta-studies. Sampleweighting allows authors of meta-analyses to compareempirical studies with small and large samplesizes. Additionally, corrections for attenuation due tomeasurement errors are recommended to calculate anunbiased combined effect size. However, the latter adjust-


ment is limited to empirical studies which report scalemeasurement reliabilities. Unfortunately, this is not oftenthe case and, thus the proportion of reliability adjustmentslags considerably behind the correction of sample-weight-ing in most of the categories, e.g., international marketing,channel management or business-to-business marketing.Noticeable exceptions are studies on marketing strategyand product management. All of them completely followthe attenuation recommendations.

The meta-analysis technique has been criticizedbecause it is subject to publication bias. Critics claim thatresearch synthesis of primary studies is biased due to thefact that non-significant studies are less likely to bepublished. Most authors address this issue by calculatingthe number of insignificant studies that are necessary toreverse the findings. Again, methodological accuratenessdiffers between the sub disciplines of marketing. Althoughsome sub areas show low frequencies (e.g., channel man-agement, consumer behavior), the calculation of fail-safeN can be regarded as a widely accepted approach includedin most of the recent meta-analyses.

Future recommendations include the increase of meta-analyses in all sub areas of marketing. The number of

studies done in other disciplines (e.g., medicine 7,856,psychology 3,558) is quite large compared to marketing.In particular, the authors would like to see more meta-analyses on channel management, e-commerce, and orga-nizational marketing. From a methodological perspec-tive, the rigor of conducting a meta-analysis has beenimproved a lot. However, some studies still do not followstandard procedures nor do they report relevant param-eters of the study. The latter aspect holds true for primarystudies as well. Some corrections of the combined effectsize, e.g., reliability adjustments, can only be calculated ifthe data basis covers all relevant information.

Overall, meta-analysis is a valid method used tosystematically organize accumulated knowledge of a dis-cipline. It has been frequently used in disciplines such asmedicine, education, and psychology. This study showsthat meta-analysis plays an important role in marketing aswell. A further increase in the use of meta-analysis willensure that marketing is viewed as a science and helpmanagers to keep track what has been discovered to date,and therefore, make better marketing decisions.

For further information contact:Arne Floh

Institute for Marketing-ManagementWU Wien

Augasse 2-61090 Wien

AustriaPhone: +43.1.31336.4626Fax: +43.1.31336.90.626


Jagdish ShethCharles H. Kellstadt Professor of Marketing

Goizueta Business SchoolEmory University

1300 Clifton Road N.E.Atlanta, GA 30322

Phone: 404.727.7603Fax: 404.727.6313



CUSTOMER-FOCUSED LEARNING IN GOVERNMENTORGANIZATIONS

Elena Kiryanova Bernard, University of Portland

SUMMARY

Organizations are social systems, and by their verynature they are conducive to learning (DiBella, Nevis, andGould 1996). Hence, organizational learning is a naturalphenomenon that occurs in all organizations regardless oftheir structural and demographic characteristics. Govern-ment organizations have been under enormous pressure toadopt market behaviors and become more flexible andcustomer-focused (Paarlberg 2007). This transformation,however, can be challenging for government organiza-tions known for their hierarchical culture and bureau-cratic control systems, which dramatically inhibit marketinformation processing (Batac and Carassus 2009;Moorman 1995; Vinyard 2004). This paper proposes amodel of customer-focused learning in government orga-nizations. Drawing on organizational learning literature,customer focused learning is conceptualized as a functionof top management customer orientation, customer feed-back, and individual employee learning orientation. Themodel posits a series of relationships among customer-focused learning, employees’ job perceptions (job ambi-guity and self-efficacy) and job outcomes (job satisfactionand organizational citizenship behaviors).

Organizational learning refers to the development ofnew knowledge or insights that could potentially producebehavior change leading to improved performance (Slaterand Narver 1995). The roots of organizational learningare grounded in individual employees’ interactions withthe environment (Argyris and Schön 1978). Employees’interpretations of environmental responses provide theinput for the creation of organizational memory thatreflects shared beliefs, assumptions, and norms. Thisorganizational memory evolves independently ofemployees’ tenure with the organization and ensures thepreservation of knowledge to be used by individuals otherthan its progenitor (Sinkula 1994).

Consistent with this cyclical view of organizationallearning, customer-focused learning refers to a process ofcontinuously improving actions through better knowl-edge and understanding of customer needs (Cummingsand Worley 1997; Fiol and Lyles 1985). The effectivenessof customer-focused learning depends on organizationalvalues, customer feedback and learning orientation ofemployees. Organizational values influence the propen-sity of the organization to create and use knowledge(Moorman 1995; Sinkula, Baker and Noordewier 1997;

Slater and Narver 1995) while unrestricted access tocustomer feedback facilitates information processingbehaviors that transform information into knowledge(Morgan 2004). Yet, even with customer-focused organi-zational values and ready access to customer feedbacksome employees may be averse to changing their estab-lished routines. Hence, learning orientation of individualemployees is an important enabling factor that is likely toinfluence the extent and the rate of organizational cus-tomer-focused learning.

Customer-focused learning has important implica-tions for employees’ job perceptions and job outcomes.First, management plays a critical role in reducing roleambiguities that inevitably arise as employees are encour-aged to become more customer focused (Singh and Rhoads1991). In turn, reduced role ambiguities enhance employ-ees’ self-efficacy, which is an important predictor of joboutcomes such as job satisfaction and OCB.

Similarly, customer feedback allows employees toanalyze task requirements, make attributions about theresults of the service delivery experience, and assesspersonal and situational resources and constraints. In turn,these evaluations result in updated self-efficacy beliefsthat influence employees’ service delivery behaviors (Gist1987; Gist and Mitchell 1992).

Finally, employee learning orientation is likely tohave a strong positive influence on employees’ role per-ceptions and self-efficacy beliefs (Button, Mathieu, andZajac 1996; Ford et al. 1998; Kozlowski et al. 2001;Phillips and Gully 1997). Learning-oriented employeeswelcome challenging tasks regardless of their confidencein having the adequate skills and abilities to producesuccessful results. They believe that necessary abilitiescan be acquired as a result of learning and, therefore, theypersist despite difficulties and a prospect of failure (Amesand Archer 1988; Elliott and Dweck 1988). Hence,employees with high levels of learning orientation arelikely to hold broader role perceptions that allow them tostep outside the boundaries of their job description inorder to acquire new skills and experiences. In addition,learning-oriented employees do not interpret negativefeedback as a failure message (Elliott and Dweck 1988;Hong et al. 1999; Sujan, Weitz, and Kumar 1994). In-stead, they consider it a signal that the strategy for a giventask requires changes (Anderson and Jennings 1980;Porter and Tansky 1996).


The model was tested with structural equation mod-eling. Cross-sectional survey data was collected fromemployees of a state service organization participating ina workplace environment study. The results of the analy-sis provided strong support for the proposed model ofcustomer-focused learning. One interesting finding thatmerits further investigation deals with the effect of roleambiguity on OCBs. In this paper, role ambiguity had anegative effect on OCBs, suggesting that the more uncer-tain an employee feels about his role expectations, the lesslikely he or she is to engage in extra-role behaviors.Another interesting finding refers to positive effects of

self-efficacy on civic virtue and conscientiousness OCBs.Although not hypothesized, these effects are actuallyquite intuitive. People who hold strong beliefs about theircompetence and personal effectiveness are likely to takeactive part in organizational governance because it offersthem an opportunity to demonstrate leadership, which isboth intrinsically and extrinsically rewarding. Similarly,they are likely to be more conscientious in regard to theirin-role and extra-role behaviors because it reinforces theirbelief in personal effectiveness. References are availableupon request.

For further information contact:Elena K. Bernard

Pamplin School of Business AdministrationUniversity of Portland

5000 N. Willamette Blvd.Portland, OR 97203–5798



ENHANCING CUSTOMER PURCHASE LIKELIHOOD THROUGHMARKET DRIVING STRATEGIES

Goran Vlasic, Bocconi University, ItalyAjay Kohli, Georgia Institute of Technology, Atlanta

Gabriele Troilo, Bocconi University, Italy

SUMMARY

Market driving is an important strategy for achievingcompetitive advantage, complementing existing predomi-nant view that firms should be driven by markets, i.e.,understanding and responding to preferences and behav-iors of market players within a given market structure(Jaworski, Kohli, and Sahay 2000; Carpenter andNakamoto 1994). Jaworski et al. (2000, p. 45) definemarket driving in a broad manner as “influencing thestructure of the market and/or the behaviors(s) of marketplayers in a direction that enhances the competitive posi-tion of the business.” On the other hand, Carpenter andNakamoto (1994, p. 572) focus on product attributes anddefine market driving as “shaping consumer tastes.” Ourwork contributes to current literature on market driving inthree main ways. First, we define market driving from theperspective of customer decision-making process thusdistinguishing between three different levels at whichmarket driving can be done: need, category, and attributelevel. Second, this contribution distinguishes betweenmarket driving efforts, i.e., orientation of an organizationto drive markets, and market driving outcomes, i.e., theactual change induced in customer decision-making pro-cess which makes focal product perceived as providinggreater utility. As an outcome, market driving is analyzedin much the same way as innovation (with the componentsof change and its meaningfulness, as in, e.g., Chandy andTellis 1998), and thus can even be referred to as marketinnovation. The third contribution is that we relate com-ponents of market driving outcomes (i.e., the actual change)to customer purchase likelihood increases, which is closelyrelated to revenue increase (DeKinder and Kohli 2008). Inthat sense, we provide propositions about possible differ-ential effectiveness of market driving components underspecific consumer-, product-, and environment-relatedcontingencies.

In economics and predominantly in business, con-sumer preferences generally are taken as relatively fixedand exogenous, although literature shows growing evi-dence that this is not the case (Carpenter and Nakamoto1994; Schmalensee 1982; Tversky, Shmuel, and Slovic1988). Seeing customer preferences as manageable, mar-ket driving was conceptualized as a way to obtain com-petitive advantage by altering customer preference struc-ture to make it “favorably asymmetric” (Jaworski, Kohli,

and Sahay 2000; Carpenter and Nakamoto 1989). Litera-ture streams that have contributed to the market drivingdiscussion are primarily environmental management lit-erature (e.g., Clark, Varadarajan, and Pride 1994; Zeithamland Zeithaml 1984) and the customer education literature(e.g., Staelin 1978).

Following the stages of generic decision makingprocess (Kotler and Armstrong 2009) and taking intoaccount research on categorization (Viswanathan andChilders 1999) we see the consumer’s decision-makingprocess as going through three distinct phases: beingmotivated to satisfy the need, searching and evaluatingalternative categories, and selecting product alternativeshaving favorable attributes. So, product choice is theoutcome of a problem (need or want) that the decisionmaker must solve (Amir and Levav 2008; Simon 1957). Inthat sense, the customer is assumed to follow his/herutility maximization when making the choice, but allow-ing for existence of different utility functions acrosscustomers. Thus, customer utility from the focal productcan be expressed as1

.

Need importance change (as one aspect of marketdriving outcomes) is the extent to which a firm increasesconsumers’ motivation to satisfy needs that can beaddressed by the firm’s products. This also encompassesthe possibility of introducing new needs to the consumermind-set, i.e., increasing the motivation from zero to acertain positive level. Need driving also implies that thefirm can reduce the relative importance of needs for whichits products are not applicable. Category importance changeis defined as the extent to which a firm increases customer’sperceived opportunity to satisfy a particular need better byusing the category to which firm’s product belongs ascompared to using any alternative category. The thirdaspect is attribute driving leading to attribute importancechange which is defined as the extent to which firmincreases the relative weight that a customer gives toattributes on which the firm’s product outperforms itswithin-category competitors. In that sense, attribute driv-ing changes the attributes that customer considers crucialwithin the focal category when assessing product’s abilityto satisfy her need.


The idea expressed in the equations below is thatchanges in purchase likelihood increase differ dependingnot only on the actual utility increase, but also on theexpected customer responsiveness to utility change (basedon ex-ante attitudes). Building on this, we propose thatconsumer-related variables (brand loyalty, categoryusage, need recognition), product-related variables (domi-nance, life-cycle stage, essentiality) and environmentalvariables (within-category competitive intensity, cross-

category competitive intensity) imply assumptions aboutex-ante values of focal attribute, category and need impor-tance. Consequently, these variables imply different effec-tiveness in terms of translating driving outcomes intoactual purchase likelihood (i.e., behavior)..

For managers this research indicates how to approachmarket driving. In that sense, different aspects should bedriven based on the target customers, focal products andenvironmental characteristics. In that sense, needs impor-tance change is expected to be the most effective in caseof customers who do not represent “traditional” marketsegments, and for non-essential products which are in thelate stage of their life cycle. Category importance changeis the most effective for enhancing purchase likelihood ofloyal customers for dominant products early in their lifecycle facing low within-category and high cross-categorycompetition. Attribute driving is the most appropriate forcustomers who recognize the category but are not loyal tothe focal brand; for products that are non-dominant and intheir intermediate life cycle stage facing low cross-cat-egory and high within-category competition.

where, e.g. ,

and äi represents a small increase

ENDNOTE

1 Ufocal

– is the utility customer perceives from the focalproduct, with ; N

i – is the motivation to

satisfy the i-th need, where , N* is set offocal needs, and with M being thetotal number of needs (which can even be consideredto be infinity with some needs having weight of 0);C

ij – is the opportunity to satisfy i-th need using j-th

category, where , C* is set of focal catego-

ries, and with P being the total numberof categories satisfying the focal need i; A

ijk – is the

importance of attribute k for choosing products withinj-th category used for satisfying i-th need, where

, A* is set of focal attributes, and

with Q being the total number ofattributes characterizing products within the focalcategory j which addresses the focal need I.

References are available upon request.

For further information contact:Goran Vlasic

Universita Commerciale Luigi BocconiVia Roentgen 1

20136 MilanItaly

Phone: +385.98.271.512Fax: +385.1.3638.000



THE IMPACT OF UNPROFITABLE CUSTOMER ABANDONMENT ONCURRENT CUSTOMERS’ EXIT, VOICE, AND LOYALTY

INTENTIONS: AN EMPIRICAL ANALYSIS

Michael Haenlein, ESCP Europe, FranceAndreas M. Kaplan, ESCP Europe, France

SUMMARY

In recent years, the topic of (unprofitable) customerabandonment has received increasing interest in aca-demic research, fueled by several studies showing thatsuch customers can represent a significant share of acompany’s client base (e.g., Haenlein et al. 2007; Nirajet al. 2001). Haenlein et al. (2006) showed, for example,that the value of the real option of abandoning unprofit-able customers can be substantial and needs to be consid-ered when calculating customer lifetime value to avoidbiased results. Unsurprisingly, the idea of “firing” unprof-itable customers also received attention in the businesspress (e.g., Haenlein and Kaplan 2009; Mittal et al. 2008).Adopting the (unprofitable) customer’s perspective, itseems likely that such abandonment will be perceived asnegative and create a certain level of dissatisfaction whichmay result in negative publicity and potentially lead to theinvoluntary loss of other (profitable) clients the companywould like to retain. However, until now, the processesunderlying such mechanisms have not been formallyinvestigated. Our manuscript intends to provide a contri-bution in this area. Based on a survey conducted among385 customers, we investigate the reactions of the aban-doning firm’s current customers toward unprofitable cus-tomer abandonment and the influence that the strength ofthe relationship between the abandoned customer and thecurrent customer under investigation has on these reac-tions. In doing so, our work represents a first step towardestimating the likely indirect cost associated with aban-doning unprofitable customers.

The conceptual framework, which builds the founda-tion for our empirical study, combines Hirschman’s (1970)exit-voice-loyalty theory with literature in the area ofsocial influence, specifically the concept of tie strength asintroduced by Granovetter (1973). Within this generalconceptual framework, we use three constructs to charac-terize the reaction current customers can show towardunprofitable customer abandonment. These are the cur-

rent customer’s exit, voice and loyalty intention. In addi-tion to these three outcome variables we also investigatethe role of relationship characteristics (i.e., overall satis-faction) and structural constraints (i.e., alternative attrac-tiveness, switching cost) in driving these reactions. Final-ly we analyze the moderating impact of tie strength onthese processes.

Our empirical analysis results in the following fourfindings: First, none of the three behavioral intentions forexit, voice, and loyalty differ significantly across the fivelevels of tie strength manipulated in our study. Thisprovides an indication that the reactions toward unprofit-able customer abandonment may not be significantlyaffected by the strength of relationship between the aban-doned customer and the current customer under investiga-tion. Second, the mean absolute intention score for loyaltyis significantly lower than those for exit and voice. Thisimplies that current customers are least likely to respondto unprofitable customer abandonment passively by re-maining attached to the organization and waiting untilsomeone else acts to improve matters, but rather tend toreact actively either by leaving the abandoning firm or byraising their voice against unprofitable customer aban-donment. Third, none of the paths between structuralconstraints and exit/ voice/ loyalty intentions is signifi-cant which implies that satisfaction and alternative attrac-tiveness do not significantly influence the choice betweenexit, voice, and loyalty. Fourth, all remaining structuralpaths in our model are significant and show signs alongwith our expectations, except for the relationship betweenalternative attractiveness and loyalty intention where weobserve a positive instead of a negative correlation. Com-bined, this provides full support for the assumed relation-ship between alternative attractiveness and exit/ voiceintention as well as the stage-like sequence betweenloyalty, voice, and exit intentions and partial support forthe link between alternative attractiveness and loyaltyintention.

REFERENCES

Granovetter, Mark S. (1973), “The Strength of WeakTies,” The American Journal of Sociology, 78 (6),

1360–80.Haenlein, Michael, Andreas M. Kaplan, and Detlef Schoder

(2006), “Valuing the Real Option of AbandoningUnprofitable Customers When Calculating Customer


Lifetime Value,” Journal of Marketing, 70 (3), 5–20.____________, ____________, and Anemone J. Beeser

(2007), “A Model to Determine Customer LifetimeValue in a Retail Banking Context,” European Man-agement Journal, 25 (3), 221–34.

____________ and ____________ (2009), “Unprofit-able Customers and Their Management,” BusinessHorizons, 52 (1), 89–97.

Hirschman, Albert O. (1970), Exit, Voice and Loyalty:

Responses to Decline in Firms, Organizations andStates. Cambridge: Harvard University Press.

Mittal, Vikas, Matthew Sarkees, and Feisal Murshed(2008), “The Right Way to Manage UnprofitableCustomers,” Harvard Business Review, 86 (4), 95–102.

Niraj, Rakesh, Mahendra Gupta, and ChakravarthiNarasimhan (2001), “Customer Profitability in aSupply Chain,” Journal of Marketing, 65 (3), 1–16.

For further information contact:Michael Haenlein

Department of MarketingESCP Europe

79, Avenue de la République75011 Paris

FrancePhone: +33.1.49.23.26.02

Fax: +33.1.49.23.22.48E-Mail: [email protected]


WHY DO USERS OF SECOND LIFE BUY VIRTUAL PRODUCTS?ANTECEDENTS OF BRAND PURCHASING INTENTIONS AND

BEHAVIORS IN THE 3D VIRTUAL ENVIRONMENT

Jason Gabisch, University of Massachusetts, Amherst

ABSTRACT

Online virtual environments have been identified asemerging marketing channels, whereby consumers canlearn about brands through the trial or purchasing ofvirtual products. Using survey data from a panel of regis-tered users (N = 209) of the popular virtual world SecondLife, this research extends the theory of reasoned actionmodel for examining drivers of purchase intentions andbehaviors in the virtual world. The study results suggestthat social influences may vary depending on the market-ing channel, virtual product trial is useful for predictingpurchase intentions and behavior, and consumers’ per-ceived behavioral consistency across channels affects theintention-behavior relationship.

INTRODUCTION

A number of technologies have developed that allowconsumers to search for information, interact with brands,communicate with other consumers, try out products, andbuy real and digital products over the internet. Someexamples of these types of technologies include onlineproduct trial (e.g., test driving automobiles, virtuallytrying on clothes), and the use of avatars (“animatedrepresentations of the user”; see Holzwarth et al. 2006) forengaging in virtual worlds (e.g., Second Life) or chattingwith online sales-representatives. Many consumers usemultiple marketing channels, and employ the internet asan information source before making purchases in the realworld (Venkatesan, Kumar, and Ravishanker 2007).Online virtual environments have been identified as anemerging marketing channel (Li, Daugherty, and Biocca2002), whereby consumers can engage in informationsearch, trial, and purchasing.

An inexpensive and low-risk way that consumers canlearn about brands and products is through the trial orpurchasing of virtual products in 3D virtual environ-ments. The purchasing of virtual products and brands invirtual worlds, such as Second Life, is not only a source ofadditional revenue for the firm, but more importantly,they can provide valuable experiences and interactionswith the brand (Arakji and Lang 2008). Firms are particu-larly interested in how these virtual world brand experi-ences may impact consumer attitudes and purchasingbehavior in other marketing channels. Several companies,such as Adidas, Reebok, Coca-Cola, Sony, IBM, and

Dell, have begun to use virtual worlds as a potentialplatform for understanding and modifying purchasingbehavior in real world marketing channels.

The primary research question in this study looks atwhy users of virtual worlds purchase virtual world brandedproducts (i.e., digital representations of real world brandedproducts). The Theory of Reasoned Action (TRA; Fishbeinand Ajzen 1975), an established model for predictingbehavior in the consumer psychology literature, is used asa theoretical framework and is extended through expan-sion of the social influences construct and addition ofvariables pertaining to the virtual environment context.Specifically, this paper explores the effects of cognitions(i.e., attitudes, behavioral intentions), social influences(i.e., social norms, self-image congruency), and the vir-tual trial of brands on consumer purchasing of brands inthe virtual world. While a handful of papers in the market-ing and e-commerce literature have looked at several ofthese constructs independently (Li et al. 2002; Suh andLee 2005), this paper develops and tests a frameworkwhich considers each of these factors simultaneously todetermine the relative effect of the antecedents on pur-chase intention and behavior in the virtual environment.

The extended TRA model is tested using a uniquedataset from users of the popular 3D virtual world, SecondLife, to investigate how brand experiences and socialinfluences affect purchase intentions and behavior. Todate, we are not aware of any published papers that havecollected data from a real-time virtual world environment.Previous literature has employed laboratory experimentswhere buyers interact with 3D objects on corporatewebsites, or interact with computer-operated sales avatars(Li et al. 2002; Suh and Lee 2005; Jiang and Benbasat2005; Wang et al. 2007; Holzwarth et al. 2006).

This paper makes several contributions to the com-puter-mediation, product trial, and social influence litera-ture. In order to determine if both real world and virtualworld social influences have an effect on virtual worldpurchasing intentions and behavior, real world and virtualworld dimensions of social norms and self-image congru-ence are examined. The perceived diagnosticity construct(i.e., how useful the virtual world brand experience is forevaluating the brand), an established construct in thevirtual product trial literature (Jiang and Benbasat 2005),is examined as a potential moderator of the relationship


between virtual world brand trial and virtual world pur-chase intentions. A new scale, perceived behavioral con-sistency, is developed and tested. The perceived behav-ioral consistency construct is defined in this paper as aconsumer’s belief that their actions are consistent acrossmarketing channels. The perceived behavioral consis-tency scale is examined as a potential moderator of therelationship between virtual world purchase intentionsand behavior.

The results of this study find that virtual world socialnorms and product trial are significant predictors of virtualworld purchase intentions. Virtual world purchaseintentions and product trial were found to be significantpredictors of virtual world purchase behavior. Perceivedbehavioral consistency (a user’s belief that their behavioris consistent across the real world and virtual world), wasfound to significantly moderate the relationship betweenvirtual world purchase intentions and virtual worldpurchase behavior. These findings offer several importantinsights: (1) the impact of social influences may varydepending on the marketing channel, (2) virtual producttrial is useful for influencing and predicting both purchaseintentions and behavior for virtual world products, and (3)consumer beliefs, and consciousness, about whether theyact consistently across channels can determine whetherbehavioral intentions translate into actual purchasingbehavior.

The paper is organized as follows: first, the concep-tual framework for this paper is presented using theTheory of Reasoned Action model. After developinghypotheses for the relationships in the model, the method-ology and results of the study are presented. The finalsection discusses the study’s findings, contributions, andimplications for managers and theory.

CONCEPTUAL FRAMEWORK ANDHYPOTHESES

The Theory of Reasoned Action (TRA) is used as anorganizing framework for developing and testing therelationships between potential antecedents of virtualworld purchase intentions and behavior. In addition to thetheory’s basic predictors (e.g., attitudes, social norms,intentions, behavior), the extended model includes self-image congruence, virtual world brand trial, perceiveddiagnosticity, and perceived behavioral consistency toimprove the model’s predictive power for examining thedrivers of purchase intentions and behaviors in the virtualworld (see Figure 1).

The Theory of Reasoned Action (TRA) is based onthe idea that decisions to carry out a behavior are delibera-tive, and reasoned, rather than spontaneous. The theoryspecifically suggests that behavior is directly influencedby behavioral intentions, and behavioral intentions are

determined by two antecedents: attitudes and subjectivenorms. Attitudes and subjective norms are derived fromunderlying behavioral and normative beliefs. Each of theantecedents of intention may have different levels ofimportance depending on the behavior and situation (Ajzen1991). In the TRA model behavioral intentions fullymediate the impact of attitudes (i.e., overall evaluationtoward performing the behavior) and subjective norms(i.e., perceived social pressure as to whether the behaviorshould be performed) on behavior. When people developan attitude toward a behavior, and perceive whetherpeople important to them approve of the behavior, thedeterminants should provide good predictions of inten-tion (Conner and Sparks 2005). In accordance with theTRA model it is hypothesized that attitudes toward thebrand will positively influence the intention of consumersto purchase virtual world brands, and virtual world pur-chase intentions will positively influence actual virtualworld purchasing behavior.

H1: There is a significant, positive relationship between

attitudes toward the brand and purchase intentions.


purchase intentions and purchase behavior.

Social Influences (i.e., Social Norms and Self-ImageCongruency)

The TRA model suggests that subjective norm is anantecedent to purchase intentions. However, several au-thors have expressed uncertainty regarding the subjectivenorm construct. For instance, Thorbjornsen et al. (2007)has suggested that the subjective norm concept is insuffi-cient for explaining how identity and social influencesimpact behavioral intentions. In a meta-analysis, Armitageand Conner (2001) found the subjective norm construct tobe a relatively weak predictor of intentions. Thus, severalresearchers have advocated that models such as the Theoryof Reasoned Action and Theory of Planned Behaviorshould include other social and identity-related variables,beyond the subjective norm construct

In addition to the social norm construct, this studyattempts to expand the breadth of social influences con-sidered by including the self-image congruency construct(Sirgy et al. 1997), and examining both the real world andvirtual world dimensions of social norms and self-imagecongruency. The self-image congruency construct hasbeen identified in the marketing literature as the degree towhich a product or brand matches a consumer’s self-concept (Kressmann et al. 2006). Self-image congruencytheory suggests that individuals prefer to interact withproducts and brands that fit with their self-image, and theytend to hold more favorable attitudes toward these prod-ucts and brands (Sirgy et al. 1997). Consumers also preferproducts and brands that they perceive to be used by


people who they can relate to and who match their self-identity. Consumers who perceive product and brandexperiences in the virtual world to be consistent with theirself-image are likely to have higher purchase intentions.

Real world and virtual world dimensions of the socialnorm and self-image congruency constructs are measuredin order to determine which types of social pressures aremore accessible and influential in the virtual environmentchannel. It is hypothesized that both real world and virtualworld dimensions of the social norm and self-imagecongruency constructs positively influence a consumer’svirtual world purchase intentions.

H3a

: There is a significant, positive relationship betweenreal world social norms and purchase intentions.

H3b

: There is a significant, positive relationship betweenvirtual world social norms and purchase intentions.

H4a

: There is a significant, positive relationship betweenreal world self-image congruency and purchaseintentions.

H4b

: There is a significant, positive relationship betweenvirtual world self-image congruency and purchaseintentions.

Virtual World Product Trial and Perceived Diagnosticity

Based on the differentiation in the literature betweendirect (e.g., physical) experience and indirect experience

(Hoch and Ha 1986; Fazio 1990; Kempf and Smith 1998),research has found virtual experience to be similar todirect experience, because it offers a high degree ofinteractivity with products and other avatars (Li et al.2002). Studies have shown that consumer interactionswith products in the virtual environment have the poten-tial to increase product knowledge, purchase intentions,leads to more confident brand attitudes, and reducesperceived risk (Li et al. 2002; Suh and Lee 2005).

The perceived diagnosticity construct is a key vari-able in the product trial and virtual reality literature.Perceived diagnosticity has been defined as “the extent towhich a consumer believes that a particular shoppingexperience was helpful for evaluating the product,” andhigher perceived diagnosticity has been shown tostrengthen consumers’ beliefs and belief confidence (Jiangand Benbasat 2005). Because online product and brandexperiences are limited by certain sensory informationsuch as physical inspection of the product, it is importantto understand how helpful an online experience has beenfor a consumer in evaluating a product (Kempf and Smith1998). It is likely that the more helpful a virtual experiencehas been for evaluating a product, the more confident theconsumer will hold their beliefs and attitudes, and thegreater influence the virtual experience will have on realworld purchase intentions and behavior. It is hypoth-esized that virtual world product trial will lead to highervirtual world purchase intentions, and that perceiveddiagnosticity moderates the relationship between trial andintentions.

FIGURE 1Conceptual Framework



virtual world trial of the brand and purchase intentions.

H6: Perceived diagnosticity positively influences the

strength of the relationship between virtual worldtrial of the brand and purchase intentions.

Perceived Behavioral Consistency

It has been empirically established in the consumerpsychology literature that behavioral intentions do notalways lead to actual behavior, as people tend to overes-timate their willingness to engage in a behavior (Ajzenet al. 2004; Brown et al. 2003). Thus, identifying vari-ables that strengthen the intention-behavior relationshipis crucial, especially for firms that rely on purchaseintention data for developing marketing strategies. Inorder to improve the relationship between virtual worldpurchase intentions and behavior, this paper proposes anewly developed construct, perceived behavioral consis-tency, as a potential moderator. The perceived behavioralconsistency construct is defined as a user’s belief that theirbehavior is consistent across marketing channels (i.e., thereal world and virtual world channel), and was developedbased on the metacognitive, and multiple selves literature.Some users of the virtual world may behave similar to howthey act in the real world, and they may even consider theiravatar to be an extension of themselves. Other users mayuse the virtual world to behave in ways that they normallywould not in the real world, and to try out new personas inthe virtual world. People have multiple selves, whichsometimes contradict each other (Mandel 2003). Yetpeople have a need for behavioral consistency in order toestablish a strong sense of self. People are more likely torely on attitudes and intentions when they are based onmultiple instances of past behavior (Petty 2006). When aperson’s intention is not compatible with their self-identity,and their prior behavior, they tend to have a weak inten-tion-behavior relationship (Granberg and Holmberg 1990).It is hypothesized that consumers who perceive theirbehavior to be highly consistent across the real world andvirtual world are more likely to act on their purchaseintentions.

H7: Perceived behavioral consistency positively influ-

ences the strength of the relationship between pur-chase intentions and purchase behavior.

METHOD

Design and Procedure

An online questionnaire was administered to regis-tered users of the virtual world, Second Life. Each respon-

dent had to be 18 years or older, had a Second Life accountfor at least three months, and had encountered real lifebrands in Second Life. These criteria were established inorder to verify that respondents would have at least a basiclevel of experience using Second Life.

The survey asked respondents to think about real lifebrands that they have encountered in Second Life, and tolist up to five of these brands. Respondents were theninstructed to select one brand, from the brands that theylisted, that they have had the most memorable experiencewithin Second Life. Throughout the remainder of thequestionnaire, respondents were asked to answer thequestions with this brand in mind. A single wave of datawas collected using the questionnaire. A pre-test wasconducted prior to full administration of the online sur-vey, to identify and revise any questions that were reportedto be confusing. As this study collected data for bothdependent and independent variables using the samemethod (self-report scales), Harmon’s one-factor test(Podsakoff et al. 2003) was conducted to test for commonmethod bias. A substantial amount of common methodvariance was not evident.

Respondents

Respondents were 209 registered users of the virtualworld, Second Life. Gender and nationality (U.S. vs.international) was counterbalanced in order to obtain arepresentative sample of Second Life users. The gender,age, and nationality statistics for the sample were similarto the average reported statistics for the overall populationof registered users in Second Life (KZero Consulting2007).

Measures

The components of the extended TRA model weremeasured with multi-item scales. To reduce the effects ofresponse bias, some of the measures consisted of a numberof negatively worded items, which were reverse-scoredprior to constructing the scales.

Dependent Variables

Virtual World Purchase Intentions. A single cat-egorical (e.g., yes/no) item was used to assess arespondent’s intention to purchase the selected brand inthe virtual world.

Virtual World Purchase Behavior. A single categori-cal (e.g., yes/no) item was used to assess a respondent’spurchasing behavior of the selected brand in the virtualworld.


Independent Variables

Attitude. Attitude toward the brand, that respondentsselected as having the most memorable experience with inSecond Life, was assessed by four 7-point semantic dif-ferential scales (α = .969). The four attitude items wereadapted from Smith et al. (2007).

Real World Subjective Norm. Subjective norm wasassessed by two items on a 7-point scale (α = .832, scoredstrongly disagree [1] to strongly agree [7]). The two itemswere adapted from Bamberg, Ajzen, and Schmidt (2003).Respondents were asked to indicate the extent to whichimportant referents in the real world think they should buyproducts with the selected brand.

Virtual World Subjective Norm. Respondents werealso asked to indicate the extent to which importantreferents (other avatars) in the virtual world think theyshould buy products with the selected brand (e.g., “Myfriends (other avatars) in Second Life think that I shouldbuy products with this brand.” “My friends (other avatars)in Second Life have bought products with this brand.”) (α =.685, scored strongly disagree [1] to strongly agree [7]).

Real World Self-Image Congruency. Self-image con-gruence was assessed with two items on a 7-point scale(α = .834, scored strongly disagree [1] to strongly agree [7]).The two items were adapted from Sirgy et al. (1997).

Virtual World Self-Image Congruency. Similar to thesubjective norm construct, respondents were also asked toindicate the extent to which they identified with users ofthe brand in the virtual world (e.g., I am very much like thetypical avatar that uses products with this brand in SecondLife.”) and self-concept (“The brand image is highlyconsistent with how I see myself in Second Life.”) (α = .823,scored strongly disagree [1] to strongly agree [7]).

Virtual World Brand Trial. A single categorical (e.g.,yes/no) item was used to assess a respondent’s trial of theselected brand in the virtual world.

Perceived Diagnosticity. Perceived diagnosticity wasassessed by two items on a 7-point scale (α = .886, scorednot helpful at all [1] to extremely helpful [7]). The twoitems were adapted from Jiang and Benbasat (2005) andKempf and Smith (1998).

Perceived Behavioral Consistency. Perceived be-havioral consistency was assessed with four items on a 7-point scale (α = .763, scored strongly disagree [1] tostrongly agree [7]). Respondents were asked how consis-tently they perceived their actions to be across the realworld and virtual world (e.g., “My actions in the virtualworld reflect how I behave in the real world,” “I considermy avatar to be an extension of myself,” “My behavior in

the virtual world is consistent with my behavior in the realworld.”).

Control Variables

Virtual World Usage. Respondents were asked toindicate on average, how many hours they spend inSecond Life each week.

Virtual World Purchase Frequency. Respondentswere asked to indicate how often they purchase any typeof virtual product in Second Life. This was assessed usinga 7-point scale (scored very rarely [1] to very frequently [7]).

RESULTS

Two hierarchical regression analyses were performed.The first regression analysis tested the ability of brandattitudes, social influences (e.g., social norms, self-imagecongruency), and virtual brand trial to predict virtualworld purchase intentions, in addition to exploring themoderation effect of perceived diagnosticity. The secondregression analysis was conducted to test the ability ofpurchase intentions to predict virtual world purchasebehavior, in addition to exploring the moderation effect ofperceived behavioral consistency. Virtual world usageand purchase frequency were controlled for in each model.

Preliminary Analyses

The means, standard deviations, correlations, andreliabilities of the variables are reported in Table 1. Asshown in Table 1, several significant low-to-moderatecorrelations were found among the variables. The follow-ing discussion presents descriptive statistics for the sampledata. The average number of years that respondents hadbeen a member of Second Life was 1.71 years (SD = .937).On average, users reported spending 25.9 hours (SD =23.4) in Second Life per week. About 66 percent of therespondents reported that they frequently purchase virtualproducts in Second Life, and 55 percent of respondentsreported trying out products with the brand that theyselected. When asked to select the real life brand that theyhave had the most memorable experience with in SecondLife (e.g., from the list of real life brands that they recalledfrom memory), the following show the frequency thateach real life brand was selected: IBM (11.9%, n = 25),Adidas (6.2%, n = 13), Coca-Cola (5.26%, n = 11), Dell(4.78%, n = 10), and Nike (4.30%, n = 9.4%).

Regression Analysis Predicting Virtual World Pur-chase Intentions

A hierarchical multiple regression analysis was con-ducted in which virtual world purchase intentions wasregressed onto the revised TRA model (see Table 2).


Attitudes, along with the control variables virtual worldusage and purchase frequency, were entered into Step 1;with virtual world social norms, real world social norms,virtual world self-image congruency, and real world self-image congruency at Step 2. Virtual world brand trial andperceived diagnosticity were entered at Step 3, and theinteraction between brand trial and perceived diagnosticitywas entered in Step 4. All continuous variables weremean-centered before computing the interaction in orderto minimize problems of multicollinearity.

The Step 1 variables (e.g., attitudes, virtual worldusage, purchase frequency) accounted for only 3.6 per-cent of the variance in virtual world purchase intentions,F(3, 205) = 2.51, p = .059, with all three of the predictorsreported as not significant. The Step 2 variables signifi-cantly accounted for an additional 13.1 percent of thevariance in purchase intentions, F(4, 201) = 5.73, p < .000,with virtual world social norms and virtual world self-image congruency found to be positively and signifi-cantly related to purchase intentions. The Step 3 variablessignificantly accounted for an additional 21 percent of thevariance in purchase intentions, F(2, 199) = 13.33, p <.000, with virtual world social norms and virtual brandtrial found to be significant predictors. Inclusion of theinteraction term between virtual brand trial and perceiveddiagnosticity at Step 4 did not produce a significantincrease in variance explained, F(1, 198) = 12.00, p =.522. The final model accounted for 37.7 percent of thevariance in virtual world purchase intentions, with virtualworld social norms, and virtual brand trial, as significant

predictors. Thus, hypotheses H3b

and H5 are supported,

but H1, H

3a, H

4a, H

4b, and H

6 are not supported.

Regression Analysis Predicting Virtual World Pur-chase Behavior

A second hierarchical multiple regression analysiswas conducted in which virtual world purchase intentionswas regressed onto virtual world purchase behavior (seeTable 3). Virtual world purchase intentions, along withthe control variables virtual world usage and purchasefrequency, were entered into Step 1. Virtual world socialnorms, real world social norms, virtual world self-imagecongruency, real world self-image congruency, virtualworld brand trial, perceived diagnosticity, the interactionterm between brand trial and perceived diagnosticity, andperceived behavioral consistency, were entered at Step 2.The interaction between virtual world purchase intentionsand perceived behavioral consistency was entered atStep 3.

As shown in table 3, the Step 1 variables explained38.5 percent of the variance in virtual world purchasebehavior, F(3, 205) = 42.77, p < .000, with virtual worldpurchase intentions, and the control variable virtual worldusage, reported as significant. The Step 2 variablesaccounted for an additional 10.5 percent of the variance invirtual world purchase behavior, F(9, 196) = 15.69, p <.000. Virtual world purchase intentions, virtual worldbrand trial, and the control variable virtual world usage,were found to be significant and positive predictors.

TABLE 1Construct Properties

Table 1 Means, Standard Deviation, and Bivariate Correlations for attitude, RW social norms, VW socialnorms, RW self-imate congruencey (SIC), VW self-image congruency, brand trial, purchase behavior,perceived diagnosticity, and perceived behavioral consistency.

Variable M SD 1 2 3 4 5 6 7 8 9 10

1. Attitude 5.90 1.21 (.96)

2. RW Social Norms 4.69 1.59 .47** (.83)

3. VW Social Norm 4.08 1.51 .24** .51** (.68)

4. RW SIC 4.35 1.55 .54** .63** .44** (.83)

5. VW SIC 4.16 1.42 .52** .58** .51** .72** (.82)

6. VW Brand Trial 5.60 .498 .19** .22** .32** .21** .32** 1

7. VW Purchase Intention .540 .500 .13 .21** .35** .21** .32** .57** 1

8. VW Purchase Behavior .360 .482 .09 .16* .32** .13* .27** .59** .60** 1

9. Perceived Diagnosticity 3.96 1.86 .34** .33** .33** .40** .48** .23** .18** .12 (.88)

10. Behavioral Consistency 4.77 1.51 .15** .08 -.01 .10 .09 -.01 -.12 -.014 .09 (.76)

Note. Mean scores in the current study are based on 7-point scales (1–7), except for virtual world brand tries, purchase intention and

behavior which are categorical variables. The reliabilities (e.g., alpha coefficients) appear in the parentheses on the diagonal. Pearson’s

correlation coefficient (and significance) is reported *p < .05, **p < .01, (two-tailed).


Inclusion of the interaction term between virtual worldpurchase intentions and perceived behavioral consistencyat Step 3 accounted for a further 1.1 percent of the variancein purchase behavior, F(1, 195) = 15.03, p < .05. The finalmodel accounted for 50.1 percent of the variance in virtualworld purchase behavior, with virtual world brand trial,virtual world intentions, the interaction between inten-tions and perceived behavioral consistency, and the con-trol variable virtual world usage, found to be significantpredictors. Thus, hypotheses H

2 and H

7 are supported.

DISCUSSION AND IMPLICATIONS

Using an extended version of the Theory of ReasonedAction model, this study tested several antecedents ofconsumer’s intentions to purchase virtual world productsfrom a focal brand. The relationship between virtualworld purchase intention and actual purchasing behaviorwas also examined. Based on a unique dataset from actualusers of an online virtual world (i.e., Second Life), virtualworld social norms, and virtual trial of products with thefocal brand, were found to be significant and positivepredictors of consumer’s intentions to purchase the vir-tual product. Virtual world self-image congruency wasalso found to be a significant predictor of purchase inten-tions throughout portions of the regression analysis.

It is interesting that virtual world social influenceswere found to be significant predictors of virtual worldpurchase intentions, while real world social norms and

self-image congruency were not significant. This sug-gests that researchers may need to consider the particularcontext and marketing channel when measuring socialinfluences. For both marketers and researchers, this find-ing may indicate that social influences in the virtual worldare more accessible, and have a greater influence ondecision making when consumers are immersed in thevirtual world environment.

In accordance with the TRA model, virtual worldpurchase intentions was found to be a significant andpositive predictor of virtual world purchase behavior.Although, not hypothesized ex ante, virtual world trial ofproducts with the focal brand was also found to be asignificant predictor of purchase behavior. In predictingboth virtual world purchase intentions and behavior,virtual product trial was consistently a significant predic-tor. This finding suggests that online product trial may beessential for influencing consumer purchase intentionsand behavior. When virtual product trial was entered intothe regression model, the attitude construct changed frombeing moderately significant to not significant. This maysuggest that the trial construct explains more variancethan attitudes in predicting purchase intentions, and thusmay be a more useful predictor of future consumer inter-action with the brand.

While virtual product trial was found to significantlypredict both virtual world purchase intentions and behav-ior, perceived diagnosticity was not found to interact with

TABLE 2Results of Hierarchican Regression Analysis

Hierarchical Multiple Regression Analysis Predicting Virtual World (VW) Purchase Intentionas

Predictor R2 R2 Change Fchange

Beta

Step 1 Attitudes .036 .036 2.516 .100VW Usage (hrs. per week) .131VW Purchase Frequency .030

Step 2 VW Social Norms .167 .131 7.902*** .236**RW Social Norms -.024VW Self-Image Congruency .254**RW Self-Image Congruency -.061

Step 3 VW Brand Trial .376 .210 33.430*** .498***Perceived Diagnosticity -.022

Step 4 VW Brand Trial x Diagnosticity .377 .001 4.11 -0.020

Note. VW = virtual world, RW = real world. Continuous interaction perdictors were mean centered.*p < .05. **p < .01. ***p < .001.


the product trial construct. This finding is surprising, aspast research, using a controlled laboratory study, hasfound this interaction to be significant (Jiang and Benbasat2005). This finding may suggest that future researchshould seek to understand consumer’s preferences forvirtual world product experiences that use realistic simu-lation strategies or fantasy-based strategies.

This study provides a pretest of the perceivedbehavioral consistency construct, a new measure developedin this paper, which may help strengthen the relationshipbetween purchase intentions and actual purchasingbehavior in the online marketing channel. It is wellrecognized by both researchers and marketers thatbehavioral intentions frequently do not translate intoactual behavior. The results of this study show that theperceived behavioral consistency construct was found tosignificantly moderate the relationship between virtualworld purchase intentions and purchasing behavior. Thisfinding suggests that the relationship between purchaseintentions and purchasing behavior in the online marketingchannel may be affected by a user’s general perceptions

about their behavior. This effect may have parallels toBem’s self-perception theory (1972), where people lookto their own behaviors to determine their current attitudesand behavioral intentions. In future research, marketersand researchers should seek to understand how consumersview their behavior across offline and online marketingchannels.

This paper has several limitations which should benoted. First, the study relied on self-report measureswhich may have impacted responses. However, severalattempts were made to reduce common method varianceor response bias by including negatively worded itemsand through the use of multi-item measures. Also, itemsassessing constructs that might be prone to commonmethod variance bias were not presented together on thesurvey. A second limitation of the study is the inability todemonstrate causality. It is suggested that future studiesemploy either longitudinal (multi-wave data) or experi-mental designs to convincingly demonstrate causal effectsbetween variables.

TABLE 3Results of Hierarchican Regression Analysis

Hierarchical Multiple Regression Analysis Predicting Virtual World (VW) Purchase Intentionas

Predictor R2 R2 Change Fchange

Beta

Step 1 Attitudes .385 .385 42.774*** .5.93***VW Usage (hrs. per week) .138***VW Purchase Frequency .021

Step 2 Attitudes .490 .105 4.483*** -.037VW Social Norms .079RW Social Norms .005VW Self-Image Congruency .088RW Self-Image Congruency -.081VW Brand Trial .367***Perceived Diagnosticity -.042Diagnosticity x Brand Trial -.025Behavioral Consistency .032

Step 3 Intentions x Behavioral Consistency .501 .011 4.130* .370*

Note. VW = virtual world, RW = real world. Behavioral Consistency = perceived behavioral consistency.Continuous interaction perdictors were mean centered.*p < .05. **p < .01. ***p < .001.

REFERENCES

Ajzen, Icek (1991), “The Theory of Planned Behavior,”

Organizational Behavior and Human Decision Pro-cesses, 50 (2), 179–211.

____________, T.C. Brown, and Carval (2004), “Explain-


ing the Discrepancy Between Intentions and Actions:The Case of Hypothetical Bias in Contingent Valua-tion,” Personality and Social Psychology Bulletin,30 (9), 1108–21.

Arakji, Yahya Reina and Karl Reiner Lang (2008), “Ava-tar Business Value Analysis: A Method for the Evalu-ation of Business Value Creation in Virtual Com-merce,” Journal of Electronic Commerce Research,9 (3), 207–18.

Armitage, C.J. and M. Conner (2001), “Efficacy of theTheory of Planned Behavior: A Meta-AnalyticReview,” British Journal of Social Psychology, 40,471–99.

Bamberg, Sebastian, Icek Ajzen, and Peter Schmidt (2003),“Choice of Travel Mode in the Theory of PlannedBehavior: The Roles of Past Behavior, Habit, andReasoned Action,” Basic and Applied Social Psy-chology, 25, 175–87.

Bem, J. Daryl (1972), “Self-Perception Theory,” Advancesin Experimental Social Psychology, San Diego, CA:Academic Press, 1–62.

Brown, T.C., I. Ajzen, and D. Hrubes (2003), “FurtherTests of Entreaties to Avoid Hypothetical Bias inReferendum Contingent Valuation,” Journal ofEnvironmental Economics and Management, 46 (2),353–61.

Conner, Mark and Paul Sparks (2005), “Theory of PlannedBehavior and Health Behavior,” Predicting HealthBehavior, Buckingham, U.K.: Open University Press,121–62.

Fazio, H. Russell (1990), “Multiple Processes by WhichAttitudes Guide Behavior: The MODE Model as anIntegrative Framework,” Advances in ExperimentalSocial Psychology, San Diego, CA: Academic Press,75–109.

Fishbein, Martin and Icek Ajzen (1975), “Strategies ofChange: Active Participation,” Belief, Attitude, Inten-tion, and Behavior: An Introduction to Theory andResearch, Reading, MA: Addison-Wesley, 411–50.

Granberg, Donald and Soren Holmberg (1990), “TheIntention-Behavior Relationship among U.S. andSwedish Voters,” Social Psychology Quarterly, 53(1), 44–54.

Hoch, J. Stephen and Young-Won Ha (1986), “ConsumerLearning: Advertising and the Ambiguity of ProductExperience,” Journal of Consumer Research, 13(September), 221–33.

Holzwarth, Martin, Chris Janiszewski, and Marcus M.Neumann (2006), “The Influence of Avatars on OnlineConsumer Shopping Behavior,” Journal of Market-ing, 70 (October), 19–36.

Jiang, Zhenhui and Izak Benbasat (2005), “Virtual Prod-uct Experience: Effects of Visual and FunctionalControl of Products on Perceived Diagnosticity andFlow in Electronic Shopping,” Journal of Manage-

ment Information Systems, 21 (3), 111–47.Kempf, Deanna S. and Robert E. Smith (1998), “Con-

sumer Processing of Product Trial and the Influenceof Prior Advertising: A Structural ModelingApproach,” Journal of Marketing Research, 35 (3),325–38.

Kressmann, Frank, M. Joseph Sirgy, Andreas Herrmann,Frank Huber, Stephanie Huber, and Dong-Jin Lee(2006), “Direct and Indirect Effects of Self-ImageCongruence on Brand Loyalty,” Journal of BusinessResearch, 59 (9), 955–64.

KZero Consulting (2007), [http://www.brandchannel.com/images/papers/359_5_rules_v1.pdf].

Li, Hairong, Terry Daugherty, and Frank Biocca (2002),“Impact of 3-D Advertising on Product Knowledge,Brand Attitude, and Purchase Intention: The Mediat-ing Role of Presence,” Journal of Advertising, 31 (3),43–57.

Mandel, Naomi (2003), “Shifting Selves and DecisionMaking: The Effects of Self-Construal Priming onConsumer Risk-Taking,” Journal of ConsumerResearch, 30 (1), 30–40.

Petty, E. Richard (2006), “A Metacognitive Model ofAttitudes,” Journal of Consumer Research, 33 (1),22–24.

Podsakoff, P., S. MacKenzie, J. Lee, N. Podsakoff (2003),“Common Method Biases in Behavioral Research: ACritical Review of the Literature and RecommendedRemedies,” Journal of Applied Psychology, 88 (5),879–903.

Sirgy, M. Joseph, Druv Grewal, Tamara F. Mangleburg,Jae-ok Park, Kye-Sung Chon, C.B. Claiborne, J.S.Johar, and Harold Berkman (1997), “Assessing thePredictive Validity of Two Methods of MeasuringSelf-Image Congruence,” Journal of the Academy ofMarketing Science, 25 (3), 229–41.

Smith, R. Joanne, Deborah J. Terry, Antony S.R. Manstead,Winnifred R. Louis, Diana Kotterman, and JacquelineWolfs (2007), “Interaction Effects in the Theory ofPlanned Behavior: The Interplay of Self-Identity andPast Behavior,” Journal of Applied Social Psychol-ogy, 37 (11), 2726–50.

Suh, Kil-Soo and Young Eun Lee (2005), “The Effects ofVirtual Reality on Consumer Learning: An EmpiricalInvestigation,” MIS Quarterly, 29 (4), 673–97.

Thorbjornsen, Helge, Per E. Pedersen, and HerbjornNysveen (2007), “This is Who I am: Identity Expres-siveness and the Theory of Planned Behavior,” Psy-chology and Marketing, 24 (0), 763–85.

Venkatesan, Rajkumar, V. Kumar, and Nalini Ravishanker(2007), “Multichannel Shopping: Causes and Conse-quences,” Journal of Marketing, 71 (April), 114–32.

Wang, Liz C., Julie Baker, Judy A. Wagner, and KirkWakefield (2007), “Can a Retail Web Site be Social,“Journal of Marketing, 71 (3), 143–57.


For further information contact:Jason A. Gabisch

Department of MarketingIsenberg School of Management

University of Massachusetts – Amherst121 President’s DriveAmherst, MA 01003Phone: 603.547.5895



EFFECTS OF ONLINE STORE ATMOSPHERIC QUALITIESON CONSUMER BEHAVIOR: HOLISTIC

VS. EXPERIMENTAL APPROACHES

George J. Avlonitis, Athens University of Economics & Business, GreeceSpiros P. Gounaris, Athens University of Economics & Business, GreeceChristos D. Koritos, Athens University of Economics & Business, Greece

SUMMARY

Despite the slow pace of growth in online retail sales,academic and industry experts, as well as online retailers,agree that online retailing has already become an integralpart of the marketing strategy of most major players inmost retail sectors. In fact, a growing number of custom-ers are becoming loyal users of the online distributionchannel, both for finding information and purchasingspecific types of goods and services. As a result competi-tion among retailers for attracting these customers hasincreased. However, as the sophistication of recommen-dation agents improves, product benefits and pricingincentives fail to deliver sustainable competitive advan-tages to online retailers. Most importantly, competition isa click away. This situation makes an imperative foronline retailers, to design their websites in such a way thatwill not only attract but engage and keep customers in thesite for the longest possible time. However, research in therole of online store atmospherics qualities remains scarce,and online retailers are left experimenting with singleguidance the normative tastes of website designers.

To start feeling this gap academic research has bor-rowed theoretical frameworks from the field of consumerbehavior in offline retail contexts. More specifically the“Stimulus-Organism-Response” model (SOR – Mehrabianand Russell 1974) which has dominated research on therole of atmospheric elements in a variety of offline retailshopping contexts (Turley and Milliman 2000) has beenempirically validated across a number of websites (e.g.,Dailey 2002; Eroglu, Machleit, and Davis 2003; McKinney2004; Mummalaneni 2005; Oh, Fiorito, Cho, and Hofacker2008; Wu, Cheng, and Yen 2008).

However, one limitation of the majority of thesestudies is the use of experimental designs. Although usingexperimental designs to study customer reactions to varia-tions of one AQ contributes importantly to our under-standing of consumer online behavior, it ignores the factthat consumer’s perceptions result from the combinedeffects of AQ (Bitner 1992; Kotler 1973–1974; Mehrabianand Russell 1974; Turley and Milliman 2000). In otherwords, experimental designs fail to capture changes inparticipants’ perception with regard to other atmospheric

elements that are being affected by changes of themanipulated atmospheric element(s).

Thus, in order to capture the role of online shoppingatmosphere in a holistic way, one needs to take intoconsideration consumer perceptions of an almost exhaus-tive number of online store AQ. To this end, Eroglu,Machleit, and Davis (2001) have developed a frameworkfor organizing the totality of online AQ. However, despitethe important contribution of Eroglu, Machleit, and Davis(2001), the empirical validation of this framework has anumber of limitations. Firstly, Eroglu, Machleit, andDavis (2003) use only fictitious online stores. Secondly,they consider a retailer from a single industry (apparel).Third, they incorporate attitude in the SOR frameworkwith no theoretical justification. Forth, they have notdeveloped a battery of items representing their categoriza-tion of online store AQ. Last but not least, the study resultshave not been validated within service contexts (i.e.,service websites). Thus, the major goal of this paper is toprovide a thorough examination of Eroglu, Machleit, andDavis (2003) framework, by adequately filling the abovementioned gaps.

To do this, we chose two types of services websites(i.e., resorts and banks) representative of the hedonic/utilitarian continuum (Babin et al. 1994). Two indepen-dent judges examined the websites of all banks in Greeceand all resorts located in the wider area of the capital city(Athens) and selecting two website as the most represen-tative of each category. This process revealed a 50 percentinter-rater agreement. In other words, authors agreed onone resort and one bank as being representative of the totallisted. In addition to this, following standard scale devel-opment processes, authors developed a measurementinstrument for capturing the importance of a number ofwebsite atmospheric elements representative of the threecategories suggested by Eroglu, Machleit, and Davis(2001). Via mall intercepts respondents were invited tobrowse the website of either resort or bank, and fill in thesurvey questionnaire. This process yielded 460 usefulquestionnaires for each website.

Data analyses via SEM validate once more the sig-nificance of the relationships suggested by the SOR


framework across both websites. Furthermore, the three-category organization of the website atmospheric quali-ties suggested by Eroglu, Machleit, and Davis (2001)proved significant antecedents of the SOR framework.Importantly, with, the exception of a direct effect from theproduct-related information AQ to satisfaction, theseeffects follow the same route observed in offline shoppingcontexts, that is, via the customer affective (pleasure/arousal) and cognitive-attitudinal (satisfaction) responses.The partial mediation of affective responses (pleasure/arousal) on the relationship between product-relatedinformation and satisfaction, is in line with Belk’s (1976)

contention regarding the inability of this psychologicaltaxonomy of responses to account for all types of situ-ational stimuli. In other words, in online shopping con-texts, product-related information AQ do not affect atti-tudes and behavioral intentions only via affectiveresponses. Other types of organismic responses such ascognitive, learning and motivational (Belk 1976) can alsomediate the effect of certain online store atmosphericstimuli on consumer behavioral responses, an issue whichwarranties further investigation. References are availableupon request.

For further information contact:Christo D. Korito

Department of Marketing & CommunicationRoom 104

Athens University of Economics & Business47A, Evelpidon Str. & 33, Lefkados Str.

AthensGreece 113 62

Phone: +30210.8203852Fax: +30210.8203851



EFFECTS OF WORD-OF-MOUTH AND E-WORD-OF-MOUTHON INNOVATION USE IN THE PRESENCE OF

NETWORK EXTERNALITIES

Tomoko Kawakami, Kansai University, JapanKazuhiro Kishiya, Kansai University, Japan

Mark E. Parry, University of Missouri – Kansas City

SUMMARY

For many years studies of new product diffusion havetied new product adoption to positive word-of-mouthcommunication about innovative products (Mahajan andMuller 1979; Mahajan, Muller, and Bass 1990; Rogers1995). A complementary perspective examines the diffu-sion of innovation use among adopter populations (Chuan-Fong and Venkatesh 2004). Extrapolating from insightsin the adoption literature, the authors examine the impactof innovation use of two types of word-of-mouth commu-nication: personal communication with friends andacquaintances (pWOM) and electronic communicationwith consumers who are not friends and acquaintances(eWOM). The authors argue that, in addition to directlyinfluencing product use, both types of word-of-mouth caninfluence consumer perceptions of network externalityvariables, which have been linked in recent research toperceptions of product advantage and ease-of-use (Song,Parry, and Kawakami 2009).

The paper develops a conceptual model that incorpo-rates both pWOM and eWOM as antecedent variablesassociated with the intensity and variety of new productuse. In addition to this direct relationship between WOMvariables and product use, the authors hypothesize thatword-of-mouth indirectly affects product use through itsimpact on consumer perceptions of network externalityvariables: (1) an innovation’s installed base and (2) theavailability of complementary products. In turn, theseperceptions of network externality variables are hypoth-esized to influence both the intensity and variety ofinnovation use. In this way the conceptual model inte-grates elements from the network externalities and word-of-mouth literatures into a single model designed to pre-dict the dimensions of new product use.

The authors test their hypotheses using data describ-ing consumer perceptions of recently introduced, newgeneration portable game players such as the NintendoDS and the Sony Play Station Portable. Survey develop-ment followed the procedures recommended by Dillman(1978) and Douglas and Craig (1983). The finalizedquestionnaire was sent to a marketing research company,who administered the survey in February 2009. Thisprocess yielded 639 usable responses, including 247 fromconsumers who had adopted a new generation portablegaming device (an adoption rate of 38.7%). The conceptsand measures of this study are available from the authorsupon request.

Having established the validity of our measurementmodel, the authors use structural equation modeling to testtheir research hypotheses (Anderson and Gerbing 1988).Model coefficient estimates are reported in Figure 1. Theassociated fit statistics meet well-established thresholds(Chi-square = 373.00, d.f. = 154, CFI = .91, IFI = .91,RMSEA = 0.076). Results indicate that, relative to eWOM,pWOM has a stronger relationship with post-adoption usebehavior. In particular, pWOM has a positive and signifi-cant relationship with both intensity of use and variety ofuse. In addition, pWOM has a positive indirect relation-ship with both usage variables that is mediated by con-sumer perceptions of the availability of complementaryproducts. In constrast, eWOM is positively and signifi-cantly related with variety of product use, but not withintensity of product use. The latter result arises from theinsignificant relationship between eWOM and consumerperceptions of the availability of complementary productsis insignificant. From an academic perspective, the resultscontribute to the innovation use literature by examiningthe impact of word-of-mouth on innovation use anddiscriminating between the effects of personal and elec-tronic word-of-mouth.


ENDNOTE

Acknowledgement: The authors acknowledges financialsupport by MEXT KAKENHI Grant-in-aid for Sci-entific Research (B) #20330093 (2008–2011) ofJapan Ministry of Education, Culture, Sports, Sci-

FIGURE 1Relationship Between pWOM/eWOM, Network Externalities and Innovation Use

pWOM

eWOM

Intensity of Use

Variety of Use

Age

Internet Skill

Perceived Installed Base

Size

Perceived Availability of

Complementary Products

Innovation Use

Word-of-Mouth Network Externalities

.40

.31* .31*

.34**

.28**

.32*

.28** -.38**

.52*

.33**

.15**

Note: ÷2 = 373.00, d.f. = 154, CFI = 0.91, IFI = 0.91, RMSEA = 0.076. **p < .01, *p < .05.

ence and Technology and Strategic Project to Sup-port the Formation of Research Bases at PrivateUniversities Matching Fund Subsidy from MEXT(2009–2013). The author also thanks JMR ScienceCo., Ltd. for their assistance with data collection.

For further information contact:Tomoko Kawakami

Faculty of CommerceKansai University

Suita, Osaka 564–8680Japan

Phone: (06)6368.0145Fax: (06)6339.7704



HOW BRAND-SPECIFIC INTERNET SEARCHES AFFECT PURCHASEBEHAVIOR: AN EMPIRICAL STUDY OF INTERRELATIONS

IN ONLINE ADVERTISING

Sebastian Hild, RWTH Aachen University, GermanyMalte Brettel, RWTH Aachen University, Germany

Andreas Engelen, RWTH Aachen University, Germany

SUMMARY

The question how advertising works and how effec-tive it is has been the topic of research for decades.Various models have been developed to explain the per-suasion process from advertising input to consumerbehavior. With the advent of the internet and its growingimportance as a place to do business, advertising on theinternet has become a topic in academic research. Internetadvertisers have since moved from classic banner adver-tising to new advertising models (e.g., “on-demand”advertising such as search engine advertising and pricecomparison websites) that make better use of the internet’sunique potential for interaction between consumer andadvertiser, often blurring the line between advertising,promotion and sales as defined in the offline world.Research into how these new advertising channels inter-act, e.g., how banner advertising affects search advertis-ing, is still missing. In our study, we address the questions:How do online advertising channels work and interact?How do clicked and unclicked ads affect sales, and eachother? And in particular, what is the role of price compari-son websites in online advertising?

Drawing on the established elaboration likelihoodmodel we develop an advertising effects model in thepresence of interrelations between advertising channels.We then apply this model using a comprehensive databaseobtained from the .com-website of a leading online-platform for used and antiquarian books. We include fourchannels in our analysis: e-mails, affiliate advertising,search-engine advertising, and advertisements on pricecomparison websites. The sample spans a period of 365days, eliminating seasonal effects in the books market. Itprovides, on a day-by-day basis, indicators on advertisingactivity and sales. In total, the data contain more than 2.8million purchases and 25 million website visits.

For data analysis we use structural equation model-ing (SEM), as it allows for the assessment of structures incomplex research models. We apply AMOS 16.0 as animplementation of a covariance-based SEM approach,enabling us to assess the overall model quality. As all ourdata are observed variables, our model is comprised onlyof the structural model with no associated measurement

models to be tested. The model quality assessment indi-cates a very good fit of the model.

We find that there are significant interrelations amongadvertising channels, as advertising affects not only salesbut other advertising channels. In particular, consumer-initiated ad exposures on search engines and price com-parison websites are a likely effect of advertising. WhileCho in his 1999 article predicts virtually no attitudechange from unclicked ads, our findings clearly show animpact of ad impressions on sales and on further adexposures. These results confirm the findings of previousstudies. A second interesting finding is the weak negativeeffect of affiliate ad impressions on search ad clicks,together with their positive effect on search ad impres-sions. A possible explanation for this is that users whoreacted to the banner exposure by using a search-enginerather than clicking the banner, thus causing a search adimpression, are apparently equally reluctant to click thisself-initiated advertisement. Similarly, the strong effect ofunclicked affiliate ad impressions on purchases can beexplained by this wide-spread reluctance to click banners.With respect to our research question on the role of pricecomparison websites, we find that they are an importantstep in the decision-making process of online consumers.In addition to extant research showing that brands matterin price comparison shopping, we find that previous adexposures also matter and increase clicks.

Our finding of significant advertising channel inter-relation has wide-ranging implications for practitioners,as the common tracking approach using cookies orhyperlinks does not account for these effects. Typically,neither tracking of unclicked ad exposures nor of multiple(clicked) exposures is possible. Therefore, the total saleseffect of advertising channels affecting other channels issystematically underestimated, as only the last channel inthe navigation sequence will be measured by direct track-ing. Further, we find that a large part of ad effects is causedby unclicked exposures, and thus cannot be measuredwith current tracking approaches. The limitations in directtracking necessarily cause advertisers to question theircurrent policies, and lead to suboptimal allocation ofresources to advertising channels due to missing andmisleading data. We therefore suggest that advertisers


apply our methodology or a similar approach to identifythe total effect of impressions and clicks in each advertis-ing channel. Combining these total effects with the cost

associated with generating the corresponding ad expo-sures allows practitioners to optimize their online adver-tising spending. References are available upon request.

For further information contact:Malte Brettel

Centre for EntrepreneurshipRWTH Aachen University

Templergraben 6452062 Aachen

GermanyPhone: +49.241.8096177



HOW CUSTOMER SATISFACTION AND LOYALTY PROGRAMSINFLUENCE CUSTOMERS’ CASH FLOW VARIABILITY:

IMPLICATIONS FOR MARKET SEGMENTATION

Crina Tarasi, Central Michigan University, Mount PleasantRuth Bolton, Marketing Science Institute, BostonAnders Gustafsson, Karlstad University, SwedenBeth Walker, Arizona State University, Tempe

SUMMARY

The cash flow a company is expected to generatedetermines its economic value. Customer relationshipsrepresent the essential source of business cash flow.Customer relationships can increase shareholder value byincreasing the level and reducing the variability of a firm’scash flow (Srinivastava, Shervani, and Fahey 1998).Under these circumstances, besides return, the variabilityassociated with specific market segments should be con-sidered when managers allocate marketing efforts.Despite its importance, customer revenue variability andits drivers have not been addressed in research focused onidentifying the optimal portfolio of customers to be tar-geted with marketing investments (Dhar and Glazer 2003;Tirenni et al. 2007). This omission is surprising becausethe expected value of a customer portfolio is substantiallyaffected by variability.

This study investigates how variation in customerrevenues (specific cash flow patterns), can be explainedusing a combination of customer satisfaction and demo-graphic, relational and market characteristics. We developa model that links variability in the cash flow of individualcustomers to actionable market segmentation variables.

In this paper we develop propositions regarding theeffects of customer characteristics on cash flow. Wedevelop propositions as opposed to hypotheses because –being the first to test the relationship between customercharacteristics and cash flow – we are unsure regardingthe form of the relationship (linear versus nonlinear,simple effects versus interaction). We will test severalfunctional forms: simple linear effects, interaction model,where the relationship between customer characteristicsand customer cash flow is moderated by satisfaction,multiplicative, exponential and semi-log. We test ourpropositions in two study contexts: financial services andtelecommunications. In both studies we focus on a randomsample of current customers, for which we gather surveydata which we combine with company’s records thatinclude customers’ demographic information andpurchasing data. In both contexts, according to all threecriteria for model comparison (Akaike Information

Criterion (AIC); log-likelihood; and R-Square) the semi-log model outperformed the other models. Consistent inboth contexts, our results indicate that satisfaction reducesthe variability of cash flows, while participation in customerloyalty programs increases it. Interestingly, characteristicsassociated with high levels of customer cash flow (youngage, lengthy relationship, and high cross-buying) are alsoassociated with an increase in cash flow variability.

The contribution of our research to the marketingliterature and practice is three fold. First, we show that bymanaging the composition of the customer base andrelationships with individual customers, firms can gobeyond the simple management of customer wealthaccording to the expected level of cash flow, by consider-ing the variability and therefore correctly estimating theexpected value. A key outcome of our study is that itencourages marketers to look beyond the level of cus-tomer purchases to consider variability over time. Second,we show that satisfaction, besides positively impactingthe level of cash flow, it also increases the expected valueof future cash flows by reducing variability. Third, weshed light on the true implications of employing loyaltyprograms: even though participation in loyalty programsis generally characterized by an increase in the level ofrevenue (Seiders et al. 2005), the variability increases inboth contexts. Probably, we still have to come up witheffective loyalty programs.

Currently, firms and marketers assume that if a firmhas a large enough customer base, variability is not anissue, but this assumption is invalidated by the largevariability and seasonality observed in most companies’sales. The assumption that by adding new customers thevariability can be reduced does not hold if the newcustomers exhibit similar behavior patterns with theextant customers. Through our research we offer insighton how to assess and manage the variability of the cus-tomer base, and open the door toward future research touncover how to build efficient customer portfolios, thatmight maximize return for a certain level of risk (variabil-ity) or minimize risk for a desired level of return. Refer-ences are available upon request.


For further information contact:Crina O. Tarasi

MarketingCentral Michigan University

100 Smith HallMount Pleasant, MI 48859



INVESTIGATING THREE-DIMENSIONAL MODEL OF CUSTOMERCOMMITMENT IN THE CELL PHONE SERVICE SETTING

Yaser F. Alabdi, Manchester Business School, United KingdomJikyeong Kang, Manchester Business School, United Kingdom

SUMMARY

Customers’ bargaining power makes it particularlyimportant that marketers investigate the distinctive natureof customer commitment. The emerging role of technol-ogy has become a source of pressure to companies, whichgives customers access to a wide range of alternatives, andurges companies to intensify their customer retentionefforts. Customers become active players in dynamic andfast-growing marketplace, and more powerful than everbefore in customer-firm relationship.

Although the extent literature in relationship market-ing investigated customer commitment as unidminsionalor two-dimensional concept, this research, based on com-mitment theory from the organizational behavior, extendour understanding of the nature of customer commitmentand conceptualize it as a three-dimensional concept;affective desire-based, calculative cost-based, and nor-

mative obligation-based in an attempt to contribute to theliterature of relationship marketing area. Their impactshave been investigated on a range of loyalty outcomes;intention to stay, word-of-mouth and willingness to pay.A sample of 525 individual contractual customers in thecell phone service setting was used to test the researchmodel using structural equation modeling.

The affective desire-based had positive impacts onloyalty outcomes, while the cost-based had detrimentaleffects. Normative commitment, surprisingly, had nega-tive effects in relation to loyalty outcomes. It seemscustomers stay with the company because they like stay-ing, however, those who committed because of obliga-tions develop negative attitude toward the company as amatter of reactions to treatments they received and mostlikely defer when the obligations end. References areavailable upon request.

For further information contact:Yaser Alabdi

Manchester Business SchoolBooth Street West

Manchester, M15 6PBUnited Kingdom

Phone: +44(0)161.275.6378Fax: +44(0)161.275.6596



THE REFERRAL LIKELIHOOD REFERRAL OF REWARD PROGRAMS:THE MODERATING EFFECT OF COMMUNICATION MEDIA AND

PRODUCT INVOLVEMENT

Yimin Zhu, Sun Yat-sen University, China

SUMMARY

Recently, the practice of rewarded referral program iswidespread in China. Based on the research of GangseogRyu and Lawrence Feick (2007), we also adopt anexchange theory framework for examining consumers’responses to referral reward programs. This article reportsthe results of two studies that investigate the impact ofreferral reward programs on referral likelihood. Accord-ing to our studies, we examine the effect of the presenceor absence of a reward, reward size, and reward recipient(i.e., does the existing customer, or both and existing andnew customers get the reward?). We also examine themoderating impact of communication media (i.e., tradi-tional or online media) and of product involvement (i.e.,high or low involvement).

In Study 1, we propose three hypotheses as following:H1: Compared with no reward, the presence of a rewardwill increase referral likelihood. And the referral likelihoodwill be enhanced with the increase of referral rewards.H1a: The impact of offering a reward on referral likelihooddepends on communication media. The presence of areward (compared with no reward) increases referrallikelihood more through online communication media(i.e., MSN) than through traditional communication media(i.e., word-of-mouth). H1b: The impact of offering areward on referral likelihood depends on productinvolvement. The presence of a reward (compared with noreward) increases referral likelihood more for low-involvement products than for high-involvement products.

Study 1 was a 3×2×2 between-subjects factorialexperiment in which we varied the reward size (no reward,smaller, larger), the communication media (traditional,online), and the product involvement (high, low). Threehundred fifty undergraduate students from a major uni-versity in South China were randomly assigned to theexperimental conditions.

In study 1, we found that if the company choosestraditional communication media in its reward program,rewards may increase the referral likelihood, but there isno significant difference between smaller rewards andlarger rewards. This means a challenge for the marketingmanager. Even you increase the promotional budget inthis condition, the effect of programs may not be higher asyou expect before. However, when the company chooses

online communication media in its reward program, thereis a significant difference between different rewards size.Providing a small reward may increase the referrallikelihood greatly through online media because theconvenience usage of internet and the less social-psychological costs. Furthermore, the impact on referrallikelihood between rewards size are also different forhigh- and low-involvement products. For example, forhigh-involvement product, consumer would like to makethe recommendation even without rewards. But for low-involvement product, the reward program will increasethe consumers’ referral likelihood. In other words, rewardprograms are more important for low-involvement productsthan for high-involvement products. When company wantsto develop a promotional program, the marketing managershould consider the involvement level of their products.

In Study 2, we propose three other hypotheses asfollowing: H2: Reward allocation scheme will have impactson referral likelihood. Consumers are most likely to makea referral in the Reward Both condition, followed byReward Me. H2a: The impact of reward allocation schemeon referral likelihood depends on communication media.Through traditional communication media (word-of-mouth), consumers are most likely to make a referral in theReward Both condition, followed by Reward Me. Throughonline communication media (MSN), consumers are mostlikely to make a referral in the Reward Me condition,followed by Reward Both. H2b: The impact of rewardallocation scheme on referral likelihood depends on prod-uct involvement. For high-involvement products, con-sumers are most likely to make a referral in the RewardBoth condition, followed by Reward Me. For low-involvement products, consumers are most likely to makea referral in the Reward Me condition, followed by RewardBoth.

Study 2 was a 2×2×2 between-subjects factorialexperiment in which we varied the scheme of the rewarddistribution (Reward Me, Reward Both), the communica-tion media (traditional, online), and the product involve-ment (high, low). Two hundred thirty-five undergraduatestudents from a major university in South China wererandomly assigned to the experimental conditions.

In study 2, we found that the rewards scheme have animpact on the referral likelihood. If the company uses thetraditional media in its reward program, they may just


provide the rewards to both existing and new consumer(Reward Both) other than to existing consumer (RewardMe). Furthermore, for low-involvement products, RewardMe (providing rewards to the existing consumer) will bemore effective than Reward Both. And for high-involvement product, things will be the opposite. There-fore, the marketing manager should choose scheme likeReward Both if their products are high-involvement onesand choose scheme like Reward me if their products arelow-involvement ones.

In general, our results show that offering a reward andReward Both will increase referral likelihood. However,in order to design more effective promotional program,company should consider the communication media theychoose in the program and the involvement level of theirproducts at the same time.

Our researches only focus on the measurement ofconsumer referral likelihood. Although the results areconsistent with exchange theory, we haven’t gain greaterinsight into the mechanism underlying the effects. There-fore, future research may add new experiment to measurethe socialpsychological costs and benefits of making areferral. Secondly, our experiment designs are from theaspect of recommender, future research may design theexperiment from the aspect of receiver. In other words,whether providing a referral reward program will influ-ence the real purchasing behavior of receiver or not?Finally, our studies only use college students as theparticipants, future research may employ other groups ofparticipants (such as professionals, housewives), choosedifferent product categories (such as mobile phone, sham-poo), and other communication media (such as BBS,telephone).

ENDNOTE

Acknowledge: this research is supported by Philosophyand Social Science Foundation of the Ministry of

Education of China under Grant 08JZD0019 and theYoung Teacher fundation of Sun Yat-sen Universityunder Grant 3171915.

REFERENCES

Biyalogorsky, Eyal, Eitan Gerstner, and Barak Libai(2000), “Customer Referral Management: OptimalReward Programs,” Marketing Science, 20 (Winter),82–95.

Bolton, Ruth N., R.K. Kannan, and Matthew D. Bralnlett(2000), “Implications of Loyalty Program Member-ship and Service Experiences for Customer Reten-tion and Value,” Journal of the Academy of Market-ing Science, 28 (Winter), 95–108.

Bone P.F. (1992), “Determinants of Word of MouthCommunication During Product Consumption,”Advances in Consumer Research, 19, 579–83.

Bristor, Julia M. (1990), “Enhanced Explanations of Word-of-Mouth Communications: The Power of Relation-ships,“ [J]. Research in Consumer Behavior, (4), 51–83.

Brown, J.J. and P.H. Reingen (1987), “Social Ties andWord-of-Mouth Referral Behavior,” Journal of Con-sumer Research, 14, 350–62.

Bussiere, D. (2000), “Evidence and Implications of Elec-tronic Word-of-Mouth,” Developments in MarketingScience, 23, 361–62.

Buttle, Francis A. (1998), “Word of Mouth: Understand-ing and Managing Referral Marketing,” Journal ofStrategic Marketing, 6, 241–54.

Celsi, Richard and Jerry C. Olson (1998), “The Role ofInvolvement in Attention and Comprehension Pro-

cesses,” Journal of Consumer Research, 15 (Septem-ber), 210–24.

Chatterjee, P. (2001), “Online Reviews: Do ConsumersUse Them?” Advances in Consumer Research, 28,129–33.

Chen, Yuxin and Mengze Shi (2001), “The Design andImplications of Customer Recommendation Pro-grams,” working paper, Stern School of Business,New York University.

Duhan, D.F., S.D. Johnson, J.B. Wileox et al. (1997),“Influences on Consumer Use of Word-of-MouthRecommendation Sources,” Journal of AcademicMarketing Science, 25 (4), 283–95.

Ekeh, Peter P. (1974), Social Exchange Theory: The TwoTraditions. Cambridge, MA: Harvard UniversityPress.

Gangseog, Ryu and Lawrence Feick (2007), ”A Penny forYour Thoughts: Referral Reward Programs andReferral Likelihood,” Journal of Marketing, 71 (Janu-ary), 84–94

Garbarino, Ellen and Mark S. Johnson (1999), “TheDifferent Role of Satisfaction, Trust, and Commit-ment in Customer Relationships,” Journal of Mar-keting, 63 (April), 70–87.

Gatignon, Hubert and Thomas S. Robertson (1986), “AnExchange Theory Model of Interpersonal Communi-cation,” in Advances in Consumer Research, RichardJ. Lutz, ed. Provo, UT: Association for ConsumerResearch, 13, 534–38.


Helm, S. (2000), “Viral Marketing: Establishing Cus-tomer Relationships by ‘Word-of-Mouse,’” Elec-tronic Markets, 10 (3),159–16.

Hennig-Thurau T. and G. Walsh (2003), “ElectronicWord-of-Mouth: Motives for and Consequences ofReading Customer Articulations on the Internet,”International Journal of Electronic Commerce, 8 (2),51–74.

Hennig-Thurau, T., K.P. Gwinner, G. Walsh, and D.D.Gremler (2004), “Electronic Word-of-Mouth via Con-sumer-Opinion Platforms: What Motivates Consum-ers to Articulate Themselves on the Internet?” Jour-nal of Interactive Marketing, 18 (1), 38–52.

Hogan, John, Katherine N. Lemon, and Barak Libai(2003), “What Is the True Value of a Lost Cus-tomer?” Journal of Service Research, 5 (February),196–208.

Kelly, Harold H. (1972), “Attributions in Social Interac-tion,” in Attribution: Perceiving the Causes ofBehavior, E.E. Jones, D.E. Kanouse, H.H. Kelly,R.E.Nisbett, S.Valins, and B. Weiner, eds. Norristown,NJ: General Learning Press.

Kivetz, Ran and Itamar Simonson (2002), “Earning theRight to Indulge: Effort as a Determinant of Cus-tomer Preferences Toward Frequency Pro-gram Reward,” Journal of Marketing Research, 39(May), 155–70.

Lee, Thae Min and Cheol Park (2006), “Effects of Direc-tion and Type of Electronic Word of Mouth Informa-tion on Purchase Decision: Cross-Cultural Compari-son Between Korea and U.S.,” Korean MarketingReview, 21 (1), 29–56.

Reingen, Peter H. and Jerome B. Kernan (1986), “Analy-sis of Referral Networks in Marketing: Methods andIllustration,” Journal of Marketing Research, 23 (4),370–78.

Rosen, E. (2001), The Anatomy of Buzz: Create Word-of-Mouth Marketing. 1st New Doubleday Currency.

Rust, Rol and Katherine N. Lemon, and Valarie A. Zeithaml(2004), “Return on Marketing: Using Customer Equityto Focus Marketing Strategy,” Journal of Marketing,68 (January), 109–127.

Thorsten Hennig-Thurau, Kevin P. Gwinner, GianfrancoWalsh, and Dwayne D. Gremler (2004), “ElectronicWord-of-Mouth via Consumer-Opinion Platforms:What Motivates Consumers to Articulate Themselveson the Internet?” Journal of Interactive Marketing,18 (1), 38–52.

Wellman, Barry, Janet Salaff, Dimitrina Dimitrova, LauraGarton, Milean Gulia, and Carloine Haythornthwaite(1996), “Computer Networks as Social Networks:Collaborative Work, Telework, and Virtual Commu-nity,” Annu.Rev.Sociol. 22, 213–38

Westbrook, R.A. and R.L. Oliver (1991), “The Dimen-sionality of Consumption Emotion Patterns and Con-sumer Satisfaction,” Journal of Consumer Research,18 (1), 84–91.

Zeithaml, Valarie A. (1988), “Consumer Perceptions ofPrice, Quality, and Value: A Means-End Model andSynthesis of Evidence,” Journal of Marketing, (3),2–21.

____________, L.L. Berry, and A. Parasuraman (1996),“The Behavioral Consequenees of Service Quality,Journal of Marketing, 60 (2), 31–46

For further information contact:Yimin Zhu

School of BusinessSun Yat-sen University

No. 135 Xingang West RoadGuangzhou, 510275

ChinaPhone: 8620.8411.4103



ASYMMETRY OF CONSUMER UNCERTAINTY AND ITS IMPACT ONBRAND DECLINE

Adina Barbulescu, The University of Tennessee, KnoxvilleAjay Kohli, Georgia Institute of Technology, Atlanta

SUMMARY

A brand becomes increasingly weak when consum-ers expect low levels of brand performance in the future(i.e., point expectation) or considerable variability in thelevel of these performances (i.e., range expectation) (Rustet al. 1999). We extend this stream of literature and arguea brand’s decline, measured as low likelihood of brandrepurchase, is also determined by the distribution ofconsumers’ range expectation above and below the pointexpectation of brand performance.

Asymmetric distributions arise when consumers’range expectation is not symmetrically distributed aroundthe point expectation – that is, consumers have unevenexpectations that a brand will exceed or underperformtheir expectation for brand performance (i.e., point expec-tation). We propose that asymmetric distributions of therange expectation (i.e., asymmetric uncertainty) impactthe longevity of consumers’ relationship with the brand.We argue that elements of temporal sequences of brandexperiences, such as the presence of a trend and peak aresources of asymmetric uncertainty.

We suggest that consumers who experience anincreasing trend of brand encounters estimate a lowerabove range as compared to participants exposed to a non-trended sequence of experiences, which leads to theformation of negative asymmetric uncertainty. On theother hand, decreasing trends tend to be associated witha low below range of expectation as compared to a non-trended series of experiences, leading to the formation ofpositive asymmetric uncertainty.

Peaks, the most extreme values of series of experi-ences are the most positive experiences a consumer haswith a brand. We suggest that the distinctiveness of a peakdepends on its location in the sequence relative to the otherexperiences of the sequence. When a peak is preceded andfollowed by experiences of a much lower level, theresulting discrepancy accentuates the peak’s high valueand its distinctiveness from the series’ other experiences.When experiencing a high contrast peak, we expect con-sumers to be influenced by the easier recall and theperceived positive valence of the peak experience andexpect future performances to likely exceed the pointestimation (Kahneman and Tversky 1973). This in turnresults in a larger above range of expectation as compared

to a series of experiences with a low contrast peak, leadingto the formation of positive asymmetric uncertainty.

We argue that holding the point expectation andrange expectation constant, consumers’ asymmetricuncertainty influences their brand choices. When con-sumers hold negative asymmetric uncertainty, they expect adisproportionately lower level of brand performance (vis-à-vis their point expectation), and discard the possibilityof an equally high level of brand performance. Thus, aconsumer who holds negative asymmetry of expectationsis less likely to continue to purchase the brand.

We argue above that negative asymmetric uncer-tainty jeopardizes the longevity of the brand relation(holding the point expectation and the range expectationconstant). We are thus interested in identifying ways toreduce negative asymmetry of uncertainty and help con-sumers continue their relationships with brands. Here, weidentify a commonly used practice by companies – pro-viding promises of better future performance along withexplanations for past performances. We argue that whenconsumers experience an increasing trend of experiences,advancing company promises of improved future perfor-mance result in expectations of even higher levels ofperformance above the point expectation; that is, weexpect to see a larger above range of expectation, evenafter a potential increase in the point expectation. The nowlarger above range of expectation reduces the inequalitybetween the above and below ranges, leading to lowernegative asymmetry.

We argue that when consumers experience a decreas-ing level of brand performance, a company’s promisesmay result in a trade-off: consumers will expect a higherpoint expectation (which increases the longevity of thebrand relation) but at the same time will lower the positiveasymmetry of uncertainty and its benefits. Consumersmay interpret the promise and explanation as a desperateattempt to hide a real cause of the decline in performance;consumers may thus become more uncertain about thebrand’s future performance and estimate a larger belowrange, even after a potential increase in their point expec-tation. The now larger above range of expectation reducesthe inequality between the above and below ranges, lead-ing to lower positive asymmetry. We conducted twostudies which validated the hypotheses presented above.


Our findings contribute to the expectation literature.We argue here that expectations should be conceptualized

as a range and not solely as a point in order to understandthe consequences of expectations on brand choices.

REFERENCES

Kahneman, D. and A. Tversky (1973), “Psychology ofPrediction,” Psychological Review, 80 (4), 237–51.

Rust, R.T., J.J. Inman, J.M. Jia, and A. Zahorik (1999),“What You Don’t Know about Customer-PerceivedQuality: The Role of Customer Expectation Distribu-tions,” Marketing Science, 18 (1), 77–92.

For further information contact:Adina Barbulescu

Department of Marketing and LogisticsUniversity of Tennessee

Knoxville, TN 37996–0570Phone: 865.974.1657



UNINTENDED CONSEQUENCES: HOW BRAND EXTENSIONS MAKEBRANDS MORE VULNERABLE TO CONSUMER CONFUSION

Eric D. DeRosia, Brigham Young University, ProvoThomas R. Lee, Brigham Young University, Provo

Glenn L. Christensen, Brigham Young University, Provo

SUMMARY

One of the topics at the intersection of marketing,psychology, and the law is the frequently studied topic ofsource confusion (e.g., Foxman, Berger, and Cote 1992;Howard, Kerin, and Gengler 2000), which is defined as aconsumer’s perception that an imitator brand (the juniormark) is made by or associated with the maker of anoriginal brand (the senior mark). We evaluate two poten-tial antecedents of source confusion: (1) a brand extensionin which a competitor of the senior mark has extended intothe junior mark’s product category, and (2) the extent ofa consumer’s motivation to carefully make the purchasedecision. We identify an interaction between these twofactors – an effect that is opposite to that generallyidentified in the literature and counterintuitive to theassumptions that guide much of the law in this area.

Rather than the extended brand becoming more vul-nerable to source confusion, this research demonstrates itis the competitors of the extended brand that are mademore vulnerable to source confusion. This circumstanceis commonplace. For example, the 2008 brand extensionof Froot Loops breakfast cereal into the clothing categorycaused all the competitors of Froot Loops to become morevulnerable to source confusion from similar junior marksin the clothing category. A review of trademark recordsshows 24 senior marks in the breakfast cereal categorythat were identical to junior marks in the clothing category(e.g., Honeycomb breakfast cereal and Honeycomb cloth-ing). In this way, brand extension has a wider sphere ofinfluence than previously considered in the literature – asingle brand extension makes all competitors of theextended brand more vulnerable to source confusion.

In the search for possible antecedents of sourceconfusion, previous researchers have not tested brandextension, but they have tested motivation to deeplyprocess information. The findings have been mixed, withFoxman, Muehling, and Berger (1990) finding less sourceconfusion among consumers with higher motivation, andHoward, Kerin, and Gengler (2000) finding more sourceconfusion among such consumers (for brands that aresimilar in semantic meaning).

We propose a broad theoretical approach to explainthe role of motivation in source confusion. From a schema-

theory perspective, the cognitive process undertaken by aconsumer when he or she encounters a product andidentifies its brand (i.e., identifies the product’s source) isa categorization process. This suggests that when con-sumers are offered a junior-marked product that is stronglyincongruent with the senior mark (e.g., a Cadillac brandnotebook computer), consumers are unlikely to subtypeand therefore unlikely to suffer source confusion.

However, we propose that if one of the senior mark’scompetitors in the original product category extends itsbrand into a new product category, consumers will viewas more plausible a similar extension of the senior mark(or a junior mark that appears to be the senior mark).

H1: If a competitor of the senior mark extends its brandinto a new product category, the likelihood ofsource confusion from a junior mark in the newproduct category will increase (vs. no brand exten-sion by a competitor of the senior mark).

The theoretical rationale underlying H1 describesconsumers as undertaking a number of cognitive pro-cesses. Models of categorization commonly suggest thatsuch schema retrieval and consideration requires cogni-tive effort (Cohen and Basu 1987). As is frequentlyobserved in the literature (e.g., Luna and Peracchio 2002),consumers expend the effort to perform cognitive pro-cesses only when motivated to do so. The effectsdescribed in H1 will be slight or even absent amongconsumers who lack the motivation to update their schemasand carefully categorize the junior mark. Two separatesources or forms of motivation are consideredhere: enduring involvement and need for cognition. Wepropose that either form could fully motivate consumersto diligently update their schemas and carefully performthe categorization task.

H2a: If a competitor of the senior mark extends its brandinto a new product category, consumers with higherenduring involvement in the original product cat-egory are more likely to be confused by a juniormark in the new product category (vs. consumerswith lower enduring involvement).

H2b: If a competitor of the senior mark extends its brandinto a new product category, consumers with higher


need for cognition are more likely to be confusedby a junior mark in the new product category (vs.consumers with lower need for cognition).

In an experiment, participants (n = 206) viewed oneof three stimuli: an ad for a typical Dell laptop computer,an ad for a new Dell laptop computer with luxuriouspositioning (e.g., with leather trim and titanium body),and an ad for a Mercedes-Benz brand extension into thelaptop computer category (with the same luxurious posi-tioning). Participants then viewed a text-only ad for a“Cadillac brand notebook computer.” Source confusionwas measured (as in Simonson 1994), as well as enduringinvolvement and need for cognition.

H1 was supported by an observation that the likelihoodof source confusion in condition 3 (the Mercedes-Benzbrand extension) was higher than in conditions 1 and 2 (nobrand extension). This indicates that a brand extension byone of the senior mark’s competitors into the juniormark’s category causes an increase in source confusion.

H2a and H2b were also supported. Among partici-pants who viewed the Mercedes-Benz brand extension,the results of a logistic regression model indicate thatconfusion is more likely if the participant has high moti-vation of either form: enduring involvement or need forcognition. This suggests that motivation moderates theeffect of brand extensions on source confusion. Consum-ers who are highly motivated (i.e., either high in enduringinvolvement or high in need for cognition) expend thecognitive effort necessary to carefully update the relevantmemory schemas with information about the brand exten-sion and carefully categorize the junior-marked product.After performing these cognitive processes, these con-sumers are more likely to suffer source confusion. Con-sumers who lack motivation (i.e., low in enduringinvolvement and low in need for cognition) are less likelyto perform the cognitive processes necessary for the brandextension to have its effects, making such consumers lesslikely to suffer source confusion.

Taken together, these findings expand the collectiveunderstanding of source confusion at the intersection ofmarketing, psychology, and the law. This study is themarketing literature’s first empirical examination of brand

extension as a potential antecedent of source confusion.Furthermore, this is the first application of schema theoryto the source confusion question, thereby identifying amoderating role of motivation in source confusion that isopposite to the effect usually assumed in the literature.

This research has many implications for marketingpractice. Our findings suggest that even if consumers havehigh motivation, managers of established brands shouldbe vigilant to protect their brand from infringement byjunior marks in other product categories. Likewise, ourfindings suggest that every brand extension by a competi-tor makes the senior mark more vulnerable to infringe-ment. Therefore, such competitor moves should be con-sidered in the context of litigation that protects the estab-lished brand from infringement.

Our findings also have implications for marketingmanagers who are choosing names for new brands. Suchmanagers should already be aware that branding deci-sions are limited by “legal boundaries” in the law of“passing off” (Hennessey, Bell, and Kwortnik 2005,p. 65) and of the trademark infringement standard of the“likelihood of confusion.” Our findings add importantnuances to these legal boundaries for brand managersoperating in the context of brand extension. In thatincreasingly commonplace context, our findings suggestthat such managers should avoid choosing a brand namethat is similar to a senior mark that consumers view asplausibly extending the brand, particularly if the consum-ers are high in involvement.

This research also has implications for the law. Forthe many trademark infringement cases that are broughtbefore the courts each year, previous brand extensions bycompetitors (embraced in the legal notion of “bridging thegap”) should be considered in the judicial evaluation oftrademark infringement. In addition, most courts considerconsumer motivation (part of the legal notion of “con-sumer sophistication”) to be negatively associated withsource confusion. Our findings suggest that the truth ismuch more complicated, with consumer motivation some-times positively associated with source confusion. Thisprovides grounds for the courts to adopt a more nuanced –and more realistic – approach in trademark confusioncases. References are available upon request.

For further information contact:Eric D. DeRosia

Marriott School of ManagementBrigham Young University

671 TNRBProvo, UT 84663



LABOR PRODUCTIVITY IN RETAILING: A MODERN ASSESSMENT

Charles A. Ingene, The Hong Kong Polytechnic University, Hong KongJianfeng Jiang, Northeastern Illinois University, Chicago

Rahul Govind, University of Mississippi, Oxford

SUMMARY

Labor productivity – output per unit of labor input –is a key driver of (i) higher profitability for business and(ii) better living standards for society. Historically, thegrowth of labor productivity growth has been slower inretailing than in manufacturing which has caused retailingto take the larger share of the blame for stagnant livingstandards. Although over the past two decades, productiv-ity growth in retailing as a whole has improved, the fourworst performing retail sectors (motor vehicle dealers,gasoline stations, grocery stores, and restaurants) accountfor over half of all retail establishments, sales, andemployment. Understanding what constrains productiv-ity growth in these four lagging retail sectors may be ofassistance in guiding the enhancement of labor productiv-ity in all of retailing.

Somewhat surprisingly, published research in thisarea has used data that is at least 30 years old. Givendramatic changes in retail structure over the past threedecades which includes the emergence of warehouseclubs and erosion in the base of mom and pop stores, it iscertainly time to update our knowledge. Indeed, there is adistinct possibility that past inferences about the drivers oflabor productivity may no longer be valid. Furthermore,earlier research only examined individual retail sectors(e.g., grocery stores) without considering interactionsbetween sectors. In contrast, the four sectors we analyzeinclude a complementary pair (motor vehicle dealers andgasoline stations) and a competing pair (grocery storesand restaurants). Thus, we are able to assess whethermacro-environmental impacts on labor productivity arecommon for complementary and competing retail sectors.

Central to our argument is a belief that micro-levelactions taken by retail firms must be compatible with themacro-environment within which they operate. Our studyis targeted toward understanding the macro-level deter-minants of labor productivity. We use data from the most

recent U.S. censuses that have complete information – the2002 Economic Census and the 2000 Census of Popula-tion and Housing Characteristics – to investigate themacro-level determinants of cross-sectional variations oflabor productivity in the afore-mentioned four key retailsectors. Our database covers 379 Metropolitan StatisticalAreas (MSAs) which accounts for more than 83 percent ofthe U.S. population.

We derive nine hypotheses from the extant literatureto predict the impacts of a set of socio-economic, geo-graphic, competitive, and marketing-mix factors on laborproductivity (measured as dollar sales per employee).This provides us with a good comparison of our resultsand those of the previous studies; thus, we can drawinferences whether macro-environmental variables exertthe same impact on retail labor productivity that they didin earlier studies.

Our OLS regression results indicate that our indepen-dent variables explain between 52 percent and 71 percentof the cross-sectional variation in labor productivity acrossthe four lines of trade. The average explanatory power is57 percent; this is comparable to previous studies. Addi-tionally, three-fourths of our hypotheses are validated.Since some beta coefficients are of (significantly) differ-ent signs from earlier studies, there is evidence thatmacro-environmental changes have altered the impact onlabor productivity. Of course, differences in effect direc-tions across lines of trade are indicative of differentialimpacts of environmental factors across retail sectors.

This study is our first step toward a series of investi-gations of retail labor productivity. We conclude ourpaper by proposing three directions for future research:(1) analyzing more retail sectors, (2) conducting a time-series study to explore the historical changes of retaillabor productivity across MSAs, and (3) including moremacro-determinant variables in the analysis. Referencesare available upon request.


For further information contact:Charles A. Ingene

M801, Li Ka Shing TowerDepartment of Management and Marketing

The Hong Kong Polytechnic UniversityHung Hom, Kowloon

Hong KongPhone: 852.2766.7342



TEAM STEWARDSHIP IN CUSTOMER SERVICE TEAMS:ANTECEDENTS AND CONSEQUENCES

Jeroen Schepers, Eindhoven University of Technology, The NetherlandsAd de Jong, Eindhoven University of Technology, The Netherlands

Ko de Ruyter, Maastricht University, The Netherlands

SUMMARY

Introduction

To effectively handle customer complaints and usecustomer feedback to their advantage, more and morecompanies install customer service teams dedicated tosolving customer problems and collecting new insights.Previous research has indicated that customer serviceteams achieve greater flexibility and adaptability in theperformance of service activities if the delegation ofauthority is moved to the frontline. In addition, a “com-mand and control” structure is difficult to implement incustomer service teams, as the team members are oftengeographically dispersed from each other and the com-pany. With reduced possibilities to monitor and controlservice performance, managers face the question of howthey can be confident that employees will act in the bestinterest of the organization and the customer. We intro-duce team stewardship as a governing principle. Wedefine team stewardship as a collectively held sense ofresponsibility to oversee and improve performance in theservice team’s area of responsibility in congruence withthe best interests of the organization and the customer. Webuild on self-determination theory to identify four driversof team stewardship and consider frontline employeeservice performance as the ultimate outcome. In addition,following stewardship proponents, we examine whethercontrol structures from agency theory and intrinsic moti-vation from stewardship theory display synergies ordissynergies in affecting employee performance.

We selected a large international car manufacturer asthe empirical setting to test our conceptual model. Wedistributed 437 questionnaires to customer serviceemployees working in Western Europe; our final sampleconsisted of 234 respondents, organized in 30 teams.Next, we asked 30 supervisors who managed these teamsto complete a survey that would rate the performance ofeach of their supervised employees over the last six

months. After verifying the validity and reliability of themeasures, we apply multilevel regression analysis to testour conceptual model. As a first important insight, weshow that team stewardship is a vital factor in influencingservice employee behavior. Most important, a sharedunderstanding to oversee and improve performance incongruence with the best interests of the organization andthe customer can drive extra-role service performance.However, at the same time, team stewardshipreduces in-role service performance. Perhaps the teammembers are so involved in doing their utmost best for thecustomer that they tend to forget about the formal serviceguidelines. Alternatively, the employee and the customermay have developed a long-lasting relationship that temptsa service employee to cross formal guidelines. Second, wefind that the control mechanism of contingent rewardsmooths the effects of stewardship on service employeeperformance. More specifically, both the negative effectof team stewardship on in-role service performance andthe positive effect on extra-role service performance areweakened when a team manager rewards the team fortheir efforts in achieving performance targets. Third, teamstewardship is most strongly driven by structural related-ness. That is, when a team has goals that are compatiblewith the goals of the organization, the team members actresponsible and are willing to work in the best interests ofthe organization. Besides structural relatedness, affectiverelatedness is also a driver of team stewardship. Finally,an interesting finding is the nonsignificance of autonomyas an antecedent of team stewardship

The results suggest that adequate group rewardstructures should be used to strike the right balancebetween the counteracting effects of team stewardship onemployee performance. Managers of customer serviceteams should therefore be both empowering as well ascontrolling to achieve optimal results. An effective methodto build stewardship is to enhance structural relatedness,for instance by involving team members in backofficeoperations.


For further information contact:Jeroen Schepers

Innovation, Technology Entrepreneurship & Marketing GroupEindhoven University of Technology

Den Dolech 2, Pav. L.07P.O. Box 513, 5600 MB Eindhoven

The NetherlandsPhone: +31 40 247 4384

Fax: +31 40 246 8054E-Mail: [email protected]


THE EFFECTS OF CUSTOMER ORIENTATION AND IDENTIFICATIONON THE SERVICE EMPLOYEE: COMMITMENT AND JOB BURNOUT

D. Todd Donavan, Colorado State University, Fort CollinsBrad D. Carlson, Saint Louis UniversityTom Dewitt, University of Hawaii – Hilo

SUMMARY

Service marketers have long known about the impor-tance of hiring good service employees. The personalinteraction between the service provider and the customeris critical to achieving satisfying exchanges (Heskett et al.2008). Service employees have been creditedwith increasing customer satisfaction (Rust, Zahorik, andKeiningham 1996; Donavan and Hocutt 2001) as well ascustomer loyalty and retention (Heskett et al. 2008; Jonesand Sasser 1995).

On the employee side, research demonstrates thatoutstanding service personnel have higher job satisfac-tion and commit more organizational citizenship behav-iors – those behaviors that go beyond specific job require-ments by promoting positive work outcomes (Donavan,Brown, and Mowen 2004). However, a question remainsunanswered regarding the effect of the service environ-ment on employee retention. In many cases, psychologi-cal withdrawal (e.g., dissatisfaction) from a job or organi-zation tends to occur prior to behavioral withdrawals (e.g.,turnover). Thus, understanding psychological factorswithin a service environment such as organizational com-mitment and job burnout that may influence turnover isimportant for service managers and researchers. Thepurpose of this study is to examine both the individual andorganization elements that may affect two variablesrelated to retention: organizational commitment and jobburnout. We investigate two levels of customer orienta-tion: (1) the organizational service orientation (i.e., OSO),and (2) the individual employee’s customer-oriented dis-position. Additionally, this study investigates the effect ofthe service worker’s identification (i.e., identity overlap)with the firm.

OSO addresses customer orientation at the organiza-tional level, whereas customer orientation has been stud-ied at the individual level as well. Brown et al. (2002)defined customer orientation as an employee’s tendencyor predisposition to meet customer needs in an on-the-jobcontext. From an employee’s perspective, CO is associ-ated with higher levels of job satisfaction, commitmentand altruism toward fellow service workers (Donavan,Brown, and Mowen 2004). Service orientation, at theorganizational level, is defined as an organization-wideembracement of a basic set of relatively enduring organi-

zational policies, practices and procedures intended tosupport and reward service-giving behaviors that createand deliver “service excellence” (Lytle, Hom, and Mokwa1998, p. 459). While CO is defined as a disposition, basedpartly on the personality characteristics of the individual,the environment may also play a role in determining aservice employee’s customer orientation. Thus, organiza-tional service orientation should positively affect a ser-vice worker’s customer orientation. Additionally, as “ser-vice excellence” is rewarded within the OSO culture,organizational commitment should increase and job burn-out should decrease.

Organizational identification is defined as the degreeto which the individual defines him/herself with the sameattributes as the organization (Dutton, Dukerich, andHarquail 1994). In this sense, organizational identifica-tion represents the employee’s relationship with the focalfirm. Strong organizational identification is associatedwith attendance (Bhattacharya, Rao, and Glynn 1995) andemployee helping behaviors (Bergami and Bagozzi 2000).We propose that as the overlap between individual self-schema and organizational schema becomes stronger, theservice workers will become more customer-oriented,resulting in a decrease in employee burnout and an in-crease in organizational commitment.

A field study was conducted with a food service firmin a Midwestern city in the United States. Employees ofthe food service firm completed a survey that includedmeasures of customer orientation (Donavan, Brown, andMowen 2004), organizational service orientation (Lytle,Hom, and Mokwa 1998), organizational identification(Bergami and Bagozzi 2001), organizational commit-ment (Mowday, Steers, and Porter 1079), and job burnout(Singh, Goolsby, and Rhoads 1994). The theoretical frame-work was tested using structural equation modeling. Theresults of the analysis largely supported the anticipatedstructure of relationships (χ2 = 272.96, df = 142, p < .01;CFI = .97; TLI = .96; and RMSEA = .05). Overall, thefindings demonstrate that, while previous research sug-gests that basic personality traits predict CO, employee’sCO is also predicted by environmental issues such asorganizational service orientation and organizational ID.These findings suggest that while firms may not be able toalter the individual’s personality, firms can work to changethe organizational culture which may improve service


performance. Surprisingly, a positive relationship betweenOSO and burnout was revealed. This finding suggests thatperhaps a firm can over-emphasize the “service” culture

to a point of annoying employees. References are avail-able upon request.

For further information contact:D. Todd Donavan

Colorado State University113 Rockwell Hall

Fort Collins, CO 80523Phone: 970.491.6897



COMPANIES’ INNOVATION ORIENTATION AND PRODUCT PROGRAMINNOVATIVENESS: THE MODERATING ROLE OF

INTERFUNCTIONAL CONSENSUS

Ruth Maria Stock-Homburg, Darmstadt University of Technology, GermanyBjoern Six, Darmstadt University of Technology, Germany

SUMMARY

The pathways to companies’ innovativeness mostlyremain a black box for both practitioners and academics(Cooper and Kleinschmidt 1995; Szymanski, Kroff, andTroy 2007). In order to shed more light on this phenom-enon, fundamental research has focused on the drivers ofinnovation (Chandrashekaran et al. 1999), such as “soft”drivers like an innovation-oriented organizational culture(e.g., Tellis, Prabhu, and Chandy 2009) and “hard” driv-ers like an innovation-oriented organizational structure(e.g., Hauser, Tellis, and Griffin 2006). Researchers andpractitioners alike further emphasize the importance of aninterfunctional approach to support a company’sinnovativeness (Siguaw, Simpson, and Enz 2006). Inparticular, interfunctional consensus has been discussedas an important mechanism to support the realization ofstrategic orientations (e.g., Atuahene-Gima, Slater, andOlson 2005) and to foster companies’ innovativeness(e.g., Xie, Song, and Stringfellow 2003).

Extant research on strategic consensus, which hasdefined the phenomenon as the “[. . .] agreement amongtop-, middle-, and operating-level managers on the funda-mental priorities of the organization” (Floyd andWooldridge 1992, p. 194), largely focuses on top man-agement team members (e.g., Homburg, Krohmer, andWorkman 1999). Less attention has been drawn on con-sensus between middle-level managers of different func-tional units (e.g., Bowman and Ambrosini 1997). How-ever, a better understanding of the role of consensusbetween marketing and R&D is in particular important inorder to grasp the pathways to companies’ innovativeness:Marketing and R&D perhaps play the most crucial rolesfor companies’ innovativeness (Atuahene-Gima andEvangelista 2000), their relationship is nevertheless oftenportrayed as far from harmonious (e.g., Griffin and Hauser1996). Thus, we need to learn more about how inter-functional consensus affects companies’ innovativeness.

The purpose of this paper is to provide a deeperunderstanding of the role of interfunctional consensus forthe implementation of companies’ innovation-orientedculture and structure. The basic structure of our frame-work comprises direct links between “hard” and “soft”aspects of companies’ innovation orientation and its prod-uct program innovativeness (PPI) under varying condi-

tions of interfunctional consensus. More specifically, weargue that both innovation orientation of organizationalculture and structure positively affect different dimen-sions of PPI (i.e., newness, frequency, and value), withorganizational culture being of higher importance.Interfunctional consensus between marketing and R&D ishypothesized to strengthen these relationships.

Analysis and Results

We empirically test our framework based on a studycomprising data from 126 companies, representing themanufacturing and the services sector. From these com-panies, dyadic data from marketing and R&D managerswere collected. In order to rule out common method bias(Scandura and Williams 2000), R&D managers wereasked to assess the antecedents of companies’ innovative-ness, while marketing managers assessed companies’actual innovativeness. Following suggestions to validatemarketing managers’ assessments of innovativeness froma customer’s perspective (Szymanski, Kroff, and Troy2007), we further gathered customer data for 49 of thesecompanies.

Our research hypotheses were tested using hierarchi-cal moderated regression analysis (Aiken and West 1991).Results indicate that the innovation orientation of organi-zational culture positively affects the newness, frequency,and value of companies’ PPI. Interestingly, we find thatthe innovation orientation of structure affects PPI charac-teristics differently. While organizational structure has apositive impact on frequency of innovation, we fail to findstatistical support for a positive impact on PPI newnessand value. Thus, organizational culture is indeed moreimportant for a company’s overall PPI. Furthermore, ouranalysis shows that the evidenced direct effects of cultureand structure on the different dimensions of PPI aresignificantly strengthened under conditions of higherconsensus among marketing and R&D managers.


Our present study provides theoretical reasoning andempirical evidence concerning antecedents of a company’sinnovativeness and the conditions under which theseantecedents are most effective. It has several importantimplications. From a conceptual perspective, our study is


one of few to consider innovativeness at the programlevel. In contrast to the majority of extant research, weintegrate three meaningful dimensions of companies’innovativeness (newness, frequency, and value) into ourframework in order to gain a more comprehensive pictureof this complex phenomenon. From a methodologicalperspective, this study is among the first study in the fieldof innovation marketing which empirically integratesthree different perspectives on companies’ innovativeness.We collected data from marketing and R&D managers.The use of customer data to triangulate managerial

responses allows gaining a more valid picture of compa-nies’ innovativeness. Finally, from a content-related per-spective, our findings reveal that organizational culture ispreeminent for companies’ innovativeness, in particularwhen compared to organizational structure. We find thatinterfunctional consensus is an important condition underwhich companies’ innovation orientation can flourish.These findings are also of managerial relevance as man-agers have to make the “right” choices when designingtheir organizations for innovation. References are avail-able upon request.

For further information contact:Ruth Maria Stock-Homburg

Darmstadt University of TechnologyHochschulstr. 1

64289 DarmstadtGermany

Phone: +49.6151.16.7322Fax: +49.6151.16.7320



WHEN DOES OPEN INNOVATION IMPROVE PERFORMANCE? ANINVESTIGATION OF RELATIONAL CAPABILITY

AND FLEXIBILITY

Sanjay R. Sisodiya, University of Idaho, MoscowJean L. Johnson, University of Amsterdam, The Netherlands

Yany Grégoire, Washington State University, Pullman

SUMMARY

The process of leveraging sources external to the firmfor new product development (NPD), which Procter &Gamble calls Connect and Develop, has been termedgenerally as open innovation. Open innovation means thatfirms actively embrace the use of external ideas andexternal paths to markets as the firm advances its tech-nologies (Chesbrough 2003, 2006). Despite its appeal, weunderstand little about the performance implications ofopen innovation. Does it always lead to superior firmperformance? What, if any, internal or external conditionsenhance or inhibit firm performance gains from openinnovation? Drawing on resource-based views where thefirm is depicted as a bundle of resources and capabilities(e.g., Barney 1991; Dutta, Narasimhan, and Rajiv 1999;Wernerfelt 1984), we explore these questions and arguecertain resources and capabilities enhance the perfor-mance gains realized from open innovation.

Open innovation involves a deliberate programmaticapproach underpinned by complex routines (Nelson andWinter 1982) built to seek, access, combine and deployexternal sources of ideas, knowledge, and technologies,thus we characterize it as a dynamic capability (Eisenhardtand Martin 2000). We argue that open innovation resultsin superior economic rewards in terms of forward lookingindicators of the firm’s economic performance (i.e.,Tobin’s q).

Open innovation hinges on a firm’s connection toexternal sources for NPD inputs such as intellectualproperty, technology, ideas, and etc. This connectionhappens through a firm’s boundary spanning activitiesthat take a variety of forms, and can involve outsourcingrelationships, strategic alliances, deep partner-style rela-tionships with upstream suppliers or downstream custom-ers, or even arms length relationships with upstreamsuppliers or downstream customers. We argue that theability to make and manage relationships is vital in realiz-ing the potential rewards from open innovations. Thisability to make and manage interfirm relationships, that isthe firm’s relational capability (e.g., Dyer and Singh1998; Johnson, Sohi, and Grewal 2004; Sivadas andDwyer 2000), entails learned ways of behaving including,

for example, procedures, interaction patterns, and opera-tional issues in interfirm relationships.

Building on the extant literature (Johnson, Sohi, andGrewal 2004) we conceptualize relational capability interms of relevant interfirm knowledge including interac-tional and functional knowledge stores, and introduceinitiation knowledge stores. We argue that relationalcapability – the amalgam of these three interfirm knowl-edge stores – positively moderates the impact of openinnovation on firm performance (H

1). Specifically we

suggest that relational capability will work to enable openinnovation because of the firm’s enhanced capacity toengage in effective external search and connect withexternal sources for new product inputs.

We believe that other elements in a firm’s resourcebundle also come into play when combined with relationalcapability, namely flexibility. Flexibility has been con-ceptualized from a number of perspectives (e.g., Johnsonet al. 2003). Here, consistent with recent research (e.g.,Fang, Palmatier, and Steenkamp 2008; Lee and Grewal2004), we focus on resource slack because such slackinstills flexibility in the firm (Lee and Grewal 2004). Fromthis perspective, flexibility involves a cushion ofresources that allows firms to adapt and respond to oppor-tunities and to ebb and flow in the innovation process(Nohria and Gulati 1996).

While relational capability enables open innovationthrough an increased capacity for external connecting andboundary spanning, resource slack serves as an additionalenabling mechanism that allows firms to take advantageof and act on potential opportunities derived from externalsearches. This relational capability and flexibilityempower open innovation capacity and further increasethe speed and sustainability of new product generationand introduction, which leads in greater firmperformance (H

2).

We test our hypotheses with primary and secondarydata from 204 high tech firms. As the bases for thecomplex moderated effects we hypothesized, weexpected and found that open innovation positivelyimpacted performance. To support H

1 we found that


relational capability moderated the influence of openinnovation on Tobin’s q. While we did not find evidencefor the linear relationship for resource slack combiningwith relational capability to further enhance performancegains from open innovation, we did find evidence of anon-linear relationship.

Our findings show that open innovation is related tofinancial performance, and firms following open innova-tion achieve superior performance. While this evidence isconsistent with P&G’s open innovation success, moreimportantly we find firms can be even more successful ifthey have the ability to connect with their environment.Hence interfirm relationships are critical to the success offirms following open innovation.

While investigating the role of flexibility in moderat-ing the effect of relational capability and open innovation,we explore the impact of maintaining slack resources onenhancing firm performance. We find evidence of anonlinear relationship for resource slack in increasing theeffectiveness of relational capability and open innovationin improving financial performance. Those firms follow-ing open innovation, maintaining a relational capability,and either lean or rich in resource slack outperformedfirms possessing moderate levels of resource slack. Thisthreshold-like effect suggests firms must maintain someminimal level of resource slack in order to be flexibleenough to respond to external opportunities. Measures,tables, figures, and references are available upon request.

For further information contact:Sanjay R. SisodiyaUniversity of Idaho875 Campus DriveP.O. Box 443161

Moscow, ID 83844–3161Phone: 208.885.0267



WHICH INFORMATION SOURCES HELP COMPANIES TO GAINSUPERIOR PRODUCT PROGRAM INNOVATIVENESS?INSIGHTS FROM COMPANY AND CUSTOMER DATA

Ruth Stock-Homburg, Darmstadt University of Technology, GermanyLars Herrmann, Darmstadt University of Technology, Germany

SUMMARY

“Innovate or die” has been a slogan in many U.S.companies in order to gain sustained competitive advan-tage. While incumbents that fail to innovate face the riskof losing their industry leadership position, relativelysmall companies can increase their market share throughproduct innovations (Srinivasan, Lilien, and Rangaswamy2002; Tellis, Prabhu, and Chandy 2009). The acquisitionof information is seen as important driver for the genera-tion of product innovations (Li and Calantone 1998;Rindfleisch and Moorman 2001). Wadhwa and Kotha(2006) argue that companies can acquire informationfrom different sources regarding product innovations.However, empirical research on the link between infor-mation and product program innovativeness has focusedmainly on isolated sources of information, for examplecustomers (e.g., Fang 2008) or cooperation between com-panies (e.g., Rindfleisch and Moorman 2001). Further-more, there is little known under which conditions infor-mation acquisition from different sources affects productprogram innovativeness.

Transactions cost analysis (TCA) provides a theoreti-cal framework for studying information acquisition fromdifferent sources. In terms of information acquisition,companies can choose between three alternative coordi-nation mechanisms: market, hierarchy, and hybrid forms.In the context of innovation generation, customers (Fang2008; Veldhuizen, Hultink, Griffin 2006) and experts(Shu,Wong, Lee 2005; Zahra and Nielsen 2002) areidentified as important market sources. Regarding hierar-chy, employees have been examined as relevant source ofinnovation-related information (Zahra and Nielsen 2002).The hybrid mechanism is typically represented by coop-eration between the company and other organizations(White and Lui 2005).

Transaction cost analysis provides valuable insightabout conditions, under which the aforementioned sourcesof information affect product program innovativenesspositively or negatively. Specifically, efficiency of thethree coordination mechanisms market, hierarchy, and

hybrid form depends on the degree of uncertainty andasset specificity (Rindfleisch and Heide 1997). Thoughexchange is a key task of marketing (Rindfleisch andHeide 1997), TCA has been applied scarcely in researchon product innovations. Thus, an important contributionof the paper is the application of TCA on product innova-tion research.

The paper entails a model consisting of antecedentsand performance outcomes of product programinnovativeness. The antecedents comprise customers,experts, cooperation, employees as sources of informa-tion acquisition. Additionally, our framework includesmoderating variables, namely market turbulence andinnovation-related human asset specificity. Drawing ontransaction cost analysis, it is argued that market turbu-lence and innovation-related human asset specificitymoderate the link between different sources of informa-tion acquisition and product program innovativeness.

To test our hypotheses, we conducted a large-scale,quantitative survey among companies in manufacturingand service industries. A particular contribution of thispaper is the triadic design of our survey. We collected datafrom marketing managers, R&D managers as well ascustomer companies, whereas marketing managers andR&D managers were selected from the same suppliercompany. We were able to obtain 109 triads. This is oneof the first studies providing an analysis of triadic data inthe context of product innovation.

We used structural equation modeling to analyze ourconceptual model. Results from 109 triads indicate that inthe case of high market turbulence, customers and expertsare the most promising sources of information acquisitionin order to achieve superior product programinnovativeness. In contrast, results reveal that informationacquisition from employees improves product programinnovativeness when innovation-related human assetspecificity is high. Furthermore, analysis shows asignificant effect of product program innovativeness onmarket performance. References are available uponrequest.


For further information contact:Ruth Maria Stock-Homburg

Marketing and Human Resource ManagementDarmstadt University of Technology

Hochschulstr. 1Germany

Phone: +49.6151.16.7322Fax: +49.6151.16.7320



HE WHO DIES WITH THE MOST ALTERNATIVE FUEL WINS: ANALTERNATIVE USE OF TERROR MANAGEMENT THEORY

Christine Page, Skidmore College, Saratoga SpringsKrystina Smith, Skidmore College, Saratoga Springs

SUMMARY

Since the Industrial Revolution, the burning of fossilfuels such as coal and oil has significantly increased theemission of greenhouse gasses including carbon dioxide,water vapor, methane, nitrous oxide, and ozone (Lentonand Schellnhuber 2007). In turn, the rise in greenhousegasses has caused the “greenhouse effect” which is chang-ing Earth drastically with severe effects including theretreat of glaciers, increased intensity of extreme weatherevents (e.g., tsunamis), drought, and species extinction.To curb the effects of global warming, climate expertsargue that it is critical that humans change their energyconsumption from fossil fuels to alternative sources ofenergy (Davidson and Robson 2007).

In this study, the efficacy of terror managementtheory as an effective promotional tool to influence con-sumers’ attitudes and behaviors toward an alternative fuelis examined. For the purposes of this study, BioHeat isused as the alternative fuel of interest as BioHeat, (1) canbe used efficiently in residential heating boilers, with noadditional parts needed, and (2) provides good qualityheat and performs consistently (National Biodiesel Board2009).

Terror Management Theory

Terror management theory (TMT) is a psychologicaltheory that studies how the anxiety of one’s own deathaffects individuals. According to TMT people managetheir death-related anxiety with an anxiety buffer, madeup of two components, an individual’s cultural worldview(an individuals concept of reality constructed by a groupof people to which the individual belongs) and his/her selfesteem (e.g., Greenberg et al. 1986).

The two components of the anxiety buffer haveresulted in two basic hypotheses. The first is the anxiety-buffer hypothesis which hypothesizes that strengtheningan individual’s self esteem should relieve anxiety due todeath related thoughts. The second hypothesis, and theone that is manipulated in this study, is the mortalitysalience hypothesis. This hypothesis states that remindingpeople of their death will increase the need for structure intheir cultural worldview, leading the individual to havemore positive reactions toward people and things thatsupport their worldview (Pyszczynski et al. 1997).

In Western societies, where “cash is king,” it has beenargued that materialism defines many people’s worldview;possessing luxury items bolsters one’s sense of mortalityand value in their culture (Solomon et al. 2004). Corrobo-rating this, studies have found that after experiencing amortality salience induction, people increase their materi-alistic pursuits as a way of sustaining their worldview orbelief that they are worthy within their culture (e.g.,Mandel and Heine 1999; Choi et al. 2007).

While alternative fuels may not necessarily be viewedas a luxury item, when faced with mortality consumersmay still desire to purchase such fuels as they can beviewed as “green” by their peers. “Being green” hasbecome a trendy lifestyle that many look to as an increasein status. Hence, we hypothesize here that individualswho are manipulated to think of about death will reactmore favorably to a materialistic representation of BioHeat,than those who are not directed to contemplate death.

Experiment

The experiment employed a 2 (mortality salience vs.control) X 2 (materialistic BioHeat ad vs. non materialis-tic BioHeat ad) design. Participants were 80 adults, 20 percondition (who were randomly assigned). The averageage range of participants was between 46–50, and theaverage annual household income group ranged from$90,000 – $99,000.

Results

In brief, results revealed no significant main for eitherbetween subjects factor; mortality salience and ad type.The results did reveal, however, a significant mortalitysalience X ad type interaction. As predicted, mortalitysalient participants’ interest in BioHeat was significantlygreater when they viewed the materialistic BioHeat adthan when they viewed the non-materialistic BioHeat ad.In contrast, non-mortality salient participants expressedgreater interest in BioHeat when viewing the non-materialistic BioHeat ad than when they viewed the ma-terialistic bio-heat ad.

These results are interesting in that they align withpast TMT research regarding the positive impact of mor-tality salience on a consumer’s desire to purchase luxurygoods (e.g., Choi et al. 2007; Fransen et al. 2008). Yet in


this instance, the luxury good was alternative fuel, aninconspicuously consumed (more utilitarian) product. Ona theoretical level, this study adds to the large body ofresearch on TMT’s effect on consumer behavior. To theauthors’ knowledge, no studies have investigated theimpact of TMT on commodities or more inconspicuouslyconsumed products. Hence, a contribution of this study is

that it strongly suggests that TMT, and specifically induc-ing thoughts of one’s own mortality can significantlyinfluence consumer’s purchase intentions of all items,luxury and non luxury. This opens up an entire sector ofpurchasing behavior that has not previously been exam-ined using TMT. References are available upon request.

For further information contact:Christine Page

Skidmore College815 North Broadway

Saratoga Springs, NY 12866Phone: 518.580.5107



CUSTOMER CONFUSION AND PRODUCT LINE SIZE: IMPLICATIONSFOR PRODUCT LINE MANAGEMENT

Sabine Kuester, University of Mannheim, GermanySebastian Buys, University of Mannheim, Germany

SUMMARY

In a competitive environment, product portfolio man-agement remains a critically important challenge for com-panies (Hauser, Tellis, and Griffin 2006). Often, compa-nies try to handle this challenge by expanding theirproduct lines (Hoch, Bradlow, and Wansink 1999). How-ever, the advantageousness of large product lines has beenquestioned by researchers for customers and companiesalike (Gourville and Soman 2005). Extended productlines can overload customers with information. This oftenmakes it more difficult for them to find their favoredproduct (Fasolo et al. 2009). In this regard customerconfusion can appear. Customer confusion is predomi-nantly the consequence of information processing errorswhich are caused by information overload (e.g., Leek andChansawatkit 2006). It can be the reason for undesirablecustomer behavior such as rescheduling and cancellationof purchase decisions (e.g., Mitchell and Papavassiliou1999).

Despite the relevance of customer confusion forcompanies’ economic success research with regard tonegative post-purchase consequences of customer confu-sion is still lacking. Therefore, this study will empiricallyanalyze the occurrence of customer confusion within apurchase process. Furthermore, we will analyze its mostcrucial consequences for companies, more specifically,its influence on customers’ purchase intention and wordof mouth intention. Additionally, we will test a tool thatshould help companies to reduce customer confusion.

The environmental psychological Mehrabian-Russelmodel provides an explanatory basis for several customerbehavior responses to customer confusion. It consists of atwo-stage process of customer behavior (Mehrabian andRussell 1974). In the first stage, exogenous (i.e., environ-mental stimuli as perceived by individuals) and endog-enous effects induce a positive or negative emotionalreaction of an individual. In the second stage, a positiveemotional reaction leads to an approach behavior while anegative reaction results in avoidance behavior. Thesereactions manifest themselves in the degree of theindividual’s motivation to act or communicate (Donovanet al. 1994). The concept of information overload isappropriate to explain the emergence of customer confu-sion (e.g., Leek and Chansawatkit 2006). The exceedanceof humans’ limited information processing capacities due

to the large number of information carrying products canlead to confusion and suboptimal decision making(Malhotra 1982).

In our experiment participants were asked to imaginea scenario in which they decided to buy a jar of jam. Thestimuli were different product lines of jam. We manipu-lated these product lines regarding the product numberand the product complexity. Based on a literature reviewand the results of 20 store-checks we chose 6, 12, and 27products to represent the different size categories ofproduct lines. The degree of product complexity wasmodeled by using two different numbers of attributes (3,6) which describe the products (Burnham, Frels, andMahajan 2003). We also manipulated the extent of infor-mation about the product line which was provided to thecustomers. One group of the respondents was given anattribute-based consumption vocabulary which containedall product attributes including the corresponding valuesprevious to the actual purchase decision. The other groupdid not receive any assistance in the purchasing process.The two groups consisted of an equal number of respon-dents. This procedure is suggested by West, Brown, andHoch (1996).

With our study we considerably add to existingresearch on the phenomenon of customer confusion.Analyzing experimental data of 1,128 consumers wefound that with increasing product line size customerconfusion increases as well. We further detected that theconfusion causing effect of product line size increaseswith increasing product complexity. Moreover, we iden-tified customer confusion as a total mediator of the nega-tive effects of product line size on purchase intention andpositive word of mouth intention. Furthermore, we foundthat customers with a high level of product knowledgeperceived less customer confusion than those with lowerlevels. On the contrary, customers who tend to maximizewere more strongly suffering from confusion than thosewho are satisficing. Finally, a consumption vocabularyturned out to be a suitable instrument to reduce customerconfusion. In addition, this managerial tool worked moreeffectively in the experimental conditions with large prod-uct lines.

Our findings have important implications for productline management. First of all, product line managers haveto be aware of the potential problematic of customer


confusion when expanding their product lines. Conse-quently, product line decisions have to be reviewed froma customer’s perspective to investigate undesirable out-comes. However, if customer confusion occurs, counter-measures must be taken by managers. In this context adownsizing of the product line size can be recommended.Finally, measures which help customers to obtain a suit-

able amount of product information within the purchasesituation and to learn their preferences regarding theproduct attributes should be pursued (Huffman and Kahn1998). Since our study is restricted to the product categoryof jam, future research could enhance the generalizabilityof our findings by studying other product categories.References are available upon request.

For further information contact:Sebastian Buys

Department of MarketingUniversity of Mannheim

68131 MannheimGermany

Phone: +49.621.181.2092Fax: +49.621.181.2398



HOW DO CONSUMERS REACT TO SERVICE MERGERS?

Melinda Andrews, The University of Southern Mississippi, HattiesburgRonald Goldsmith, Florida State University, Tallahassee

SUMMARY

Mergers are a common worldwide strategic businessalternative. Merger deals are especially prevalent in theservices sector, with several multi-billion dollar servicemergers occurring in the past decade (e.g., Sirius and XMRadio in 2008; Northwest Airlines and Delta in 2008;Wachovia and Wells Fargo in 2008; Alltel and TPG in2007; AT&T and Bellsouth in 2006). Often during themerger process, managers become overly focused onelements such as ongoing negotiations, legal and regula-tory issues, and financial details and overlook the con-cerns and perceptions of important stakeholder groupssuch as consumers (Balmer and Dinnie 1999; Kumar andBlomqvist 2004). As a result, consumers may experienceuncertainties about their future relationship with the merg-ing firm and eventually switch to a competitor. Therefore,it is vital to understand the effects of service mergers onconsumers. The present study examines pre-merger brandvalence in the context of balance theory. Pre-mergerbrand attitudes toward the consumer’s own brand andservice quality expectations are hypothesized to affect theconsumer’s perception of a merged service brand.

The influence of pre-merger brand valence on post-merger brand attitudes and service quality expectations isassessed via an online experiment. Perceived fit of themerging brands is also measured. In an effort to maintainexperimental control in this study, fictitious telecommu-nications brands are used in the study to control for priorexperience. Balance theory provides the foundation forthe 2 x 2 between-subjects design in which the valence ofbrands is manipulated via fictitious consumer reviews.MANOVA is used to analyze the data and proposedhypotheses.

As expected, the main analyses indicate consumersreact more positively toward service mergers involvingtwo positive brands than mergers involving a negativebrand. Vice-versa, consumers generally reacted morenegatively toward service mergers involving two nega-tive brands than mergers involving a positive brand.Based on balance theory, it is hypothesized that when aconsumer has a negative brand, a merger might signal apossible positive change even when the merger is withanother negative brand because of the balanced nature ofthe attitudes toward the merging firms. In addition, theconsumer could hypothetically think, “could the service

get any worse?” Although the main analyses are not fullyconsistent with balance theory, the subsequent analysesinvolving perceived fit do appear to support the previ-ously mentioned notions when a consumer’s own nega-tive brand is involved in a merger. In contrast, customersreact more negatively when their own positive brand isinvolved in a merger.

Obviously, consumers have varying reactionsdepending on the valence of the merging brands. Amerger involving a negative brand damages the image ofthe merged brand regardless of how positive the otherbrand is perceived before the merger. The results indicatethat mergers involving two positive valence brands out-perform mergers involving a negative brand. Moreover,when two positive brands merge it leads to higher levelsof post-merger brand attitudes and service quality expec-tations. Thus, mergers of two positive service firms do notsuffer as much as mergers involving a negative firm,which supports Basil and Herr’s (2006) notion of a “bal-ance boost.” In addition, the results for the mergersinvolving the negative valence brands also support theiridea of a “contamination effect.” Contrary to balancetheory, the “contamination” is even more extreme in themergers involving two negative brands with the respon-dents having the lowest levels of post-merger brandattitude and service quality expectations. Althoughrespondents perceive two negatively valenced servicefirms as fitting together, which indicates balance, themerging firms are definitely not deemed a pleasant pair.

Service mergers involving a negative and positivevalence brand may be more common in the real worldbecause the successful firm is in a position to take over theflailing firm. Thus, it is critical to understand the influenceof a merger on the current customers of either a negativeor a positive firm with a firm of opposite valence. Theresults of the mergers involving the opposite valencebrands provide vital managerial information. Customersfond of their service provider have adverse reactions to amerger with a negative brand with lower levels of post-merger brand attitudes and service quality expectations.Thus, the pre-merger positive brand suffers as a conse-quence of the merger. On the other hand, the pre-mergernegative brand fares better when merged with a positivebrand. These customers have more positive post-mergerbrand attitudes and service quality expectations.


REFERENCES

Balmer, John M.T. and Keith Dinnie (1999), “CorporateIdentity and Corporate Communications: The Anti-dote to Merger Madness,” Corporate Communica-tions: An International Journal, 4 (4), 182–92.

Basil, Debra Z. and Paul M. Herr (2006), “Attitudinal

Balance and Cause-Related Marketing: An Empiri-cal Applications of Balance Theory,” Journal ofConsumer Psychology, 16 (4), 391–403.

Kumar, Shailendra and Kristiane Hansted Blomqvist(2004), “Mergers and Acquisitions: Making BrandEquity a Key Factor in M&A Decision-Making,”Strategy & Leadership, 32 (2), 20–27.

For further information contact:Melinda L. Andrews

The University of Southern Mississippi118 College Drive #5091

Hattiesburg, MS 39406–0001Phone: 601.266.4629



SLEEPING WITH THE ENEMY? ATTRIBUTIONS TRIGGERED BY AN“OPEN ARCHITECTURE” SALES MODEL

Winfried Daun, University of St. Gallen, SwitzerlandDaniel Wentzel, University of St. Gallen, Switzerland

Torsten Tomczak, University of St. Gallen, Switzerland

SUMMARY

In recent years, many companies have introduced asales model that is commonly referred to as an “openarchitecture” model. In this model, companies that areboth manufacturers and distributors of their own productsand services open their distribution channels to their owncompetitors (Kelleher and Wood 2007). For instance,many retail banks do not only sell their own investmentproducts (such as mutual funds) to their clients, but alsothose of other banks. Integrating competitor products intoa company’s offering bears obvious risks, such as loss ofmarket share, dilution of one’s own value proposition, anderosion of brand image. These potential downsides wouldneed to be outweighed by a strong and positive overallcustomer reaction, such as an increase in satisfaction orloyalty. However, the current literature has not examinedconsumers’ reactions to open architectures in detail. Inour paper, we address this gap in the literature and exam-ine whether an open architecture model triggers positiveor negative perceptions. Specifically, we (1) map out a“typology”specific cues that are likely to cause or influ-ence these perceptions.

Attributions Triggered by an “Open Architecture”Model

Attribution theory has been used to account for thedifferent views customers might form of an “open archi-tecture” offering. If a company has been selling only itsown products, customers will have formed an expectationabout the company’s offering – namely, that it does notinclude any products from other brands or manufacturers.If the company then shifts to an open architecture model,this will not be in line with what customers have come toexpect. Researchers have shown that “disconfirmedexpectancies” are one of the main causes for attributionalthinking (e.g., Hastie 1984; Hunt, Domzal, and Kernan1982; Weiner 1985). We can therefore assume that theintroduction of an open architecture triggers attributionalthinking.

We expand this idea by arguing that customers maygenerate both favorable and unfavorable attributions inresponse to an open architecture model. On the positiveside, research has often found that a violation of expect-ancies has favorable effects. For instance, disconfirmed

expectancies may lead to higher acceptance of a salesmessage (Hunt, Domzal, and Kernan 1982) and morefavorable attributions regarding the motives of a salesper-son (Smith and Hunt 1987). However, attributions canalso be unfavorable – this is frequently the case whenconsumers are suspicious about a salesperson’s behaviorand start thinking about ulterior, undisclosed motives(Fein 1996; Friestad and Wright 1994). This line ofreasoning extends the range and complexity of potentialattributions. Rather than attributing the cause for an openarchitecture to benevolent motives (e.g., “better service tothe client”), some customers might perceive the offeringof third-party products as part of a sophisticated scheme.

In order to improve their sales effectiveness,companies will want to foster the instigation of favorableattributions and to avoid or counteract unfavorableattributions. With this goal in mind, we sought to understandwhether there are specific “cues” that influence whetherattributions are positive or negative. For instance, Campbelland Kirmani (2000) suggest that persuasion motives cantranspire in the sales episode itself, through the agent’sbehavior. In a similar vein, Fein (1996) argues that acustomer might recognize cues in a salesperson’s behaviorthat seems to hint directly or indirectly at certain ulteriormotives. We also have reason to believe that customersmay treat the very fact that a company includes third-partyproducts as a powerful cue – namely, when they perceivethis behavior either as non-typical (Sujan, Bettman, andSujan 1986) or as opposing the company’s best interest(e.g., Koeske and Crano 1968; Smith and Hunt 1978;Etgar and Goodwin 1982). Overall, it seems very likelythat cues will have an influence on whether unfavorable orfavorable attributions prevail in the attitude-formingprocess.

Method, Results, and Discussion

We conducted a qualitative study to develop a typol-ogy of the attributions caused by an open architecturemodel and the cues that might influence the polarity ofthese attributions (favorable vs. unfavorable). Thisexploratory step is a critical precondition for the quantita-tive, confirmatory research we will undertake at a laterpoint. In our research, we focused on the Swiss bankingindustry which has recently introduced open architecturemodels. Specifically, we conducted 20 semi-structured


interviews that lasted between 40 to 70 minutes. Allrespondents were banking clients who had been told thatthey participated in a research study on their generalexperiences with banking services.

The results provide support for our theoretical rea-soning. Firstly, the vast majority of our respondentsexpected that their bank would only offer them its owninvestment products to choose from. When this expect-ancy was disconfirmed, respondents actively searched forreasons behind the open architecture and produced very

different attributions. Finally, our qualitative results sug-gest a number of cues that seem to impact the strength,location and polarity of attributions. These results are notonly of theoretical interest, but are also of relevance topractitioners. By actively managing the cues identified inour research, managers may present their own and theircompetitors’ products in a way that fosters favorableattributions and inhibits unfavorable ones. This, in turn,may lead to higher customer satisfaction and loyalty.References are available upon request.

For further information contact:Winfried Daun

Center for Customer InsightUniversity of St. Gallen

Rosenbergstr. 51 / CH-9000 St. GallenSwitzerland

Phone: +41(0)44.234.5798Fax: +41(0)44.234.5337



A SUSTAINABILITY TYPOLOGY FOR NONPROFIT ORGANIZATIONS:INHERENT CHALLENGES AND NORMATIVE STRATEGIES

Robert E. McDonald, Texas Tech University, LubbockJay Weerawardena, University of Queensland, Australia

Sreedhar Madhavaram, Cleveland State UniversityGillian Sullivan Mort, La Trobe University, Australia

SUMMARY

Nonprofit organizations emerge to serve a societalneed that has not been met by either the business orgovernment sectors. Insufficient profitability and insuffi-cient public benefit are likely reasons for the failure ofthese two larger sectors of society to address the need.Therefore, nonprofits enter into environments that bytheir nature present barriers to organizational success. Inorder to survive and thrive, a nonprofit organization mustachieve fiscal sustainability and remain relevant byaddressing a significantly important social need. Thispaper proposes a diagnostic typology of nonprofit organi-zation sustainability based on these two dimensions andprescribes strategies to achieve sustainability.

Nonprofit organizations typically emerge to satisfyneeds that neither the business nor public sectors satisfy(Etzioni 1972; Kotler and Murray 1975; Pestoff 1992).Usually, the business sector does not satisfy the needbecause it cannot do so profitably. Additionally, govern-ment cannot satisfy the need because there is not enoughpublic support to do so. Therefore, nonprofit organiza-tions often pursue missions that are neither financially norpolitically viable. Despite the fundamental challengesthat the sector faces, the evidence indicates that thenumber of nonprofits and their collective eco-nomic impact is increasing. However, there is a consensusamong researchers, policy planners, and professionalmanagers that nonprofits operate in an increasingly com-petitive environment (Chetkovich and Frumkin 2003).

Given the inherent challenges of nonprofits and thisincreased competitiveness, how do these organizationssustain themselves as they address societal needs thatneither the private nor government sectors are able toaddress? As nonprofits cannot rely solely on operationalprofit, and they lack government’s taxing authority(Hansmann 1980), nonprofits employ a unique opera-tional model in which success depends on multiple stake-holders to provide necessary resources to function.

Long-term success requires nonprofit organizationsto adopt strategies to ensure that they satisfy significant

social needs in a fiscally sustainable manner. This paperpresents a typology of sustainability based on the twodimensions that lead to the emergence of nonprofits: fiscalsustainability and significant social need. Then, drawingfrom the diverse extant literature, we suggest normativestrategies to help nonprofits move toward the positiveends of the typology’s two dimensions.

The typology presented here is intended to be usedstrategically as a diagnostic and prescriptive tool based onorganizational performance measured both financiallyand socially, i.e., mission outcome. These are the “doublebottom line” metrics commonly demanded of nonprofits(Chetkovich and Frumkin 2003; Sawhill and Williamson2001). We suggest that nonprofit organizations can beclassified based on the long term financial viability of theorganization’s operations and on its social viability, i.e.,the perceived enduring significance of the social need it isattempting to address.

In the proposed typology, nonprofit entities can becategorized by the strength or weakness of financialresources and performance. In addition to the fiscaldimension, nonprofit organizations can be ranked as highor low along the dimension of addressing a perceivedsignificant social need. Based on the organization’s cat-egorization, we suggest normative strategies that will helpit to improve or sustain its operations.

Proposed strategies to achieve fiscal sustainabilityinclude both revenue enhancing and cost reducing strate-gies. Other proposed strategies focus on a need-basedsustainability that is achieved by remaining relevant tosociety, through effective service delivery and the man-agement of society’s perceptions of the organization’scontribution. By selecting the appropriate normative strat-egies based on where the organization fits in the typology,the nonprofit may hope to improve its long term survivaland success.

Implications for professional nonprofit managers andresearchers are presented, as are suggestions for futureresearch in the area. References are available upon request.


For further information contact:Robert E. McDonald

Rawls College of Business, MS 42101Texas Tech University

Lubbock, TX 79409–2101Phone: 806.742.3162



SUSTAINABILITY AND INSURANCE: AN APPROACH TOSUSTAINABILITY SEGMENTATION

Mareike Bodderas, University of St. Gallen, Switzerland

ABSTRACT

This study integrates customers’ environmental andsocial concerns into an insurer’s pricing and marketingprocess. The established procedure measures customers’sustainability: (1) perceived knowledge of sustainability;(2) attitude toward sustainability; and (3) sustainablysensitive behavior. Insurers can respond to sustainablysensitive customers and consider their (1) security orien-tation, (2) insurance orientation, and (3) claims history.

INTRODUCTION

Sustainability is more than a recent trend. Persons areincreasingly expressing their opinion and lifestyle throughsustainable attitude and behavior. They demonstrate theirposition in their sustainable sensitive beliefs, in the prod-ucts that they buy, or in the transportation that they choose(Autio et al. 2009; Krystallis et al. 2008; Dunlap andCatton 1979; Schwepker and Cornwell 1999; Larocheet al. 2001; Getzner and Grabner-Kräuter 2004). In thiscontext, sustainability seems to become a very importantbusiness issue (Kirkpatrick 1990). However, it is a chal-lenging task for companies to consider sustainable atti-tude of their customers. They need to identify customers’sustainable concerns, respond to them, and manage theirneeds effectively.

In the context of the insurance industry, securityorientation, insurance orientation, and claims history areelements that companies must take into account whenmanaging sustainable customers (Anderson and Ander-son 2009). By predicting the influence of sustainableattitude on customers’ behavior, insurance companies canadapt their customer relationship strategy and adjust prod-ucts and services. Accounting for customers’ sustainableconcerns and employing sustainability attributes can cre-ate a competitive advantage for insurers. Sustainabilityattributes differ from standard actuarial variables and,therefore, can provide additional useful underwritinginformation. Thus, a reliable and valid segmentation frame-work based on sustainable factors can assist insurers indeveloping a strong sustainable image, identifying sus-tainable profile, managing sustainable customers, andpredicting losses.

Although several studies investigate alternative at-tributes such as, e.g., personal characteristics, as classifi-cation variables (e.g., Brocket and Golden 2007), noempirical approach addresses these opportunities andchallenges for insurers dealing with sustainable custom-

ers. The present paper aims to adapt the, in scientificliterature established sustainability measurement, to theinsurance industry. In particular, our study, first, predictsthe impact of sustainable knowledge, attitude, and behav-ior on customers’ insurance orientation and risk profile,and second, groups insurance customers with similarsustainability and loss profile. A sustainability modelbased on the theory of planned behavior (TPB) (Fishbeinand Ajzen 1975; Ajzen and Fishbein 1980) is developedand extended by insurance-specific elements. The finalmodel comprises six dimensions: (1) perceived knowl-edge of sustainability, (2) attitude toward sustainability,(3) sustainably sensitive behavior, (4) security orienta-tion, (5) insurance orientation, and (6) claims history.

So far, no empirical study has identified the sustain-able attributes of customers in the insurance industry.Thus, this paper investigates the sustainable profile ofinsurance customers in terms of their knowledge, attitude,and behavior. Therefore, this study empirically tests theproposed model of insurance and sustainability on a largedataset. We conduct an empirical survey of insurancecustomers in Germany. This study starts by investigatingthe relationship among customers’ sustainable concerns,security orientation, insurance orientation, and claimshistory. The relationship between the sustainability andinsurance dimensions is predicted by linear regressionanalysis. Latent Class Analysis (LCA) is then performedto identify the response patterns of individuals with simi-lar sustainability profiles. Moreover, a regression analysisfor four classes is conducted, using claims statistics asindependent variables and sustainability indicators aspredictors. Four classes were labeled as “attitude-oriented,”“mobility-oriented,” “behavioral-knowledge,” and“mobility-rhetoric.” The quality of derived measures isassessed by means of model fit and by identification ofsustainable predictors that discriminate among the classes.

The paper is structured as follows. The next sectionsummarizes the research on customer segmentation basedon sustainability characteristics, and proposes an insur-ance-specific sustainability model. The proposed modelis empirically tested in the third section. A summary ofresults is given in the final section.

TOWARD SUSTAINABILITY SEGMENTATIONIN THE INSURANCE INDUSTRY

The rationale for segmentation is that customers canbe grouped based on their needs and behavior (McDonald1995; Piercy and Morgan 1993). Members of groups are


likely to show similar responses to marketing activities orsimilar behavior in creating losses. According to eco-nomic pricing theory, through market segmentation, profitcan be maximized when pricing levels discriminate amongsegments (Frank et al. 1972). Empirical research showsthat companies that apply segmentation perform better,because they can fine-tune their customer offerings (Beaneand Ennis 1987; Wind 1978). Marketing distinguishesqualification from attractiveness as segmentation criteria(Dibb 1999). These criteria are relatively simple to inter-pret, and they allow for an evaluation of the segmentationprocess: robustness of segmentation output and assess-ment of business issues, such as the profitability of asegment. Despite the attention that the segmentation ofsustainable concern receives in the marketing literature,this approach has not yet found its way into the insuranceliterature. However, an important development in theinsurance industry is the use of alternative attributes aspsychobehavioral variables, to classify insurance custom-ers (Brocket and Golden 2007).

Alternative Segmentation Variables in the InsuranceIndustry

Recent developments in the field of predicting lossesfocus on additional alternative attributes, such as biologi-cal, psychological, and behavioral attributes to forecastrisky automobile driving and insured losses (Brockett andGolden 2007; Monaghan 2000; Kellison et al. 2003;Miller and Smith 2003). Brockett and Golden (2007)argue why individuals’ bio-psychological constitutioninfluences driving performance and financial behavior.As early as 1949, Tillman and Hobbs (1949) examinedlifestyle variables and stated”: . . . a man drives as he lives”(p. 329). They identified the relationship between thepersonality characteristics, social background, and acci-dent rates of taxi drivers. High-accident drivers had simi-lar life histories, including emotional instability, disre-spect for authority, school-related problems, juvenilecriminal behavior, and short-term employment. Similarresults are obtained from other studies in this field (Shiharand Compton 2004; Sommers et al. 2000; Arnett 1990;Grey et al. 1989). High-accident drivers are characterizedby aggressive driving, impulsive behavior, disregard forothers, and disrespect for authority. They take more riskswhile driving, and are more likely to participate in sociallydeviant behavior, such as having poor personal relation-ships, driving for emotional reasons, and having notice-ably more moving violations. Further studies identifyseveral personal characteristics that influence drivingbehavior and involvement in driving accidents. Forexample, risk perception is correlated with anxiety, sensa-tion seeking, and aggression (Jonah and Dawson 1987).Thus, drivers with high-risk acceptance are more likely tobe involved in accidents than are responsible drivers.Finally, young drivers are found to perceive less danger in

driving situations and take greater risks than older drivers(Wison and Jonah 1988).

Demographic Segmentation Variables in Marketing

In the literature on sustainable consumer behavior, or“green” marketing, the profiles of sustainable customersare based on traditional variables (Kilbourne andBeckmann 1998). Empirical studies of sustainable seg-mentation attempt to identify the demographic variablesthat determine environmental and social attitude andbehavior (Dunlap and Catton 1979; Schwepker andCornwell 1999; Laroche et al. 2001; Getzner and Grabner-Kräuter 2004). These variables allow for efficient seg-mentation of customers. For example, the variable of“age” is explored, and the assumption that “green” con-sumers are younger has been supported (e.g., Andersonand Cunningham 1972; Kinnera et al. 1974; Aaker andBagozzi 1982). Nevertheless, the goodness of these vari-ables is controversial because the results of the empiricalstudies are inconsistent. For instance, Roberts (1996)identifies the environmentally conscious consumer asolder than the average consumer. Gender, as anotherdemographic variable, has influence on ecological aware-ness; women are more ecologically conscious than menare (Anderson and Cunningham 1972; Stern et al. 1993;Laroch et al. 2001). MacDonald and Hara (1994) disprovethis. Nevertheless, numerous authors state that demo-graphic variables do not sufficiently explain the ecologi-cally friendly concerns of customers. As a result, theyfavor psychographic variables such as attitude, knowl-edge, and behavior (e.g., Schlegelmilch et al. 1996;Schwepker and Cornwell 1991; Straughan and Roberts1999).

Psychographic Variables for Sustainability Segmen-tation

Employing psychographic variables to classify sus-tainable customers is founded on TPB. According to thistheory, personality traits, personal values, and lifestyleare predictors of actual behavior (Fishbein and Ajzen1975; Ajzen and Fishbein 1980). Simply claiming to be“sustainable” is no longer enough to predict a consistentsustainable profile. In this study, the sustainability profileconsists of three components: (1) knowledge of sustain-able issues, (2) attitude toward sustainability, and (3)sustainably sensitive behavior.

Knowledge

Knowledge is an important predictor in decision-making. In order to be environmentally and socially“sustainable,” people need to understand the consequencesof their behavior. A knowledge component therefore hasto be factored into any accurate measure of sustainability.


Several studies have found that knowledge is a consider-able predictor of environmentally friendly behavior (e.g.,Vining and Ebreo 1990; Chan 1999). However, otherstudies refute these findings. For example, Schlegelmilchet al. (2001) do not find a statistically significant relation-ship between respondents’ knowledge about environ-mental and social topics and their environmental andsocial friendly purchasing behavior. The authors attributethese results to the difficulty of objectively capturingenvironmental knowledge. Another explanation is thatknowledge and belief have distinct influences on behav-ior. For example, customers are unlikely to consume“green” products because they are aware of the shortcom-ings (Getzner and Grabner-Kräuter 2004).

Attitude

Several studies have identified attitude as a goodpredictor of sustainable behavior (Schwepker and Cornwell1991; McCarty and Shrum 1994; Roberts 1996; Chan1999; Laroche et al. 2001). In general, a positive attitudetoward environment leads to respective behavior. In con-sequence, attitudes and values, which refer to a specificperception about environmentalism, are closer to actualbehavior (Dembkowski and Hanmer-Lloyd 1994). Forexample, McCarty and Shrum (1994) find that the moreinconvenient respondents believe that recycling is, theless likely they are to recycle. Krystallis et al. (2008) haveconfirmed consumer values and beliefs as predictors oforganic purchasing.

Behavior

Several studies segment sustainable groups based onthe membership in environmental and social organiza-tions (Murphy 1989), on the recycling of materials, on thepurchase of environmentally friendly products (Larocheet al. 2001; Bohlen et al. 1993), and on investment in“green” shares (Grabner-Kräuter 2004). For example,environmentally conscious behavior is supposed to be animportant predictor for customers’ willingness to pay ahigher price for environmentally friendly products or toabstain from unecological consumption. However, theempirical findings unexpectedly indicate the opposite(Laroche et al. 2001): although respondents agree torecycle and buy environmentally friendly products, theyare not willing to pay a higher price for them. Autio et al.(2009) have examined the ways in which young consum-ers construct their images of green consumerism. Theyfind that young people need to believe that their actionsreally make a difference.

This ambiguous result might be attributed to socialpressure and to the researcher response bias, for example,the desire to receive an environmentally friendly answerfrom respondents (Laroche et al. 2001). However, Getzerand Grabner-Kräuter (2004) argue that the desirable

response bias is a general problem when investigating therelationship between attitude and behavior. Therefore, thewillingness of respondents to behave in a certain wayshould be interpreted as a general preference rather thanas a predictor. For this reason, the present study accountsfor objectively verifiable behavior (e.g., actual use ofpublic transport) in addition to the general behavioralattitude (e.g., expressed intention to use public transports).

Conceptual Model of Sustainability and Insurance

This empirical study explores the influence ofsustainability on insurance customers by: (1) investigat-ing the influence of distinct sustainability dimension oninsurance dimension; and (2) identifying the target groupof sustainable customers (see Figure 1). As outlinedabove, this study employs psychographic criteria such asknowledge, attitude, and behavior to predict customers’sustainability profile. Therefore, the final conceptual modelis developed as a TPB model, that the assumingsustainability dimension is an important predictor of cus-tomers’ insurance profile (Fishbein and Ajzen 1975;Ajzen and Fishbein 1980). Thus, the effect of customers’sustainable orientation on their behavior, which is ofrelevance to insurers, is determined (e.g., readiness toassume risk, insurance orientation, product individualiza-tion, and claims history). Further, by segmenting indi-viduals who exhibit similarities in sustainable knowledge,attitude, and behavior (sustainably sensitive mobility andsustainably sensitive behavior), companies can managethe target group, respond to its needs, and predict losses ofsustainable customers.

SUSTAINABILITY AND INSURANCE

Our findings are reported in three sections. First, theempirical procedure and the participants of the study aredescribed. Second, the influence of sustainability on insur-ance customers is investigated. We determine customers’risk and insurance orientation and their claims history byusing linear regression analysis. Third, in order to identifythe target group in the insurance industry, we segmentinsurance customers according to their sustainability pro-file by using Latent Class Analysis (LCA).

Participants and Procedure

Our sample is taken from German insurance custom-ers for automobile and household insurance. Of the 15,000prospective respondents, 1,226 returned completed ques-tionnaires (response rate = 8.2%). The participants’ ageranged from 27 to over 60 (27–30 = 17.5%; 31–40 =28.7%; 41–50 = 23.2%; 51–60 = 16.5%; more than 60 =14.1%). The proportion of females was 44.9 percent.

The questionnaire comprises six measures: knowl-edge, attitude, behavioral intention, verifiable behavior,


insurance orientation, and claims history. All measuresconform to common practice in this field of research (e.g.,Diamantopoulos et al. 2003; Dunlap et al. 2000). Influ-ence of Sustainability on Risk and Insurance Orientation

First, we employ explorative factor analysis to testthe internal consistency of the construct variables and toverify the conceptual model, consisting of the sustainabilityand insurance dimensions. As a result of the statisticalidentification of the factor loadings, each of the modeldimensions requires at least two indicators (McDonaldand Ho 2002). In the first step, we include all items of thesurvey, and in the second step, we perform factor analysiswith fewer items. The indicators with low reliabilities,low factor loading, and no apparent relationship to theconstruct dimensions are excluded from further analysis.

Table 1 presents an overview of the included items ofthe final construct and factor weighting. Finally, wecondense 31 selected items of the survey to 11 factors.Thus, the sustainability comprises four factors (sustain-able behavior, sustainable mobility, sustainable attitude,and sustainable knowledge). Three factors represent cus-tomer attitude toward risk and insurance (readiness toassume risk, insurance orientation, and product individu-alization). Furthermore, four factors represent customers’claims history (liability insurance, comprehensive insur-ance coverage, part insurance coverage, and householder’sinsurance coverage). The final model integrates custom-

ers’ sustainable profile, customers’ risk, insurance atti-tude, and their claims history. The sustainability dimen-sions represent the independent variables as drivers towardsecurity orientation, insurance orientation, and claimshistory (details for the result of factor analysis appear inTable 1). We perform linear regression analysis to predictthe causal relationship between sustainability and insur-ance. As hypotheses in the conceptual model (see Fig-ure 1), we determine the effect of customers’ sustainabilityon their insurance profile. In addition, we confirm thegoodness of fit of the model and the statistical significanceof the estimated parameters. We check for goodness of fitby the R-squared. For statistical significance, we check byt-tests of individual parameters. To measure the effect ofindependent variables on dependent variables, we refer tostandardized ß-coefficients.In general, different anteced-ents drive customers’ attitude toward security and insur-ance (see Table 2). More specifically, customers’ readi-ness to assume risk is determined by sustainable attitude(ß = .078). However, the effects of sustainable mobilityand sustainable knowledge are insignificant. In the case ofinsurance orientation, sustainable knowledge appears tobe a negative driver (ß = .089). Testing the influence ofsustainability on product individualization, sustainablevalue (ß = .063) and sustainable knowledge (ß = .07)appear to be positive drivers.Overall, the results confirmthat sustainable orientation of customers’ needs to beconsidered separately. Whereas sustainable mobilityreduces customers’ readiness to assume risk, sustainable

Figure 1: Conceptual model of sustainability and insurance

Demographic criteria

Security and Insurance Orientation

Claims History

Independent Var iables Dependent Variables

Intention Impact

Psychographic criteriaCustomer att itude and behavior toward sustainability: (1) perceived knowledge about sustainability, (2) att itude toward sustainability, (3) sustainable-sensitive mobility (4) sustainable-sensit ive behavior

Sustainability profile Ident ificat ion of dist inct sustainable groups given the response patterns on psychographic criteria

FIGURE 1Conceptual Model of Sustainability and Insurance


TABLE 1Overview of Factor Loading for Sustainability and Insurance Dimensions

Construct Underlying Items Factor Loading KMO

Independent Variables

Sustainability dimensions .721

1 Sustainable behavior Ecological purchase .744Fair-trade purchase .718Shopping in ecostore/local market (obj.) .64Engagement in social/environmental projects (obj.) .484Donation (obj.) .497

2 Sustainable mobility Public transport network .644Sustainable travelling .71Hold travel card (obj.) .483Use local public transport (obj.) .776

3 Sustainable attitude Media reports about climate change .586Governmental contribution to environment .674Abstain from sth. because of environment .636

Individual contribution to environmental protection .736

4 Sustainable knowledge Alternative fuel .612Erosion of ozone layer .631Ecological footprint .498Provision .497

Dependent variables

Insurance dimensions .496

5 Readiness to assume risk Individuals’ risk appraisal .784Individuals’ risk behavior in free time .789

6 Insurance orientation Organization of personal insurance issues .746Trust in insurance companies .778

7 Product individualization Information of personal driving behavior .774Demand of individualized automobile insurance .674

Claims history .528

8 Liability insurance Number of claims .959Amount of claims .961

9 Comprehensive insurance coverage Number of claims .948Amount of claims .950

10 Part insurance coverage Number of claims .949Amount of claims .921

11 Householder’s insurance coverage Number of claims .945Amount of claims .946

Notes: Extraction method: principal component analysis. Rotation method: varimax with Kaiser Normalization. Rotationconverged in four iterations. All factor loadings are significant at the 0% level.

behavior increases it. Customers with high level of sus-tainable knowledge we assume as to be more independent,and thus, in general, less insurance-oriented. On the onehand, they are more likely to organize their insurancethemselves, which is in line with less trust in insurancecompanies. On the other hand, sustainable knowledge and

sustainable mobility increase customers’ demand for in-dividualized products and services. When testing theimpact of customers’ sustainability on their claims his-tory, the results indicate a distinct impact of sustainablebehavior and attitude. Whereas sustainable attitude hassignificant impact on low claims amounts and frequencies


in liability insurance (β = -.099), sustainable behaviorinfluences the claims history for comprehensive insur-ance coverage (β = .177). The results of sustainability andclaims history suggest that, overall, sustainable behaviorstimulates customers’ activity and leads to higher fre-quency and larger claims in motor insurance.

Sustainable Target Groups in the Insurance Industry

In this section, we conduct LCA to identify theresponse patterns of people with similar sustainable atti-tude and behavior. LCA assumes that each individualbelongs to only one class. However, class membershipcannot be observed. It is treated as probabilistic. In gen-eral, LCA assumes that answers to a set of items areindependent of the class membership, but class member-

ship is correlated to individual responses. We run regres-sion models for four classes using claim statistics asindependent variables, sustainability indicators as predic-tors, and demographic items as covariates. First, we testwhether sustainability variables adds useful informationbeyond what is already captured through demographicdata. Therefore, we run three models: First, we run a fullmodel, which includes both, demographic variables ascovariates and attitudinal variables as predictors. Further,we run a restricted model, which excludes the demo-graphic variables as covariates. Finally, we estimate atraditional model, which excludes attitudinal variablesbut accounts for demographic variables as predictors.

The likelihood ratio test reported in Table 4 showsthat the restricted and traditional models, which consis-

TABLE 2Results of the Regression Analysis

Dependent Independent Stand. β CoefficientsVariable Variable (T-Value) F-Value Adj. R2

Security and insurance orientation

Readiness to assume risk Sustainable behavior -.08 (-2.245)** 2.569* .008Sustainable mobility -.028 (-0.775)Sustainable attitude .078 (2.175)**Sustainable knowledge -0.003 (-0.079)

Insurance orientation Sustainable behavior -.037 (-1.024) 2.077** .006Sustainable mobility -.035 (-0.965)Sustainable attitude .017 (0.466)Sustainable knowledge -.089 (-2.482)**

Product individualization Sustainable behavior .033 (0.916) 2.036** .005Sustainable mobility .063 (1.759)*Sustainable attitude .026 (0.725)Sustainable knowledge .07 (1.95)**

Claims history

Liability insurance Sustainable behavior .05 (0.948) 1.204 .013Sustainable mobility .037 (0.694)Sustainable attitude -.099 (-1.874)*Sustainable knowledge 0.0 (-0.01)

Comprehensive insurance coverage Sustainable behavior .177 (3.406)*** 2.077 .011Sustainable mobility -.048 (-0.915)Sustainable attitude -.032 (-0.61)Sustainable knowledge .081 (1.542)

Part insurance coverage Sustainable behavior .039 (0.735) .458 .005Sustainable mobility .002 (0.03)Sustainable attitude -.057 (-1.066)Sustainable knowledge .026 (0.489)

Householder’s insurance Sustainable behavior -.028 (-0.534) .132 .001Sustainable mobility .005 (0.096)Sustainable attitude -.015 (-0.273)Sustainable knowledge -.02 (-0.384)

Notes: *Regression is significant at the 0.1 level, **Regression is significant at the 0.05 level, ***Regression is significant at the 0.001level.


tently exclude attitudinal variables and/or demographicvariables, provide a significantly worse fit. Demograph-ics do not fully capture the same information as attitudinaldata. However, including demographics in addition to theattitudinal model provides a better fit indicated by L2. Thelarger the value, the poorer the model fits the data and theworse the observed relationships are described by thespecific model. We identify four classes based on whetherrespondents agree or disagree with the statement. Asindicated in Table 3, about 41 percent of the respondentsbelong to Class 1, 23 percent to Class 2, about 18 percentto Class 3, and the remaining 17 percent to Class 4.

The results for the estimated model are presented inTable 4. R2 indicates how well the model explains thedependent variable. The R2 is similar to the explainedvariance in analysis of variance and to item communali-ties in factor analysis. However, the Wald statistic test

associated with a significant p-value means that the indi-cator does discriminate among the clusters. Overall, thedimensions of sustainable behavior, sustainable mobility,and sustainable knowledge can distinguish the four classes.Except for the statement fair-trade purchase, all items aresignificant predictors of the classes on a level of 1 percentrather than 5 percent. Concerning sustainable attitude, theitems discriminate less significantly among the classes:the attitude item governmental contribution is a signifi-cant predictor on a level of 10 percent only.

If attitudinal groups are different, the claim statistic isexpected to be significantly different across the foursustainability groups. The Wald statistic test is associatedwith a significant p-value support that the claims historydiffers among the groups. For example, the meanresponse on claims is lowest in Class 1 (mean = 0.10). Themean claim value in Class 2 is 0.11. Class 3 indicates a

TABLE 3Model Summary, Classification of Group Membership (N = 715)

Classification Modal

Class 1 Class 2 Class 3 Class 4Probabilistic (41.03%) (23.66%) (18.18%) (17.13%) Total

Class 1 260.76 16.38 8.81 7.54 293.49Class 2 15.63 143.65 7.68 2.17 169.13Class 3 6.59 4.09 117.68 1.60 129.96Class 4 19.02 3.88 2.84 96.69 122.42Total 302.00 168.00 137.00 108.00 715.00

Model summary

Full model (including sociodemographic variables as covariates)

L-squared (L²): 460.4315 (p-value = 1)Log-likelihood (LL): -231.602Classification errors: 0.1346

Restricted model (Excluding sociodemographic variables as covariates)

L-squared (L²): 538.5158 (p-value = 0.068)Log-likelihood (LL): -407.094Classification errors: 0.4998

Traditional model (Including sociodemographic variables as predictors and excluding attitudinal variables)

L-squared (L²): 753.857 (p-value = 0.0016)Log-likelihood (LL): -487.3451Classification errors: 0.4333

Notes: Classification modal indicates the highest membership probability, whereas the classification probabi-listic indicates the probability to belong to alternative groups.


Std.Class 1 Class 2 Class 3 Class 4 Wald p-value Mean Dev.

R2 0.84 0.91 0.98 0.94

Dependent

Claim (mean response ; 0=no claim; 1=claim) 0.10 0.21 0.18 0.68No 51.44 -27.18 2.19 32.67 16.37 0.00 20.67 31.89Yes -51.44 27.18 -2.19 -32.67 -20.67 31.89

Predictors

Sustainable Behavior

Ecological purchase 8.68 30.32 -63.01 12.97 19.68 0.00 9.96 0.02Fair-trade purchase 0.78 -4.83 21.20 -3.54 7.81 0.10 7.78 0.05Shopping in ecostore/local market (obj.) 9.80 -9.70 32.14 -14.40 18.44 0.00 18.43 0.00Engagement in social/environmental projects (obj.) 0.67 -19.53 -20.01 11.97 11.76 0.02 11.31 0.01Donation (obj.) 22.67 -21.64 30.97 -12.77 19.80 0.00 18.32 0.00

Sustainable Mobility

Public transport network 10.85 -16.58 -5.61 21.16 16.83 0.00 16.59 0.00Sustainable travelling -9.30 -46.55 24.28 39.12 20.06 0.00 19.52 0.00Hold travel card (obj.) -16.17 29.29 -4.90 -13.86 11.83 0.02 5.85 0.12Use local public transport (obj.) -14.75 26.11 -16.72 -14.11 18.92 0.00 11.04 0.01

Sustainable attitude

Media reports about climate change 27.01 -20.47 2.83 0.78 10.04 0.04 9.92 0.02Governmental contribution to environment 16.88 -3.98 -7.99 -2.26 9.02 0.06 7.13 0.07Abstain from sth. because of environment -0.77 6.65 -11.91 20.33 11.42 0.02 10.26 0.02Individual contribution to environmental protection -8.80 1.08 14.86 -11.16 10.22 0.04 8.60 0.04

Sustainable knowledge

Alternative fuel 4.22 -10.99 3.83 9.46 12.42 0.01 8.13 0.04Erosion of ozone layer -14.11 26.66 -79.94 8.75 19.55 0.00 18.78 0.00Ecological footprint -5.46 -44.45 26.34 -28.21 11.78 0.02 7.53 0.06Intercept 2.19 -1.24 -2.27 1.32 2.10 0.55<Covariates

Driving history

-10,000 km -0.78 1.25 0.68 -1.15 35.01 0.0010 - 15,000 km -0.30 0.20 0.79 -0.7015 - 20,000 km 2.34 -0.85 -2.94 1.4520 -25,000 km -0.99 1.93 -0.07 -0.8625,000 + km -0.27 -2.53 1.54 1.26

Age

- 30 years 0.27 -2.07 -0.29 2.09 36.79 0.0031 - 40 years -1.49 0.93 0.88 -0.3241 - 50 years 0.55 -2.29 -0.02 1.7651 - 60 years -0.22 1.61 -2.96 1.58+ 60 years 0.90 1.82 2.39 -5.10

Urbanity

- 5,000 inhabitants 0.31 -0.46 -1.39 1.54 36.00 0.005,001 - 50,000 inhabitants 1.15 -0.29 -2.52 1.6650,001 - 150,000 inhabitants -2.41 3.06 4.45 -5.10150,001 - 500,000 inhabitants 0.04 1.77 -2.02 0.21+ 500,000 inhabitants 0.91 -4.08 1.48 1.70

TABLE 4Identification of Different Classes by Means of Sustainable Indicators (N = 715)


mean value of 0.18. Class 4 shows the highest meanvalue (0.68).

As expected, the response pattern differs among thefour groups. The beta effect estimates below the columnlabeled “Class 1” suggest that Class 1 is influenced bysustainable behavior. For example, ecological purchase(beta = 8.68), shopping in ecostore/local market (obj.)(beta = 9.80), and donation (obj.) (beta = 22.67) havepositive effects on this class. In terms of sustainablemobility, Class 1 indicates a positive effect of knowledgeabout public transport network (beta = 10.85), whereasverifiable use of public transport (hold travel card, beta =-16.17) and use of public transport (beta = -14.75) havenegative influence. In terms of sustainable attitude, mem-bers of Class 1 favor media reports about climate change(beta = 27.01) and governmental contribution to environ-mental protection (beta = 27.01), but they do not feelindividual responsibility (beta = -8.80). Knowledge ofsustainable issues, in particular, erosion of ozone layer(beta = -14.11) is lacking. In summary, Class 1 is labeledthe “attitude-oriented-class.”

Members of Class 2 differ from the other classes bydisapproving of sustainable behavior, e.g., engagement insocial/environmental projects (obj.) (beta = -19.53) anddonation (obj.) (beta = -21.64). Concerning sustainablemobility, the results indicate strong effects of verifiablemobility. Members of Class 2 are the only ones which holdtravel card (obj.) (beta = 29.29) and use local publictransport (obj.) (beta = 26.11). Having a look at knowl-

edge about public transport network and sustainabletraveling the effect is strongly negative, which indicatesthat sustainable mobility is not due to beliefs, but toconvenience. Overall, sustainable attitude and sustain-able knowledge are not a differential attributes. Membersof Class 2 are the only ones who state knowledge aboutecological footprint (beta = 26.34). To summarize, Class 2is the mobility-oriented-class.

Class 3 suggests a strong positive influence of verifi-able behavior. Respondents are distinguished by sustain-able purchase, indicated by a positive effect of fair-tradepurchase (beta = 21.20) and shopping in ecostore/localmarket (obj.) (beta = 32.14), although they reject thestatement ecological purchase (beta = -63.01). Further,members of this class differ from other classes by dona-tion (obj.) (beta = 30.97). In terms of sustainable mobility,Class 3 prefers sustainable traveling (beta = 24.28),although members of this class do not approve this byindicating knowledge about public transport net (beta =-5.61), hold travel card (obj.) (beta = -4.90), and use localpublic transport (obj.) (beta = -16.72). Class 3 is summa-rized as “behavioral-knowledge-class.”

Members of Class 4 favor sustainable mobility con-firmed by the predictors knowledge about public trans-port network (beta = 21.16) and sustainable traveling(beta = 39.12). Since they cannot approve this, the influ-ence of verifiable predictors is negative, e.g., hold travelcard (obj.) (beta = -13.86) and use local public transport(obj.) (beta = -14.11). The results of sustainable attitude

TABLE 4 (CONTINUED)Identification of Different Classes by Means of Sustainable Indicators (N = 715)

Std.Class 1 Class 2 Class 3 Class 4 Wald p-value Mean Dev.

Education

High school -0.99 -0.16 0.86 0.29 33.92 9.20E-05Secondary school -0.87 0.49 1.57 -1.19A-level 2.21 0.63 -4.02 1.18University -0.36 -0.96 1.60 -0.28

Income

- 1,000 -3.13 2.29 3.63 -2.79 33.46 0.001,001 – 2,000 0.65 1.88 -0.88 -1.652,001 – 4,000 -0.40 1.02 1.73 -2.344,001 – 6,000 3.01 2.23 -6.40 1.166,001 – 8,000 2.43 -2.55 -0.96 1.08+ 8,000 -2.55 -4.87 2.89 4.54

Gender

Male -0.43 -0.37 1.13 -0.33 10.99 0.01Female 0.43 0.37 -1.13 0.33


are ambiguous. The statement abstain from somethingbecause of environment (beta = 20.33) is accepted, althoughpersons do not agree with individual contribution toenvironmental protection (beta = -11.16). Sustainableknowledge exists concerning alternative fuel (beta =9.46) and erosion of ozone layer (beta = 8.75), but noknowledge about ecological footprint (beta = -28.21). Insummary, Class 4 is labeled the “mobility-rhetoric-class.”

In addition, we include sociodemographic and socio-economic attributes as covariate, and environmental atti-tudes as predictors. As indicated by the model test (seeTable 4), the demographic variables do not contributesignificantly more additional information. Nevertheless,the demographics are significant on a 0% level or 1% levelfor gender, respectively. Education appears as a highlysignificant covariate (9.20E-05). The results concerningthe covariates are quite contrary and the adaptabilitydemographic variables for marketing application appearambiguous (for details see Table 4).

SUMMARY

Discussion

Applying linear regression analysis, we identify theimpact of sustainability on customers’ security, insuranceorientation, and claims history. The results indicate thatsustainable antecedents have significant impact on cus-tomers’ attitude toward security and insurance. Sustain-able attitude is a significant driver of customers’ readinessto assume risk; sustainable knowledge drives customers’interest in individualized services and products.

By applying LCA, it is possible to identify individu-als who share a sustainability profile. For example, thesustainability indicators distinguish significantly amongthe segments and thus allow for a test of the model quality.The covariates allow for an estimation of segment attrac-tiveness, such as profitability of customers in age andincome. Taken together, the results stemming from LCAindicate that insurance customers can be classified intofour distinct sustainability groups. The findings of LCAanalysis support the limits of traditional attributes fromdemographics for determining customer behavior. Whiledemographic variables fulfill statistical significance, theylack the qualitative explanatory power of the psycho-graphic variables.

Future Research and Limitations

Marketing in the insurance industry can segmentcustomers according to sustainable characteristics.Therefore, it is meaningful to replenish the traditionalsegmentation attributes such as gender and age, withsustainable attributes. The actuary must verify and integratesustainable attributes into the premium calculation. With

respect to the variables explored in this study that indicateverifiable sustainable behavior (e.g., prove actualsustainable behavior by showing temporary publictransport card), it will be possible to consider sustainability,when predicting insurance losses. By controlling for thetraditional attributes, psychographic variables mightprovide a more useful approach in segmenting insurancecustomers and give information distinct from standardactuarial variables.

There are limitations that must be taken into accountwhen considering these findings. First, the customers’security orientation and claims history relies on individu-als’ self-report measures. While there might be method-ological bias due to subjectivity, the sustainable profileneeds to be verified based on objective claims history.Supporting these self-report measures with observationalor behavioral measures would strengthen the findings. Asecond limitation is related to the sample. The objective ofthis study was to examine sustainable conscious attitudeand behavior based on technical losses in automobile andhousehold of insurance. Using a sample from anotherinsurance sector, such as health insurance would extendthe sustainability insurance profile. Future research mightrefine segmentation variables, identify additional sustain-able criteria, and extend the findings to other domains ofsustainability (e.g., economically conscious attitude andbehavior).

Managerial Implication

Managers and researchers must request the adapt-ability of the typical sustainable consumer profile (young,mid and high income, educated, urban woman) in terms ofmarketing application. The results suggest that psycho-graphic measures in general and the dimensions, espe-cially, knowledge, attitude, and behavior discriminatesustainable conscious customer profiles. Therefore, apply-ing either psychographic variables or a mixed model(incorporating a range of demographic and psychographicvariables) should be preferred to traditional demographicprofiling methods.

Furthermore, the tested dimensions are applicablepredictors of customers’ security and insurance orienta-tion. Hence, it yields useful underwriting informationhow they will react in creating losses, such as in automo-bile insurance and household insurance. Insurers aresupposed to extend the traditional attributes and accountfor sustainable characteristics to predict customers’ riskprofile, which is vital in the field of product developmentand pricing. By managing sustainability segments effec-tively, insurers can establish products and services, whichare consistent with customers’ sustainable opinion andlifestyle. By predicting the influence of sustainable atti-tude on customers’ behavior, insurance companies canadapt their customer relationship strategy and adjust prod-


ucts and services. For example, insurers can offer indi-vidualized products to customers who differ in sustain-able knowledge. In the insurance industry, risk-taking

behavior and security perception are elements that insur-ers must take into account when dealing with sustainablecustomers.

REFERENCES

Aaker and Bagozzi (1982), “Attitudes Toward PublicPolicy Alternatives to Reduce Air Pollution,” Jour-nal of Marketing, 46 (1), 85–94.

Anderson and Anderson (2009), “Sustainability RiskManagement,” Risk Management and InsuranceReview, 12 (1), 25–38.

Anderson and Cunningham (1972), “The Socially Con-scious Consumer,” Journal of Marketing, 36 (7), 22–31.

Ajzen and Fishbein (1980), Understanding Attitudes andPredicting Social Behavior. Englewood Cliffs, NJ:Prentice-Hall.

Arcidiacono and Jones (2003), “Finite Mixture Distribu-tion, Sequential Likelihood, and the EM Algorithm,”Econometrica, 71 (3), 933–46.

Arnett (1990), “Drunk Driving, Sensation Seeking, andEgocentrism Among Adolescents,” Personality andIndividual Differences, 11 (6), 541–54.

Autio, Heiskanen, and Heinonen (2009), “Narratives of‘Green’ Consumers – The Antihero, the Environ-mental Hero and the Anarchist,” Journal of Con-sumer Behavior, 8, 40–53.

Bean and Ennis (1987), “Market Segmentation: AReview,” European Journal of Marketing, 21, 20–42.

Bohlen, Schlegelmilch, and Diamantopoulos (1993),“Measuring Ecological Concern: A Multi-ConstructPerspective,” Journal of Marketing Management, 9,415–30.

Brockett and Golden (2007), “Biological andPsychobehavioral Correlates of Credit Scores andAutomobile Insurance Losses: Toward an Explica-tion of Why Credit Scoring Works,” Journal of Riskand Insurance, 74 (1), 23–63.

Chan (1999), “Market Segmentation of Green Consumersin Hong Kong,” Journal of International ConsumerMarketing, 21 (2), 7–24.

Cohen (2005), “Consumer Credit, Household FinancialManagement, and Sustainable Consumption,” Inter-national Journal of Consumer Studies, 31(1), 57–65

Corrado and Ross (1990), “Environmental Issues in the1990s: Green Issues in Britain and the Value of GreenResearch Data,” ESOMAR Annual Congress, 43 (Sep-tember), 347–69.

Cureton and D’Agostino (1983), Factor Analysis: AnApplied Approach. Hillside, NJ: Lawrence ErlbaumAssociates.

Dempster, Laird, and Rubin (1977), “Maximum Likeli-hood from Incomplete Observations,” Journal of theRoyal Statistical Society. Series B (Methodological),

39 (1), 1–38.Diamantopoulos, Schlegelmilch, Sinkovics, and Bohlen

(2003), “Can Socio-Demographics Still Play a Rolein Profiling Green Consumers? A Review of theEvidence and an Empirical Investigation,” Journal ofBusiness Research, 56, 465–80.

Dunlap and Catton (1979), “Environmental Sociology,”Annual Review of Sociology, 5, 243–73.

Fishbein and Ajzen (1975), Belief, Attitude, and Behav-ior: An Introduction to Theory and Research. Read-ing, MA: Addison-Wesley.

Frank, Massy, and Wind (1972), Market Segmentation.Englewood Cliffs, NJ: Prentice-Hall.

Getzner and Grabner-Kräuter (2004), “Consumer Prefer-ences and Marketing Strategies for Green Shares,”The International Journal of Bank Marketing, 22 (4),260–78.

Grey, Triggs, and Haworth (1989), “Driver Aggression:The Role of Personality, Social Characteristics, Riskand Motivation,” Human Factors Group, Depart-ment of Psychology, Monash University Report No.CR 81, Federal Office of Road Safety.

IUCN/UNEP/WWF (1991), “Caring for the Earth: AStrategy for Sustainable Living,” Gland,Switzerland.|IUCN – The World Conservation Union,UNEP – United Nations Environment Programme,WWF – World Wide Fund for Nature.

Jonah and N. Dawson (1987), “Youth and Risk: AgeDifferences in Risky Driving, Risk Perception, andRisk Utility,” Alcohol, Drugs, and Driving, 3 (3/4),13–29.

Kaiser and Gutscher (2003), “The Proposition of a Gen-eral Version of the Theory of Planned Behavior:Predicting Ecological Behavior,” Journal of AppliedSocial Psychology, 30, 586–603.

____________, Schultz, and Scheuthle (2007), “TheTheory of Planned Behavior without Compatibility?Beyond Method Bias and Past Trivial Associations,”Journal of Applied Social Psychology, 37, 1522–154.

Kellison and Brockett (2003), “Check the Score: CreditScoring and Insurance Losses: Is There a Connec-tion?” Texas Business Review, Special Issue, 1–6.

Kilbourne and Beckmann (1998), “Review and CriticalAssessment of Research on Marketing and the Envi-ronment,” Journal of Marketing Management, 14(6), 513–32.

Kinnera, Taylor, and Ahmed (1974), “Ecologically Con-cerned Consumers: Who Are They?” Journal ofMarketing, 38 (4), 20–24.

Kirkpatrick (1990), “Environmentalism,” Fortune, 12,44–51.


Krystallis, Vassalo, Chryssohoidis, and Perrea (2008),“Societal and Individualistic Drivers as Predictors ofOrganic Purchasing Revealed Through a PortraitValue Questionnaire (PVQ) – Based Inventory,” Jour-nal of Consumer Behavior, 7, 164–87.

Laroche, Bergeron, and Barbaro-Forleo (2001), “Target-ing Consumers Who Are Willing to Pay More forEnvironmentally Friendly Products,” Journal of Con-sumer Marketing, 18 (6) 503–20.

McDonald and Dunbar (1995), Market Segmentation.Basingstoke: Macmillan.

____________ and Ho (2002), “Principles and Practicesin Reporting Structural Equation Analysis,” Psycho-logical Methods, 7, 64–82.

MacDonald and Hara (1994), “Gender Differences inEnvironmental Concern among College Students,”Sex Roles, 33 (5/6), 369–74.

Meffert and Bruhn (1996), “Das Umweltbewusstsein vonKonsumenten,” Die Betriebswirtschaft, 56 (5), 621–48.

Miller and Smith (2003), “The Relationship of Credit-Based Insurance Scores to Private Passenger Auto-mobile Insurance Loss Propensity, A Actuary Study,”Epic Actuaries, online at [http://www.epicactuaries.com].

Monaghan (2000), “The Impact of Personal Credit His-tory on Loss Performance in Personal Lines,” Casu-alty Actuary Society Forum, (Winter), 79–19–05.

Mulaik (1972), The Foundations of Factor Analysis. NewYork: McGraw-Hill.

Murphy (1989), “Green is the Color, Consumerism is theGame,” Survey, (Winter), 2–4.

Nelson (1989), “Shades of Green,” Survey, (Winter), 9–11.

Piercy and Morgan (1993), “Strategic and OperationalMarket Segmentation: A Managerial Analysis,” Jour-nal of Strategic Marketing, 1, 123–40.

Prendergast and Thompson (1997), “Cynical Segmenta-tion of Environmental Markets: The Product of Mar-keters’ Disposition or Corporate Demands?” Journalof European Marketing, 6 (4), 17–34.

Rivera-Camino (2007), “Re-Evaluating Green MarketingStrategy: A Stakeholder Perspective,” European Jour-nal of Marketing, 41(11/12), 1328–58.

Roberts (1996), “Green Consumers in the 1990s: Profiles

and Implications for Advertising,” Journal of Busi-ness Research, 36 (3), 217–32.

Schwepker and Cornwell (1999), “An Examination ofEcologically Concerned Consumers and Their Inten-tion to Purchase Ecologically-Packaged Products,”Journal of Public Policy and Marketing, 10 (2), 77–101.

Shihar and Compton (2004), “Aggressive Driving: AnObservational Study of Driver, Vehicle, and Situ-ational Variables,” Accident Analysis and Preven-tion, 36, 429–37.

Stern, Dietz, and Kalof (1993), “Value Orientation, Gen-der, and Environmental Concern,” Environment andBehavior, 25 (3), 322–48.

Schlegelmilch, Bohlen, and Diamantopoulos (1996), “TheLink Between Green Purchasing Decisions and Mea-sures of Environmental Consciousness,” EuropeanJournal of Marketing, 30 (5), 35–55.

Schmitt (1996), “Uses and Abuses of Coefficient Alpha,”Psychological Assessment, 8 (4), 350–53.

Straughan and Roberts (1999), “Environmental Segmen-tation Alternatives: A Look at Green ConsumerBehavior in the New Millennium,” Journal of Con-sumer Marketing, 16 (6), 558–75.

Sommers, Dyehouse, Howe, and Manharth (2000), “Drink-ing Drivers: A Cluster of Risk-Taking Behaviors,”15th International Conference on Alcohol, Drugs,and Traffic Safety, Stockholm, Sweden, May 25,2000.

Tillman and Hobbs (1949), “The Accident-Prone Auto-mobile Driver: A Study of the Psychiatric and SocialBackground,” The American Journal of Psychiatry,106, 321–31.

United Nations General Assembly (1987), “Report of theWorld Commission on Environment and Develop-ment: Our Common Future,” Transmitted to theGeneral Assembly as an Annex to document A/42/427 – Development and International Co-operation:Environment. Retrieved on: 2009–02–15.

Vining and Ebreo (1990), “What Makes a Recycler? AComparison of Recyclers and Nonrecyclers,” Envi-ronmental Behavior, 22, 55–73.

Wendel and Kamakura (2000), Market Segmentation:Conceptual and Methodological Foundations, 2nd

ed. Boston: Kluwer Academic Publishers.

For further information contact:Mareike Bodderas

Institute of Insurance EconomicsUniversity of St. Gallen

Kirchlistrasse 2, 9000 St. GallenSwitzerland

Phone: +41(0)71.243.40.61Fax: +41(0)71.243.40.40



AN EMPIRICAL ANALYSIS ON THE ADOPTION OF ALTERNATIVEFUELS: CONSUMERS’ PERCEIVED RISKS ASSOCIATED WITH

NATURAL GAS VEHICLES

Klaus-Peter Wiedmann, Leibniz University of Hanover, GermanyMartin Kassubek, Leibniz University of Hanover, GermanyNadine Hennigs, Leibniz University of Hanover, GermanyLars Pankalla, Leibniz University of Hanover, Germany

SUMMARY

The increasing world energy demand, the effect ofglobal warming, and the recent worldwide financial crisisconstitute economics’ major challenges, especially withdirect effects on private transport (IPCC 2008; KahnRibeiro et al. 2007). Particularly in Europe, compressednatural gas (CNG) is regarded as a short call alternative fortraditional combustion engines (IGU 2005; Ristovskiet al. 2004). The European Union has declared the long-termed target to disengage from fossil fuels and to employa significant amount of alternative technologies as well asfuels in private transport, especially gaseous fuels, bio-fuels, and hybrid vehicles (EurActiv Network 2008b).The principal factors that motivate European govern-ments to pass and promote such a collective long-runtraffic strategy include primarily environmental (e.g.,reducing air pollution) and geostrategic (e.g., reducingdependence on petroleum imports) reasons (Yeh 2007;Janssen et al. 2006).

Compared to conventional vehicles, CNG-drivencars can be considered as new transportation technology.The appliance of the Diffusion of Innovation Theoryprovides knowledge to investigate consumer’s attitudestoward the adoption of NGVs. Generally, diffusion ofinnovation is defined as “a process by which an innova-tion is communicated through certain channels, over time,among members of a social system” (Rogers 1995). Thediffusion process is considered to pass sequentially throughstages of knowledge, persuasion, decision, implementa-tion, and confirmation of pertaining innovation (Robertsonand Gatignon 1986; Mahajan et al. 2000).

Conceptual Model

Understanding the perceived risk phenomenon, allpotential dimensions should be integrated into one singlemodel. For the purposes of this paper, regarding allprospective and directly attributable risk dimensions, thisresearch follows the results of Stone and Grønhaug (1993)and differentiate perceived risk in case of purchasing aNGV into six highly interrelated components of perceivedrisk. Against this background, Figure 1 shows the pro-

posed conceptual model to investigate specific perceivedrisk factors in case of purchasing a natural gas vehicle forprivate use.


Within the data analysis, we first uncovered thevarious dimensions underlying the perceived risks in caseof purchasing a NGV by a factor analysis using theprincipal component method with varimax rotation in linewith a German data sample. The factor analysis produceda nine-factor structure with a Kaiser-Meyer-Olkin mea-sure of .803. All items had medium (> 0.5) up to highfactor loadings (> 0.8). For market segmentation pur-poses, profiling the cluster solutions should lead toward aclassification scheme through describing the characteris-tics of each cluster to explain how they might differ onrelevant dimensions. To develop a profile of each marketsegment, more detailed information comes from lookingat the questionnaire variables cross-tabulated by clustersegment. Comparisons among the five clusters were con-ducted on a variety of descriptive variables includingdemographic and socioeconomic characteristics. Basedon the variables from which they derived, the five clusterswere labeled as follows: Cluster 1 is referred to as theService-Oriented Skeptics, members of Cluster 2 arereferred to as the High-Involved Traditionalists, Cluster 3is referred to as the Risk-Averse Uninvolved, members ofCluster 4 are called the NGV-Interested Urbanites, andCluster 5 is referred to as the Ecology-Minded Non-Drivers.

Further Research Steps and Managerial Implications

First, the different dimensions could be used as abasis to develop a more robust measurement instrumentcovering the different types of adopters. Second, in thenext step of developing hypotheses, we should as wellemphasize the interplay between the different variables.In line with a cross-national development of alternativefuels and engines, cultural differences in customer behav-ior, various policy instruments (e.g., tax rebates anddelivery infrastructure), and economic indicators (e.g.,payback periods or resale values) have to be included in


future research. Moreover, there are wide-ranging mana-gerial implications of our research. The results offer astarting point for manufacturers and policymakers as theyinvestigate how to best manage product characteristics,

Financial risk

Social risk

Time risk

Performance risk

Physical risk

Psychological risk

Perceived risk in case of purchasing

a NGV

Car involvement

Ecological awareness

FIGURE 1The Conceptual Model

local incentives, and environmentalism sentiment in orderto attract a larger number of consumers to NGVs. Refer-ences are available upon request.

For further information contact:Martin Kassubek

Institute of Marketing and ManagementLeibniz University of Hanover

Koenigsworther Platz 130167 Hanover




DETERMINANTS OF CROSS-BUYING: A RE-EXAMINATION

Markus Blut, TU Dortmund University, GermanyDavid M. Woisetschläger, TU Dortmund University, Germany

SUMMARY

In existing literature, cross-buying is seen as animportant contributor of customer profitability (Reinartzand Kumar 2003; Venkatesan and Kumar 2004; Verhoef,Franses, and Hoekstra 2001). Typically, cross-buying isdefined as the number of categories that customers pur-chase from (Reinartz and Kumar 2003). Despite the highrelevance of cross-buying for customer profitability, rela-tively few articles empirically examine potential determi-nants of cross-buying (e.g., Hallowell 1996; Jeng 2008;Kumar, George, and Pancras 2008; Kamakura, Kossar,and Wedel 2004; Kamakura, Ramaswami, and Srivastava1991; Kamakura, Wedel, de Rosa, and Mazzon 2003; Liuand Wu 2007; Ngobo 2004; Reinartz, Thomas, and Bascoul2008; Soureli, Lewis, and Karantinou 2008; Verhoef andDonkers 2005; Verhoef, Franses, and Hoekstra 2001,2002). Existing studies on cross-buying that include cus-tomer attitudes as determinants primarily focus on threerelational factors: customer satisfaction, trust, and pay-ment equity. In turn, studies that use past purchase behav-ior to predict cross-buying do not consider customerattitudes as explanatory variables. Consequently, a com-parative assessment of its determinants’ explanatory poweris not possible. From a management perspective, itremains unclear which of the numerous potential determi-nants are the best predictors of cross-buying behavior.

Against this background, the present research con-tributes to the literature by integrating existing cross-buying determinants into one conceptual model. Therelative impact of these determinants on cross-buyingbehavior is tested empirically. In total 60,000 customerswere randomly selected from one retailer to be contactedvia e-mail, from which 10,622 (17.7%) joined the surveyand completed the questionnaire. Again 1,050 customerswere randomly selected to combine attitudinal data from

the customer survey with purchase behavior data fromcustomer data base of the examined retailer. Purchasebehavior data include each transaction made betweencustomers and the firm over a period of three years. Theexamined sample is representative for the total customerbase regarding age (M = 32.00 years), gender (male =86.1%, female = 13.9%) and income (M = 2.320 ).

Summarizing the results of the empirical study, wehave to state that prior antecedents of cross buying can beclassified into four sets of variables: (1) Customer attitudes,(2) past purchase behavior, (3) socio-demographiccharacteristics, and (4) marketing instruments. Adifferentiated examination was undertaken with the resultthat socio-demographic variables were found to have onlya weak effect on cross buying, whereas past purchasebehavior and marketing instruments have a strong impact.Regarding customer attitudes, we found affectivecommitment, attractiveness of alternative, marketplaceinvolvement to have weak effects on a customer’s crossbuying behavior. The effects were in line with ourexpectations.

Furthermore, we found sales and purchase frequencyto mediate the relationship between customer attitudesand cross buying, which has been proposed by Reinartz,Thomas, and Bascoul (2008). Therefore, companies havefirst to invest in customer loyalty. As a consequence, loyalcustomers will engage in cross-buying. Hence, customerattitudes largely have an indirect rather than a direct effecton cross buying (satisfaction, payment equity, trust, cal-culative commitment, convenience, involvement, valueshopping). Overall, past purchase behavior and informa-tion about individually employed marketing instrumentshave the strongest predictive power of a customer’s crossbuying behavior.

REFERENCES

Hallowell, R. (1996), “The Relationships of CustomerSatisfaction, Customer Loyalty, and Profitability: AnEmpirical Study,” International Journal of ServiceIndustry Management, 7 (4), 27–42.

Jeng, S.P. (2008), “Effects of Corporate Reputations,Relationships and Competing Suppliers’ MarketingProgrammes on Customers’ Cross-Buying Inten-tions,” The Service Industries Journal, 28 (1), 15–26.

Kamakura, W.A., S.N. Ramaswami, and R.K. Srivastava(1991), “Applying Latent Trait Analysis in the Evalu-ation of Prospects for Cross-Selling of FinancialServices,” International Journal of Research in Mar-keting, 8 (4), 329–49.

____________, M. Wedel, F. de Rosa, and J.A. Mazzon(2003), “Cross-Selling Through Database Market-ing: A Mixed Data Factor Analyzer for Data Aug-mentation and Prediction,” International Journal ofResearch in Marketing, 20 (1), 45–65.

€


____________, B.S. Kossar, and M. Wedel (2004), “Iden-tifying Innovators for the Cross-Selling of New Prod-ucts,” Management Science, 50 (8), 1120–33.

Kumar, V., M. George, and J. Pancras (2008), “Cross-Buying in Retailing: Drivers and Consequences,”Journal of Retailing, 84 (1), 15–27.

Liu, T.C. and L.W. Wu (2007), “Customer Retention andCross-Buying in the Banking Industry: An Integra-tion of Service Attributes, Satisfaction and Trust,”Journal of Financial Services Marketing, 12 (2),132–45.

Ngobo, P.V. (2004), “Drivers of Custom ers’ Cross-Buying Intentions,” European Journal of Marketing,38 (9/10), 1129–57.

Reinartz, Werner J. and V. Kumar (2003), “The Impact ofCustomer Relationship Characteristics on ProfitableLifetime Duration,” Journal of Marketing, 67 (1),77–99.

____________, J.S. Thomas, and G. Bascoul (2008),“Investigating Cross-Buying and Customer Loyalty,”Journal of Interactive Marketing, 22 (1), 5–20.

Soureli, M., B.R. Lewis, and K.M. Karantinou (2008),

“Factors that Affect Consumers’ Cross-BuyingIntention: A Model for Financial Services,” Journalof Financial Services Marketing, 13 (1), 5–16.

Venkatesan, R. and V. Kumar (2004), “A CustomerLifetime Value Framework for Customer Selectionand Resource Allocation Strategy,” Journal of Mar-keting, 68 (4), 106–25.

Verhoef, P.C., P.H. Franses, and J.C. Hoekstra (2001),“The Impact of Satisfaction and Payment Equity onCross-Buying: A Dynamic Model for a Multi-ServiceProvider,” Journal of Retailing, 77 (3), 359–78.

____________, ____________, and ____________(2002), “The Effect of Relational Constructs on Cus-tomer Referrals and Number of Services Purchasedfrom a Multi-Service Provider: Does Age of Rela-tionship Matter?” Journal of the Academy of Market-ing Science, 30 (3), 202–16.

____________ and B. Donkers (2005), “The Effect ofAcquisition Channels on Customer Loyalty and Cross-Buying,” Journal of Interactive Marketing, 19 (2),31–43.

For further information contact:Markus Blut

TU Dortmund UniversityOtto-Hahn-Str. 6Dortmund, 44227

GermanyPhone: +49(0)231.755.3277



HUMANIC INTERACTION IN PATIENT EXPERIENCE ASA VALUE PROPOSITION

Jie Chen, Purdue University, West LafayetteSandra S. Liu, Purdue University, West Lafayette

En-Chung Chang, Renmin University of China, Beijing

SUMMARY

Understanding value and its delivery have beenadvocated as a research priority by the leading schol-ars of the marketing field (Lindgreen and Wynstra 2005).Discussions and studies have been focusing at the busi-ness front but there has not been empirical investigation ofthe relationship between patient experience and patientvalue in healthcare, which is the largest service industry inthe United States (Herzlinger 1997). The current studyattempts to fill this gap by examining the relationshipsbetween patient experience, patient value, and the result-ing patient recommendation with a comprehensive theo-retical framework. Specifically, the current study intendsto develop the theoretical and operational definitions ofpatient experience and patient value from humanics inter-action perspective and then to construct a conceptualframework for the relationship between patient experi-ence and patient value as well as the impact of value onpatient recommendation.

The Theoretical Framework

A theoretical framework is proposed to examine therelationship between patient experience and patient valueas well as the impact of value on patient recommendation.Patient experience in the current study focuses on thepatient-staff interactions and is defined as the patient’shumanics interactions with the caregivers (Gentile, Spiller,and Noci 2007). Previous studies have suggested thatpatients and staff can interact in a variety of ways/styleswhich include verbal/nonverbal, cognitive/affective, andhigh/low controlling interactions, among others (Bradley,Sparks, and Nesdale 2001; Hausman 2004; Ong et al.1995). The current study incorporates all the interactionstyles into the measurement of patient experience. Spe-cifically, there are three proposed dimensions of patientexperiences in our framework: information connected-ness, nonverbal interaction, and personal greeting. Com-pared with experience, value is at the higher level ofabstractions for a product or service (Zeithaml 1988). Thisstudy adopts Holbrook’s definition and defines patientvalue as: “the patients’ evaluations of their appreciatedhealthcare experiences.” “Appreciated” is used herebecause the appreciated consumption experience canelicit the customer’s intrinsic value (Holbrook 1996,

1999). Holbrook (1996, 1999) summarizes the higherlevel abstractions/value into eight categories: efficiency,fun, excellence, aesthetics, status, ethics, esteem, spiritu-ality. Three dimensions are relevant with the currentstudy: excellence (function), status (social symbols), andspirituality (admiration). According to the context of thecurrent study, these three dimensions are specified asempowerment (excellence), respectfulness (spirituality),and responsiveness (status) which are abstracted from thethree dimensions of patient experience. As previouslydiscussed, patient value is created in the form of patientexperiences (Holbrook 1996; Prahalad and Ramaswamy2004; Norton, Curtiss, and Hart 2001; Berry, Carbone,and Haeckel 2002; Hill et al. 2001). Thus, we hypothesizethat “patient experiences creates the patients’ value.”

The value has also been found to directly influencebehavioral intentions, including a consumer’s tendency torecommend the product to other consumers (Cronin,Brady, and Hult 2000; Sirdeshmukh, Singh, and Sabol2002; Choi et al. 2004). Thus, our hypothesis is “recom-mendation of healthcare services is positively influencedby patient value.”

The long wait time has caused great concern forhospital administrators (Fottler and Ford 2002). We there-fore incorporate patient experience involving waiting intoour framework. Time has been demonstrated to have botheconomic value and consumer value (Leclerc and Schmitt1999). Intuitively, long wait time decreases customervalue. Thus, we hypothesize that “wait time negativelyinfluences the patient’s value.”

Methodology

Data are collected through semi-structured observa-tions and exit surveys. The emergency department of ahospital was selected as the context for study due to itstimeframe for each patient visit is well confined thatallows the duration of each observation to be manageable.Participants for this study included 550 emergencydepartment patients from three hospitals, serving similartarget segments, located in Michigan, Indiana, and Ala-bama. An availability sampling method was used in thisstudy. Participants joined this research via close observa-tion and self-reporting evaluations.



Structure equation modeling was used for data analy-sis. The results from AMOS 7 show that the model isacceptable, as indicated by the GFI, AGFI, and CFI valuesof .86, .83, and .84 respectively. The results showed thatvalue of empowerment and responsiveness significantlyinfluences patient recommendation, which is a commonlyused indicator for customer loyalty (Sirdeshmukh et al.2002). Both wait time between arrival to triage andbetween triage to being seen by the staff were shown tosignificantly drive the generation of responsiveness butwith a different extent. The finding of the fact that the waitafter triage has a stronger impact on patient perceivedvalue has an important strategic implication. The ERshould consider shortening the wait after triage moreimportant than shortening the wait before triage. As to thevalue of empowerment, only information connectednesswas shown to drive it but not with a statistically significantextent.

Both nonverbal interaction and personal greeting inan experience created values of respectfulness andresponsiveness. However, none of their relationships wasfound to be significant. These insignificant relationshipsseem to indicate that (1) the emergency department (ER)context is a unique one that their patients are typically inan acute state of ailment and hence they focus more onhaving the immediate medical problem solved. We needto take caution in generalizing the conclusions of thisstudy to other contexts of patient care, such as inpatientcare; (2) the experiences that can create patient value aregenerated from the memorable, impressive and excep-tional cues in healthcare services (Berry et al. 2002;Haeckel, Carbone, and Berry 2003). If the service meetsbut does not exceed the patient’s expectation or instigatedelight from the patients, the experience may not signifi-cantly be recognized as creating patient value. Referencesare available upon request.

For further information contact:Sandra S. Liu

Purdue University812 W. State Street, Room 318West Lafayette, IN 47907–1262

Phone: 765.494.5645Fax: 765.494.0869



PRE- AND POST-MARKET LIBERALIZATION CUSTOMER SERVICEQUALITY PERCEPTIONS OF AN INCUMBENT MONOPOLY: THE

CASE OF THE SAUDI TELECOMMUNICATIONS INDUSTRY1

Alhassan G. Abdul-Muhmin, King Fahd University of Petroleum & Minerals, Saudi Arabia

SUMMARY

Market liberalization can be viewed simply as aprocess of getting to a freer market situation (van deMerwe 1987). It describes the industrial organizationphenomenon where a previously non-competitive marketoccasioned by regulation is deregulated, allowing addi-tional suppliers. According to the structure-conduct-performance (SCP) paradigm of industrial organization(Tirole 1993), market liberalization is expected to result inpositive economic outcomes for an industry and theindividual firms that comprise it, including the incumbentmonopoly. Several such positive outcomes have beenreported in the literature, including industry growth (e.g.,Singh 2000), quality improvements (e.g., Yan 1999),increased innovation (e.g., Markard and Truffer 2006),and profitability growth (e.g., Cabanda and Ariff 2002;Park, Li, and Tse 2006; Jeng and Lai 2008). For serviceorganizations, other desirable marketing-related outcomesto strive for are customers’ quality perceptions, satisfac-tion, loyalty, and attitudes are also. A relevant question iswhether market liberalization can positively impact thesevariables as well, particularly for an incumbent monopolyprovider. The present paper addresses this question. Itexamines whether liberalization has resulted in changes incustomers’ perceptions of Saudi Telecom Company (STC),the incumbent monopoly provider.

In Saudi Arabia, liberalization of the telecommunica-tions market began in 2002 with the award of four VSATlicenses to competitors. This was followed in 2004 by theaward of two data licenses and a second mobile license toItisalat of the United Arab Emirates. In 2005, Saudisexperienced for the first time the intrigue of a secondtelecom provider as Itisalat’s Mobily began operations. In2007/2008, the entry of a third mobile provider (ZainTelecom) further consolidated the liberalization process.

Liberalization has undoubtedly increased competitionin the Saudi telecom market, and anecdotal evidence alsosuggests that STC has significantly changed its marketbehavior as a result. It has implemented massive post-liberalization tariff reductions, expanded its serviceofferings, regularly offers promotions, and undertakenmany other market-oriented initiatives. The SCP paradigmwould suggest improvement in customer-relatedperformance variables given these initiatives. The studytherefore specifically examined whether market

liberalization has impacted (1) customers’ qualityperceptions, satisfaction, loyalty, and attitudes towardSaudi Telecom (STC), and (2) structural relationshipsamong quality, satisfaction, loyalty, and attitudes. It testedhypotheses that: (1) customers’ quality perceptions,satisfaction, loyalty and attitudes toward the incumbentwill be higher post-market liberalization; (2) service qualityperceptions will positively impact satisfaction with theincumbent both pre- and post-market liberalization, andsatisfaction will positively impact loyalty toward theincumbent both pre- and post-market liberalization, and(3) attitudes toward the incumbent will positively moderatethe relationship between quality and satisfaction andbetween satisfaction and (3) attitudes will positivelymoderate these relationships

The hypotheses are tested using data from two wavesof a tracking survey conducted in 2004 before Mobilystarted operations (wave 1, pre-liberalization data set) and2008 after Zain started operations (wave 2, post-liberalization). Both surveys used the same questionnaireto measure quality, satisfaction, loyalty, and attitudes.Service quality was measured using the SERVQUALdimensions (Parasuramnan et al. 1988 and related pa-pers).

Independent samples T-tests are used to comparemean scores across the two waves and examine the impactof market liberalization. Ordinary least squares regressionis used to examine structural relationships among theconstructs in each wave and to examine changes acrosswaves. The results show statistically significant improve-ments in the key constructs between the pre- and post-liberalization periods. Structural relationships among theconstructs are also largely as expected. However, themoderating effects of attitude are not as expected. Liber-alization also appears have resulted in interesting changesin structural relationships among the constructs. Specifi-cally, the impact of responsiveness and assurance qualitydimensions on satisfaction are higher post-liberalization;that of reliability and tangibles dimensions are lower post-liberalization; and that of satisfaction on loyalty is lowerpost-liberalization.

The study concludes that, in terms of how it isperceived by consumers, liberalization has actually ben-efitted the incumbent. A general implication is that,although incumbent monopolies have an understandable


tendency to not want market liberalization, under certainconditions this can actually be good for them. The studyalso concludes that theoretically, liberalization’s positiveeffect is not only in economic variables. Perceptual

related variables also benefit. Finally, liberalization maynot change only industry dynamics; structural relation-ships among the perceptual constructs investigated in thisstudy may also change.

ENDNOTE

1 The author acknowledges King Fahd University ofPetroleum & Minerals for using its various facilitiesin the research reported in this paper.

REFERENCES

Cabanda, Emily and Mohamed Ariff (2002), “Perfor-mance Gains Through Privatization and Competitionof Asian Telecommunications,” ASEAN EconomicBulletin, 19 (3), 254–79.

Jeng, Vivian and Gene C. Lai (2008), “The Impact ofDeregulation on Efficiency: An Analysis of LifeInsurance Industry in Taiwan from 1981 to 2004,”Risk Management and Insurance Review, 11 (2),349–75.

Markard, Jochen and Bernhard Truffer (2006), “Innova-tion Processes in Large Technical Systems: Market

Liberalization as a Driver for Radical Change,”Research Policy, 35, 609–25.

Parasuraman, A., Valerie A. Zeithaml, and Leonard L.Berry (1988), “SERVQUAL: A Multiple-Item Scalefor Measuring Consumer Perceptions of ServiceQuality,” Journal of Retailing, 64 (Spring), 12–40.

Park, Seung H., Shaomin Li, and David K. Tse (2006),“Market Liberalization and Firm Performance Dur-ing China’s Economic Transition,” Journal of Inter-national Business Studies, 37, 127–47.

Singh, J.P. (2000), “The Institutional Environment andEffects of Telecommunication Privatization andMarket Liberalization in Asia,” TelecommunicationsPolicy, 24, 885–906.

Tirole, Jean (1993), The Theory of Industrial Organiza-tion. Cambridge, MA: The MIT Press.

van de Merwe, Sandra (1987), “Deregulation in Servicesand the Marketing Challenge,” The Service Indus-tries Journal, 7 (1), 24–34.

For further information contact:Alhassan G. Abdul-Muhmin

Department of Management and MarketingKing Fahd University of Petroleum & Minerals

P.O. Box 1185Dhahran 31261Saudi Arabia

Phone: +966.3.860.4265Fax: +966.3.860.2544

E-Mail: [email protected]: http://faculty.kfupm.edu.sa/MGM/alhassan/


ROADS TO CUSTOMER PERFORMANCE: INVESTING WISELY IN THESERVICE-PROFIT CHAIN

Barbara Caemmerer, University of Strathclyde, United KingdomHeiner Evanschitzky, University of Strathclyde, United Kingdom

SUMMARY

The majority of Service-Profit Chain (SPC) studieshave been carried out in the private sector. The main focusof investigation has been on the link between employeeand customer attitudes (such as satisfaction), customerbehavior (such as share of wallet) and the contribution tothe service provider’s performance (such as profitability).Our study contributes to the extant knowledge by depart-ing from this traditional line of enquiry as follows:

1. Study context. We chose a public sector organiza-tion, a national economic development agency, asthere is a strong need for a better understanding howprivate sector practices embraced in the public sectorcontribute to the delivery of public services. More-over, unlike previous studies carried out in servicesettings with low complexity, the economic develop-ment agency we chose offers complex interventionand consultancy services with its business customers,typically small- and medium-sized companies.

2. Performance measures. Unlike previous SPC stud-ies, we do not use the service provider’s performanceas the ultimate outcome variable. Instead – and due tothe context of our study – we use customer outcomesas a critical performance criterion for such publicservice providers. Moreover, we use subjective (e.g.,customer satisfaction) as well as objective (e.g., per-formance of customers’ business) measures to assesscustomer outcomes as a result of the interventionwith the public service provider.

3. Level of analysis. As the organization delivers itsservices through branches, we use hierarchical linearmodeling (HLM) to account for individual employeeand customer data being nested at a branch level(level 2). We do not only explain drivers of employeeand customer outcomes, but also systematic vari-ances across branches that occur due to contextualdifferences with regard to the implementation ofmanagement practices in and characteristics (such asfunding and staff-customer ratio) of these branches.

In order to address our research objectives, (1) toexamine how employee advocacy can be establishedthrough internal organizational initiatives and how itimpacts on customer outcomes, and (2) to assess the

relative impact of internal and external organizationalinitiatives on customer outcomes, we conduct two multi-level studies. We use objective data on organizationalinitiatives for multiple branches (n = 12) of an economicdevelopment agency as well as survey data from multipleemployees (n = 725) and multiple customers (n = 985)nested within the 12 branches. Each employee as well aseach customer can be matched to a particular outlet. Basedon this we estimate the following model:

Level 1 Model:

ADVOC = βo + β

1 (STRAT) + β

2 (VALUE)

+ β3(MS) + r

ADVOC = employee advocacy of products/services offered

STRAT = service strategyVALUE = service values

MS = management supportSP = service performance

Level 2 Model:

βo = γ

00 + γ

01 (MARKO) + γ

02 (SERVO)

+ γ03

(SOLO) + γ04

(FUNDING) + γ05

(S/CRATIO) + γ06

(INTSPEND) + μo

MARKO = external organizational initiatives that belongto a dimension of the market orientation construct.

SERVO = external organizational initiatives that belongto a dimension of the service orientation construct.

SOLO = external organizational initiatives that belong toa dimension of the solution orientation construct.

FUNDING = change in total funding between previousand this year.

S/CRATIO = staff/customer ratio.INTSPEND = total spending on internal administration

and management.

Our study provides empirical evidence for the cen-trality of the employee advocacy construct in the SPC incomplex service settings, as it is significantly related tocustomer outcomes. Moreover, average employee advo-cacy moderates the impact of service performance oncustomer outcomes. This might be due to the fact thatcomplex, highly intangible services have large degrees ofcredence qualities which prevent customers from assess-ing the quality of the advice given. Only if employees

(1)

(2)


really believe in the quality of their services they deliver,customers will accept the advice given which might leadto an increase in their performance. In support of this, theresults also reflect that external organizational initiativeshave no impact on customer outcomes, whereas internalorganizational initiatives directed at employees – medi-ated by employee advocacy – have a significant effect oncustomer outcomes. Therefore, more effective serviceimprovements may be achieved through organizationalinitiatives that are directed at employees, not at customers.Moreover, our findings suggest that from all internalorganizational initiatives that are directed at employees,those in support of solution orientation have the strongestimpact on employee advocacy. Service orientation initia-tives have little, whereas market orientation initiativeshave a negative impact. In a sense, this finding supportsthe paradigm shift to a service-dominant logic in whichcustomers are perceived as co-creators of value. Thisencompasses a much more tailored and integrated approachto services in which employees become more facilitatorsthan deliverers and may be particularly relevant in com-plex services.

The findings also show that particular branch charac-teristics have a positive impact on employee advocacy.While it is intuitive that an increase in funding and a morefavorable staff/customer ratio positively influence employ-ees’ perceptions of the organization they work for, it issurprising that the same characteristics have no directimpact on customer outcomes. Instead, this link is medi-ated by employee advocacy. This stresses the importanceof the branch environment as it influences how employeesperceive the organization they work for and the servicesthey deliver. Surprisingly, our data also suggest that anincrease in spending on internal administration and man-agement has a negative impact on employee advocacy. Itis possible that, in the context of the organization, anincrease in administrative spending is related to anincrease in internal reporting procedures and processesthat negatively impact on employee perceptions of orga-nizational service commitment. As our study is one of thefirst to investigate this modified SPC concept in a complexpublic service setting, overall more research is required tounderstand whether the findings reported here can bereplicated. References are available upon request.

For further information contact:Barbara Caemmerer

University of Strathclyde173 Cathedral StreetGlasgow, G41 3EL

United KingdomPhone: 0044.141.5483196



AN ASSESSMENT OF THE LONG-TERM ORIENTATION SCALE:A REPLICATION AND EXTENSION BASED ON FINDINGS

ACROSS 10 EUROPEAN COUNTRIES

Louise May Hassan, Heriot-Watt University, United KingdomEdward Man Kee Shiu, University of Strathclyde, United Kingdom

Gianfranco Walsh, University of Koblenz-Landau, Germany

SUMMARY

There has been great debate on the structure ofculture, its measurement and effect on behavior for manydecades. Hofstede (1980) provided one of the seminalworks of our time by developing a theory of culture basedon four dimensions and devising a means of measuringcultural orientation, with the addition of a fifth dimensionnow conceptualized as long-term orientation (Hofstede2001). However, investigations of the psychometric prop-erties of Hofstede’s scales have revealed shortcomings(Spector et al. 2001) such as the fact that Hofstede’smeasures are primarily for aggregate-level analysis andare therefore not appropriate for individual-level analysis(Bond 2002).

The criticism directed at Hofstede’s measures has ledto the development and use of alternatives. One suchexample of an alternative cultural value scale is that oflong-term orientation developed recently by Beardenet al. (2006), and for the purpose of this study termed theLTO scale. This scale developed and tested with over2,000 respondents across four countries (Argentina, Aus-tria, Japan, and the U.S.) affords the potential to captureLTO at an individual-level. The scale contains twosubscales (tradition and planning) measured with fouritems each and has been found to explain variance (R2 upto .14) in personal ethical values at the individual-level(Nevins et al. 2007). While Bearden et al. (2006) haveprovided evidence that the LTO scale replicates relativelywell across independent country samples, an independentassessment of the psychometric properties and validity ofthis scale across a wider range of countries is needed.

Based on the time-oriented conceptualization of long-term orientation, Bearden et al. (2006, p. 457) define LTOas “the cultural value of viewing time holistically, valuingboth the past and the future rather than deeming actionsimportant only for their effects in the here and now or theshort term.” Given this definition, a high score in LTOwould suggest the individual values of planning, tradi-tion, working hard and persevering for future benefits.Bearden et al. (2006) undertook initial item generationand reduction while examining content validity, they thenconducted a series of six empirical studies examiningdimensionality, scale reliability including test-retest reli-

ability, measurement invariance as well as convergent,discriminant, predictive and nomological validity.

Their results show that the LTO subscales have anadequate level of reliability where Cronbach alpha arereported to be above .71 for the dimension tradition andabove .62 for planning across their studies. Test-retestcorrelation for tradition and planning are .78 and .69respectively. Further, across the studies conducted,Bearden et al. (2006) evidenced convergent and discrimi-nant validity for the two dimensions and establishedpredictive and nomological validity in terms of the pre-dicted relationships to downstream variables frugality,compulsive buying, personal ethics and cognitiveresponse measures. They established metric invarianceacross two national samples (Japan and the U.S.) in onestudy and partial metric invariance across four nationalsamples (Argentina, Austria, Japan, and the U.S.) inanother study.

In our study a final sample of 3,491 respondentsacross ten EU countries was gained. Sample sizes rangefrom n = 169 (Lithuania) to n = 562 (Greece).

In addition to measuring LTO using the eight-itemscale developed by Bearden et al. (2006), three items areused to capture a measure of individualistic orientationwith two items capturing uncertainty avoidance. Theseitems are based on the scales reported in Mourali et al.(2005) and Reardon et al. (2006) and were used to helpestablish nomological validity.

Based on the analysis of survey data from the tencountries, the 8-item LTO scale performs reasonably wellwith results that lend support for the internal validity of theLTO scale within most of the countries examined. Spe-cifically, both LTO dimensions are reliable in all but onecountry, namely Greece where the planning dimensionhas a Cronbach alpha value below .6. Overall results fromthe Greek sample deviate from the general findings. Interms of dimensionality of the scale all country samplesexcept the Greek support a two-dimensional correlatedmodel. However, we find lower levels of internal consis-tency reliability for the planning subscale with less thanadequate AVE values for most country samples. Themeasurement invariance of the scale was tested using both


classical measurement theory and generalizability theory.Although full metric invariance was not found across theten country samples using the procedures described bySteenkamp and Baumgartner (1998), this is evidencedamong two groups consisting of five and three countriesrespectively. Furthermore, the results of the multi-facetGT analysis also reveal a low proportion of variance due

to country-level effects. Based on these results it can beconcluded that the Bearden et al. (2006) LTO scaleperforms well and is, by and large, invariant across the EUcountries studied. This finding is of paramount impor-tance given the likelihood that the scale will be usedpredominantly in cross-cultural research.

REFERENCES

Bearden, William O., R. Bruce Money, and Jennifer L.Nevins (2006a), “A Measure of Long-Term Orienta-tion: Development and Validation,” Journal of theAcademy of Marking Science, 34 (3), 456–67.

Hofstede, Geert (1980), Culture’s Consequences: Inter-national Differences in Work-Related Values. BeverlyHills, CA: Sage Publications.

____________ (2001), Culture’s Consequences: Com-paring Values, Behaviors, Institutions, and Organi-zation Across Nations. California: Sage.

Mourali Mehidi, Michael Laroche, and Frank Pons (2005),“Individualistic Orientation and Consumer Suscepti-bility to Interpersonal Influence,” Journal of ServicesMarketing, 19, 164–73.

Nevins, Jennifer L., William O. Bearden, Bruce Money

(2007), “Ethical Values and Long-Term Orienta-tion,” Journal of Business Ethics, 71, 261–74.

Reardon, James, Chip Miller, Bram Foubert, Irena Vida,and Liza Rybina (2006), “Antismoking Messages forthe International Teenage Segment: The Effective-ness of Message Valence and Intensity Across Dif-ferent Cultures,” Journal of International Marketing,14 (3), 115–38.

Spector, Paul E., Cary L. Cooper, and Kate Sparks (2001),“An International Study of the Psychometric Proper-ties of the Hofstede Value Survey Module 1994: AComparison of Individual and Country/ProvinceLevel Results,” Applied Psychology, 50 (2), 269–81.

Steenkamp Jan-Benedict E.M. and Hans Baumgartner(1998), “Assessing Measurement Invariance in Cross-National Consumer Research,” Journal of ConsumerResearch, 25, 78–90.

For further information contact:Louise M. Hassan

Heriot-Watt UniversityEdinburgh, Scotland

United KingdomE-Mail: [email protected]


CROSS-CULTURAL EXAMINATION OF ANTECEDENTS OF SALESMANAGER EFFECTIVENESS: A STUDY OF SALESPEOPLE

IN SIX COUNTRIES

Ning Li, George Mason University, FairfaxWilliam H. Murphy, University of Saskatchewan, Saskatoon

SUMMARY

The sales manager is “the critical link between thevision of an organization’s upper management and thefirm’s revenue-generating personnel, the sales force” andimpacting salesperson performance to a great extent fallson the shoulders of the sales manager (Deeter-Schmelz,Kennedy, and Goebel 2002, p. 617). Considering thecritical role played by sales managers, there is surpris-ingly limited research focused on sales manager effective-ness. Given this gap, the first purpose of our work is tobring greater clarity to the under-researched construct ofsales manager effectiveness. Based on path-goal theory(House 1971), the contingency model of leadership effec-tiveness (Fiedler 1967), and the sales management workby Deeter-Schmelz et al. (2002, 2008), five antecedents tosales manager effectiveness were explicated: trust, rela-tionship to manager, training and coaching, empower-ment, and frequency of sales managers working with theirsalespeople. Additionally, with so many companiesactively managing sales forces in multiple countries, oursecond purpose is to understand whether relationshipswould vary, or be universal, across cultures. UsingHofstede’s (1980) national culture framework, wehypothesized why particular antecedent-to-effectivenessrelationships would be moderated by power distance,uncertainty avoidance, masculinity, and individualism.

Our six country study included salespeople fromAustralia, Canada, China, Mexico, New Zealand, and theUnited States (N = 1,094). Salespeople in each countrycompleted surveys (provided in English except China andMexico). Where possible, existing measures were usedand most used 7-point scales. Analysis included tests forreliability, validity, cross-culture measurement equiva-lence, and model testing using OLS regression.

In sum, our adjusted model R2 was .77. We found thattrust (limited to the honesty and integrity dimension) hasuniversally positive effects on sales manager effective-ness. Thus, sales managers in any country must be recruitedfor their ability to lead with earning trust at the forefront,giving salespeople confidence in the honesty of salesmanagers’ comments and directions.

As to other antecedents, we found positive maineffects for closeness of a salesperson’s relationship to his

or her supervisor, training and coaching provided by salesmanagers, empowerment received by a salesperson, andfrequency the sales manager works with a salesperson.However, these antecedents were also moderated bynational culture dimensions. This suggests that seniorexecutives might be able to tap a source of competitiveadvantage by understanding how to engage sales managerresources appropriately across cultures. For instance, wefound that closeness of a salesperson’s relationship totheir sales manager has a greater positive effect on effec-tiveness in high individualism and in low power distancecountries. The rationale for finding this effect in lowpower distance countries is that in these countries thepsychological gap between managers and subordinates isexpected to be small, a condition converse to that of high-power distance countries. Thus, a closer relationship withtheir sales manager would be more expected by sales-people in low power distance countries.

Training and coaching have always been viewed asa positive sales manager role. As expected, we found thattraining and coaching is of greater importance in highuncertainty avoidance countries. This finding challengessales executives to give special emphasis to their salesmanagers in these countries so as to engage purposefullyin training and coaching activities. Additionally, wefound that the positive effect of empowerment on salesmanager effectiveness is less in high uncertainty avoidancecountries.

Frequency of a sales manager working with asalesperson was found to have an inverted-U effect oneffectiveness. Finally, we found that salespeople fromcultures with high masculinity perceive sales managersmore effective (relative to salespeople from lowmasculinity cultures) when sales managers work withthem less frequently.

Our findings provide researchers and sales execu-tives a better understanding of how five factors influencesalespeople’s perception of sales manager effectiveness.Further, the finding that culture moderates a number ofantecedent – to effectiveness relationships provides criti-cal knowledge for sales managers. This study includes sixcountries with a wide range of national culture scores,indicating that our findings may prove applicable wellbeyond the particular countries studied. We hope our


research gives other scholars direction for continuedstudy into sales manager effectiveness and the effects of

national culture when managing sales forces throughoutthe world. References are available upon request.

For further information contact:Ning Li

School of ManagementGeorge Mason University

4400 University Drive, MSN 5F4Fairfax, VA 22030

Phone: 703.993.1227Fax: 703.993.1809



CROSS-CULTURAL DIFFERENCES IN ONLINE BRANDCOMMUNITIES: AN EXPLORATORY STUDY OF INDIAN

AND AMERICAN ONLINE BRAND COMMUNITIES

Vivek Madupu, Missouri Western State University, St. JosephDelonia Minor-Cooley, Texas Southern University, Houston

SUMMARY

There is a growing interest among marketers aboutonline brand communities as they are perceived to bringmany benefits to firms. Research on brand communitieshas shown that higher level of participation and long termmembership increases the likelihood of adopting a newproduct from the preferred brand (Thompson and Sinha2008), positively influences members’ commitment to thebrand (Casalo, Flavian, and Guinaliu 2008), creates oppo-sitional loyalty where members actively reject competi-tors’ brands (Muniz and O’Guin 2001) and positivelyimpacts customers brand attitude, loyalty and attachmentto the brand (McAlexander, Schouten, and Koenig 2002;Algesheimer et al. 2005). Such information could beimmensely useful in identifying the shortcomings of theproducts and improve future product offerings.

Online Brand Communities

Muniz and O’Guinn (2001) defined brand communi-ties as “social entities that reflect the situated embeddednessof brands in the day-to-day lives of consumers and theways in which brands connect the consumer to the brand,and consumer to consumer” (p. 418).

Motives for Participation

Based on Dholakia et al. (2004), Dholakia and Bagozzi(2004), and Flanagin and Metzger (2001), we have iden-tified the following five motives for participation in onlinebrand communities. Receiving Information refers to mem-bers browsing the brand community website for informa-tion, reading postings by other members or posting mes-sages to get desired information from other members.Giving Information refers to posting messages in responseto other members’ request for information or postingmessages informing readers. Social Integration involvesestablishing and maintaining contact with others (Flanaganand Metzger 2001). Self-Discovery refers to individuals’“need to understand and deepen salient aspects of them-selves through social interactions” (Dholakia et al. 2004).Status Enhancement is the “value that a participant derivesfrom gaining acceptance and approval of other membersand the enhancement of one’s social status within thecommunity on account of one’s contributions to it”(Dholakia et al. 2004).

Consequences of Participation

Several scholars have identified three characteristicsof brand communities (Muniz and O’Guinn 2001; Schauand Muniz 2002). Consciousness of kind is defined as “theintrinsic connection that members feel toward one another,and the collective sense of difference from others not inthe brand community” (Muniz and O’Guinn 2001). Moralresponsibility is defined as “a felt sense of duty or obliga-tion to the brand community as a whole, and to itsindividual members” (Muniz and O’Guinn 2001). Sharedrituals and traditions is defined as “conventions or prac-tices that set up visible public definitions and socialsolidarity and perpetuate the brand community’s sharedhistory, culture, and consciousness” (Muniz and O’Guinn2001).

Cultural Differences Between India and USA

Hofstede (1991) defines national culture as “thecollective programming of the mind which distinguishesthe members of one group or category of people fromanother.” Culture affects how an individual thinks andacts in the world. In collectivist cultures, members tend tobe open and willing to share their knowledge with othermembers of their in-group (Chow et al. 2000).

On the other hand, members of individualistic cul-tures are motivated by self-interest (Markus and Kitayama1991). They view themselves as independent and givepreference to their own thoughts and goals.

Discussion

The purpose of this study was to understand if nationalculture has an impact on the motivations for participation,and in community characteristics that are developed be-cause of participation in online brand communities. Basedon the extant literature on Hofstede’s individualism andcollectivism dimension, we argued that for members fromcollectivistic cultures receiving information, giving infor-mation, social integration, self discovery, and statusenhancement should be stronger than members fromindividualistic cultures; however, receiving informationshould be a stronger motive for members from individu-alistic cultures. We also argued that members of onlinebrand communities in collectivistic cultures will exhibit


stronger conscious of kind, moral responsibility, andshared rituals and traditions than members of brand com-munities in individualistic cultures. We collected datafrom two different brand communities, one located inIndia and the other in USA. The results of our studyprovided empirical support for all but one of the hypoth-eses.

Although virtual communities, including online brandcommunities, are modern day socializing places, it ismore so in the case of the collectivistic cultures. For onlinebrand communities in individualistic cultures, managersshould engage in giving more information than theypossibly could receive from the members. This type ofexchange could possibly encourage repeat visits andhigher levels of participation.

Limitations and Future Research

As with any study, the present study has some limita-tions. This study focused on online brand communitiesformed around motorcycles which are technical products.As the study is based on a web survey, a selection biasmight have affected our findings. Also, almost all respon-dents in the study were male.

This study offers several suggestions for futureresearch. One area is refining the scales developed in thisstudy. Another area is to study a successful online brandcommunity that has members from different countries.Finally, the present study should be replicated with mem-bers of other online brand communities that vary in size,activity and type of the brand before the results can begeneralized. References are available upon request.

For further information contact:Vivek Madupu

MarketingSteven L. Craig School of BusinessMissouri Western State University

309J, 4525 Downs DriveSt. Joseph, MO 64507Phone: 816.271.4349


Delonia CooleyMarketing

Jesse H. Jones School of BusinessTexas Southern University

3100 Cleburne St.Houston, TX 77004

Phone: 713.313.1305Fax: 713.313.7722



WHY DO MANAGERS GIVE PRICE CONCESSIONS? THE ROLE OFORGANIZATIONAL PRICE PERCEPTIONS

Christian Homburg, University of Mannheim, GermanyDirk Totzek, University of Mannheim, Germany

SUMMARY

Prior work in economics, management, and market-ing has identified structural (i.e., objective) drivers ofprice competition – such as market growth, market con-centration, or competitive intensity. Firms’ subjectiveprice perceptions as a driver of competitive pricing actionshave largely been neglected. However, these perceptionsare often selective and biased. For example, firms oftenselectively focus on evaluating their performance relativeto their chief competitors and overemphasize beatingthem. Furthermore, managers often engage in overlyaggressive pricing actions, which might provoke retalia-tion and intensify price competition.

Firms’ subjective price perceptions should notablymatter in business-to-business markets and in situationswhere prices are negotiated and fixed for each deal.However, the questions of how and to what extent selec-tive and biased organizational price perceptions shape amanager’s intensity of price concessions in order to closedeals have largely been neglected. This lack of research isremarkable as quick price concessions in the face ofcompetitive challenges and in order to close deals areconsidered the major pricing obstacles in business-to-business markets. Thus, a deeper understanding of orga-nizational price perceptions provides important insightsas to why managers engage in price cutting.

Based on the literature on managerial and organiza-tional cognition and on institutional theory, we argue thatprice-related organizational cognitions shape price-related managerial behavior, i.e., the intensity of priceconcessions of senior managers with primary pricingresponsibility when negotiating prices. We suggest thatorganizational price perceptions consist of three vari-ables. First, we consider perceived price cutting momen-tum as the firm’s perceived external pressure to discount.Second, perceived price aggressiveness captures the extentto which a firm considers its own price behavior asaggressive and as aggressively countering price cuts ofcompetitors. Third, the institutionalization of price per-ceptions captures the degree to which the organizationalperception of the competitive pricing environment isinstitutionalized and no longer questioned. We furtheraddress how the intensity of price concessions affects firmperformance. Finally, as both structural industry factorsand organizational perceptions should affect a manager’srebating behavior, we test our hypotheses by statistically

controlling for the effects of major structural variables –such as competitive intensity, the customers’ price sensi-tivity, or market growth.

We test our framework by drawing on a cross-indus-try survey sample of 230 business units. The samplecontains a broad range of industries (among others: machin-ery, automotive, consulting, and IT/telecommunications),and most of the informants are head of sales or head of thebusiness unit/managing director.

Results of our analyses show that the firm’s per-ceived aggressiveness and the perceived price cuttingmomentum lead to a higher degree to which managerswith pricing responsibility give price concessions whennegotiating prices with customers. As a result, a firm’sprice-related success, and ultimately return on sales, arenegatively affected. The same holds for the institutional-ization of price perceptions, which also leads to a higherintensity of price concessions. These effects are observedwhile controlling for major structural drivers of pricecompetition.

The key contribution of our study is that it advancesour knowledge on pricing processes in business-to-business firms. Our study shows that managerial pricingdecisions are not only affected by objective (i.e., hard)factors but are largely determined by cultural-cognitive(i.e., soft) factors at the organizational level. More specifi-cally, shared beliefs on the competitive pricing landscapeand incrusted price perceptions affect managerial pricingbehavior.

From a managerial point of view, firms mustacknowledge that soft factors affect their senior managers’pricing decisions and that the firm’s pricing processes –notably in business-to-business markets – must accountfor these intangible and cultural factors. As a result, firmsshould frequently question them in order to prevent pricingactions which destroy margin and lead to lower firmprofits. More specifically our results imply that firms needto assess their prevailing perception of their own as wellas their competitors’ aggressiveness with respect to pricingand whether this perception is really true. Firms shouldalso carefully examine whether organizational priceperceptions are incrusted and no longer questioned bytheir executives and their salespeople. The collection anddissemination of price-related information on thecompetitive landscape should be important in this context.


For further information contact:Christian Homburg

Marketing DepartmentUniversity of Mannheim


Phone: +49.621.181.1555Fax: +49.621.181.1556



THE INFLUENCE OF A SALESPERSON’S DIALECT: IS REGIONALDIALECT A DRIVER OR AN OBSTACLE IN PERSONAL SELLING?

Robert Mai, Dresden University of Technology, GermanyStefan Hoffmann, Dresden University of Technology, Germany

SUMMARY

The regional dialect of a salesperson is commonlynegatively connoted and inherently associated with adeficient divergence from standard speech (Carlson andMcHenry 2006). Hence, there has been a recent growth ofaccent and dialect reduction services offered in order toenhance comprehensibility of employees (Rogers, Dalby,and Nishi 2004). Reviewing the literature of relevantbusiness studies reveals that virtually no study has beenundertaken to examine a speaker’s regional dialect in abusiness context. To fill this void, this study focuses on theimpact of a salesperson’s regional dialect on customer’sperception and attitude in personal selling. This approachis based on findings from the fields of linguistics andsociology. The study constitutes, in which way a dialectcan be a driver or an obstacle in persuasive communica-tion to shape customer attitudes.

The study is based on industrial buyers evaluatingactual sales pitches. Overall, ANOVA results suggest thatrespondents do not devaluate salespersons with a regionaldialect per se. When customers are interacting with a salesrepresentative who speaks with regional dialect variouseffects of dialect come into play and overlay the process-ing of the message. In this regard, speech quality is key toexplain listener’s perception of a speaker. Listeners whoare exposed to a salesperson with a low perceived qualityof speech are equally satisfied with the salespersonregardless whether or not he speaks with a regionaldialect. In accordance with confirmation/disconfirmation-paradigm (Oliver 1997) salespersons with a regionaldialect, however, benefit from a high quality of speech,since they are exceeding stereotyped expectations of thelisteners toward the dialect speakers.

The empirical study further indicates that stereotypi-cal associations with the dialect and its speakers stronglyaffect satisfaction with the salesperson: In correspon-dence with previous findings on linguistic stereotypes(Hall 1980; Puntoni et al. 2009), it is shown that afavorable dialect positively influences customer judg-ments of the salesperson. More importantly, in line withAccent Prestige Theory (Giles 1970) stereotypical beliefsabout traits of the people who generally speak a certain

regional dialect explain a great portion of the customers’satisfaction with the salesperson.

Whereas stereotypical associations with the dialectand its speakers strongly affect perception of the salesper-son, convergence of the speaker and degree of dialectstrength essentially influence customers’ purchase inten-tion and, partially, their satisfaction with the company. Incongruence with Social Identity Theory (Tajfel 1981), aspokesperson with a similar dialect outperformed otherswho are obviously being categorized as members of theout-group. The current findings also imply that strength ofperceived regional dialect has a non-linear impact (invertedU-shaped function) on purchase intention of a customer.The results highlight that a salesperson with a weakregional dialect is more favorable than a speaker with nosuch coloring or a maximally perceived regional dialect.

This study needs to be based upon additional research,because of the study’s sample size. Moreover,generalizability of the study needs to be validated forlanguages other than German and different cultures. Theauthors of this article further elaborate this field by exam-ining foreign and regional dialects.

In summary, the study offers new arguments for thedebate on reduction of regional dialects. At the mostgeneral level, it was shown that regional dialect of aspeaker is not necessarily a disadvantage in personalselling. Hence, dialect modification therapy should ratherfocus on improving quality of verbal expressions (such asvoice volume, tone, hesitation, etc.) instead of hiding aregional speech pattern or changing it. Marketers need toconsider the following five criteria when judging theemployability of a sales representatives with a regionaldialect: First, is the salesperson’s speech of high qualityalthough he speaks with a regional dialect? Second, is thedialect of moderate strength? Third, is the dialect similarto the ones of the targeted customers? Fourth, is the dialectperceived positively? And fifth, do customers have posi-tive stereotyped associations (e.g., positive image) towardthe ethnic group who speaks the regional dialect? If thesalesperson meets all those criteria he will be as persua-sive – and even more persuasive – as a sales representativewho speaks standard speech.


REFERENCES

Armstrong, Robert W. and Siew Min Yee (2001), “DoChinese Trust Chinese? A Study of Chinese Buyersand Sellers in Malaysia,” Journal of InternationalMarketing, 9 (3), 63–86.

Carlson, Holly K. and Monica A. McHenry (2006),“Effect of Accent and Dialect on Employability,”Journal of Employment Counseling, 43 (2), 70–83.

Giles, Howard (1970), “Evaluative Reactions to Accents,”Educational Review, 22 (3), 221–27.

Puntoni, Stefano, Bart De Langhe, and Stijn M.J. VanOsselaer (2009), “Bilingualism and the EmotionalIntensity of Advertising Language,” Journal of Con-sumer Research, 35 (6), 1012–25.

Rogers, Catherine L., Jonathan Dalby, and Kanae Nishi(2004), “Effects of Noise and Proficiency on Intelli-gibility of Chinese-Accented English,” Language &Speech, 47 (2), 139–54.

Tajfel, Henri (1981), Human Groups and Social Catego-ries: Studies in Social Psychology. New York: Cam-bridge University Press.

For further information contact:Robert Mai

Department of Business Management and EconomicsDresden University of Technology

01062 DresdenGermany

Phone: +49(0)351.463.36021Fax: +49(0)351.463.37176



FRANCHISING: A REVIEW OF LITERATURE FROMMARKETING PERSPECTIVE

Selcuk Ertekin, University of North Texas, Denton

ABSTRACT

The literature on franchising is replete with researchconducted in various disciplines including economics,law, management, marketing, and management science.However, there is no literature review paper that approachto the topic from a marketing perspective. In this, study weattempt to address this gap by categorizing the literature inorder to identify shortages and surpluses of research. Weidentified three major streams to categorize the relevantarticles: Price, Trust, and Authority. The overarchingfinding of this study is that franchising research is notsaturated from a marketing perspective.

INTRODUCTION

Franchising activities in the U.S. generated over$2.31 trillion of annual output in 2005, which accountedfor approximately 45 percent of total retail sales. There isa tremendous growth in the franchising market comparedto a decade ago, with numbers doubling in terms of shareand volume. More than 75 industry sectors use franchis-ing to distribute goods and services to consumers, and thisprovides direct employment for more than 11 millionAmericans (International Franchising Association 2005).Overall, franchising was among the most preferred mar-ket expansion strategies (Cateora, Graham, and Graham2006).

The literature is replete with research on franchisingconducted in various disciplines including economics,law, management, marketing and management science.Elango and Fried (1997) provided a comprehensive reviewby incorporating all of these disciplines. Their study isbroad but somewhat limited from a marketing disciplinestand-point. Also, many important articles have beenpublished in major marketing journals during the lastdecade, and there is a need to review the existing state ofmarketing research on franchising.

METHODOLOGY

The purpose of this study is to approach franchisingfrom a marketing perspective and to review the literaturein order to identify shortages and surpluses of research. Athree-part data collection process was used. First, stan-dard electronic search procedures were used to access theABI/Inform, Ebsco Host, Business Source Premier, Dis-sertation Abstract, and Science Direct databases. We usedkeyword combinations of “franchising,” “franchise sys-

tems,” and “franchisor” to drive the search. In addition,Journal of Marketing, Journal of Marketing Research,Marketing Science, Management Science, and Journal ofBusiness Research were manually searched. Finally, thereferences of the collected studies were searched for otherarticles directly contributing to our understanding offranchising from the B2B marketing perspective. Overall,forty-one articles were classified as relevant and aresummarized in Appendix 1. Table 1 lists the number ofarticles sourced from each journal searched. Brief topicalsummaries were prepared on each identified article, andthese were used to analyze the literature in order toidentify major research streams.

RESULTS

We fitted the identified articles into three groups:Price (Contracts and Ownership), Trust (Relational Con-tracting), and Authority (Conflict and Opportunism). Thesethree groups map into two major categories: Markets andHierarchies as shown in Figure 1. This categorizationscheme is grounded on theoretically based approachesthat are developed in academic research to categorizevarious forms of industrial organization, particularlyBradach and Eccles (1989). In fact, the major categoriesof “markets” and “hierarchies” stem from Coarse (1937)’sclassic article about the origins of markets and hierarchiesto describe the organization of industry. According to thisapproach “the operation of market costs something and byforming an organization and allowing some authority (anentrepreneur) to direct resources, certain marketing costsare saved” (Bradach and Eccles 1989, p. 99). The market-ing costs include the costs associated with determiningrelevant prices and negotiating and concluding contracts.The control mechanisms of price, authority and trust mapon market, hierarchy and relational contracting (relation-ships) all of which are concepts that must be treatedindependently to organize industries (Bradach and Eccles1989).

The markets and hierarchies’ dichotomy is based onthe idea that transactions are governed by the institutionalarrangement that is most efficient. The three dimensionsof transactions dictate whether markets or hierarchies aremost efficient are: uncertainty, asset specificity, and fre-quency. These dimensions affect the costs associated withwriting, executing, and enforcing contracts, when suchcosts are high, markets fail and hierarchies emerge in theirplace. Here, uncertainty refers to the uncertainty aboutindividual performance in groups and uncertainty about


future contingencies. Asset specificity dimension statesthat when highly specific assets are involved in the trans-action, this may lead to post-contractual opportunisticbehavior. As a result of all these two, frequent (andtherefore costly) recontracting occurs. In these instances,the installation of an authority relation (hierarchy) betweencontracting parties is less costly than an arms-lengthmarket transaction. Thus, markets and hierarchies aremutually exclusive means of allocating resources (Bradachand Eccles 1989; Williamson 1985).

According to Bradach and Eccles (1989), a growingbody of research later documented the nonmarket andnonhierchical organizational forms – forms typically saidto reside between markets and hierarchies (Richardson

1972; Eccles 1981; Williamson 1985). This stream ofresearch highlighted the fact that relations between inde-pendent exchange partners can be stabilized throughformal (written contracts) and informal mechanisms (trust).These arrangements enable firms to gain access to know-how unavailable in-house, spread the risks associatedwith uncertain ventures; benefit from economies of scale;enter new product and geographic markets rapidly; man-age interorganizational dependencies; and respond quicklyand flexibly to changing circumstances.

We identified all of these issues in our review ofliterature in franchising from a marketing perspective. Weobserved that Bradach and Eccles (1989) scheme waseffective to categorize the literature as shown in Figure 1.

TABLE 1List of Reference Sources and Number of Entries and Year of First Publication

Source n % Year

California Management Review 1 2.1 1987

European Journal of Marketing 5 13.0 1999

Harvard Business Review 1 2.1 2003

International Journal of Research in Marketing – – –

Industrial Marketing Management 1 2.1 1997

Journal of Advertising – – –

Journal of Advertising Research – – –

Journal of Business – – –

Journal of Business-to-Business Marketing – – –

Journal of Business and Industrial Marketing 1 2.1 2002

Journal of Business Research 3 6.5 1996

Journal of Consumer Research – – –

Journal of Marketing 4 10.8 1976

Journal of Marketing Research 2 4.3 1995

Journal of Product Innovation Management – – –

Journal of Personal Selling and Sales Management – – –

Journal of Small Business Research 1 2.1 1997

Management Science 5 10.8 1994

Marketing Science 4 10.8 1989

Sloan Management Review – – –

The Journal of Consumer Marketing 5 10.8 1998

Organization Science 1 2.1 2001

Journal of Retailing 9 19.5 1990

Total 43 100 –


TRUST (RELATIONSHIPS)

Kalnins and Mayer (2004) state that retail and servicebusiness units will enjoy reduced failure rates if they areaffiliated with experienced multiunit owners andfranchisors. As a result of their relationship with theexperience of individual owners and franchisees, theygain knowledge that is tacit and of value locally. Theyhighlighted the continued importance of such local expe-rience and tacit knowledge in all franchise businesses bystating that franchise systems benefit from their owner’slocal congenital experience but not from distantly gainedexperience.

Hadjimarcou and Barnes (1998) studied the processof international expansion by a relatively small franchisor

attempting to expand internationally using a case study.They highlighted the benefits of finding a partner in thehost country and the strategic alliances created as a resultby small firms attempting to expand internationally.Monroy and Alzola (2002) studied the behavior in thefranchise network from the perspective of quality man-agement. They defined quality in franchise networks as astrategy to focus on customer satisfaction through thecommitment between network partners. They proposed amodel for measuring transactional (short-term) and rela-tional (long-term) quality. They found that the contractdoes not completely reflect the actual conduct of franchisor-franchisee relationships. They suggest that it is necessaryto harmonize the transactional or contractual view withthe relationship view.

FIGURE 1Study Dimensions

MARKETS HIERARCHIES

Price Trust (Relationships):(Contracts): Kalnins and Mayer (2004)

Dahlstrom and Nygaard (1994) Hadjimarcou and Barnes (1998)Lal (1990) Monroy and Azola (2002)

Agrawal and Lal (1995) Doherty and Alexander (2004)Desai and Srinivasan (1995) Perry, Cavaye, and Coote (2002)

Gupta and Loulou (1998) Doherty and Alexander (2006)Shane (2001)

Azoulay and Shane (2001)Desai (1997)Light (1997)

Baroncelli and Manaresi (1997)Chen and Dimou (2005)

Hing (1999)Kaufmann and Dant (2001)

(Ownership): Authority (Conflict and Opportunism):Dant and Kaufmann (2003) Dant and Schul (1992)

Dant, Paswan, and Kaufmann (1996) Srinivasan (2006)Lafontaine and Kaufmann (1994) Hunt and Nevin (1976)

Windsperger and Dant (2006) Kaufmann and Rangan (1990)Burkle and Posselt (2008) Dahlstrom and Nygaard (1999)Anderson and Fok (1998) Gassenheimer, Baucus, and Baucus (1996)

Castrogiovanni and Justis (1998) Windsperger (2004)Schmidt and Oldfield (1999) Kalnins (2004)

Hopkinson and Hogarth-Scott (1999)Fladmoe-Lindquist and Jacque (1995)

Nault and Dexter (1994)Chow and Frazer (2003)

Michael (1999)


Doherty and Alexander (2004) examined franchisor-franchisee relationship in the context of internationalfashion retailing. They studied the mechanisms by whichinternational franchisors control their franchisees in afashion retailing context. They contended that non-coercivesources of power such as supportive, communicative andtrusting relationships are the most successful to control aninternational franchise system. Perry, Cavaye, and Coote(2002) highlighted the fact that most research aboutfranchising relationships concentrated on social bondssuch as trust and commitment. They state that little researchconsiders technical bonds and how they interact withsocial bonds within a relationship. They provided aninformation technology perspective to franchise systemrelationships and investigated how technical bonds ofinformation systems link with social bonds in the relation-ship between a franchisor and a franchisee. The contribu-tion of their study is the finding that franchisor’s increasedtechnical competence from information technologyimproves the social bonds in the relationship. Doherty andAlexander (2006) also studied the mechanisms by whichinternational franchisors control their franchisees in afashion retailing context.

FUTURE RESEARCH ON TRUST(RELATIONSHIPS)

Studies in control in international franchise systemsneeds further research attention. From a methodologicalstandpoint, more qualitative studies are needed on inter-national retail franchising because the majority ofresearch is survey based (Doherty and Alexander 2006).An interesting research question for control in interna-tional franchise systems could be “how can social andtechnical bonds in B2B franchise systems linked together?”(Perry, Cavaye, and Coote 2002).

AUTHORITY (CONFLICT ANDOPPORTUNISM)

Conflict is studied as an authority dimension bynumerous studies. Dant and Schul (1992) analyzed fran-chise channel conflict in the context of fast food restau-rants. Their results revealed a high overall incidence ofpersuasion and problem solving but a preference forbargaining and third party intervention when the disputedissues involve high stakes, complexity and policy con-notations and when the franchisee dependency is ratedhigh.

Srinivasan (2006) concentrated on the performanceof dual distribution systems where the franchisor licensethe operation of some of its units to franchisees whileowning and operating some units itself. Srinivasan (2006)considered the firm’s age, scope of vertical integration,advertising, financial leverage, and financial liquidity as

firm characteristics that influence the relationshipbetween dual distribution performance and intangiblevalue. Hunt and Nevin (1976) focus on deceptive prac-tices in the granting or selling of franchises. They explorefull disclosure laws designed to protect prospective fran-chisees from misrepresentations by franchisors. They findthat full disclosure laws greatly reduce the incidence ofdeceptive practices. Kaufmann and Rangan (1990) stud-ied the amount of marketing activity necessary to mediatethe conflict between the existing franchisees and thefranchisor. They created a model based on theconceptualization of two counteracting effects: spatialcompetition and relative system attractiveness.

Opportunism is studied as an authority issue by anumber of studies as well. Dahlstrom and Nygaard (1999)analyzed the opportunistic behavior in franchise systemsin the context of a Norwegian oil distribution network.They found that opportunistic behavior consistentlyincreases transaction costs whereas cooperative interactioncurbs bargaining costs. They suggested that cooperation,definition of complementary goals and formalization ofmutual obligations can alleviate opportunism in franchisesystems. Gassenheimer, Baucus, and Baucus (1996)combined opportunism to various parameters. They foundsupport for the hypothesis that communication not onlyincreases performance and satisfaction in a franchisesystem but it also moderates the relationship betweenopportunism and satisfaction and opportunism andperformance.

Windsperger (2004) analyzed the allocation of deci-sion rights in franchising relationships. Their main thesisis that the higher the franchisor’s portion of intangibleknowledge assets relative to franchisee’s, the moreresidual decision rights should be assigned to the franchisor,and the higher is the degree of centralization. They foundthat the franchisor’s intangible system-specific know-how and brand name assets have a stronger influence onthe allocation of decision rights in the franchising networkthan the franchisee’s intangible local market assets. Kalnins(2004) framed “encroachment,” which is the case whenfranchisors add new units of their brand proximately totheir franchisees’ existing units, as a loudly complainedform of channel conflict in franchising. Using revenuedata from U.S. lodging industry, Kalnins (2004) foundthat when franchisors approve new same-brand units inthe vicinity of incumbent units, these new units cannibal-ize the incumbents’ revenues.

FUTURE RESEARCH ON AUTHORITY(CONFLICT AND OPPORTUNISM)

For future research on conflict, Dant and Schul (1992)suggested exploring the sequential utilization of alternativebehavioral tactics by channel leaders to resolve dyadic


disputes, and examining the contingent impact of pastdisputes and their outcomes on subsequent choice behavior.They also suggested that future studies may compareconflict resolution mechanisms across alternativegovernance structures from discrete to more relational.

In order to describe the marketing performance per-spective in franchising, there is a need for studies focusedon performance implications of dual distribution systemsacross diverse industry contexts, incorporating environ-mental and organizational characteristics and using in-depth interviews and surveys (Srinivasan 2006)

For future research in opportunism, research thatsimultaneously examines formal structures, informal con-trols, and incentives has potential to develop a morecomprehensive theory for antecedents of transaction costs(Dahlstrom and Nygaard 1999).

There is also a need for a detailed study of communi-cation and opportunism inter-relationship in franchisesystems which incorporate objective measures such as;financial data, market share data, industry assessment,and expanded measures on communication, opportunisticbehavior, monitoring, and customer satisfaction(Gassenheimer, Baucus, and Baucus 1996, p. 31;Castrogiovanni and Justis 1998). Other relationship sub-jects that call for academic interest are the allocation ofdecision rights across franchising networks and the rela-tionship between decision rights and performance of afranchise system (Windsperger 2004). In addition, anotherbroad area for future research could be a systematicquantification of the economic magnitude of encroach-ment phenomenon (Kalnins 2004).

PRICE – CONTRACTS

Contracts are studied by a large number of studies.Dahlstrom and Nygaard (1994) utilized agency theory toinvestigate contractual decisions among Norwegian oilretailers. They investigated three contractual forms: cor-porately-owned and operated facilities, corporately owned/franchise operated outlets, and franchise owned and oper-ated service stations. They found that franchisor managedoutlets exhibit the highest revenues, while franchisee-owned outlets are employed in areas with lowest popula-tion density. They also found that franchisor-operatedoutlets are more likely to be utilized near freeways thancorporately-owned/franchisee-operated service stations.

Lal (1990) focused on the use of royalty fees andmonitoring activities to improve the coordination betweenchannel members and specified that these are seen morefrequently in fast food, hotel, and motel/car rental fran-chises than other industries such as automobile and gaso-line dealerships

Kaufmann and Dant (2001) also studied fees andproposed that there is a positive relationship between theinitial franchise fee and royalty rate. Agrawal and Lal(1995) expanded the discussion on royalty fees and ana-lyzed the inter-relationships between royalty rate, moni-toring frequency, and monitoring costs within a franchisesystem. They found that the royalty rate positively affectsmonitoring frequency. Agrawal and Lal (1995) also notethat monitoring costs negatively affect monitoring fre-quency and the service level. Desai and Srinivasan (1995)studied signaling strategies in the context of a contractbetween a franchisor and a franchisee. Desai and Srinivasan(1995) developed a purely mathematical model about thecost of signaling strategies, i.e., informing franchiseesabout a demand potential. They stressed the fact thatusually franchisors are not able to fully observe franchi-sees performance and franchisees are uncertain about thedemand potential signaled by the franchisors. They con-cluded that the costs incurred by information asymmetriesare lesser in a three-part scheme which allows a franchiseecan retain a higher proportion of earnings for higherquantities sold as opposed to a two-part scheme wherefranchisor forces a franchisee to provide a desired level ofservice for a fixed royalty fee.

Gupta and Loulou (1998) studied the joint implica-tions of contracts, costs and product substitutability in amarketing channel. They found that inserting indepen-dent retailers into the channel system provides manufac-turers buffering from price competition. This is especiallytrue when the products of the manufacturers are substitut-able and intra-channel contracts are observable. They alsohighlight the fact that lack of channel coordination mayresult in a reduction of manufacturer’s incentives to investin reducing production costs. Gupta and Loulou (1998)developed a preliminary mathematical model for fran-chising channels. They criticized existing franchisingchannel research by stating that marketing literature tradi-tionally focused on demand side whereas operations lit-erature on the cost in franchise systems. They pointed outto a need to combine these two in a cross-functionalapproach where horizontal and vertical dimensions ofinter-organizational relations are combined. Later, Shane(2001) studied contract efficiency and brought a marketperformance dimension. According to Shane (2001), largefirms that adopt contract policies to screen franchisees aremore likely to survive than large firms that do not adoptthese policies. Shane (2001) stressed that non-contractualmechanisms such as training and communication mecha-nisms are survival enhancing if adopted in ways consis-tent with efficient contract theory.

Azoulay and Shane (2001) studied franchise systemcontract conflicts by using evolutionary view of contractdesign in which entrepreneurs undertake policy experi-ments on the basis of the information they possess. Azoulay


and Shane (2001) emphasized the importance of contract-ing efficiency and showed that contracts which havesuperior information are rewarded with survival. Anexample of the significance of superior knowledge froma franchisees point of view is definition of exclusiveterritories in franchising contracts whereby franchiseescan protect themselves from the encroachment problem.Desai (1997) studied a different type of fee which ischarged on franchising advertisements. He developed amathematical model that discussed the effects of twotypes of advertising fees on the behavior of franchisechannel members. He highlighted the fact that althoughfixed advertising fee allows the franchisor to achieve fullchannel coordination on price and service, sales-basedadvertising fee is more profitable when franchisees aresufficiently heterogeneous.

Light (1997) suggested that franchisors should con-sider signing long-term contracts with their franchisees.The longer contracts signal to franchisees that they aremore likely to enjoy their business’s prosperity before thefranchisor can seek control of the outlet. Baroncelli andManaresi (1997) studied divestment as a change of theform of controlling retail operations, from ownership tocontractual relationships of franchising. They find thatdivestment is not merely a strategy to save unprofitablefirms but it can be used as a strategy to deploy resourcesefficiently in order to grow a business. They furthered thisposition by stressing that franchising can be considered asa hybrid form of divestment, replacing ownership-basedcontrol in any market

Chen and Dimou (2005) looked at international hotelchains and analyzed the effect of parameters such as thelevel of development of the target market, brand growthrate, international experience, and brand nationality, indetermining the selection of control and market entrymodes, such as choosing a franchise system, managementcontracts, or direct ownership. They suggested that bud-get hotels may favor franchising whereas upper anddeluxe hotels may opt for company ownership or manage-ment contracts as their international expansion strategy.

FUTURE RESEARCH ON PRICE – CONTRACTS

In terms of control, contracting issues require furtherattention. For instance, the relationship between efficientcontracting and firm survival is not studied in detail. Inaddition, in case of efficient contracts, the mechanisms bywhich contracting costs influence firm survival over timeis a fruitful avenue for future research. The effect ofinformation asymmetries in royalty rates, variance ofcontractual arrangements and monitoring intensity acrossfranchising industries are other contract related futureresearch areas (Agrawal and Lal 1995).

Another prospective research stream may be theanalysis of the effect of advertising on signaling schemeswithin the context of franchising contracts (Desai andSrinivasan 1995). Gupta and Loulou (1998) pointed out toa need for analytical studies to explain the impact of R&Dbetween competitors in franchise systems and the impactof flexible technologies on buyer supplier relationships infranchise systems and franchising contracts. Furthermore,non-contractual mechanisms to control contracting prob-lems and performance analysis for rare franchising con-tracts (i.e., contracts that are less on standard documentsand more on specific and detailed script) is not yet fullyexplored (Shane 2001).

Other miscellaneous future research streams may befocused on the role of advertising commitment in a com-petitive setting, and to explain why a large number offranchisors do not use any type of advertising fee (Desai1997). According to Baroncelli and Manaresi (1997)there is a need to approaching franchising as an interna-tional divestment strategy in future research.

In addition, empirical studies about how interna-tional franchisors support their franchisees would guidebusiness professionals interested in franchising business.There is also a lack of research in understanding expan-sion strategies in the multinational hotel industry (Chenand Dimou 2005).

PRICE – OWNERSHIP

Dant and Kaufmann (2003) examine the changes inownership patterns of franchise systems as they mature.They compare three alternative theories within the fastfood industry. Signaling theory predicts that franchisesystems will move toward a greater reliance on franchisedoutlets as systems mature, while resource acquisitiontheory (ownership redirection thesis) predicts a tendencyin the opposite direction. A third theoretical perspective,tapered integration or plural forms, suggests a tendencytoward maintaining a steady state of mixed distribution.They find that although franchisors value the benefits ofthe mix of ownership types and do maintain that mix overtime, there is some evidence of a greater tendency topermanently convert existing franchised outlets to com-pany-owned outlets as fast food systems mature and gaingreater access to resources. Dant, Paswan, and Kaufmann(1996) reports the results of a meta-analysis of the existingstudies of ownership redirection thesis – the premise thatthere may be a deliberate effort on the part of the morepowerful franchisors to convert the previously franchi-see-owned outlets into company owned outlets. Theystress the disparate operationalizations of the phenom-enon itself and the manner in which the construct mea-surement affects the determination of its existence.


Lafontaine and Kaufmann (1994) studied two sets ofcompeting theories to explain the existence of franchiseownership, one set based on resource constraints andanother on incentives issues. As franchise systems mature,these theories predict different patterns in the evolution ofthe mix of franchised and company-owned outlets. Theyfind support for the prediction of both of these theories.Windsperger and Dant (2006) studied ownership redirec-tion in franchising. They argue that the structure anddynamics of ownership patterns in franchising networksdepends on the contractibility of the franchisor’s system-specific assets and the contractibility of the franchisee’slocal market assets. They also sate that under the propertyrights view, ownership redirection will result from anincrease in the contractibility of the franchisee’s localmarket assets and the resultant increase of the franchisor’sbargaining power during the contract period. They alsoargue that informational, financial and managerial resourceconstraints are only relevant for the change of ownershipstructure if they are non-contractible. Burkle and Posselt(2008) proposed a new theory based on franchisor’soptimal risk allocation to explain why a mixture of fran-chise-owned and franchisor-owned units exists in fran-chise systems. They state that the costs of risk and control-ling franchised units explain the varying fraction of fran-chisee-owned to total selling units. They also find that theincentive to franchise decreases with an increasing frac-tion of franchisee-owned to total selling units, as well aswith decreasing costs of control. This explanation ofplural systems is consistent with the ownership redirec-tion hypothesis. Anderson and Fok (1998) analyzed theindependently owned and locally operated firms in resi-dential real estate business examined why a residentialreal estate brokerage firm would choose to franchise.Particularly, they tried to measure the productive effi-ciency levels of franchised relative to non-franchisedbrokerages. They found that franchised firms are moreefficient in the allocation of resources.

Castrogiovanni and Justis (1998) proposed and testedtheoretical perspectives relevant to franchising as a busi-ness form. They stated that franchising differs from otherorganizational forms in three areas: geographical dis-persal of organization units, replication across units, andjoint ownership. Based on these factors, they extendedboth the franchising and organization theory by showinghow the franchising phenomenon fits into a general per-spective on organizations. Schmidt and Oldfield (1999)stated that ownership of franchise stores selling coffeeand bakery products results in hard lives with workstarting at 1:00 a.m. in the morning to prepare the dough-nuts for the morning rush. The retail only concept in suchfranchise systems gives owners greater control over theproduct. Schmidt and Oldfield (1999) suggest that groupsof such franchise store owners could invest in the jointownership of central production facilities which may

result in efficiency gains. Hopkinson and Hogarth-Scott(1999) stated that there are two distinct disciplinary orien-tations and research streams that are analyzing franchis-ing, economic and behavioral. According to Hopkinsonand Hogarth-Scott (1999), economic theories argue thatfranchising offers particular advantages to the franchisorthat are not found in fully owned and integrated channels.This stream is embedded in macro-economic theory thatstates that franchising is mainly an efficient channel form.Based on this idea, this stream explains franchising mainlyby three theories: resource constraint theory, agency theory,and search cost theory. A separate research stream exam-ines behavior in franchise systems and seeks empiricalexplanation of phenomena such as conflict and trustwithin franchise relationships.

Hopkinson and Hogarth-Scott (1999) argue thatmotivation to franchise depends on the specific strategyemployed by the franchisor which in turn affects rela-tional quality. They describe a model drawn by linkingstrategic direction, franchise motivation and relationalquality.

Fladmoe-Lindquist and Jacque (1995) combinedtransaction cost theory and agency theory to analyzeownership strategies for international services franchis-ing. By running a logistic regression on data providedfrom twelve U.S. firms with operations in ninety-twocountries, they observed that the propensity to franchise isdirectly related to the franchisors international experi-ence, the geographical and cultural distance betweenfranchisor and franchisee and the degree of uncertainty inthe host country.

Nault and Dexter (1994) utilized differential equa-tions to examine electronically linked franchise channelsof distribution. These franchise structures are consideredas an emerging channel structure where exclusive territo-ries are allowed to franchisees and customers are in a wayowned by franchisees by use of electronic transactionmonitoring. They concluded that the possibility of inter-franchise transfer that is made possible by emerging ITsystems results in franchisees making greater investmentsthan traditional franchise systems. Hing (1999) stressedthe importance of communication among owners of fran-chise systems, especially for small franchisees.

Chow and Frazer (2003) analyzed the operationaldifferences between mobile franchising arrangementsand fixed-site franchises from an agency-theoretic per-spective. They concluded that start-up risk for owners islower in mobile sale units and mobile operations exhibita higher level of repeat customers than fixed-site fran-chises. Finally, Michael (1999) also studied the decisionrights focused on advertising ownership.


FUTURE RESEARCH ON PRICE – OWNERSHIP

Hopkinson and Hogarth-Scott (1999) support theneed for research that integrates the behavioral and eco-nomic streams of franchise research. Also, literature sug-gests more studies that compare the relative advantages ofequity-based market penetration versus franchising ininternational perspective (Fladmoe-Lindquist and Jacque1995).

Nault and Dexter (1994) noted the need to compareand contrast the positive cost to the franchisor of addingfranchise units vs. the negative effects of increased roy-alty due to small networks in future studies. In addition,mobile franchise systems characterized by lower riskinvestment and outcome-based contracts must be studiedin more detail (Chow and Frazer 2003).

CONCLUSIONS AND RECOMMENDATIONS

This paper reviewed the literature on franchising inthe context of marketing. It was not until 1975 thatfranchising attracted substantive academics interest. How-ever, the attention has been impressive since then, parallel

to the increasing practical prevalence and importance offranchising in business marketing and management. Theliterature in franchising is replete with research conductedin various disciplines including economics, law, manage-ment, marketing, and management science. However,there is no literature review paper that approaches to thetopic from a marketing perspective. In this, study weattempted to address this gap by categorizing the literaturein order to identify shortages and surpluses of research.

We identified three groups of articles: Price (Con-tracts and Ownership), Trust (Relationships), and Author-ity (Conflict and Opportunism). These three groups mapinto two major categories: Markets and Hierarchies. Wefound that all areas require further research attention. Thisis particularly true for Trust and Authority because exist-ing research is at a nascent stage. Studies in these twoissues list a wide range of fruitful avenues for futureresearch. In fact, the same is true for Price-Contracts andPrice-Ownership issues as we cannot state that there is aresearch surplus. The overarching finding of this study isthat franchising research is not saturated from a marketingperspective.

REFERENCES

Agrawal, D. and R. Lal (1995), “Contractual Arrange-ments in Franchising: An Empirical Investigation,”Journal of Marketing Research, 32 (2), 213–21.

Anderson, R.I. and R. Fok (1998), “The Efficiency ofFranchising in the Residential Real Estate BrokerageMarket,” The Journal of Consumer Marketing, 15(4), 386–402.

Azoulay, P. and S. Shane (2001), “Entrepreneurs, Con-tracts, and the Failure of Young Firms,” ManagementScience, 47 (3), 337–58.

Baroncelli, A. and A. Manaresi (1997), “Franchising as aForm of Divestment: An Italian Study,” IndustrialMarketing Management, 26 (4), 223–35.

Bradach, J.L. and R.G. Eccles (1989), “Price, Authority,and Trust: From Ideal Types to Plural Forms,” AnnualReview of Sociology, 15, 97–118.

Burkle, T. and T. Posselt (2008), “Franchising as a PluralSystem: A Risk-Based Explanantion,” Journal ofRetailing, 84 (1), 39–47.

Castrogiovanni, G.J. and R.T. Justis (1998), “FranchisingConfigurations and Transitions,” The Journal of Con-sumer Marketing, 15 (2), 170–89.

Cateora, P.R., J. Graham, and J.L. Graham (2006), Inter-national Marketing, 13th ed. McGraw Hill/Irwin.

Chen, J.J. and I. Dimou (2004), “Expansion Strategy ofInternational Hotel Firms,” Journal of BusinessResearch, 11 (1), 1730–40.

Child, J. (1987), “Information Technology, Organization,and the Response to Strategic Challenges,” Califor-nia Management Review, 30 (1), 33–50.

Chow, L. and L. Frazer (2003), “Servicing CustomersDirectly: Mobile Franchising Arrangements in Aus-tralia,” European Journal of Marketing, 37 (3/4),594–613.

Dahlstrom, R. and A. Nygaard (1994), “A PreliminaryInvestigation of Franchised Oil Distribution in Nor-way,” Journal of Retailing, 70 (2), 179–91.

____________ and ____________ (1999), “An Empiri-cal Investigation of Ex Post Transaction Costs inFranchised Distribution Channels,” Journal of Mar-keting Research, 36 (2), 160–70.

Dant, R.P. and P.L. Schul (1992), “Conflict ResolutionProcesses in Contractual Channels of Distribution,”Journal of Marketing, 56 (1), 38–54.

____________, A.K. Paswan, and P.J. Kaufman (1996),“What We Know About Ownership Redirection inFranchising: A Meta Analysis,” Journal of Retailing,72 (4), 429–44.

____________ and P.J. Kaufman (2003), “Structural andStrategic Dynamics in Franchising,” Journal of Retail-ing, 79 (2), 63–75.

Desai, P.S. and K. Srinivasan (1995), “Demand SignalingUnder Unobservable Effort in Franchising: Linearand Nonlinear Price Contracts,” Management Sci-ence, 41 (10), 1608–23.

____________ (1997), “Advertising Fee in Business-


Format Franchising,” Management Science, 43 (10),1401–19.

Doherty, A.M. and N. Alexander (2004), “RelationshipDevelopment in International Retail Franchising,”European Journal of Marketing, 38 (9/10), 1215–35.

____________ and ____________ (2006), “Power andControl in International Retail Franchising,” Euro-pean Journal of Marketing, 40 (1112), 1292–1316.

Eccles, R.G. (1981), “The Quasifirm in the ConstructionIndustry,” Journal of Economic Behavior and Orga-nization, 2 (4), 335–57.

Elango, B. and V.H. Fried (1997), “Franchising Research:A Literature Review and Synthesis,” Journal of SmallBusiness Management, 35 (3), 68–81.

Fladmoe-Linquist, K. and L.L. Jacque (1995), “ControlModes in International Service Operations: The Pro-pensity to Franchise,” Management Science, 41 (7),1238–49.

Gassenheimer, J.B., D.B. Baucus, and M.S. Baucus (1996),“Cooperative Arrangements among Entrepreneurs:An Analysis of Opportunism and Communication inFranchise Structures,” Journal of Business Research,36 (6), 67–79.

Gupta, S. and R. Loulou (1998), “Process Innovation,Product Differentiation, and Channel Structure: Stra-tegic Incentives in a Duopoly,” Marketing Science,17 (4), 301–16.

Hadjimarcou, J. and J.W. Barnes (1998), “Case Study:Strategic Alliances in International Franchising,” TheJournal of Consumer Marketing, 15 (6), 598–97.

Hing, N. (1999), “Maximizing Franchisee Satisfaction inthe Restaurant Sector,” The Journal of ConsumerMarketing, 16 (5), 502–61.

Hopkinson, G.C. and S. Hogarth-Scott (1999), “Fran-chise Relationship Quality: Micro-Economic Expla-nations,” European Journal of Marketing, 33 (9/10),827–48.

Hunt, S.D. and J.R. Nevin (1975), “Tying Agreements inFranchising,” Journal of Marketing, 39 (3), 20–26.

____________ and ____________ (1976), “Full Disclo-sure Laws in Franchising: An Empirical Investiga-tion,” Journal of Marketing, 40 (2), 53–62.

International Franchise Association (2004), “FranchisingEconomic Impact Study,” Washington, D.C., [URL:www.franchise.org].

Kalnins, A. (2004), “An Empirical Analysis of TerritorialEncroachment within Franchised and Company-Owned Branded Channels,” Marketing Science, 23(4), 476–89.

____________ and K.J. Mayer (2004), “Franchising,

Ownership, and Experience: A Study of Pizza Res-taurant Survival,” Management Science, 50 (12),1716–28.

Kaufmann, P.J. and V.K. Rangan (1990), “A Model forManaging System Conflict During Franchise Expan-sion,” Journal of Retailing, 66 (2), 155.

____________ and R.P. Dant (2001), “The Pricing ofFranchise Rights,” Journal of Retailing, 77, 537–45.

Lafontaine, F. and P.J. Kaufmann (1994), “The Evolutionof Ownership Patterns in Franchise Systems,” Jour-nal of Retailing, 70 (2), 97–113.

Lal, R. (1990), “Improving Channel Coordination throughFranchising,” Marketing Science, 9 (4), 299–318.

Light, D.A. (2003), “Franchising: Getting It Right fromthe Start,” Harvard Business Review, (May/June),13–15.

Michael, S.C. (1999), “Do Franchised Chains AdvertiseEnough?” Journal of Retailing, 75 (4), 461–78.

Monroy, F.M. and L.M. Alzola (2005), “An Analysis ofQuality Management in Franchise Systems,” Euro-pean Journal of Marketing, 39 (5/6), 585–605.

Nault, B.R. and A.S. Dexter (1994), “Adoption, Trans-fers, and Incentives in a Franchise Network withPositive Externalities,” Management Science, 13 (4),412–23.

Perry, C., A. Cavaye, and L. Coote (2002), “Technical andSocial Bonds within Business-to-Business Relation-ships,” The Journal of Business & Industrial Market-ing, 17 (1), 75–88.

Richardson, G.B. (1972), “The Organization of Indus-try,” The Economic Journal, 82, 883–96.

Scmidt, R.A. and B.M. Oldfield (1999), “Donkin’Donuts: The Birth of a New Distribution and Fran-chising Concept,” The Journal of Consumer Market-ing, 16 (4), 376–89.

Shane, S. (2001), “Organizational Incentives and Organi-zational Mortality,” Organization Science, 12 (2),136–60.

Srinivasan, R. (2006), “Dual Distribution and IntangibleFirm Value: Franchising in Restaurant Chains,” Jour-nal of Marketing, 70, 120–35.

Williamson, O.E. (1985), The Economic Institutions ofCapitalism. New York: Free Press.

Windsperger, J. (2004), “Centralization of FranchisingNetworks: Evidence from Austrian Franchise Sec-tor,” Journal of Business Research, 68 (7), 1361–69.

Windsperger, J. and R.P. Dant (2006), “Contractibilityand Ownership Redirection in Franchising: A Prop-erty Rights View,” Journal of Retailing, 82 (3), 259–72.


APPENDIX 1List of Reviewed Articles

Quantit (N)/ Research Designs Sample Size StatisticalReference Context Qualitative (L) Utilized (Response Rate) Country Techniques Group

Agrawal & Lal – N Theoretical/ – U.S. Correlation Price-(1995) non-empirical Contracts

Anderson & Fok Real estate N Survey Data (National Data from U.S. Data Envelopment Price-(1998) Assoc. of Realtors) 276 firms Analysis Software Ownership

Azoulay & Shane All types N Secondary data 170 (28.6%) U.S. Correlation Price-(2001) (Franchise Annual) Contracts

and survey

Baroncelli & Manaresi Retail N Mail survey, 72 (31%) Italy Descriptive Price-(1997) phone calls Statistics Contracts

Burkle & Posselt Retail L Theoretical/ – – Mathematical Price-(2008) non-empirical Ownership

Castrogiovanni & L Non-empirical, – U.S. – Price-Justis (1998) theory based Ownership

Chen &Dimou Hotel N Secondary Data – Int. Logistic Price-(2005) (public stats at regression Contracts

firm level)

Chow & Frazer All types N&L Survey and 186 (27%) Australia MANCOVA / Price-(2003) Interview ANOVA Ownership

Dahlstrom & Nygaard Oil N Mail survey and 179 (61%) Norway SEM Authority(1999) distribution Interview and 216 (50%)

Dahlstrom & Nygaard Oil Retail N Mail survey 181 (61%) Norway Logit Price-(1994) Regression Contracts

Dant & Kaufmann Fast Food N Mail survey 152 (24%) U.S. Mult. Discriminant Price-(2003) Analysis Ownership

Dant, Paswan & All types N Meta Analysis – – Correlation Price-Kaufman (1996) Ownership

Dant & Schul Fast food N Survey and interview 176 (20%) U.S. Discriminant Authority(1992) analysis, F test,

Desai (1997) – L Non-empirical – – – Price-Contracts

Desai &Srinivasan – L Mathematical Model – – – Price-(1995) Contracts

Doherty &Alexander Fashion L Multiple case design – – – Trust(2006) retail

Doherty & Alexander Fashion L Case study – – – Trust(2004) retail

Elango &Fried (1997) All types L Literature Review – – – All

Fladmoe-Lindquist Hotel, N Secondary data 10302 International Logistic Price-& Jacque (1995) restaurant, goods (UN Int. Services service units regression Ownership

and food retail Stats)

Gassenheimer, Baucus Fast food N Mail survey 195 (7.2 %) U.S. Correlation Authority& Baucus (1996)


APPENDIX 1 (CONTINUED)List of Reviewed Articles


Gupta & Loulou – L Non-empirical, – – – Price-(1998) theoretical model Contracts

Hadjimarcou & Restaurant L Case Study (Silver – Mexico – TrustBarnes (1998) Steak company data)

Hopkinson & General L Non-empirical, – – – Price-Hogarth-Scott (1999) distribution theoretical model Ownership

systems

Hing (1999) Restaurant N Survey 127(57%) Australia Correlation Price-Contracts

Hunt &Nevin Fast foods N Survey (study from 664 (–) U.S. Descriptive All(1975) Univ. Wisconsin-

Maddison)

Hunt &Nevin All types N Survey (Franchise 102 (45%) U.S. Descriptive Authority(1976) Opportunity

Handbook)

Kaufmann & Dant Fast Food N Survey 152 (not U.S. Regression Price-(2001) specified) Contracts

Kaufmann & Rangan – L Non-empirical, – – – Authority(1990) theoretical model

Kalnins (2004) Hotel N Secondary data – U.S. Regression Authority(Texas Comptroller)

Kalnins & Mayer Restaurant N Secondary data – U.S. Correlation, Trust(2004) (pizza) (Texas Comptroller) regression

Lal (1990) – L Non-empirical, – U.S. – Price-theoretical model Contracts

Lafontaine & All types N Survey 130 (30%) International Regression Price-Kaufmann (1994) Ownership

Light (1997) All types L Non-empirical, – – – Price-discussion Contracts

Michael (1999) Restaurant N Secondary Data – U.S. Regression Price-35 restaurant chains Ownership

Monroy & Alzola – L Non-Empirical – U.S. – Trust(2005)

Nault & Dexter – L Non-empirical, – U.S. Differential Price-(1994) mathematical model equations Ownership

Perry, Cavaye & All types N Mail survey 174 ( 25%) Australia SEM TrustCoote (2002)

Schmidt & Oldfield Fast food L Non-empirical, – U.S. – Price-(1999) conceptual discussion Ownership

Shane (2001) All types N Secondary Data 2997 U.S. Logistic Price-businesses(Compustat, Regression Contracts

firm files)


APPENDIX 1 (CONTINUED)List of Reviewed Articles


Srinivasan (2006) Restaurant N Secondary Data 55 firms, U.S. Correlation Authority(Compustat, firm files) 394 firm

years

Windsperger (2004) All types N Mail survey 83 (38%) Austria Regression Authority

Windsperer & Dant All types N Mail survey 83 (39) Austria Regression Price-(2006) Ownership

For further information contact:Selcuk Ertekin

College of Business Marketing & LogisticsUniversity of North Texas

1155 Union Circle #311277Denton, TX 76203–5017



THE ROLE OF MARKETING ORGANIZATION IN THE MARKETORIENTATION-PERFORMANCE RELATIONSHIP

Lancy Mac, University of Macau, China

ABSTRACT

This study is undertaken to gauge the effect of mar-keting organization on the market orientation-performancerelationship. Results show that a formal marketing depart-ment does help enhance the MO-performance relation-ship. For those with a formal marketing department, lessintegration of the marketing department will contribute tothe MO-performance relationship.

INTRODUCTION

Recent decades have seen increasing emphasis onshared responsibility of marketing within the firm ashighlighted by Drucker’s (1954) assertion that marketingis everybody’s job and McKenna (1991)’s HBR article“Marketing is Everything.” Following this trend, theimportance of marketing is seen to be declining withvarious accusations like its “lack of accountability” (Shethand Sisodia 2002), “lack of relevance” (Webster 2005),“losing its seat at the table” (Webster, Malter, and Ganesan2003) and “gets no respect in the boardroom” (Schultz2003). It is regarded that Marketing has entered its mid-life crisis (Brady and Davis 1993) with trails on itscontribution to business success. Gradually, key market-ing thinkers move to rejuvenate the discipline with studiesand commentaries on sustaining the viability of marketingas an indispensible management function (Webster 1997;Workman, Homburg, and Gruner 1998; Varadarajan1992), most influential of which perhaps is Moorman andRust’s (1999) empirical study. Their argument is thatwhile market orientation is important to organizationalperformance, the marketing function should co-exist whichcan further enhance the effectiveness of a market orienta-tion. Other studies suggest that marketing organizationshould be studied with a broader perspective encompass-ing not only its structure but also its non-structural aspectslike relationship with other organizational functions(Workman, Homburg, and Gruner 1998).

This study aims to demonstrate the effect of market-ing organization in facilitating the implementation of themarketing concept so as to enhance firm-level perfor-mance. It is argued in this paper that being properlyorganized for marketing can help enhance the marketorientation-performance relationship. Market orientationis widely acknowledged to have instrumental effect onfirm performance (Slater and Narver 1994; Jaworski andKohli 1993; Keh et al. 2007). We suggest that its contri-

bution can further be enhanced by a complementarymarketing organization. To this end, a detailed literaturereview was undertaken on studies related to differentperspectives of the marketing organization which is pre-sented next. This leads to the proposition of the researchmodel of this study and the methodology used. Finally, theresults and discussions of the findings will be elaborated.

MARKETING ORGANIZATION

As noted by Workman, Homburg and Gruner (1998),marketing organization is an area under-studied with mostearly studies mainly descriptive (e.g., Piercy 1986) andconceptual (e.g., Achrol 1991; Day 1997; Webster 1992,1997). It is far ignored when compared with studies on thecultural aspects of marketing, specifically market orienta-tion and its effect on performance. This may be attributedto the diversity in scope that marketing organizationencompasses including structure, interactions andbureaucratic procedures (Workman, Homburg, and Gruner1998).

The relevance of marketing organization is welldocumented in various marketing textbooks with duediscussions on the way how the marketing department canbe organized to facilitate customers’ need fulfillment(e.g., Kotler et al. 2006). It is widely acknowledged as akey facilitator of how well strategy is implemented (McKee,Varadarajan, and Pride 1989). Some studies suggestedthat there should be a fit between marketing organizationand its business strategy in order to instill firm perfor-mance (Olson, Slater, and Hult 2005) and marketingperformance (Vorhies and Morgan 2003). Surprisingly,very few studies are undertaken on matching it with othermarketing factors like marketing strategy. Very recentstudies found strong instrumental effect of marketingorganization on market orientation (Verhoef and Leeflang2009) and in facilitating the market orientation-performance relationship (Merlo and Auh 2009).

Existing studies on marketing organization can beclassified into two broad perspectives, namely functionalgroup perspective and activity-based perspective usingthe labels of Workman, Homburg, and Gruner (1998).(Detail reviews can be found in Workman, Homburg, andGruner 1998; Harris and Ogbonna 2003). Functionalgroup approach views marketing as a distinct entity withinthe organization and research mainly focus on investiga-tion of the location of marketing department and its


structural organization as well as its characteristics includ-ing formalization and centralization. Activity-basedapproach on the other hand views marketing as a set ofactivities and normally is investigated by the dispersion ofmarketing activities, relative power/influence of market-ing and its interaction with other functional groups.

We classified the literature review below in terms ofstructural and non-structural aspects of marketing organi-zation taking into account the classification of Workman,Homburg, and Gruner (1998) as well as our conceptualmodel.

STRUCTURAL ASPECTS OF MARKETINGORGANIZATION

Early studies on marketing organization are descrip-tive with focus on whether there is any marketing depart-ment (Piercy 1986), where it is located (Hise 1965),structural location of marketing activities (Ruekert, Walker,and Roering 1985) and the organizational form that bestfit different environmental conditions (Achrol 1991, 1997).The basic assumption of these studies is that having astrong marketing department can help in the implementa-tion of marketing. Having a formal marketing departmentis found to be a reflection of the “adoption of the market-ing concept” (Hise 1965). In fact, there is a need to adaptorganization structure to facilitate responsiveness to cus-tomers’ needs (George, Freeling, and Court 1994; Day1994) and be more market-oriented (Slater and Narver1995). Homburg, Workman, and Jensen (2000) evensuggested the movement toward building a “customer-focused organizational structure” in order to enhanceperformance. The way how marketing is being organizedand its relationship with in particular the sales departmentare important and relevant as found by the field study ofWorkman, Homburg, and Gruner (1998). There is how-ever limited research on testing this empirically. Ourresearch is formulated in response to their call for studiesto gauge the performance implications of marketing orga-nization. We therefore hypothesize that:

H1: Firms with a formal marketing organization willenhance the market orientation-performancerelationship.

NON-STRUCTURAL ASPECT OF MARKETINGORGANIZATION

A formal marketing structure refers to one where aformal marketing department exists in the organization.However, sole existence of a specialized department doesnot guarantee that marketing is done well. Thus, a lot ofstudies were being initiated on the activity-based perspec-tive of marketing organization. There are different streamsof study including responsibilities of the marketing depart-ment, marketing’s influence, and relative power as well asinteraction with other functional units.

Dispersion of Marketing Activities

Not much has been done on the impact of crossfunctional dispersion of marketing activities except forPiercy (1986) and Tull et al. (1991)’s descriptive studies.They found that the marketing department is normally notin charge of all traditional marketing activities and thatmany companies have a dispersed organization ratherthan one that is integrated. This is consistent with thefindings of Workman, Homburg, and Gruner (1998) whofound in their field study that marketing departments indifferent firms have varying responsibilities and thatmany traditional marketing activities are performed byother functional units. This stream of study is also acknowl-edged as the marketing mix paradigm which originatedfrom McCarthy (1960) as it includes examination of thepractice of various marketing activities which are gener-ally classified into the four Ps including product, place,price, and promotion. According to Webster, Malter, andGanesan (2005), marketing should be shared by differentdepartments and “solos” can be detrimental to firm perfor-mance (as in the case of Sony). Sharing of marketingfunctions by other departments actually reflects that unitsare allocated with the work they can do best. The higherthe dispersion of activities, the more likely that other unitsshare and understand the marketing concept leading to anorganization-wide adoption of marketing or being marketdriven (Day 1994, 1999). We therefore hypothesize that

H2: Higher cross functional dispersion of marketing activi-ties will enhance the market orientation-performancerelationship.

Marketing’s Influence and Power Within the Firm

Another non-structural aspect of marketing organi-zation that is populous in the research arena is the inves-tigation of the functional influence of the marketingdepartment. It is generally acknowledged that the moreinfluence or power the marketing department has, themore likely the organization will have the marketingconcept being embraced and implemented. Moorman andRust (1999) found empirical support with his study of the“value of marketing function” on performance. In hisstudy, he demonstrated that a stronger marketing depart-ment can lead to better firm performance over and abovethe contribution of an organizational-wide market orien-tation. Market orientation is important but the marketingfunction turns out to be more influential. Their worksparkled a series of studies on the role of the marketingfunction in the organization with consistent results(Krohmer, Homburg, and Workman 2002; Verhoef andLeeflang 2009). If marketing has higher influence in thefirm, it is likely that people in the firm will have more“marketing-mindedness” (Greyser 1997) and so facili-tates the adoption of the market orientation (Merlo andAuh 2009; Verhoef and Leeflang 2009). Verhoef and


Leeflang (2009) found that market orientation is a fullmediator of marketing influence and performance. On theother hand, Merlo and Auh (2009) found that marketorientation and marketing subunit influence are not bythemselves significant contributors of performance. Bothmust exist and interact in order to create a positive perfor-mance impact. These results suggest that both structuraland cultural dimensions of marketing are indispensiblefor performance. Therefore the hypothesis that follows is:

H3: Higher marketing power within the firm will enhancethe market orientation-performance relationship.

Marketing’s Cross Functional Interaction

Marketing’s interaction with other functional units isevident to produce desirable outcomes including betterimplementation of marketing strategy (Anderson 1982)and business success (Ruekert and Walker 1987). Themarketing department is always in interaction with cus-tomers externally and other functions internally in theprocess of implementing the marketing programs. Mostof these studies are done specifically referring to one otherfunctional unit with R&D being the major unit beingstudied as closer cooperation between them lead to newproduct success (Griffin and Hauser 1996; Gupta, Raj,and Wilemon 1986). More communication in transferringcustomer information and higher integration result inbetter tailoring products to the needs of the customers andthus ensure better performance. We therefore hypothesizethat:

H4: Higher marketing interaction with other functionalunits enhance the market orientation-performancerelationship.

DATA COLLECTION

A survey was being conducted in Guangdong Prov-ince in South China targeting enterprises of differentindustries, ownership types, and sizes. Personal interviewwas the primary contact method although some question-naires were dropped-off and collected after completion. Atotal of 923 companies were approached but eventuallyonly 425 questionnaires were received, yielding a responserate of 46 percent. After screening the questionnaires foromissions and irregularities, only 409 questionnaires wereuseable for final analysis. Majority of the sampled firmswere local firms (48.6% of the total sample) in the servicesector (71.6%), small with less than 100 employees (75.9%)and not so experienced (in operation in China for 5 yearsor less) (49.2%).

MEASUREMENTS

Marketing organization is being classified into thestructural and non-structural aspects. The structural as-

pect of marketing organization is being measured by theexistence of a marketing department, a sales department,a marketing and sales department, a department equiva-lent to a marketing department and having neither market-ing nor sales department. This is adopted and modifiedfrom Workman, Homburg, and Gruner (1998)’s study.All of them are dummy-coded. The non-structural aspectsof marketing organization is being measured by theextent of cross functional dispersion of marketing activi-ties, power of the marketing function and marketing’scross functional interaction (Workman, Homburg, andGruner 1998). Cross functional dispersion of marketingactivities is measured as the percentage of the ten market-ing activities obtained from the literature (Piercy 1986;Tull et al. 1991) that are performed by the marketingdepartment. These ten activities include selling, pricing,strategic planning, product design, new product develop-ment, marketing research, advertising and promotion,customer service, exporting, and distribution. Power ofmarketing function is measured by asking respondents torate the status of marketing over three other functionalunits (Sales, R&D, and finance) on a 3-point scale rangingfrom lower to higher. Interaction with other functionalunits is measured by asking respondents to rate the extentof interaction with the three functional units on a 3-pointscale ranging from high to low.

Market orientation is measured using Narver andSlater (1990)’s 15-item scale measuring customer orien-tation, competitor orientation and interfunctional coordi-nation. Corporate performance is measured by subjec-tively assessing the company’s performance relative to itsmajor competitors in terms of sales, profit, market share,and return on investment. Subjective measures are deemedto be as good as objective ones (Dess and Robinson 1984)and are widely applied in many studies involving bothdeveloped and developing countries (Siu 2000a, 2000b;Akimova 2000; Deshpandé, Farley, and Webster 1993).All items are measured on a 5-point Likert scale.

A number of control variables are added into theanalysis including the size of the organization measuredby the number of employees, the number of years operat-ing in China, the type of industry (secondary or tertiary)and the type of ownership (with and without foreigninvestment).

DATA ANALYSIS METHOD

To achieve the objectives stated earlier, hierarchicalregression analysis is employed in two stages. In the firststage, the dependent variables (firm performance) areregressed on the control variables, market orientation(MO) and the structural marketing organization variableswith the total sample of 409 companies. The purpose ofthis stage is to investigate whether a structural organiza-tion is contributive to firm performance. The variables are


entered in the following steps: (1) control variables, (2)three MO factors, and (3) structural marketing organiza-tion. In the second stage, only those with a marketingdepartment are included in the analysis. Most existingstudies on marketing organization are based on the unani-mous a priori assumption that there is a marketing depart-ment. By selecting only firms with a marketing depart-ment, results will be more accurate. The independentvariables are again entered in three steps: (1) controlvariables, (2) three MO factors and (3) non-structuralorganizational factors.

FINDINGS

Stage 1: Structural Marketing Organization

Results show that a formal marketing department hasa significant contribution to overall performance of thefirm (Table 1). This confirms previous assertion of theinstrumental role of marketing in an organization(Moorman and Rust 1999). Other forms of marketing

organization including sales department, integrated mar-keting and sales department as well as any equivalent-to-marketing department are found not to have significantimpact. On the other hand, only competitor orientation isfound to be the only significant MO component leading toperformance, with contribution increased slightly in step 3than step 2 when structural marketing organization isentered into the equation. There is also a significantchange in the R-square from step 2 to step 3. All theseindicate that having a formal marketing department has acontributive effect on the market orientation-performancerelationship thus supporting H1.

A clearer picture emerged when we categorized com-panies into high and low MO firms by using the median ofthree as the break off. The effect of MO on performanceis plotted in Figure 1. It is shown that when the firm is highon MO, firms with a formal marketing department havehigher performance. However, at low MO, having amarketing department will have lower performance thanthose who do not. This signifies that the formation of a

TABLE 1Regression Results of Structural Marketing Organization

Step 1 Step 2 Step 3

Beta T Beta T Beta T

Control variables:Size of company 0.164 3.002** 0.145 2.670** 0.116 2.128*

Years in China -0.008 -0.153 0.002 0.045 0.009 0.178

Type of industry 0.065 1.257 0.058 1.141 0.052 1.029

Type of ownership 0.010 0.199 0.009 0.180 -0.002 -0.049

Market orientation:Customer orientation 0.091 1.366 0.100 1.505

Competitor orientation 0.162 2.441* 0.164 2.508*

Interfunctional coordination -0.014 -0.210 -0.029 -0.429

Structural marketing organization:Have a Marketing dept 0.222 3.143**

Have a Sales dept -0.108 -1.491

Have a mktg & sales dept 0.031 0.380

Have a dept equivalent to mktg 0.023 0.407

Have no mktg nor sales dept -0.014 -0.184

F 2.653* 4.347** 4.168*

R-square 0.027 0.073 0.117

R-square change – 0.047** 0.043**

* p < 0.05; **p < 0.01


marketing department will not always lead to better per-formance. It is only when the organization is market-oriented, a separate marketing function will help facilitateperformance. On the contrary, if the firm is low in marketorientation, a marketing department will even drag downthe performance of the entire organization.

Stage 2: Non-Structural Marketing Organization

Given that a formal marketing organization is rel-evant and instrumental to firm performance, a secondstage of analysis is done to further investigate how mar-keting can be better structured to facilitate performance.Only firms with a formal marketing department areincluded in the analysis with 278 firms making up 68percent of the sample. Results in Table 2 show that crossfunctional dispersion of marketing activities and power ofmarketing are significant predictors of performance whileinteraction is not.

Cross functional dispersion of marketing activities ispositively related while power of marketing is negativelyrelated. These show that organizations with more inte-grated marketing department are more likely to have otherdepartments being involved and performance will belower. This seems to mean that too much concentration ofmarketing functions in one department is not a good thingand dispersion of marketing functions have an instrumen-tal effect on performance. The negative beta of the powerof marketing indicates that having a powerful marketingfunction does not guarantee performance but on the con-trary may not be good for the organization. Again, onlycompetitor orientation is found to be significantly relatedto performance and this is the case only when the non-

structural marketing organizational factors are enteredinto the equation. Thus, H2 is supported while H3 and H4are not.

DISCUSSIONS

Results of this study lend support to our argument thatnot only is marketing organization important and instru-mental to firm performance, it is also crucial in cultivatinga market orientation culture. This is consistent with previ-ous research (Moorman and Rust 1999; Merlo and Auh2009; Verhoef and Leeflang 2009) and possibly providesadditional empirical evidence to counter the accusationthat marketing is losing its importance. This study alsoprovides another empirical context for testing this rela-tionship so can contribute to cross-cultural comparison.

Our results suggest that there is still need for a formalmarketing organization that can serve as the hub forcultivating expertise in understanding and serving cus-tomers. Other forms of marketing structure seem not asuseful like having a sales department, an integrated mar-keting and sales department or a department equivalent tomarketing. The common misperception that Chinese man-agers have limited understanding of marketing seems tobe at fraud. Many of them can actually distinguish market-ing and sales as separate functions, and that each havedifferent implications on their business success.

While a formal marketing department is relevant,there is also a need to disperse marketing responsibilitiesto other functions in the organization appropriately sopeople are assigned the tasks they do best. This will alsoensure that different functions have sufficient under-

FIGURE 1Effect of Marketing Department on MO-Performance Relationship


standing of marketing and its values thus creating a trulymarket-oriented organization. Our results here actuallycoincide with the contention of George, Freeling, andCourt (1994) who propose that the marketing organiza-tion should be reinvented to incorporate integrators andfunctional specialists within the same organization arena.It is essential to integrate all company efforts to “servingeach distinct consumer, channel or product segment”(p. 46) while at the same time marketing can specialize indeveloping their capabilities on gathering market intelli-gence, promotion, advertising, etc. in order to createcustomer value.

In this study, we employed a two-stage unsophisti-cated analysis approach where all companies are analyzedin the first stage and only those with a marketing depart-ment in the second stage. Prescreening for a formalmarketing structure enables us to accurately delineate theeffect of the non-structural aspects of marketing organiza-tion. Other studies did so assuming that there is a market-ing department but this may not often be the case particu-larly with smaller organizations. This method allows us tofirst identify the effect of a structural organization and,second, how this formal structure can be further designedin terms of assignment of responsibilities, its status versus

other functions and the amount of cross functional inter-actions, in order to further enhance its contribution. To thebest of the author’s knowledge, no study has been done toisolate the performance impact of a formal marketingstructure. This study seeks to fill in this research gap.

Another key implication of this study is that marketorientation should be studied with a disaggregatedapproach. The three components of market orientation(customer orientation, competitor orientation andinterfunc-tional coordination) originated from Narver andSlater (1990) do not behave similarly and have differentimplications on performance. Only competitor orienta-tion is found to be the only significant predictor ofperformance in our study. This is consistent with thefindings of Noble, Sinha, and Kumar (2002) who studiedthe effect of strategic orientation on firm performance.Although a lot of studies view market orientation inaggregate, significant differences exist in the way how thethree components influence performance (Lukas andFerrell 2000; Zhou, Brown, and Dev 2008) so worthindividual assessment. While customer orientation refersto the ability to understand and respond to customersneeds so as to create superior value, competitor orienta-tion focuses on doing so better than others through a clear

TABLE 2Regression Results of Nonstructural Marketing Organization

Step 1 Step 2 Step 3

Beta T Beta T Beta T

Control variables:

Size of company 0.164 2.458* 0.140 2.128* 0.164 2.536*Years in China -0.006 -0.094 -0.013 -0.207 -0.015 -0.233Type of industry -0.030 -0.471 -0.056 -0.911 -0.055 -0.923Type of ownership 0.001 0.010 0.001 0.008 0.017 0.289

Market orientation:Customer orientation 0.101 1.261 0.055 0.698Competitor orientation 0.155 1.947 0.158 2.046*Interfunctional coordination 0.046 0.554 0.092 1.136

Non-structural marketing organization:Cross functional dispersion 0.204 3.377**Power of marketing -0.126 -2.113*Interaction with marketing 0.075 1.283

F 1.864 3.773** 4.680**R-square 0.029 0.095 0.159R-square change – 0.067** 0.064**

* p < 0.05; **p < 0.01


understanding of competitors’ strengths/weaknesses aswell as its capabilities and strategies (Narver and Slater1990). Given the highly competitive environment of China,this result is hardly surprising.

Evidence suggests that organization system and struc-ture and their impact on performance deserve more atten-tion than it currently has. Results of this study provide

support that there is both direct and indirect impact on keyissues in the organization. More research is needed toexplore the process through which marketing organiza-tion can enhance the implementation of marketing. Morehas to be done also on understanding different forms ofmarketing structure. In this way, synergy can be achievedwith aligning all aspects of the organization for superiorperformance.

REFERENCES

Achrol, Ravi S. (1991), “Evolution of the MarketingOrganization: New Forms for Turbulent Environ-ments,” Journal of Marketing, 55 (4), 77–93.

____________ (1997), “Changes in the Theory of Inter-Organizational Relations in Marketing: Toward aNetwork Paradigm,” Academy of Marketing ScienceJournal, 25 (1), 56–71.

Akimova, Irina (2000), “Development of Market Orien-tation and Competitiveness of Ukrainian Firms,”European Journal of Marketing, 34 (9/10), 1128–48.

Anderson, Paul F. (1982), “Marketing, Strategic Plan-ning, and the Theory of the Firm,” Journal of Market-ing, 46 (Spring), 15–26.

Brady, John and Ian Davis (1993), “Marketing’s Mid-Life Crisis,” McKinsey Quarterly, 14 (2), 17–28.

Day, George S. (1994), “Capabilities of Market-DrivenOrganizations,” Journal of Marketing, 58 (October),37–52.

____________ (1997), “Aligning the Organization to theMarket,” in Reflections on the Futures of Marketing,D.R. Lehmann and K.E. Jocz, eds. Cambridge, MA:Marketing Science Institute, 67–93.

____________ (1999), “Creating a Market-Driven Orga-nization,” Sloan Management Review, 41 (1), 11–22.

Deshpandé, Rohit, John U. Farley, and Frederick E.Webster (1993), “Corporate Culture, Customer Ori-entation, and Innovativeness in Japanese Firms: AQuadrad Analysis,” Journal of Marketing, 57 (Janu-ary), 22–27.

Dess, Gregory G. and Richard B. Robinson (1984), “Mea-suring Organizational Performance in the Absence ofObjective Measures: The Case of the Privately-HeldFirm and Conglomerate Business Unit,” StrategicManagement Journal, 5 (3), 265–73.

Drucker, Peter (1954), The Practice of Management. NewYork: Harper & Row.

George, Michael, Anthony Freeling, and David Court(1994), “Reinventing the Marketing Organization,”McKinsey Quarterly, 4, 43–62.

Greyser, Stephen A. (1997), “Janus and Marketing: ThePast, Present, and Prospective Future of Marketing,”in Reflections on the Futures of Marketing, D.R.

Lehmann and K.E. Jocz, eds. Cambridge, MA: Mar-keting Science Institute.

Griffin, Abbie and John R. Hauser (1996), “IntegratingMechanisms for Marketing and R&D,” Journal ofProduct Innovation Management, 13 (3), 191–215.

Gupta, Ashok K., S.P. Raj, and David L. Wilemon (1986),“A Model for Studying R&D: Marketing Interface inthe Product Innovation Process,” Journal of Market-ing, 50 (2), 7–17.

Harris, Lloyd C. and Emmanuel Ogbonna (2003), “TheOrganization of Marketing: A Study of Decentral-ized, Devolved, and Dispersed Marketing Activity,”Journal of Management Studies, 40 (2), 483–512.

Hise, Richard T. (1965), “Have Manufacturing FirmsAdopted the Marketing Concept,” Journal of Mar-keting, 29 (July), 9–12.

Homburg, Christian, John P. Workman, and Ove Jensen(2000), “Fundamental Changes in Marketing Orga-nization: The Movement Toward a Customer-Focused Organizational Structure,” Journal of theAcademy of Marketing Science, 28 (4), 459–78.

Jaworski, Bernard J. and Ajay Kohli (1993), “MarketOrientation: Antecedents and Consequences,” Jour-nal of Marketing, 57 (July), 53–70.

Keh, Hean Tat, Thi Tuyet Mai Nguyen, and Hwei Ping(2007), “The Effects of Entrepreneurial Orientationand Marketing Information on the Performance ofSMEs,” Journal of Business Venturing, 22 (4), 592–611.

Kotler, Philip, Swee Hoon Ang, Chin Tiong Tan, KevinLane Keller, and Siew Meng Leong (2006), Market-ing Management: An Asian Perspective. U.S.: PrenticeHall.

Krohmer, Harley, Christian Homburg, and John P. Work-man (2002), “Should Marketing Be Cross-Func-tional? Conceptual Development and InternationalEmpirical Evidence,” Journal of Business Research,55 (6), 451–65.

Lukas, B. and O.C. Ferrell (2000), “The Effects of MarketOrientation on Product Innovation,” Journal of theAcademy of Marketing Science, 28 (Spring), 239–47.

McCarthy, E. Jerome (1960), Basic Marketing: A Mana-gerial Approach. Homewood IL: Irwin.

McKee, Daryl O., Rajan Varadarajan, and William M.


Pride (1989), “Strategic Adaptability and MarketPerformance: A Market-Contingent Perspective,”Journal of Marketing, 53 (3), 21–35.

McKenna, Regis (1991), “Marketing is Everything,”Harvard Business Review, 69, 79.

Merlo, Omar and Seigyoung Auh (2009), “The Effects ofEntrepreneurial Orientation, Market Orientation, andMarketing Subunit Influence on Firm Performance,”Marketing Letters, (25 March).

Moorman, Christine and Roland T. Rust (1999), “TheRole of Marketing,” Journal of Marketing, JM/MSISpecial Issue on Fundamental Issues in Marketing,63 (Special Issue), 180–97.

Narver, John C. and Stanley F. Slater (1990), “The Effectof a Market Orientation on Business Profitability,”Journal of Marketing, 54 (4), 20–35.

Noble, Charles H., Rajiv K. Sinha, and Ajith Kumar(2002), “Market Orientation and Alternative Strate-gic Orientations: A Longitudinal Assessment of Per-formance Implications,” Journal of Marketing, 66(4), 25–39.

Olson, Eric M., Stanley F. Slater, and G. Tomas M. Hult(2005), “The Performance Implications of Fit AmongBusiness Strategy, Marketing Organizational Struc-ture and Strategy Behavior,” Journal of Marketing,69 (July), 49–65.

Piercy, Nigel (1986), “The Role and Function of the ChiefMarketing Executive and the Marketing Department:A Study of the Medium-Sized Companies in theU.K.,” Journal of Marketing Management, 1 (3),265–89.

Ruekert, Robert W. and Orville C. Walker (1987),“Marketing’s Interaction with Other Functional Units:A Conceptual Framework and Empirical Evidence,”Journal of Marketing, 51 (January), 1–19.

____________, ____________, and Kenneth J. Roering(1985), “The Organization of Marketing Activities:A Contingency Theory of Structure and Performance,”Journal of Marketing, 49 (Winter), 13–25.

Schultz, Don E. (2003), “Marketing Gets No Respect inthe Boardroom,” Marketing News, (November 24).

Sheth, Jagdish N. and Rajendra S. Sisodia (2002), “Mar-keting Productivity: Issues and Analysis,” Journal ofBusiness Research, 55, 349–62.

Siu, Wai Sum (2000a), “Marketing and Company Perfor-mance of Chinese Small Firms in Hong Kong,”Marketing Intelligence and Planning, 18 (5), 292–307.

____________ (2000b), “Marketing Philosophies andCompany Performance of Chinese Small Firms in

Hong Kong,” Journal of Marketing Theory and Prac-tice, 8 (1), 25–37.

Slater, Stanley F. and John C. Narver (1994), “DoesCompetitive Environment Moderate the Market Ori-entation-Performance Relationship?” Journal ofMarketing, 58, 46–55.

____________ (1995), “Market Orientation and the Learn-ing Organization,” Journal of Marketing, 59 (July),63–74.

Tull, Donald S., Bruce E. Cooley, Mark R. Philips, Jr., andHarry S. Watkins (1991), “The Organization of Mar-keting Activities of American Manufacturers,” Work-ing Paper No. 91–126, Cambridge, MA: MarketingScience Institute.

Varadarajan, Rajan (1992), “Marketing’s Contribution toStrategy: The View from a Different Looking Glass,”Journal of the Academy of Marketing Science, 20 (4)(Fall), 335–43.

Verhoef, Peter C. and Peter S.H. Leeflang (2009),“Understanding the Marketing Department’s Influ-ence Within the Firm,” Journal of Marketing, 73(March), 14–37.

Vorhies, Douglas W. and Neil A. Morgan (2003), “AConfiguration Theory Assessment of MarketingOrganization Fit With Business Strategy and ItsRelationship with Marketing Performance,” Journalof Marketing, 67 (1), 100–15.

Webster, Frederick E., Jr. (1992), “The Changing Role ofMarketing in the Corporation,” Journal of Market-ing, 56 (October), 1–17.

____________ (1997), “The Future Role of Marketing inthe Organization,” in Reflections on the Futures ofMarketing, D.R. Lehmann and K.E. Jocz, eds. Cam-bridge, MA: Marketing Science Institute, 39–66.

____________ (2005), “A Perspective on the Evolutionof Marketing Management,” Journal of Public Policy& Marketing, 24 (1), 121–26.

____________, A.J. Malter, and S. Ganesan (2005), “TheDecline and Dispersion of Marketing Competence,”MIT Sloan Management Review, (Summer), 35–43.

Workman, John P., Christian Homburg, and Kjell Gruner(1998), “Marketing Organization: An IntegrativeFramework of Dimensions and Determinants,” Jour-nal of Marketing, 62 (3), 21–41.

Zhou, Kevin Z., James R. Brown, and Chekitan S. Dev(2008), “Market Orientation, Competitive Advan-tage, and Performance: A Demand-based Perspec-tive,” Journal of Business Research, doi:10.1016/j.jbusres.2008.10.001.


For further information contact:Lancy Mac

University of MacauAvenida Padre Tomas Pereira

Taipa, MacauChina

Phone: 853.8397.4758Fax: 853.2883.8320



SUSPICION OF ULTERIOR MOTIVES IN SALES INTERACTIONS:EXAMINING THE CASE OF VALUE CO-CREATION

Nicole Kirpalani, Hofstra University, Hempstead

SUMMARY

The notion of value co-creation has received muchinterest in recent years. At the heart of the value co-creationconcept is the idea that “value is defined by and co-createdwith the consumer rather than embedded in output” (Vargoand Lusch 2004, p. 6). Not much is known so far abouthow consumers perceive the interaction with a co-cre-ation agent, particularly with respect to their persuasionknowledge (Friestad and Wright 1994). Even though theyare presumed to be cooperative in nature, co-creationinteractions may be more complex and ambiguous thaninteractions involving traditional sales agents. While thereare studies of persuasion knowledge in sales interactions(e.g., Campbell and Kirmani 2000; DeCarlo 2005), extantresearch thus far has not addressed the area of co-creation.This topic has importance in light of research suggestingthat sales interactions are subject to a “sinister attributionbias” – once consumers categorize someone as a salesper-son, they infer sinister ulterior motives, even when nosuch motive is plausible (Main, Dahl, and Darke 2007).Little is known thus far about the extent to which consum-ers have such negative biases in co-creation interactions.

As suggested by Kirmani and Zhu (2007), persuasionknowledge generally entails suspicion – a state in whichconsumers have suspended judgment and are activelyengaged in attributional thinking regarding possible ulte-rior motives (Fein 1996). Understanding how suspicionevolves in a sales encounter is important, as suspicion mayact as an impediment to the delivery of value. This isespecially critical in a value co-creation situation, whereboth the process and the outcome are important compo-nents of value.

The main goal of the present research is to explorewhen and how a consumer becomes suspicious that aco-creation agent she/he is interacting with has an ulteriormotive for his or her behavior. As suggested by Marchandand Vonk (2005), suspicion is proposed to be a gradualprocess that perceivers undergo as more situational infor-mation becomes available. Of special interest is the pat-tern of suspicion as it evolves during the course of the co-creation interaction. Consistent with this goal, a processoriented study involving a scenario-based co-creationsituation has been developed. Building on Marchand andVonk’s (2005) research design and procedure, the presentresearch examines the process of suspicion in an unfold-ing co-creation situation and aims to provide insights into

the nature of consumers’ cognitive responses during theprocess.

Pretests were conducted in terms of context andprocedure. Based on the pretest results, a scenario involv-ing the customized design of a diamond engagement ringwas developed. In this scenario, a jewelry designer assistsa couple to custom design an engagement ring. Consistentwith Marchand and Vonk’s (2005) research design, thescenario contains ten statements that provide informationof the designer’s part of the dialogue. All ten statementswere pretested to be equally informative. The statementswere developed with the natural progression of a realisticsales encounter in mind. Throughout the interaction, thedesigner makes recommendations for elements the cus-tomers may like.

Sixty-four undergraduate students participated in thestudy and were given booklets containing the scenario andinstructions. The task was to imagine oneself in theinteraction with the jewelry designer and form an impres-sion of the situation. Participants were then asked to writedown their thoughts for each statement for two minuteseach. Participants’ cognitive responses to each of the tenstatements were transcribed and coded for suspicion.Marchand and Vonk’s (2005) coding scheme was appliedin order to assess participants’ level of suspicion for eachof the ten statements. Accordingly, levels of suspicionwere coded with a suspicion score of “3” for maximumsuspicion, a score of “2” for moderate suspicion, and ascore of “1” when the participant was not suspicious(when she/he felt certain that the designer had eithergenuine motives or ulterior motives).

Of the 64 participants, 14 (22%) did not becomesuspicious. The remaining 50 participants (with suspicionmeans > 1.00) were subject to further analysis. A repeatedmeasures ANOVA with suspicion levels for the ten state-ments (1 – 10) as within-subjects factor was performed.Results showed that the level of suspicion first increasedin statements 1 through 3, consistent with the initiationphase of suspicion proposed by Marchand and Vonk(2005). Suspicion then dropped (statements 4 and 5),before increasing slightly for statements 6 and 7 (“attribu-tion phase”). Finally, suspicion leveled off for statements8 through 10 (“resolution phase”). More formally, therepeated measures ANOVA showed a significant lineartrend for level of suspicion (F (1, 49) = 15.98, p = .000).Further regression analysis on the linear model showed


that as the interaction progressed (statements 1 through10), suspicion decreased (coefficients: α = 1.533, ß =-.39.) The linear model produced an R2 of .52 (F (1, 8) =8.52, p = .019).

In conclusion, the research provided several insightsinto the natural process of suspicion in a co-creationcontext. Without explicitly manipulating possible trig-gers of suspicion, the analysis of participants’ cognitiveresponses showed a negative linear trend, with suspicionfirst increasing, but then decreasing as the interactionprogressed. This pattern was consistent with previousresearch by Marchand and Vonk (2005). Overall, thestudy suggests that suspicion in itself does not necessarily

lead to the conclusion that the agent has ulterior motives.While suspicion may be triggered, it is not always anegative inference, but may resolve itself in a positivedirection instead.

The process oriented view of the co-creation interac-tion has its limitations. The insights gained from the studyrelied on the analysis of the participants’ cognitive re-sponses. Future research could experimentally manipu-late specific factors that might trigger suspicion in thiscontext (for example, a consumer’s expectation of theinteraction, as well as salesperson behavior that mayaffect suspicion) in order to be able to explain suspicionpatterns in co-creation in more detail.

For further information contact:Nicole Kirpalani

Department of Marketing and International BusinessFrank G. Zarb School of Business

Hofstra University222 Weller Hall

Hempstead, NY 11549Phone: 516.463.5706



SALESPERSON’S KARMA ORIENTATION: A CONCEPTUALFRAMEWORK AND RESEARCH PROPOSITIONS

Ramendra Singh, Indian Institute of Management Ahmedabad, IndiaRakesh Singh, Times of India, XLRI Jamshepur, India

SUMMARY

Facing a role that is dominated by stressful customerengagements, and chasing sales quotas, salespeople arelikely to be guided by their own personal values andorganizational environment, and in the long term arelikely to develop spiritual thoughts that allow them toview their work with a sense of duty that brings more joy,commitment, satisfaction to their job (Giacalone andJurkiewicz 2003). Indian philosophy of life views allindividual actions having latent power to bring joy if theaction is good, or sorrow otherwise (Dasgupta 1991 ascited in Mulla and Krishnan 2006), and popularly knownas an individual’s karma. In this paper, we propose a newconstruct, salesperson’ Karma Orientation, conceptual-ized at the individual salesperson’s level. Karma orienta-tion of salespeople influences their ability and willingnessto perform sales related tasks. Such clarity onself-awareness is essential in relationship developmentwith customers (Badrinarayanan and Madhavaram 2008),and enhances their selling effectiveness, ethical behav-iors, and spiritual well-being, providing the salespersonswith a sense of duty toward their work. Karma orientationmotivates an individual to make efforts toward work withdetachment to work, and no expectation of rewards inreturn.

Karma Orientation-Conceptualization and Dimen-sions

Based on the above theoretical development for thisnew construct, we define the karma orientation of sales-persons as their behaviors resulting from a sense ofselfless action, performed while doing their work as a dutytoward customers in particular, and society in general.We propose four dimensions of karma orientation1 ofsalespersons:

1. Work as selfless action,

2. Work as duty toward others,

3. Detachment from work-related rewards,

4. Equanimity under environmental influences,

Conceptual Framework

We propose the following conceptual framework forsalesperson’s karma orientation with key antecedents andconsequences, as shown in the illustration below.

Antecedents Consequences

EthicalClimate

SupervisoryBehaviors

PersonalValues

OrganizationalValues

Salesperson’sEffectiveness

SpiritualWellbeing

EthicalBehaviors

OrganizationalCommitment

JobSatisfaction

Karma OrientationWork as selfless action.

Work as duty towards others.Detachment from work-related rewards.

Equanimity under environmentalinfluences.


Prpositions

Proposition 1: Higher ethical climate in organizationswould lead to higher karma orientation of its salespersons.

Proposition 2: Positive Supervisory behaviors emphasiz-ing a sense of duty and service will have positive impacton salesperson’s karma orientation.

Proposition 3: Salesperson’s personal values such assense of calling, duty and need to contribute to society,will have positive impact on his/her karma orientation.

Proposition 4: Organizational values emphasizing indi-vidual dignity and personal growth will have positiveimpact on salesperson’s karma orientation.

Proposition 5: Salesperson’s with high karma orientationwould show higher selling effectiveness.

Proposition 6: Higher salesperson’s karma orientationwould lead to their higher spiritual wellbeing.

Proposition 7: Higher karma orientation of salespersonsleads to their higher ethical behaviors.

Proposition 8: Higher karma orientation of salespersonsleads to their higher organizational commitment.

Proposition 9: Salespersons with high karma orientationare likely to have high job satisfaction.

Managerial Implications

Along with the challenges to maximize sales forceeffectiveness, sales managers face another set of complextasks of managing workplace stress and anxiety, greed forperformance rewards, and maintaining work-life balance.The spiritual needs of salespeople, such as the search fora greater meaning in their work need to be addressed bythe sales managers by offering broader definitions of thegoals that their sales team strives for. By promotingspiritual interpretations of business situations and sup-porting karmic orientation of salespeople, sales managerscan make significant improvement in the psychologicalwell-being of their people. As such individuals with highlevels of karmic orientation are likely to be high oncustomer empathy, emotional intelligence, and sense ofduty toward others. Therefore, a team of salespeople withkarmic orientation will be delivering better on goalsachievements as their efforts and work strategies will bedriven by a sense of calling, and duty. In such a scenario,sales managers need not tradeoff their short-term andlong-term objectives as the sales people will function withequanimity toward environmental situations. Further, salesmanagers need to adopt and promote humanistic values atthe firm level, to establish practices that are ethical, andthat recognize the self-worth of individual salespeople, sothat they achieve personal as well as organizational goals(Badrinarayanan and Madhavaram 2008). References areavailable upon request.

ENDNOTE

1 Dimensions 2 and 3 have been also suggested in Mulla

and Krishnan (2006) and dimensions 2, 3, and 4 alsodescribed in Mulla and Krishnan (2007). Howeverthese studies are not in salesperson’s context.

Ramendra SinghIndian Institute of Management Ahmedabad

FPM House-1IIM Ahmedabad

Vastrapur, Ahmedabad 380015India



WHEN CUSTOMER-ORIENTED SALES EFFORTS LOSE GROUND:AN EXPERIMENTAL STUDY ON THE “DILUTION EFFECT”

Peter Mathias Fischer, University of St. Gallen, SwitzerlandWalter Herzog, WHU Otto Beisheim School of Management, Germany

Sven Reinecke, University of St. Gallen, Switzerland

SUMMARY

Since customers are becoming more and more pow-erful as well as increasingly demanding (Day and Mont-gomery 1999), companies have begun to take initiativeswhich should help to retain their most valuable customers(Gummesson 1998). Most importantly, in order to betteradapt to various customer needs (George, Freeling, andCourt 1994; Johnston and Marshall 2006), marketing andsales organizations have abandoned their product-oriented or geographical organizational structures in favorof customer-oriented sales structures (Homburg, Work-man, and Jensen 2000).

Such an organizational change implies a decline offunctional boundaries, an increased importance of cross-functional teams (Montgomery and Webster 1997), astrong emphasis on key account management, and anestablishment of customer segment managers with a highlevel of strategic responsibility (Homburg, Workman,and Jensen 2000). In addition to their financial perfor-mance (e.g., sales or contribution margin), customer seg-ment managers are increasingly held responsible for thesuccessful implementation of strategic customer segment-specific projects such as new product planning, brandmanagement, complaint and churn management (Cravens1995; Neslin et al. 2006). Such projects would havehardly been possible to implement in a marketing andsales organization primarily organized by products orgeographical regions (cf., Sheth and Sharma 2008).

However, even though most sales directors contendthat such strategic projects and long-term customer rela-tionships are the core focus of their sales strategy (Cra-vens 1995) and even though necessary organizationalchanges have indeed been undertaken in many firms(Homburg, Workman, and Jensen 2000), most companiesstill lack customer centricity (Shah et al. 2006). Wepropose that one explanation for this deficit is a subtlepreference shift of the sales director when she/he evalu-ates the performance of his/her sales segment managers.More specifically, we argue that over time sales directorsincreasingly devalue achievements in strategic customersegment-specific projects. This assumption rests on aseries of studies on the so-called “common measure bias,”whereupon individuals tend to neglect dissimilar infor-mation cues due to their lower perceived comparability

(Slovic and MacPhillamy 1974). Since customer seg-ment-specific projects reflect particular needs of hetero-geneous customer segments, they are perceived as lesscomparable than financial key performance indicators.We assume that this, in turn, leads to a persistentlyreduced preference for customer segment-specific projects(cf., Kivetz and Simonson 2000). This “dilution effect,”as we call it, is obviously a major threat to the successfulimplementation of customer oriented sales structures andthe idea of customer orientation in general.

We ran two experimental studies with experiencedpractitioners in order to test if customer segment-specificprojects are devalued over time. To do so, we conducteda first (scenario-based) study in which 139 sales execu-tives were assigned to three between-subjects conditionsonly differing by the number of performance evaluations(2 vs. 4 vs. 6 evaluations) they had to carry out. Whereas,on average, in condition 1 (2 periods) participants assignedan importance weight of 45.1 percent to project achieve-ment, the importance weight decreased to 42.56 percent incondition 2 (4 periods) and to 37.70 percent in condition3 (6 periods). Not only the main effect was significant(F(2, 126) = 4.14, p = .018); a trend analysis also indicateda linear contrast (F(1, 126) = 8.217, p = .005).

Study 2 built on study 1 and replicated the effect in adifferent and more elaborate scenario-based experimentwhich allowed to better control for confounding factors(e.g., cognitive load). Beyond, study 2 showed that theeffect only occurs if customer segment-specific projectsare perceived as dissimilar: Perceived similarity wassuccessfully manipulated by randomly asking partici-pants (145 sales executives) to either elaborate on differ-ences or similarities of two customer segment-specificprojects. Results showed that only participants assignedto the dissimilarity condition devalued customer seg-ment-specific projects over time.

The results of the studies have important implica-tions. First, even though sales directors aim to considerlong-term, strategic, and customer segment-specificprojects when evaluating their sales segment managers,they devalue these measures over time and hence partlyjeopardize the advantages of the customer oriented salesorganization. Second, the more heterogeneous the salesorganization’s customer segments, the more the dilution


effect seems to be a major concern. Hence, in order tomitigate this negative effect, the challenge of such orga-nizations is to assign customer segment-specific projectswhile emphasizing the commonness of the different

projects. Developing and empirically testing practicalstrategies in doing so is a fruitful avenue for furtherresearch. References are available upon request.

For further information contact:

Peter Mathias FischerInstitute of Marketing

University of St. GallenDufour Street No. 40a

St. Gallen, 9000Switzerland

Phone: +41(0)71.224.2888Fax: +41(0)71.224.2835



FROM A DISTANCE WE ALL HAVE ENOUGH: TEMPORALPERSPECTIVE AND THE DUAL ROLE OF PRICE

Christian Homburg, University of Mannheim, GermanyTorsten Bornemann, University of Mannheim, Germany

SUMMARY

Products are often evaluated from different temporalperspectives. Consumers exposed to the preannouncementof a new product, for example, may already evaluate thisproduct for consumption from such a distant perspective,long before it is available for purchase. When laterencountering the launched product, they may reevaluate itfor consumption from a proximal perspective. Other con-sumers that have not been exposed to the preannouncementmay form their first evaluation of the product when it islaunched and available for purchase, thus from a proximalperspective only. In either case, consumers mentally tradeoff the costs and the benefits associated with the consump-tion of the product. During this evaluation process, theproduct’s price can be diagnostic for both the costs asso-ciated with the purchase of the product (via the price-perceived sacrifice relationship) as well as the benefits ofusing it (via the price-perceived quality relationship).

This research introduces the temporal perspective ofa product evaluation as a context factor influencing con-sumers’ interpretation and use of price information.Construal level theory argues that over time, differentaspects of an action assume salience. Evaluations of anaction from a distant perspective are primarily based onaspects related to desirability and the benefits of thataction. However, with greater temporal proximity, con-sumers’ focus shifts to feasibility- and cost-related aspects.We conceptualize the price-perceived quality relation-ship as related to the desirability of using a product and theprice-perceived sacrifice relationship as related to thefeasibility associated with the purchase. In line withconstrual level theory, we find that the degree to whichconsumers interpret price information according to itsdesirability- or feasibility-related implications dependson the temporal perspective of the evaluation context(experiment 1). Specifically, we show that when a productis evaluated for distant future (vs. proximal) consumption,the price-perceived quality relationship is more pronouncedwhereas the reverse holds true for the price-perceivedsacrifice relationship. The same high-priced product wasconsequently evaluated more favorably when consideredfor distant future compared to proximal consumption.

However, many consumers may initially receiveinformation about a product’s price from a distant per-spective (e.g., in the form of a preannouncement) beforethey evaluate the same product for proximal consumption

at a later point in time. Experiment 2 shows that whenprice has already been used as an input for productevaluation from a distant perspective, its evaluative impli-cations as an indicator of quality may become “immortal-ized” and price information may not be reinterpretedaccording to its feasibility-related implications in a laterevaluation for proximal consumption. As a consequence,later product evaluations for proximal consumption aresimilar to initial evaluations from a distant perspective.Process insights suggest that this effect stems from anattenuated focus on price information when price hasalready been considered in a previous evaluation com-pared to when it constituted a new piece of informationduring judgments for proximal consumption.

These findings provide valuable insights for researchand managerial practice. First, this research underlines theimportance of time for consumers’ evaluation and use ofprice information. Whereas previous research found thattime pressure influenced consumers’ use of price infor-mation as an indicator of quality and monetary sacrifice,our study reveals that also the temporal perspective of theevaluation context has an impact on consumers’ interpre-tation of price information. Second, our research followscalls to examine effects of temporal construal not onlybetween but also within subjects. Specifically, we extendprevious research on construal-carryover effects by ex-amining the repeated exposure to and evaluation of infor-mation that can be interpreted as a desirability-related andas a feasibility-related aspect (experiment 2). Extendingprevious research on sequential decision making, we findevidence for the neglect of information that has been usedto complete a task at time one in subsequent evaluations.

Finally, our study has important implications for thedesign of pre-launch communication activities, which arecommon especially in high technology industries. In theseindustries, achieving the highest possible adoption rates atthe time of market launch is important in order to recoupR&D costs. Moreover, many products in these industriesare launched with a high price strategy (e.g., Sony’s newPlayStation Portable). Our results show that evaluationsof a high-priced product for immediate consumption atlaunch are significantly more favorable when price infor-mation was already considered in a previous productevaluation from a distant perspective. Thus, price mayconstitute an important piece of information in new prod-uct preannouncements.


For further information contact:Christian Homburg

Marketing DepartmentUniversity of Mannheim


Phone: +49.621.181.1555Fax: +49.621.181.1556



PRICE CONTEXT, PRIOR KNOWLEDGE, AND WORKINGMEMORY CAPACITY

Shan Feng, Drexel University, PhiladelphiaRajneesh Suri, Drexel University, Philadelphia

SUMMARY

Prior product knowledge has been shown to influ-ence the cues that consumers use to assess product quality(Rao and Monroe 1988; Rao and Sieben 1992). Knowl-edgeable consumers tend to use intrinsic cues, whereasless knowledgeable consumers tend to use extrinsic cuessuch as price to infer a product’s quality. However,judgments of a product are affected not only by its owncharacteristics but also by characteristics of other prod-ucts that are judged concurrently (Farley, Katz, andLehmann 1978) or retrieved from memory (Biernat, Manis,and Nelson 1991; Kahneman and Miller 1986; Upshaw1962; Urbany and Dickson 1990). Such shopping behav-ior raises a question as to how other products viewed whenevaluating a target product, i.e., price context, mightinfluence consumers’ use of price to infer a product’squality. Will consumers use price to infer a product’squality more when competitive products in an evaluationcontext are priced relatively lower than or higher than atarget product?

Rao and Monroe (1988) suggest that in quality evalu-ation the attention paid to price will first decline and thenincrease, as prior knowledge increases. They argue thathigh knowledgeable consumers, relative to moderateknowledgeable consumers, are more likely to use price toinfer quality when they believe there is a positive price-quality relationship in the market because price informa-tion is easier to process and interpret. Given the dual roleof price to evaluate monetary sacrifice and to infer quality(Monroe 2003) when consumers become more concernedabout monetary sacrifice associated with a product theyare more likely to rely on their prior knowledge structure.It is argued that the utilization of prior knowledge could becontext-driven. When the target product is relativelyexpensive in an evaluation context, concerns about mon-etary sacrifice makes highly knowledgeable and lessknowledgeable consumers get more overwhelmed bytheir knowledge (or lack thereof) when assessing mon-etary sacrifice. Consequently these consumers will useprice more to infer a product’s quality compared to thosewith moderate prior knowledge. On the other hand, whenthe target product’s price is relatively low, concerns aboutmonetary sacrifice are low and consequently consumersirrespective of their knowledge are likely to evaluate theproduct information thoroughly. As a result, price willless likely be used to evaluate a product’s quality.

In Study 1, the target product (women’s blazer,$109.99) was evaluated in one of the two price contexts:low ($19.99, $49.99, $79.99) vs. high ($139.99, $179.99,$219.99). The results showed a significant interactioneffect of evaluation context and prior knowledge onconsumers’ quality perceptions. Contrasts indicated thatwhen the target product is relatively expensive in theevaluation context, product quality will be perceivedhigher for participants with low and high prior knowl-edge, but not for those with moderate knowledge. Inter-estingly, however, there were no significant differenceson quality perceptions among consumers with differentlevels of prior knowledge when the target product’s priceis relatively low in the context. These results suggest thatprice is used more to infer quality when the target product’sprice was relatively high. Evaluating a target product in acontext requires consumers combine information pro-cessing and storage simultaneously. Working memory isoften used to describe the ability to simultaneously main-tain and process goal-relevant information (Conway et al.2007). Furthermore, working memory capacity interactswith prior knowledge to affect people’s cognitive perfor-mance, such as the ability to process information moredeeply. For example, Hambrick and Engle (2002) sug-gests a “rich get richer” hypothesis by arguing that knowl-edgeable people with high levels of working memorycapacity derived a greater benefit from their knowledgethan those with lower levels of working memory capacity.It is likely that working memory capacity of consumersmight impact the utilization of prior knowledge in qualityjudgment. For consumers with high prior knowledge,higher working memory capacity will increase their use ofprior knowledge by focusing more on the intrinsic at-tributes of the product in quality perception.

Study 2 replicated Study 1 with a different product(digital camera) as well examined the role of workingmemory capacity and prior knowledge on the evaluationof relatively high or low prices. The target product($159.99) was evaluated in one of the two price contexts:low ($49.99, $79.99, $109.99) vs. high ($209.99, $249.99,$289.99). Following Hambrick and Engle (2002), work-ing memory capacity measured both the operation spantask (Turner and Engle 1989) and the counting span task(Engle et al. 1999). Results showed an interaction be-tween prior knowledge and working memory capacitywhen target price was relatively high in the context.Consumers with high prior knowledge with high working


memory capacity were less likely to use price to infer atarget product’s quality than those with low workingmemory capacity. No interactions were observed in highprice context.

The results from both studies are consistent in that therelationship exhibited between prior knowledge and rela-

tive attention paid to price is U-shaped only when the priceof a target product is relatively high compared to otherproducts in the evaluation context. Furthermore, workingmemory capacity impacts the use of prior knowledge butprimarily for consumers with high prior knowledge.

For further information contact:Shan Feng

Drexel UniversityP. O. Box 34729

Philadelphia, PA 19101Phone: 215.895.2145



THE IMPACT OF FAVORING NEW CUSTOMERS OVER LOYALCUSTOMERS ON LOYAL CUSTOMERS’ PERCEPTIONS

OF PRICE FAIRNESS

Martin Fassnacht, WHU – Otto Beisheim School of Management, GermanyJochen Mahadevan, WHU – Otto Beisheim School of Management, Germany

SUMMARY

Behavioral pricing literature identifies the issue ofprice fairness as an important factor for customers’ reac-tions to prices. Several studies show that customers tendto react negatively toward a company if they perceive aprice as unfair (e.g., Campbell 1999; Xia, Monroe, andCox 2004). We define price fairness as a customer’sassessment of whether the comparison of a perceivedprice with a reference price is reasonable, acceptable orjustifiable. Reference prices can be past prices or competi-tor prices (Bolton, Warlop, and Alba 2003). While theissue of price fairness has only been started to be investi-gated in the recent past, the issue of price discriminationhas already been studied for a long time. In this study wefocus on personal price discrimination in terms of favor-ing new customers over loyal customers. This representsa common strategy, especially in the service sector. Forexample, in the telecommunication sector price decreasesare often only available for new customers but not forloyal customers.

However, previous research does not provide a link-age between perceived price fairness and personal pricediscrimination between new customers and loyal custom-ers. Thus, the objectives of our research are to examine (1)how favoring new customers over loyal customers interms of personal price discrimination affects loyal cus-tomers’ perceived price fairness and (2) which moderat-ing variables influence this relationship. The investiga-tion is based on a research model by Xia, Monroe, and Cox(2004), which has not yet been tested empirically.

On the basis of Equity-Theory, we hypothesize thatfavoring new customers over loyal customers will lead tolower loyal customers’ perceived price fairness. In addi-tion, we expect that similarity of transactions, perceivedmotive fairness of a company’s prices, loyal customers’trust in the company and consideration of social normswill moderate this relationship.

The telecommunication sector was identified for theempirical investigation due to the fact that new customersin this sector are often favored in price over loyal custom-ers in terms of personal price discrimination. After apreliminary study, we conducted four experiments with a2x2 mixed within- and between-subject design using ascenario approach which is often applied in price fairnessresearch (e.g., Haws and Bearden 2006; Homburg, Hoyer,and Koschate 2005). Cell size was 25 for each experimen-tal condition. We collected the data as part of an onlinesurvey. Participants were all real customers of telecom-munication products, which highly adds to thegeneralizability of the results (e.g., Vaidyanathan andAggarwal 2003). The dependent variable loyal custom-ers’ perceived price fairness was measured using threeitems adapted from Bolton and Alba (2006), Bolton,Warlop, Alba (2003), Darke and Dahl (2003) andVaidyanathan and Aggarwal (2003). All items were mea-sured on seven-point Likert scales ranging from com-pletely agree (1) to completely disagree (7). All data wereanalyzed employing a mixed factor ANOVA with loyalcustomers’ perceived price fairness as dependent variable.

The results of our study report a negative relationshipbetween favoring new customers over loyal customers interms of personal price discrimination and loyal custom-ers’ perceived price fairness. This relationship is shown tobe influenced by several moderating variables which aresimilarity of transactions, perceived motive fairness of acompany’s prices, loyal customers’ trust in the companyand consideration of social norms. Overall, our resultshighlight the importance of behavioral aspects in pricing.If companies find that they have to discriminate pricesbetween new customers and loyal customers, loyal cus-tomers’ perceived price fairness should be considered. Inaddition, the moderating variables identified in this studyprovide starting points for implementing different pricesfor new customers and loyal customers.


For further information contact:Martin Fassnacht

Marketing and CommerceWHU – Otto Beisheim School of Management

Burgplatz 2, 56179 VallendarGermany

Phone: +49.261.6509.441Fax: +49.261.6509.449



AGAINST THE GREEN: AN EXAMINATION OFNON-GREEN CONSUMERS

J. Joseph Cronin, Jr., Florida State University, TallahasseeMark R. Gleim, Florida State University, Tallahassee

Stephanie J. Lawson, Florida State University, Tallahassee

SUMMARY

This paper utilizes a qualitative study to investigatenon-green consumers. The marketing literature focusessolely on green consumers, not yet addressing the barriersthat lead to non-green behaviors. As firms take note of theattention paid to the environment, the opportunity tocapitalize has never been greater. The number of greenproducts introduced has nearly doubled each of the pastthree years (Makower 2009). However, there are stillapproximately 160 million U.S. consumers that are reluc-tant to purchase green products (Stengel 2009). The goalof this paper is to identify the barriers that impede theadoption of green products. To accomplish this goal, theresearch presents a qualitative study of non-green con-sumers to better identify attitudes and perceptions thathinder consumers’ consumption of green products. Theemerging themes suggest that distrust in “green” firms,lack of familiarity with green products, the influence ofsocial and personal norms, and economic barriers impedeconsumers’ adoption of sustainable goods and services.

Introduction

As consumers’ become increasingly aware of theissues surrounding climate change, they are placing greateremphasis on environmentally friendly1 product purchases.While past research suggests that 75 percent of Americansclaim to be environmentalists (Mackoy, Calantone, andDröge 1995), it is estimated that nearly 160 million adultconsumers in the U.S. have yet to purchase a greenproduct this year (Stengel 2009). For example, in 2008 theorganic food market made up only 3.5 percent of all thefood product sales in the U.S. (Organic Trade Association2009). While there has been an increase in the number ofgreen products made available (Makower 2009), and anincrease in green consumption, little is known in regardsto why consumers continue to make non-green purchas-ing decisions.

Discovering why so many consumers are hesitant topurchase products that are purported to be better for theenvironment is the objective of this research. The goal isto better understand why consumers’ are reluctant topurchase environmentally friendly products. Results of aqualitative study are presented in order to better understandnon-green consumers’ perceptions of green products.Non-green consumers may be defined as consumers that

make purchases with little or no concern for the physicalenvironment. These consumers choose to purchase non-green products even though environmentally friendlyalternatives are available. Semi-structured interviews wereconducted with twelve individuals. The participants rangedin age from 20 to 64, with varying degrees of socio-economic status. The results garnered from the interviewswere compared to current pro-social consumer influencetheory (Osterhus 1997). Analysis of the data yielded manyinteresting findings and greater insights regarding theattitudes and perceptions of non-green consumers.

Findings

The first issue raised by the participants was a lack oftrust regarding “green” firms. Specifically, the generalbelief was that many firms only go green for monetarybenefit. A lack of familiarity with green products wasidentified as a barrier. The participants generally notedthat they could not recall seeing advertisements for greenproducts or recall seeing them prominently displayed instores. The resulting lack of awareness thereby created abarrier to the adoption of green products.

Social norms were also identified as a barrier asparticipants noted the detrimental impact of social normson green consumption. Specifically, the actions of theparticipants’ families, and mentality of society, werebelieved to negatively impact green consumption as theparticipants noted a lack of positive social normativeinfluence regarding green consumption.

Personal norms were another factor found to impactnon-green consumption. Awareness of the need for greenproducts was considered low by participants who sug-gested that they did not believe climate change was takingplace. If climate change, or any negative environmentalimpact, was not expected, the motivation for change wasweak. When examining awareness of the consequences ofpurchasing green products, the participants were extremelyconsistent in their responses as they all had low levels ofperceived consumer effectiveness (Webster 1975). Theparticipants did not attribute blame for the problem, orresponsibility for correcting the problem, to themselves.As such, it appears that consumers do not believe they arelikely to be seriously affected by environmental degrada-tion issues.


Lastly, the economic investment (time and effort)needed to find and purchase green products were alsonoted as barriers to adoption. The participants suggestedthat they found themselves purchasing the same productsrepeatedly and were not willing to try anything new sincethey had already found something that worked well enough.Consumers indicated that they were comfortable withwhat they knew and were not likely to be willing to makethe effort to find new green products and determine theireffectiveness. Another economic obstacle for green prod-ucts is price. With the exception of one, all of the subjectsnoted that price is an obstacle in their adoption of greenproducts. The comparative high price of green products

thus appears to be a barrier to the adoption of greenproducts.

The research presented examines the barriers facingnon-green consumers. As non-green consumers representa majority of adult consumers in the U.S., it is critical thatacademics and practitioners better understand the barriersfacing non-green consumers. Through a better under-standing of the barriers non-green consumers’ face, firmswill be able to develop marketing strategies designed toeliminate or minimize the factors that discourage this vastsegment of consumers from purchasing green goods andservice.

ENDNOTE

1 For the purposes of this paper, the terms: green, eco-friendly, environmentally friendly, and sustainableare used interchangeably.

REFERENCES

Mackoy, R.D., R. Calantone, and C. Dröge (1995),“Environmental Marketing: Bridging the DivideBetween the Consumption Culture and Environmen-talism,” in Environmental Marketing: Strategies,Practice, Theory, and Research, Michael Jay Polonskyand Alma T. Mintu-Winsatt, eds. New York: Haworth

Press, 37–54.Makower, J. (2009), “Taking Care of Business,”

[www.greenbiz.com], (August).Organic Trade Association (2009), “Market Information

and Industry Research,” [www.ota.com].Osterhus, T.L. (1997), “Pro-Social Consumer Influence

Strategies: When and How Do TheyWork?” Journalof Marketing, 61 (October), 16–29.

Stengel, R. (2009), “Doing Well by Doing Good,” TIMEMagazine, (September 21).

Webster, F.E. (1975), “Determining the Characteristics ofthe Socially Conscious Consumer,” Journal of Con-sumer Research, 2 (December), 188–96.

For further information contact:Mark R. Gleim

Department of MarketingFlorida State UniversityRovetta Business Annex

821 Academic WayP.O. Box 3061110

Tallahassee, FL 32306E-Mail: [email protected]


DOES GREEN REALLY HELP BEING GREEN? THE ROLE OF COLORIN PROCESSING FLUENCY

Joon Yong Seo, University of Utah, Salt Lake CityDebra L. Scammon, University of Utah, Salt Lake City

SUMMARY

Although color is an essential element of marketingcommunications, we do not know much about the role ofcolor in marketing (Garber and Hyatt 2003). For example,can color as part of an information presentation enhancemessage persuasiveness? Or does color have only aes-thetic value with little impact on message persuasiveness?If there is an effect, how can that be explained? Thecurrent literature does not provide clear answers to thesequestions. We address these issues by linking processingfluency theory to the color literature. Research on pro-cessing fluency suggests that consumer judgments areoften influenced by the ease or difficulty with which newinformation can be processed (Schwarz 2004). Buildingon this research, we propose that one potential source ofprocessing fluency is color. There is very little researchthat explores whether and how color can influence thefluency with which people process target information. Wepropose that background color that matches the content oftarget information makes the information easier to pro-cess conceptually, thereby increasing the appeal of thetarget information. We investigated this in the context ofpro-environmental messages, specifically testing the roleof green color in pro-environmental messages and itsimpact upon evaluations of the target information.

Experiment 1

We examined the effect of different backgroundcolors (green and yellow) on attitudes toward householdrecycling tips. We hypothesized that a green backgroundwould lead to more favorable attitudes toward the recy-cling tips, presumably because the conceptual relatednessbetween the color green and recycling will facilitateprocessing of the information. Participants read house-hold recycling tips on either a green or yellow back-ground. They then evaluated the information on multipleitems. Participants also reported their liking of the back-ground color, current mood, and how easy it was to readthe article. As predicted, evaluations of the recycling tipswere more positive in the green (vs. yellow) condition.There were no differences between the two conditions incolor preference, mood, or ease of reading. These resultssuggest that color can enhance ease of processing, whichin turn leads to more favorable attitudes toward the targetinformation.

Experiment 2

In experiment 1, the green background led to morepositive evaluations of the recycling tips. To show thatthis effect is driven by the conceptual fit between thetarget information and the background color, in experi-ment 2 we tested whether this effect is limited to theevaluations of target information that is conceptuallyrelated to the color green. Further, we provide more directevidence of conceptual fluency. We hypothesized thatparticipants would experience more conceptual fluencywhen there is conceptual fit (vs. non-fit) between targetinformation and a background color. A 2 (green vs. red)X 2 (environmental vs. non-environmental message)between-subjects design was used. Participants first readenergy saving tips on either a green or red background.For some participants the goal of energy saving wasprotecting the environment while for others it was savingmoney. Protecting the environment is assumed to be moreconceptually related to the color green than is savingmoney. To measure conceptual fluency, participantsreported how much effort they spent to understand thearticle, how clear the meaning of the article was, and howquickly they got the idea from the article. Participantsexperienced more conceptual fluency with the environ-mental message in the green (vs. red) condition, but notwith the non-environmental message. The 2-way interac-tion was significant. Consistent with these results, partici-pants’ willingness to follow the energy saving tips washigher with the green (vs. red) background in the environ-mental message condition, but not in the non-environmental message condition. These results provideadditional evidence that conceptual fluency can enhancemessage persuasiveness, ruling out other explanationssuch as the influence of color on mood or perceptualfluency.

Experiment 3

Color research suggests that longer wavelength col-ors (e.g., red, yellow) are arousing and evoke a preventionfocus, whereas shorter wavelength colors (e.g., green,blue) are calming and induce a promotion focus (Stoneand English 1997; Mehta and Zhu 2009). These findingssuggest the possibility that the results from experiments 1and 2 were driven by regulatory fit, rather than conceptualfluency. To test this possibility, we used the revised NewEnvironmental Paradigm (NEP) Scale (Dunlap et al.2000). This scale measures attitudes toward the environ-ment and consists of two dimensions that map onto


different regulatory foci: promotion- and prevention-focus. Participants completed items representing bothdimensions on either a green or yellow background,resulting in a 2 (green vs. yellow: between-subjects) X 2(prevention- and promotion-focused items: within-sub-ject) design. If regulatory fit drives participants’ responses,scores on promotion-focused (prevention-focused) itemswill be higher in the green (yellow) condition, becausethere is better regulatory fit between promotion-focus(prevention-focus) and the color green (yellow). Accord-ing to the conceptual fluency account, however, scoreswill be higher in the green (vs. yellow) condition regard-less of regulatory fit. Supporting the conceptual fluencyaccount, participants’ scores on both dimensions werehigher in the green (vs. yellow) condition. There was nointeraction effect, ruling out the regulatory fit account.

Discussion

In summary, we proposed and demonstrated that theconceptual fit between target information and background

color can make target information easier to process,resulting in higher message persuasiveness. Our researchadds to the color and processing fluency literature inimportant ways. First, we suggest a previously unex-plored route by which conceptual fluency confers favor-able attitudes to the target information, namely, the con-ceptual fit between the target information and backgroundcolor. Second, prior research suggests that visual featuresof information can enhance perceptual fluency (e.g., easeof reading). Our research is the first to demonstrate thatcolor as a physical feature can also enhance conceptualfluency (e.g., ease of understanding). The finding thatconsumer perceptions of pro-environmental tips are sig-nificantly influenced by the color green has importantimplications for public policy and social marketing. Impor-tantly, our findings suggest that the role of color inmarketing is not limited to aesthetic values. Rather, theyshow that a carefully chosen color can enhance messagepersuasiveness. References are available upon request.

For further information contact:Joon Yong Seo

University of Utah1645 East Campus Center Drive

Salt Lake City, UT 84112Phone: 801.587.5047



THE PSYCHOLOGICAL IMPACT OF NUTRITION INFORMATION:THE ROLE OF GUIDELINE DAILY AMOUNTS

Louise May Hassan, Heriot-Watt University, United KingdomEdward Man Kee Shiu, University of Strathclyde, United Kingdom

Nina Michaelidou, University of Birmingham, United Kingdom

SUMMARY

In recent years there has been increasing public andgovernmental debate about the harmful effects of obesityand what causes individuals to become overweight andobese. Figures from the World Health Organization (WHO)show that there are currently more than one billion over-weight adults globally, of whom 300 million are classedas obese (WHO 2004). Obesity rates in many areas of theworld (e.g., North America, Europe, the Middle East,Australasia, and China) have risen three-fold since 1980(WHO 2004). To make healthier food choices consumersmust have the appropriate information to decide whatfood is healthier for them. To this end, nutrition labelinghas been made mandatory on packaged foods in manycountries (e.g., Nutrition Labeling and Education Act[NLEA] 1990 in the U.S.). More recently, as governmentsput increasing pressure on manufacturers and retailers intheir efforts to tackle obesity, Guideline Daily Amounts(GDA) information is beginning to appear on the front-of-packs in addition to back-of-pack nutrition information.The GDA format typically shows the amount and percent-age of calories, fat, saturated fat, sugar, and salt as apercentage of the guideline daily amount for adults or forchildren depending on the product type. Limited researchhas explored the effects of the provision of GDA informa-tion particularly in a restaurant or café setting. Feunekeset al. (2008) tested the effectiveness of a range of front-of-pack labeling formats and conclude that simpler formatstake less time and effort to evaluate and thus are better fordecision making within a shopping environment. Theyfurther find that the GDA format takes longer to processand is harder to comprehend than other simpler visualformats. This stream of research is of increasing impor-tance given the trend in legislation regarding restaurantmenu labeling. Research in consumer psychology hasfound that temptation, conflict, and self-control are keyfactors influencing consumer choice (e.g., Dholakia et al.2006; Luomala, Laaksonen, and Leipamaa 2004). But thequestion remains, in what way does GDA informationmoderate the impact of these psychological factors (temp-tation, conflict, and self-control) on choice? This researchexamines the effect of the provision of varying levels ofGDA information on consumers’ propensity to choose to

eat a cake in a café scenario based on an experimentalapproach.

A between-subject experiment was conducted at theBritish Broadcasting Corporation (BBC) Good Food showin Birmingham, U.K. during November/December 2007.Subjects were randomly assigned into three treatmentgroups, namely control group with no GDA informationpresent, moderate GDA group and high GDA (higherpercentages of calories and sugar) group. Respondentswere asked to choose whether or not they would have achocolate cake in a café when taking a break from shop-ping. A monochrome GDA format was presented under acolor photograph of a chocolate cake. The same photo-graph of the chocolate cake was used across all subjects.The moderate GDA values reflect real GDA informationon standard cakes widely available in the marketplacesuch as coffee stores. The higher GDA figures are alsobased on cakes from similar coffee stores but represent the“least healthy” options. Of the 299 female respondentswho took part in the experiment 97 (32%) of the partici-pants are in the control condition, 99 (33%) participantsare in the moderate GDA condition and the rest (103 or35%) are in the high GDA condition. Manipulation checks(t tests) show that participants in the high GDA conditionthought the cake to have higher calories (Means = 5.19and 4.42; t = 4.10, p < .001) and more sugar (Means = 5.31and 4.71; t = 3.46, p < .001) than those in the moderateGDA condition and found effects in the expected direc-tion. Generalized Linear Modeling (GZLM) was used totest the main research model.

The GZLM results confirm a direct effect of GDA onconsumer choice. In particular, specification of high lev-els of calories and sugar contents has a direct effect inreducing the proportion of subjects choosing to eat thecake. In line with previous research, this study alsoconfirmed the direct effects of temptation (e.g., Baumeisteret al. 1994), conflict (e.g., Luomala et al. 2004) and self-control (e.g., Dholakia et al. 2006) on choice. Moreimportantly, this study has established the relevance of theGDA information in terms of its moderating effect. Ourresults show that the presence of high GDA values has asignificant influence in weakening the relationship between


conflict and choice, as well as strengthening the impact ofself-control on choice. This would suggest that high levelsof perceived “unhealthiness” leads the consumer to choosehealth rather than indulgence and therefore reject the

cake. Thus timely presentation of objective nutritioninformation can provide an actionable target to resolve theconflict and activate self-control in favor of a healthierchoice.

REFERENCES

Baumeister, Roy F., Todd F. Heatherton, and Dianne M.Tice (1994), Losing Control: How and Why PeopleFail at Self-Regulation. San Diego, CA: Academic.

Dholakia, Utpal M., Mahesh Gopinath, Richard P. Bagozzi,and Rajan Nataraajan (2006), “The Role of Regula-tory Focus in the Experience and Self-Control ofDesire for Temptations,” Journal of Consumer Psy-chology, 16 (2), 163–75.

Feunekes, Gerda I.J., IIse A. Gortemaker, Astrid A.Willems, Rene Lion, and Marcelle van den Kommer(2008), “Front-of-Pack Nutrition Labeling: TestingEffectiveness of Different Labeling Formats Front-

of-Pack in Four European Countries,” Appetite, 50(1), 57–70.

Luomala, Harri T., Pirjo Laaksonen, and Hanna Leipämaa(2004), “How Do Consumers Solve Value Conflictsin Food Choices? An Empirical Description andPoints for Theory-Building, in Advances in Con-sumer Research, Barbara E. Kahn and Mary F. Luce,eds. Provo, UT: Association for Consumer Research,564–70

National Labeling and Education Act (1990), Public Law101–535, 104 Stat. 2355.

WHO (2004), “Global Strategy on Diet, Physical Activityand Health,” in Fifty-Seventh World Health Assem-bly, WHA57.17.

For further information contact:Louise M. Hassan

Heriot-Watt UniversityEdinburgh

Scotland, United KingdomE-Mail: [email protected]


LINKING SERVICE FAILURE TYPES TO COGNITIVE APPRAISAL

Jiraporn Surachartkumtonkun, University of New South Wales, AustraliaPaul G. Patterson, University of New South Wales, Australia

Janet R. McColl-Kennedy, University of Queensland, Australia

SUMMARY

While zero defects is a noble goal, services are oftenmanufactured and delivered in real time, leaving no timefor quality inspection prior to delivery to the customer.Hence service failure is inevitable from time to time.Based on equity and social exchange theories, a servicefailure situation represents a disproportionate exchangeof resources, where customers experience a loss ofresources such as money, time, self-esteem, etc. On theother hand, service recovery is viewed as the company’sattempt to provide a gain to customers in order to make upfor the loss experienced in the service failure. To developan effective service recovery strategy that returns a cus-tomer to a state of satisfaction, the company or servicepersonnel must provide customers with a resource that isequal to or exceeds what they have lost in the servicefailure. For example, a customer who had their self-esteem damaged in a service failure would prefer toreceive an explanation and a genuine apology from theorganization.

To explore the possible resources “lost” in a servicefailure situation, we employ cognitive appraisal theory.The theory proposed by Lazarus and Folkman (1984)shows that at the cognitive appraisal stage a personevaluates whether a particular (stressful) encounter withthe environment (e.g., service encounter) is relevant totheir well-being, and if so, if it is likely to create a threat,or harm/loss to them. The resources which are at stake arebroadly categorized as psychological (e.g., self-esteem,sense of control) and personal (tangible) resources (e.g.,money, time, physical harm). In other words, we examinewhat is at stake in a service failure from the perspective ofcore human values (need for self-esteem, justice, eco-nomic resources, and control) rather than focusing onexpectation surrounding product/service attributes. Thefundamental purpose of our study is to examine whetherparticular categories of service failure situations (e.g.,inappropriate behavior or unresponsive staff, core servicefailure, etc.) are associated with different cognitiveappraisal (e.g., threats to self-esteem, fairness, etc.). Sec-ond, we then examine if these associations are consistentacross East-West culture.

Using critical incident technique to examine cogni-tive appraisal process, our findings shed new light on theappraisal processes of customers in service failure cir-cumstances. Five dimensions of cognitive appraisal

emerged from the data (435 respondents from Thailandand the U.S.): (1) Threats to physical well-being refer apotential harm to the person safety (or loved one): (2)Threats to self-esteem refer to a potential damage to aperson’s view of his or her self worth: (3) Threats toeconomic resources involve a potential loss of money andtime in which people put into each transaction: (4)A threatto fairness needs is a feeling of a sense of injustice, or ofbeing cheated or treated unfairly: (5)Threats to controlrefer to a potential loss of the person’s ability to achievea goal, deal with a problem, or control a situation. Theservice failure situations which lead to different types ofappraisal were classified into (1) core service failure, e.g.,billing error, unsafe practice, (2) employee unresponsivebehavior, e.g., uncaring, unresponsive, (3) speed of ser-vice, e.g., waiting time, (4) employee inappropriate e.g.,being rude, impolite, and (5) ethical problems, e.g., cheat-ing, hard sell.

The results from logistic regression showed whatresources are at risk (cognitive appraisal) in differentservice failure situations. Customers’ self-esteem is likelyto be damaged in a circumstance that involves employeeinappropriate behavior. Economic resources (money andtime) are under threat in service failure situations causedby core service failure, or speed of service (slow). Fair-ness or need for justice is invoked when customers expe-rience employees’ unresponsive behavior and unethicalcircumstances. Threats to control needs did not show asignificance occurrence in any particular situation.

Eastern and Western customers seem to have differ-ent cognitive appraisals in some given situations. Westerncustomers put more importance on economic resources(money and time) than Eastern customers in most servicefailure circumstances (e.g., core service failure, inappro-priate behavior, and unethical behavior). Physical well-being seems to be a greater concern among Thais thanAmericans who experienced core service failure. Thismay be explained by the situational context where theincidents occurred. Most of the situations reported byThais involve public transportation and public healthcare, which are known for their low quality services, andas a result Thai customers felt unsecure using theseservices.

It is important to recognize that customers possess aset of core human needs and values. Violate these needsand gross dissatisfaction is the result. Meet or enhance


these needs and the levels of customer satisfaction willemerge. It is a case of organization using a different lens

and viewing customers as people first and customerssecond. References are available upon request.

For further information contact:Jiraporn Surachartkumtonkun

University of New South WalesSydney, NSWAustralia 2052

Phone: +61(2)9385.3615Fax: +61(2)9663.1985



SERVICE RECOVERY PARADOX: MYTH OR REALITY?

Christian Brock, University of Strathclyde, United KingdomMarkus Blut, University of Dortmund, Germany

Heiner Evanschitzky, University of Strathclyde, United KingdomClay Voorhees, Michigan State University, East Lansing

SUMMARY

The consequences of service failure and the impact ofservice recovery on customers’ post-failure attitudes havegrown into an important research stream in the marketingliterature (Andreassen 1999; Blodget and Hill 1997;Davidow 2003; De Matos et al. 2007; de Witt and Brady2003; Homburg and Fürst 2005; Maxham and Netemeyer2002; McCollough et al. 2002; Richins 1983; Tax andBrown 1998; Tax et al. 1998). Research generally agreesthat a satisfactory recovery provided by the company can“turn angry and frustrated customers into loyal ones”(Hart, Heskett, and Sasser 1990, p. 148). However, thereis less agreement when considering claims that success-fully recovering can “create more goodwill than if thingshad gone smoothly in the first place” (Hart, Heskett, andSasser 1990, p. 148). This second claim is the essence ofthe service recovery paradox (SRP) and to date, empiricalresearch has provided conflicting evidence in support ofand against this paradoxical viewpoint. Due to the confu-sion throughout the marketing literature a simple, butcritical question remains unanswered:

“Do customers that experience satisfactory recoveryefforts exhibit even more favorable attitudes andbehavior than customers who do not experience afailure in the first place?”

The goal of this research is build upon currentunderstanding of the SRP by investigating the SRP usinga field study and leveraging data from a series of self-report measures and actual purchase behavior. We addresssuppositions in the literature by testing the followinghypotheses:

H1: There is a significant, positive SRP effect for (a)satisfaction (SAT), (b) repurchase intentions (RPI),and (c) word-of-mouth (WOM).

H2: There is a significant positive SRP effect for (a)cognitive (CL), (b) affective (AFL), (c) conative(CNL), and (d) action loyalty (AL).

H3: The effects associated with the SRP (higher loyaltylevels ((a) cognitive (CL), (b) affective (AFL), (c)conative (CNL), and (d) action loyalty (AL)) forconsumers that experience a successful recovery

relative to consumers that experience no servicefailure) will be stronger when past customerexperience with the provider is favorable rather thanunfavorable.

Results

Based on analysis of variance we do not find supportfor H1. The level of satisfaction (H1a) is higher forcustomers without a service failure, our control group(CG), than for satisfied complainers (SC). The tests forH1b, also do not provide support for the hypothesis andare in line with findings from De Matos et al.’s (2007)meta-analysis. Similarly, the word-of-mouth intention(H1c) for customers without a service failure is alsohigher than for satisfied complainers. Ultimately, theresults of H1b and H1c are consistent with the findings ofDe Matos et al. (2007), but our findings provide evidencethat the SRP does not exist and that the opposite effects aretrue for repurchase intentions and word-of-mouth. Withrespect to satisfaction, our findings also suggest the oppo-site effects, which directly counter the findings of DeMatos et al. (2007).

Regarding hypothesis H2, our results indicate noempirical support. More interestingly, we also assess thepost-recovery behavior (“action loyalty”) as measured bythe average time-span between two consecutive pur-chases before and after the recovery attempt using objec-tive purchase data (H2d). Results indicate that satisfiedcomplainers exhibit a higher purchase frequency beforethe service failure than the control group with no failure.However, after the failure, successfully recovered cus-tomers (satisfied complainers) do not change their post-recovery behavior significantly. In fact, they even slightly(yet insignificantly) reduce their repurchase frequency.These results largely disconfirm the existence of the SRPfor actual behavior. Hence, we argue that the SRP shouldbe more likely to occur if a customer has had favorableprior experiences.

The extension of H2 by integrating the moderatingimpact of past experience leads to some interesting find-ings. We analyzed data using a 2 (Prior Experience:Favorable (FPE) and Unfavorable past experience (UPE))x 2 (Type of Customer: Satisfied complainer (SC) and NoService Failure (CG)) design that included the four types


of loyalty as dependent variables. Results for cognitiveloyalty show a significant interaction effect between “pastexperience” and “type of customer.” An inspection of themeans for cognitive loyalty, however, reveal an effect thatis opposite of that proposed in H3a. More specifically, theantithetical effects to the SRP are stronger and significantunder conditions of favorable prior experience. Underunfavorable conditions, there are no significant differ-ences across the customer groups. Thus, when consumershave favorable prior experience with the firm, the effectsthat counter the SRP are more likely to occur than whenconsumers have relatively unfavorable prior experiencewith the firm. For both the affective loyalty and conativeloyalty stages, the interactions were not significant, fail-ing to support H3b and H3c. Findings based on thebehavior (action loyalty) after the recovery provide nosupport for an SRP: There is no significant interactioneffects between the group of satisfied complainers and thecontrol group could be detected. Therefore, we find no

empirical proof for H3d. However, we also analyze thepurchase behavior before the service failure to take con-trol of behavior changes. The group of satisfied com-plainer with a positive prior experience reduces their levelof repurchase behavior, but not significant.

Concluding Remarks

We can best sum up our results, by answering aquestion posed by Michel and Meuter (2008) on the SRP:“Is it true but overrated?” Our results suggest that it is notonly untrue, but its antithetical effects exist where strate-gies based on the SRP may provide worse outcomes thanstrategies that simply focus on getting service right thefirst time. In conclusion, it appears that successful recov-ery fails in comparison to simply providing good servicethe first time and the SRP is both untrue and overrated.References are available upon request.

For further information contact:Christian Brock

Department of MarketingUniversity of Strathclyde

173 Cathedral StreetGlasgow G4 0RQUnited Kingdom

Phone: +49.173.7040049E-Mail: [email protected]


SETTING IN VERSUS EXPECTING MORE: MODERATING EFFECTS OFRELATIONSHIP LENGTH AND BRAND EQUITY ON CONSUMER

JUSTICE PERCEPTIONS

Jeremy S. Wolter, Florida State University, Tallahassee

SUMMARY

The importance of justice evaluations within thecontext of service failures has been firmly established(Blodgett, Hill, and Tax 1997). Simply put, people desirejustice when they perceive they have been wronged. Acustomer’s justice evaluation of a firm’s behavior affectsthe customer’s satisfaction with the firm and their loyaltyto the firm. However, the exact type of justice a customermay prefer has not had the same consensus. Both distribu-tive justice and interactional justice have interchangedtheir leading roles throughout the literature. Thus, theremust be unmeasured moderators that are influencing theprevious findings.

In this research, the effects of moderators on therelationship between justice perceptions and recoverysatisfaction as well as firm satisfaction were evaluated. Afirm’s brand equity and a customer’s relationship lengthwere postulated as affecting the justice dimensions asexplicated by social exchange theory and rewards expec-tations states theory. More specifically, it was expectedthat a habituation effect would attenuate the relationshipof interactional justice with satisfaction with the recoveryas well as overall firm satisfaction for customers who havea long-standing relationship with a firm. It was alsoexpected that a firm’s brand equity would trigger a

customer’s reward expectation. For firm’s with high brandequity, the triggered reward expectation states wouldstrengthen the effect of distributive justice while attenuat-ing the effect of interactional justice. The opposite effectwas expected for firm’s with low brand equity.

A form of critical incident technique was utilized togather data from 583 respondents and between-groupanalyses in structural equation modeling was used forhypothesis testing. A conceptual model of a service recov-ery evaluation based on Maxham and Netemeyer (2002)was used as the framework in which to test the moderatingeffects. In addition to the primary analysis, commonmethod variance and a confound check of the effect ofpricing was also assessed.

The results confirmed the strengthening effect ofbrand equity on the relationship between distributivejustice and satisfaction with the recovery. The attenuatingeffect of relationship length on interactional justice wasnot supported by the data. Thus, the results suggest thatfirms with low brand equity may be able to recover fromservice failures by simply emphasizing the polite treatmentof customers. In contrast, it would seem that firms withhigh brand equity are expected to provide bothcompensation and polite treatment during a servicerecovery. References are available upon request.

For further information contact:Jeremy S. Wolter

Department of MarketingCollege of Business

Florida State University821 Academic Way

Tallahassee, FL 32306–1110Phone: 850.644.4091



CONCEPT PROVIDING IN THE WHOLESALING INDUSTRY: DOSTRUCTURAL CUSTOMER BONDS HELP OR DO THEY HARMCOMPETITIVE ADVANTAGE AND RELATIONSHIP QUALITY?

Sandra Pocsay, Saarland University, GermanyHanna Schramm-Klein, University of Siegen, Germany

Joachim Zentes, Saarland University, GermanyNikolaus Hohl, University of Siegen, GermanyJulia Naskrent, University of Siegen, Germany

SUMMARY

Studies have most often analyzed business relation-ships from a manufacturer’s perspective. If wholesalers’downstream relationships have been considered at all, thefocus has been on the possibilities of streamlining manu-facturers’ production processes through integrated sup-ply strategies. The perspective of wholesalers themselveshas largely been neglected. The aim of this study thereforeis to analyze the nature of the relationship between whole-salers and their customers. More specifically, the effectsof structural customer bonds on relationship quality arebeing studied as well as the role of the competitiveadvantage that wholesalers’ customers gain due to therelationship. The effects of various bonds on the competi-tive advantage of the customers on the one hand as well ason the stability of the relationships implemented by thewholesaler on the other hand are analyzed (see Figure 1for the conceptual model).

In this research, structural customer bonds are re-garded as consisting of the dimensions “relationship-specific investments,” “operational linkages,” “contrac-tual bonds,” and “information exchange.” Relationship

quality has been proven to have a significant influence onperformance outcomes of the buyer as well as of the sellerin a dyad and therefore is chosen as the dependent variablein this study. It is defined as an overall assessment of thestrength of a relationship and conceptualized as a multidi-mensional construct capturing the different but relatedfacets of a relationship. We understand relationship qual-ity as a composite measure of commitment and trust.

For our study, we chose the automotive aftermarketas an industry in which wholesalers have developed abroad spectrum of different structural bonding measuresand frequently offer customers the possibility to benefitfrom these. The relationships were evaluated from abuyer’s perspective, since in a buyer’s market the focus onthe buyer’s point-of-view may explain more variance intypes of relationships. We therefore addressed repair shopowners belonging to one of four wholesaler-managedcooperative systems in Germany and by this included thecustomer relationships of over 150 different wholesalecompanies.

A total of 202 usable questionnaires were availablefor the investigation. To test our hypotheses, we con-

FIGURE 1Conceptual Model

Structural Bonds

Relationship QualityContractual

Bonds

InformationExchange

OperationalLinkages

Relationship-SpecificInvestments

CompetitiveAdvantage

Commitment

TrustH(1)

H(2) H(3a)


ducted PLS regression. We proposed a positive effect ofstructural customer bonds on the quality of the whole-saler-customer-relationship as well as on the competitiveadvantage that customers gain through the relationship.Our results show highly significant positive effects bothon relationship quality and on competitive advantage.Additionally, we could confirm a positive effect of com-petitive advantage on relationship quality.

Competitive advantage was identified as a (partial)mediator of the relation between structural bonds andrelationship quality. In order to gain more insights into thenature and effects of structural bonds in wholesaler-customer-relationships, in an explorative way, we furtheranalyzed the four distinct dimensions of structural bondsincluded in the research. Only information exchange andrelationship-specific investments had a significant directeffect on relationship quality. All four dimensions,however, showed a significant effect on competitiveadvantage.

These results show that the quality of a wholesaler-customer relationship mainly depends on relationship-specific investments and information exchange with infor-

mation exchange yielding a significantly higher influencethan relationship-specific investments. This shows thatalthough wholesalers have started binding their custom-ers by requesting specific investments in the relationship,the stronger bonding effect still stems from the wholesal-ers’ market knowledge and the information they can passon to their customers.

Our study has shown that structural bonds betweenwholesalers and customers are important in order to buildcompetitive advantage of the customers as well as toincrease the quality of the relationship. Wholesalers canbe assured that it really is worth the effort to invest scarceresources into the development of structural bondingmeasures. This study is among the first to empiricallydifferentiate between different dimensions of structuralbonds and their respective impact focusing on the net ofcustomers from a wholesaler’s point-of-view. While alldimensions of structural bonds have positive effects oncompetitive advantage, only relationship-specific invest-ments and information exchange benefit relationship qual-ity. These findings imply consequences for managerialpractice as well as for future research. References areavailable upon request.

For further information contact:Hanna Schramm-KleinUniversity of SiegenHoelderlinstrasse 3

57068 SiegenGermany

Phone +49.271.740.4281Fax +49.271.740.2724



PRODUCTION OF TRUST IN BUYER-SELLER RELATIONSHIPS: THECONTINGENT ROLES OF PRODUCT QUALITIES AND

LIFE-CYCLE STAGE

Oliver Schilke, University of California, Los Angeles/RWTH Aachen University, GermanyMalte Brettel, RWTH Aachen University, Germany

Gunnar Wiedenfels, ProSiebenSat.1 Media AG, Germany

SUMMARY

Among the success factors in relating with customers,trust has been proposed to play a unique role (Saparito,Chen, and Sapienza 2004; Zaheer, McEvily, and Perrone1998), as trust may significantly foster customers’ purchaseintentions (Doney and Cannon 1997; Stewart 2003). Whiledifferent disciplines have adopted varying perspectives instudying the phenomenon, trust in buyer-seller relationshipscan be broadly defined as the buyer’s willingness to bevulnerable to the actions of the supplier (Mayer, Davis, andSchoorman 1995; Rousseau et al. 1998).

Mayer et al. (1995) suggest that the supplier’strustworthiness is key to customer relationshipscharacterized by high trust. As trustworthiness can thus beregarded as a firm attribute important to the performanceof customer relationships and as firms differ in theirdegree of trustworthiness, it has been proposed thattrustworthiness can be a source of competitive advantage(Barney and Hansen 1994).

Despite the importance of understanding the role oftrustworthiness in customer relationships, the antecedentsto trustworthiness have been at times obscured. First, fewsystematic frameworks exist specifying a comprehensivetypology of key drivers of suppliers’ trustworthiness.When distinctions between different types of trustworthi-ness sources have been made conceptually, they have notbeen tested empirically (e.g., Mayer, Davis, and Schoorman1995; Zucker 1986), or they have not been tested empiri-cally in the same study (e.g., McKnight, Choudhury, andKacmar 2002; Stewart 2003). As noted by Ferrin andDirks (2003), empirical research on the antecedents oftrust has lagged behind the theory. Second, among thecomparatively small number of studies focusing on thedrivers of trustworthiness, most have exclusively investi-gated main effects. That is, little systematic attention hasbeen paid to the conditions that determine whether spe-cific antecedents will be particularly (in)effective inincreasing suppliers’ trustworthiness. Finally, there isvery little integrative work on trustworthiness of suppliersconsidering both antecedents and consequences simulta-neously. Resolving these issues is important to the devel-opment of theories of buyer-seller relationships and trust.

This paper explores a comprehensive set of anteced-ents to trustworthiness of suppliers. Building on Zucker’s(1986) three modes of trust production (process, charac-teristic, institutional), we identify a variety of factors andtest for their relative impact on trustworthiness as per-ceived by buying firms. Enriching the analysis, we intro-duce two moderators influencing the effectiveness ofthese antecedents. Drawing from information economics,we hypothesize differences in the effects of trust buildingmodes depending on the qualities of products sold (cre-dence vs. experience goods). Moreover, turning to thequestion whether specific modes of trust production areunique for NEVs, we consider the moderating impact ofsupplier lifecycle stage (NEVs vs. established firms). Inaddition to exploring antecedents to trustworthiness, wealso incorporate customer relationship outcomes by com-bining Zucker’s (1986) framework of trust productionmodes with Mayer et al.’s (1995) widely cited integrativemodel of organizational trust.

In order to test our research model, we conducted alarge-scale questionnaire survey. We applied three crite-ria in selecting a population for our study: First, wefocused on B2B purchasing situations, which led us tostudy the trust perceptions of corporate customers (asopposed to consumers). Second, we focused on firmsfrom two different industries: (1) public relations agen-cies (PR) and (2) manufacturers of photovoltaic modules.This selection was guided by our objective to analyzedifferences between credence and experience goods. Fol-lowing the definition by Darby and Karni (1973), PRservices can be considered typical credence goods in thatthey are “utilized either in combination with other goodsof uncertain properties to produce measurable output or ina production process in which output, at least in a subjec-tive sense, is stochastic, or where both occur” (p. 69).Photovoltaic modules, in contrast, feature the characteris-tics of experience goods, as a buyer will be able to judgetheir output and handling ”costlessly only after purchase”(Darby and Karni 1973, p. 69). Third, we decided to askthe participating buying firms to respond to our question-naire with reference to a pre-specified list of suppliers.This procedure allowed us to ensure that supplying firmswere in the same industry as the customer and that suffi-cient variance exists with regard to the life cycle stage of


the suppliers. Thus, we conducted extensive interviewswith industry experts, aimed at compiling a list of promi-nent suppliers that were clearly either an established firmor an NEV. We then contacted these suppliers and askedfor their consent. This process resulted in the selection of12 suppliers in the PR industry (five established firms andseven NEVs) and 13 suppliers in the photovoltaic mod-ules industry (six established firms and seven NEVs).

We contacted a total of 6,887 firms satisfying ourselection criteria, using purchasing decision makers askey informants who were provided with a link to ouronline questionnaire. 1,178 firms responded to our inquiry,yielding a response rate of 17.1 percent. The first questionin the survey asked the participant to rate his familiaritywith each of the suppliers from our list. The five-pointanswer scale ranged from “I don’t know this supplier.” to“We have used this supplier.” When the participantresponded with three (“I know this supplier and havedetailed knowledge about it”) or higher for at least one ofthe suppliers from the list, an algorithm selected onesupplier that was used for the rest of the survey. In total,436 firms qualified for completing the entire question-naire. Forty-four cases were dropped from further analy-sis primarily because of incomplete responses, leaving392 complete responses in the final sample.

We analyzed the data using structural equation mod-eling and found overall support for our research model. As

predicted, we found that the supplier’s trustworthiness isstrongly linked to the extent of trust in that supplier, whichsignificantly enhances the customer’s intention to pur-chase from this supplier. In addition, the majority ofZucker’s (1986) trust building modes significantly enhancethe trustworthiness of the supplier. More specifically,reputation (a process-based source of trustworthiness),social similarity and personal relationships (as character-istics-based sources of trustworthiness), and institutionalembeddedness and external references (institution-basedsources of trustworthiness) all had a positive and signifi-cant (p < .05) relationship with the supplier’s perceivedtrustworthiness.

Consistent with our moderating hypotheses, we notonly found that measures of risk reduction are morestrongly related to trustworthiness when analyzing agree-ments involving credence goods, but also that trust is astronger driver of purchase intention when experiencegoods are traded. However, no significant differenceswere detected between NEVs and established firms withregard to the trust-purchase intentions relationship, whichled us to reject the respective hypothesis. We did find,however, that NEVs realize the institution-based trustproduction mode to a lesser extent than established firms.This means that reputation can be realized even by youngfirms. References are available upon request.

For further information contact:Oliver Schilke

University of California, Los Angeles264 Haines Hall

375 Portola PlazaLos Angeles, CA 90095–1551

Phone: +1.310.825.1313Fax: +1.310.206.9838



EFFECTS OF SUPPLIER’S MARKET ORIENTATION ON CHANNELRELATIONSHIP: A CROSS-CULTURAL STUDY

Trang Phuc Tran, University of North Texas, Denton

SUMMARY

Kohli and Jaworski’s (1990) seminal study empha-sizes the importance for many manufacturers to know theexpectations and preferences not only of their end-cus-tomers, but also of their distributors in the marketingchannel. In the channel relationships, the proclivity towardcooperation and collaboration is consistent with strongercommitments and trusts between channel members. Sincemarket orientation and relationship marketing share thesame philosophical network, some studies have con-verged and integrated two concepts either directly orthrough intermediate variables such as satisfaction andbrand value (Clark 1999).

Despite the prominent role of market orientation onchannel relationships and its ensuing influences on mar-ket-oriented behaviors in dyadic relationships, there hasbeen a paucity of research attention to date. Therefore,researchers need to focus on studies related to marketorientation in the channel relationship context and shouldbe able to explain the implications of the adoption ofmarket orientation in distribution channels (Frazier 1999;Avlonitis et al. 1999). In addition, internationalizationand globalization have spawned greater competitive chal-lenges across channel structures and environments. Anadoption of market-oriented behaviors is likely to becomea viable strategy that firms think to ease channel tensionsbut defend themselves from potential environmental threatsfrom international competition. However, there is littleresearch investigating the moderating effect of cross-cultures on this relationship. The purpose of the researchis to investigate the effects of market orientation on achannel relationship between suppliers and distributorsand demonstrate the role that cultural dimensions play inthis channel relationship by developing a set of proposi-tions through an extensive review of literature in therelevant marketing field.

Market Orientation

The marketing concept has been defined over timehas resulted in the differences in definitions of marketorientation. Kohli, Jaworski, and Kumar (1993)’s studyfocused on the firm’s activities and behaviors regardingcustomer needs, competitive information, market intelli-gence, and the sharing of such knowledge. Narver andSlater (1990) suggested that market orientation consistedof three behavioral components (customer orientation,competitor orientation, and inter-functional coordination).

Deshpandé et al. (1993) focuses on customer orientation:“The set of cross-functional processes and activitiesdirected at creating and satisfying customers throughcontinuous needs-assessment.” In this study, Kohli andJaworski’s (1990) definition of market orientation isadopted and used because of its more comprehensiveconceptualization of the construct. They say “marketorientation is the organization-wide generation of marketintelligence pertaining to current and future customerneeds, dissemination of the intelligence across depart-ments, and organization-wide responsiveness to it.”

National Culture Dimensions

This research centers on two out of five dimensionsdefined by Hofstede (1980): collectivism/individualismand power distance because it assumes that these twodimensions are highly relevant to the cross-cultural com-parison study and that channel partners interact differ-ently in different cultures. Collectivism/individualism is adimension widely studied in cross-national business rela-tionships (Williams et al. 1998). A partner in a collectivistculture sets a higher priority to the group’s interest andnorms than to his or her interest while the counterpart inan individualist culture is inclined to disregard other’sconcerns and goes for his or her own interest (Hofstede1980). Power distance is another dimension. Power dis-tance is “the extent to which less powerful members oforganizations within a country accept the unequal distri-bution of power” (Hofstede 1980). In a lower powerdistance culture, a partner is more likely to be involved ina decision-making process and participate in the discus-sion table (Kale and McIntyre 1991), while the counter-part in a higher power distance culture seems to obeycommands by managers or supervisors who are in a higherrank of hierarchy.

Proposition Development

A model of channel relationship elements is designedto cover a dyadic context. It hypothesizes the likely effectsof a supplier’s market orientation on both the distributor’smarket orientation and other channel relationship factors,including trust, and performance, and the effects of cul-tural dimensions on this relationship. The propositions are(1) the supplier’s market orientation is positively corre-lated with the distributor’s market orientation, (2) thesupplier’s market orientation is positively correlated withthe distributor’s trust in the supplier, (3) the distributor’smarket orientation is positively correlated with its trust in


the supplier, (4a) the positive relationship betweensupplier’s market orientation and distributor’s trust willbe stronger in highly collectivistic culture than in a highlyindividualistic culture, (4b) the positive relationshipbetween supplier’s market orientation and distributor’smarket orientation will be stronger in highly collectivisticculture than in a highly individualistic culture, (5a) thepositive relationship between supplier’s market orienta-tion and distributor’s trust will be stronger in a high-powerdistance culture than in a low power distance culture, (5b)the positive relationship between supplier’s market orien-tation and distributor’s market orientation will be strongerin a high-power distance culture than in a low powerdistance culture, (6) the distributor’s trust in its supplier ispositively correlated with its financial performance, and(7) the distributor’s market orientation is positively corre-lated with its financial performance.

Methodology

American distributor names are selected from themembership list of associations affiliated with the NationalAssociation of Wholesalers. To collect data foreign com-panies in the U.S., Chamber of Commerce of South Koreais contacted. Survey questionnaires are divided into twoparts, one is business-oriented and the other one is demo-graphic-oriented. In this study, multi-item scales fromprevious studies are used to measure all constructs dis-cussed in this research. The 20 items validated by Kohli,Jaworski, and Kumar (1993) formed the basis for marketorientation measure. Ganesan (1994)’s and Naman andSlevin (1993)’s instruments are used to measure trust andperformance, respectively. References are available uponrequest.

For further information contact:Trang Phuc Tran

University of North Texas1155 Union Circle #311160

Denton, TX 76203–5017Phone: 940.594.9568



HOW HUMOR IN ADVERTISING WORKS: A TEST OFALTERNATIVE MODELS

Martin Eisend, European University Viadrina, Germany

SUMMARY

The literature provides different explanations on howhumor in advertising works which can be broadly catego-rized into cognitive and affective models. The presentstudy uses meta-analytic data in order to test these modelsas well as an integrative affective-cognitive model thatexplains additional mechanisms of how humor in adver-tising works.

Theoretical Background

There are three ways by which humor affects cogni-tive responses that, in turn, impact A

AD and A

BR

1. Humor enhances cognitions in general, whereby posi-tive cognitions outweigh negative ones (informationprocessing approach),

2. Humor reduces negative cognitions (distractioneffect),

3. Humor impacts ad related cognitions but not neces-sarily brand related cognitions (vampire effect).

As for affective processes, humor enhances positiveaffect and suppresses negative affect; both affective reac-tions impact A

AD and A

BR.

As for the integrative model, humor leads to positiveaffect that leads consumers to process congruent informa-tion by enhancing positive cognitions. The same mecha-nism should apply to the suppression of negative affectthat increases congruent (i.e., negative) information.

The relationship between ad cognitions, brand cogni-tions, A

AD, and A

BR follows the dual mediation hypothesis:

Exposure to advertising impacts ad and brand cognitions.Brand cognitions impact A

BR whereas ad cognitions impact

AAD

. AAD

impacts both brand cognitions and ABR

.

Method

The models include nine different variables in total.That is, 36 off-diagonal cells have to be filled in order toproduce the input of a correlation matrix for structuralequation modeling. For this purpose, mean correlationsfrom three meta-analyses dealing with advertising effectswere used (Brown, Homer, and Inmann 1998; Brown and

Stayman 1992; Eisend 2009). At least two effect sizes foreach cell of the matrix were included. A fixed modelapproach with sample size weighted correlations wasapplied. Correction for measurement error was consid-ered in the structural equation model by using weightedaverage reliability coefficients to compute error terms.The harmonic mean of the sample size underlying thepooled correlations was used for the analysis.

Results

The fit indices indicate an acceptable fit of the mod-els. The results of the cognitive model show that humorreduces negative ad related cognitions but does not affectbrand related cognitions directly. Ad cognitions influenceA

AD, which influences brand related cognitions and A

BR.

Brand related cognitions do not influence ABR

.

The affective model supports the assumed effects:Humor enhances positive affect and decreases negativeaffect. Affect impacts A

AD which influences A

BR.

The integrative affective-cognitive model supportsfurther paths. Humor directly impacts negative ad-relatedcognitions, and both positive and negative affective states.Affect influences A

AD, and positive affect enhances posi-

tive ad cognitions. Ad related cognitions influence AAD

.A

AD impacts A

BR and brand cognitions, but brand cogni-

tions do not impact ABR

.

A mediation analysis shows that positive ad relatedcognitions do mediate the effect from positive affecttoward A

AD. As for brand related positive cognitions,

there is no mediation of the direct effect of positive affecton A

BR because the path from brand cognitions to A

BR is

not significant.

When comparing the nested affective and cognitivemodels with the full model (integrative affective-cognitive), the full model provides a better fit than each ofthe nested models.

Discussion

Humor reduces negative ad-related cognitions, whichprovides support for the distraction hypothesis that postu-lates that humor reduces counter arguing as related to thead. The results for the affective model support the mecha-nisms of an automatic and direct humor effect as sug-


gested by mechanisms of affect transfer or the idea ofspontaneous action tendencies that are linked to affectiveexperiences. While the findings of the affective and thecognitive model both hold, the integrative affective-cognitive model reveals some additional paths from affectto cognitive responses. These results are consistent with acongruency effect or an affect-as-information heuristic.However, the model does not support any effects ofnegative affect on cognitions. The result can be explainedby the fact that positive thoughts are more easily acces-sible only in the case of positive affective states, while

memory accessibility remains constant in the case ofnegative affective states.

As for practical implications, the results show thataffective reactions triggered by humor do increase posi-tive cognitions while humor directly reduces negativeones. By this, humor may help to overcome weaknesses inadvertising messages such as weak arguments or negativeinformation. Further research is needed to provide a moredetailed account on these phenomena. References areavailable upon request.

For further information contact:Martin Eisend

European University ViadrinaGroße Scharrnstraße 5915230 Frankfurt (Oder)

GermanyPhone: +49.335.5534.2870



SE HABLA ESPAÑOL: EXPLORING LATINOS’ EMOTIONALREACTIONS TOWARD POLITICAL LANGUAGE

TAILORED-ADS

Sindy Chapa, Texas State University, San MarcosEnrique P. Becerra, Texas State University, San MarcosStephen Halbrook, Texas State University, San Marcos

SUMMARY

Latinos have become the largest ethnic minority inthe United States as the group now accounts for 15 percentof the country’s population. In addition, recent changes tothe community’s demographics have increased theimportance of the Latino vote in local, state, and nationalelections. Due in part to these developments, Democratsand Republicans alike have found the need to attract whathas been termed the “Latino voting-bloc” or the “sleepinggiant.” Not surprisingly, recent presidential campaignsand their outreach programs had to heavily target theLatino vote, tailoring their messages accordingly. How-ever, there is a lack of research in the context of televisedpolitical advertising targeted to U.S. Latinos. This studyexplores the U.S. Latino population’s emotional reactiontoward “language-tailored” TV ads. Specifically, thisstudy analyzed the emotions produced by TV ads broad-cast in English, Spanish, and English-with-Spanish-sub-titled in a Latino-themed environment.

The literature points out the importance of emotionson people’s perceptions of advertising as a whole (e.g.,Schwartz 1973). Emotions are linked to advertisementattitudes and candidate image evaluations (e.g., Chapa,Hausman, and Minor 2008; Tedesco 2002). Not surpris-ingly, research shows that emotional appeals are domi-nant features of political advertisements (Kern 1989).Substantial political research supports the contention thatemotions affect the evaluation of the candidate (e.g.,Tedesco 2002).

The U.S. Latino population is by no means a mono-lithic group, however, it does share common values andbeliefs, such as common culture (e.g., Becerra andKorgaonkar 2009), which is displayed through the use ofthe Spanish language. Based on Connaughton and Jarvis’(2004) premise and Hofstede’s (1984) framework thatLatino people are more emotional and motivated bymicro-level interactions, we expect Language-tailoredpolitical ads (whether it is the Spanish-version or English-with-Spanish-subtitles ads) to induce more positive feel-ings among Latinos than English-version ads. Older gen-eration U.S. Latinos may be influenced in a lesser extentby Spanish language ads than the younger generations;thus generational levels may impact the influence pro-

vided by the language-tailored ads have on the differentgroups.

A one-ad experimental design was employed. Forthis experiment a positive political advertisement featur-ing president-elect Obama was selected. This advertise-ment was selected because it was produced in both lan-guages; Spanish and English. Unfortunately, there wereno available advertisements that met all three require-ments and therefore the third requirement, Spanish-language subtitles, was added to the English-languageadvertisement. A total of three advertisements were usedin this experiment.

The results on the effects on each emotion indicatethat there is no difference among the emotions elicited bythe three types of ads. The Spanish version ad elicits morenegative emotions than the English ad (p < .05). Similarly,the English with Spanish subtitles ad elicits more negativeemotions than the English ad (p < .05). There is nodifference between the Spanish and the English withSpanish subtitles version. The generational level was notsignificant. As a result, the hypothesis that language-tailored ads create more positive emotions on Latinos thanEnglish ads is rejected.

Wallendorf and Reilly (1983) suggest that Latinosmay adopt their internalized concept of the American wayof life and in doing so, expect a political candidate tocommunicate his/her message in English. Therefore, usingLanguage-tailored ads may seem contradictory to theirexpectations and may produce negative emotions towardthe ads. Furthermore, Koslow, Shamdasani, and Touch-stone (1994) found that Latinos believe that communicat-ing in English is better than in Spanish, which reduces theinfluence of advertisements in Spanish. This suggests thatthe negative emotions may be elicited because the lan-guage used in the ads was that other than English. Anotherexplanation for our findings is that U.S. Latinos mayprefer non-Latinos to communicate in English and mayview communications in Spanish from non-Latinos poli-ticians as disingenuous. However, these inferences needfurther exploration to be validated.

The findings suggest that Language-tailored ads fromnon-Latino English speaking politicians may elicit nega-


tive emotions, which is the opposite effect intended bypolitical ads. Therefore, non-Latino English speakingpoliticians may need to reach Latinos in English and/or

perhaps use a Latino Spanish-speaking politician in theiradvertisements to communicate and present a more trust-worthy message.

REFERENCES

Becerra, E.P. and P. Korgaonkar (2009), “Hispanics’Information Search and Patronage Intentions Online,”Journal of Electronic Commerce Research, 10 (2),76–92.

Chapa, S., A. Hausman, and M. Minor (2008), “Cognitiveand Affective Evaluations of Political Advertising inLatin America: The Effect of Message Strategy,”Latin America Association for Consumer ResearchConference, Sao Paulo Brazil, (July).

Connaughton, S.L. and Jarvis S (2004), “Invitations forPartisan Identification: Attempts to Court LatinoVoters Through Televised Latino-Oriented PoliticalAdvertisements, 1984–2000,” Journal of Communi-cation, 54 (1), 38.

Hofstede, G. (1984), “Cultural Dimensions in Manage-

ment and Planning,” Asia Pacific Journal of Man-agement, 24 (5), 81–99.

Kern, M. (1989), 30-Second Politics: Political Advertis-ing in the Eighties. New York: Praeger.

Koslow S., P.N. Shamdasani, and E.E. Touchstone (1994),“Exploring Language Effects in Ethnic Advertising:A Sociolinguistic Perspective,” Journal of ConsumerResearch, 20 (March), 575–85.

Schwartz, T. (1973), The Responsive Chord. Garden City,NY: Anchor Press/Doubleday.

Tedesco, J.C. (2002), “Televised Political AdvertisingAffects: Evaluating Responses During the 2000 Robb-Allen Senatorial Election,” Journal of Advertising,31 (1), 37–48.

Wallendorf, M. and M.D. Reilly (1983), “Ethnic Migra-tion, Assimilation, and Consumption,” Journal ofConsumer Research, 10 (3), (December), 292–302.

For further information contact:Sindy Chapa

The Center for the Study of Latino Media & MarketsThe School of Journalism and Mass Communication

Texas State University601 University Drive

San Marcos, TX 78666–5709Phone: 512.245.3412



THE CONSUMER’S PERCEPTION OF THE EMOTIONALAUTHENTICITY OF AN ADVERTISING SPOKESPERSON:

THEORETICAL DISCUSSION ANDQUALITATIVE STUDY

Katja Leschnikowski, University of Bern, SwitzerlandHarley Krohmer, University of Bern, Switzerland

SUMMARY

Against the background of negative consumerreactions to advertising (e.g., Ha and Litman 1997),consumers increasingly demand authenticity in marketingand communications (Gilmore and Pine II 2007). Inadvertising, research on a specific type of authenticity,emotional authenticity, is relatively scarce. Our paper willaddress the following research questions: What isemotional authenticity of an advertising spokesperson?How do recipients judge emotional authenticity? Howdoes emotional authenticity affect the consumer’s responseto advertising?

By drawing upon literature from psychology, emo-tional authenticity of an advertising spokesperson – froma consumers’ point of view – can be defined as the per-ceived congruence between the spokespersons’ emo-tional expressions and her inner feelings (cf., Salmela2005; similarly Harter 2002). Emotional authenticity isgiven, when the consumer believes that the spokespersonactually feels the expressed emotions in the advertise-ment. Contrarily, emotional authenticity is low when therecipient believes that the spokesperson does not actuallyfeel the expressed emotions, but emotional expressionshad been put on.

Research Design

To answer our research questions, we conducted 12in-depth interviews (students and non-students sample).We focused on the emotion of “enjoyment” since positiveemotions play a crucial role in advertising (Goldberg andGorn 1987). Real print advertisements showing spokes-persons expressing the emotion of enjoyment, and morespecifically showing a felt/unfelt smile (high/low authen-ticity) (Ekman 1999; Ekman and Friesen 1982), were usedas stimuli. A number of methods for improving the qualityof the analysis were adopted.

Results

Concerning the authenticity judgment, most recipi-ents quickly and spontaneously decided on the emotionalauthenticity of the spokesperson. By further discussing

reasons for their judgment, participants struggled withfinding explanations. To overcome the complexity, manyrecipients might have used a process with a low degree ofcognitive effort that resulted into a spontaneous judg-ment. More specifically, respondents referred to their gutfeeling, their first impression, and heuristics. Surpris-ingly, during our discussion, the respondent’s assessmenthardly ever changed, supporting the quality of the moreaffective and unconscious authenticity judgment. None-theless, the authenticity judgment also involved somecognitive elements. So, our respondents compared theportrayed emotional expressions and the advertising con-tent with their knowledge (Grayson and Martinec 2004)of the particular emotion (Shaver et al. 1987). When theemotional content in the advertisement is thought to be(in-)congruent with the recipient’s emotional knowledge,the person is likely to be (in-)authentic (cf., Grayson andMartinec 2004). Further, respondents assigned authentic-ity when they had personal experiences with the advertis-ing content and when they could relate it to their lifeexperiences (“self-referencing,” Burnkrant and Unnava1995). Moreover, if all elements in the advertisement arecongruent with the emotions expressed by the spokesper-son, emotions are likely to be authentic. Contrarily, ifsomething irritates the consumer, emotional authenticityis likely to be denied. When assessing someone as authen-tic in terms of enjoyment, respondents also referred tosome cues, e.g., a felt smile, an open and relaxed emo-tional expression, and a spontaneous snapshot of theperson not facing the camera. In contrast, e.g., an unfeltsmile and a static or cramped body language signify lowemotional authenticity of enjoyment. Additionally, someinformants discussed contrived authenticity in advertis-ing: emotions in advertising are not entirely authentic,since they have been created by the actress to fulfill anadvertising purpose (cf., Stern 1994). However, emo-tional expressions can seem authentic or can be perceivedas authentic (cf., Gilmore and Pine II 2007).

Regarding the consumer’s response to emotionalauthenticity, our qualitative study suggests more positiveand stronger effects of emotional authenticity in compari-son to a lack of authenticity. Despite the positive advertis-ing response, the consumer’s awareness of contrivedauthenticity might slightly diminish advertising response


in comparison to a reaction toward a spontaneous genuineemotional experience in real life.

Managerial Implication and Future Research

The affective and quick authenticity judgment implieshigh managerial relevance of authenticity: in the today’sadvertising environment emotional authenticity mightplay a crucial role since it is expeditiously recognized andhighly valued by the audience. In order to create authenticadvertisements it is important to reflect the consumer’s

life experiences. Further, companies need to ensure thatthe elements in the advertisement are congruent to eachother and that nothing distracts the audience from thegenuine experience due to a lack of congruence. Werecommend the following questions for further research:firstly, emotional authenticity should get studied withother emotions. Further, our findings with regard to theconsumer’s response should get validated in a quantita-tive study. In line with this, some variables that moderatethe effectiveness of emotional authenticity should getaddressed. References are available upon request.

For further information contact:Katja Leschnikowski

Institute of Marketing and ManagementUniversity of Bern

Engehaldenstrasse 4Berne, 3012Switzerland

Phone: 011.41.31.631.80.31Fax: 011.41.31.631.80.32



TEACHING MARKETING RESEARCH BACKWARDS

Alvin C. Burns, Louisiana State University, Baton RougePeter “Danny” Weathers, Louisiana State University, Baton Rouge

Ann Veeck, Western Michigan University, KalamazooRonald F. Bush, University of West Florida, Pensacola

ABSTRACT

We advocate teaching measurement and data analy-sis at the beginning of the marketing research course (i.e.,“backwards”). This approach has distinct advantages;although, we note some disadvantages. We believe thatthe backwards approach will be superior to the traditionalapproach, and we offer six propositions as to improvedstudent learning.

INTRODUCTION

Most undergraduate marketing major curricula arecomposed of four required courses: principles of market-ing, consumer behavior analysis, marketing research, andmarketing strategy. Of these, marketing research repre-sents a unique teaching challenge because it is a blend oftopics that range from simple concepts to highly complexformulas. Typically, these topics are treated in a step-by-step sequence. Figure 1 is the 11-step sequence advocatedby Burns and Bush (2010).

As can be seen in this representative figure, themarketing research process begins with establishing theneed for research and ends with preparation of the formalmarketing research report. In between these steps areproblem definition, research objectives, research design,survey methods, measurement and questionnaire design,sampling, data collection, and data analysis. Statisticaldata analysis is logically located at the end of the market-ing research process because, with an actual survey, mostof the steps must be completed before data analysis cancommence. An inspection of leading marketing researchtextbooks (Aaker, Day, and Kumar 2006; Burns and Bush2010; Churchill, Brown, and Suter 2010; Hair, Bush, andOrtinau 2010; Malhotra 2010; McDaniel and Gates 2009;Parasuraman, Grewal, and Krishnan 2007; Zikmund andBabin 2010) reveals that all of these textbooks coverstatistical analysis in their latter chapters.

CHALLENGES OF TEACHING DATAANALYSIS

Most certainly, statistical analysis is difficult anddemanding material to teach. It relies on statistical con-cepts and theories that are myriad and complicated; itnecessitates the learning and use of statistical analysissoftware; and it requires that students thoroughly under-

stand measurement scales. Even with a statistics prerequi-site for the marketing research course, students oftenflounder in the data analysis section of the class. Someauthors have developed novel and creative pedagogies forthe teaching of data analysis (see, for example, Bridges1999; Dobni and Links 2008; Tam and Siu 1996); how-ever, its location in the sequence of steps (i.e., chapters)conspires against effective teaching for at least two rea-sons. First, it is taken up some time after measurement(Note in Figure 1, that measurement is in the 7th step whiledata analysis is in the 10th step, so, at minimum, a three-week separation), and, second, data analysis is covered atthe end of the semester often when topic coverage hasslipped behind the ideal schedule, and it is the most hectictime for students. Consequently, teachers must remind,revisit or otherwise re-educate students of basic measure-ment concepts as these often dictate allowable statisticalanalyses, and teachers face time pressure to cover a largevolume of statistical material in a short amount to time. Atthe same time, students are distracted by demands andrequirements of other courses, extracurricular activities,and gaps such as Spring Break. In this paper, we offer a“backwards” approach to the timing and teaching ofstatistical analysis in the marketing research course thatovercomes these difficulties.

TEACHING MARKETING RESEARCHBACKWARDS

Figure 2 compares the “forward” or traditionalsequence with our “backwards” flow of topic treatment inthe marketing research course. Note that we advocate theteaching of data analysis very early in the course, pre-ceded only by a section on measurement. That is, themarketing research course semester begins with a data setcomprised of a number of different measures of marketingconstructs. The instructor describes relevant measure-ment concepts as a way of introducing students to thevariables in the data set. The instructor then moves todescriptive statistics as useful tools to summarize thefindings in the data set. He/she then proceeds to inferentialand relationship statistical analyses as ways to generalizethe findings to the population represented by the data setand to determine stable associations or insightful relation-ships. With the statistical analyses mastered, students arethen taught how to present the findings with tables,figures, and accompanying text in a marketing researchreport. This approach provides students with a clear


understanding of what is ultimately produced from amarketing research project: a summary of the key findingsin the Findings section of the marketing research report.

PROS OF THE BACKWARDS APPROACH

Specifically, we advocate that statistical analysis becovered very early in the marketing research course, andwe envision at least five advantages to this approach ascompared to the traditional step-by-step approach. First,measurement and analysis are co-joined. Selection of theproper statistical technique is highly dependent on thescaling assumptions of the variables in the data set, so inour approach, students are provided with a data set, andmeasurement scales are taught with the use of this data set.We then immediately take up the various descriptive,inferential, relational, and predictive statistical techniqueswhile measurement scales are fresh in students’ minds. Asecond advantage is that statistical analysis is covered inthe marketing research course temporally closer to thestatistics prerequisite course. Often, the statisticsprerequisite course is taken the semester prior to takingmarketing research. Anyone who has taught marketingresearch knows full well that marketing students “forget”concepts that they learned in the statistics prerequisitecourse. By locating the data analysis topics early in themarketing research class, some of this mental purgingmay be forestalled.

A third advantage of the backwards approach is thatstatistical analysis is understood when questionnaire design

is taken up. By understanding the nuances of data analysis(based on knowledge of measurement scales), studentsmake better choices as to question response formats onquestionnaires. This advantage is especially importantwhen team projects for local sponsors are incorporatedinto the course, and students must create questionnairesfor the sponsor’s problem. A fourth advantage is thatstatistical analysis is taken up at a less hectic time of thesemester. The beginning of the semester is relatively freeof time and energy demands as compared to the end of thesemester when papers, exams, and projects for all coursesare competing for students’ time. So, by locating the dataanalysis material early in the semester, there is the oppor-tunity to garner more student attention and concentrationon these difficult topics. The last advantage is that earlyhands-on exposure to statistical analysis may alert stu-dents to statistical “quirks” such as low correlations thatinvariably are found in survey data, sparse data that occurswith cross tabulations with many rows and columns, howto handle missing data, treatment of outliers, and so on.These data analysis nuggets of knowledge are sometimesuseful when research objectives are being formulated orquestions being designed for the questionnaire.

CONS OF THE BACKWARDS APPROACH

Naturally, our advocated approach is no panacea, andwe readily admit to some disadvantages. First, it beginsthe marketing research course with the most difficultmaterial. As we noted earlier, students tend to have anaversion to statistics, so beginning the marketing research

FIGURE 1 Sequence of Marketing Research Steps*


class with a very heavy dose of statistical analysis isdefinitely not to a student’s liking. In fact, this firstimpression may taint students’ perceptions of marketingresearch toward the negative. Another disadvantage isthat this approach is contrary to the chapter sequence invirtually every marketing research textbook. Even if theinstructor can cleverly “create” a textbook by rearrangingthe chapters of an existing text, there are invariablyconcepts referred to in the data analysis treatments that arecovered in the earlier, traditional textbook chapter sequencethat will be foreign or at least murky to students who donot encounter these chapters until after the data analysischapters are covered. Finally, we believe that ourapproach will best succeed with the use of a comprehen-sive, expository data set. If students are to learn aboutmeasurement concepts and the data analysis implicationsof these concepts plus run and interpret data analyses atthe beginning of the course, it is best to have a single dataset that accommodates all of these issues and facilitateslearning of the data analysis concepts. Some textbooksinclude comprehensive integrated case data sets that may

be useful, but it is incumbent on the instructor to scrutinizethe data set to ensure that it facilitates teaching of all of thesalient concepts and issues.

PROPOSITIONS: BACKWARDS VERSUSFORWARD MARKETING RESEARCH

Admittedly, our experience with teaching marketingresearch backwards is not deep. Moreover, we have nothad the opportunity to empirically test the veracity of ouradvocated approach. Nonetheless, we do have some intui-tive beliefs as to how our advocated backwards approachwould perform compared to the traditional, forwardapproach in various learning assessments. We offer thesein the form of the following propositions along with therationale for each proposition in the following Table.

In conclusion, we invite teachers of marketing researchto critically examine our recommended approach and tovoice concerns or to seriously consider teaching market-ing research backwards.

FIGURE 2

Topic Flow with Forward and Backwards Marketing Research Forward B ackwards

Measures

Analysis

Problem Definition

Presentation

Surveys

Questionnaires

Sampling

Collection

Problem Definition

Surveys

Questionnaires

Measures

Sampling

Collection

Analysis

Presentation


REFERENCES

Aaker, David A., V. Kumar, and George S. Day (2006),Marketing Research, 9th edition. Wiley Higher Edu-cation.

Bridges, Eileen (1999), “Learning and Customer Needs inthe Undergraduate Marketing Research Course,”Journal of Marketing Education, 21 (1), (April), 51–58.

Burns, Burns, and Ronald Bush (2010), MarketingResearch, 6th edition. Prentice Hall.

Churchill, Gilbert A., Tracy A. Suter, and Tom J. Brown(2010), Basic Marketing Research, 7th edition.Cengage Learning.

Dobni, Dawn and Graham Links (2008), “PromotingStatistical Intelligence in Marketing Research Stu-

TABLE 1Propositions Pertaining to Backwards Versus Traditional Marketing

Research Pedagogy

Proposition:

Backwards will effect… Rationale for Proposition

1: Better understanding of statistical Statistical material is covered when students are more analysisfocused and less distracted, plus temporally closer to the statisticsprerequisite course.

2: Better operational definitions More and deeper understanding of measurement-and-analysisco-dependencies will enhance students’ abilities to relate tomeasurement scales for constructs.

3: Less scale type-analysis errors Early and deeper learning of measurement scale dictates onstatistical analyses will result in greater learning of these topics.

4: Better questionnaire designs Experience and understanding of data analyses will result infewer blunders (such as open-ended questions or improperscales) in questionnaire design.

5: Better research objectives Experiencing data analysis and how findings are presented in theFindings section of the final report will guide students towarddoable research objectives.

6: Less need for hand-holding during Having experienced data analysis early on and as a function ofproject analysis Propositions 1–5, students will operate more autonomously and

correctly during team projects.

dents: Best Practice,” Marketing Education Review,18 (1), (Spring), 61–64.

Hair, Jr., Joseph, Robert Bush, and David Ortinau (2008),Marketing Research. McGraw Hill.

Malhotra, Naresh K. (2010), Marketing Research: AnApplied Orientation, 6th edition. Prentice Hall.

McDaniel, Carl and Roger Gates (2009), MarketingResearch, 8th edition. Wiley Higher Education.

Parasuraman, A., Dhruv Grewal, and R. Krishnan (2007),Marketing Research, 2nd edition. Cengage Learning.

Tam, Kam-Chuen and Wai-Sum Siu (1996), “TeachingStatistical Analysis in The Market Research Courseat Tertiary Institutions in Hong Kong,” Journal ofEducation for Business, 71, (5), (May/June), 300–4.

Zikmund, William G. and Barry J. Babin (2010), ExploringMarketing Research, 10th edition. Cengage Learning.


For further information contact:Alvin C. Burns

Department of MarketingE.J. Ourso College of Business

Louisiana State University3124 Taylor Hall, Nicholson Extension

Baton Rouge, LA 70803Phone: 225.578.8786



SELECTING A FREE VENDOR-HOSTED ONLINE SURVEYTOOL FOR STUDENT USE

Matt Elbeck, Troy University, Dothan

SUMMARY

Use of the Internet to conduct market research has notescaped the attention of instructors using this channel forstudent assignments. The purpose of this study is to helpeducators select a free online vendor-hosted survey toolfor student use by identifying, ranking, classifying andevaluating the services appropriate for student exposureto online marketing research, questionnaire design, andcontrol over analysis by accessing the raw survey respon-dent data. The benefits of this study are to offer students24–7 access to manage their survey instrument(s) andrespondent data from any computer.

Method

A mixed methods approach was used to identifyonline vendors and rate them according to ease of use.First, multiple online searches and vendor website exami-nation identify the population of free vendor-hosted onlinesurvey tools which are then described in detail. Online in-links data were used to develop a popularity-importance(P–I) ranking of the vendors as a proxy for vendor qualityto help orient the reader to the collection of 14 onlinevendors. Each of the 14 online survey vendor sites wasvisited and by examining the pricing tab which details thevarious pricing plans and their benefits resulted in theassignment of vendor services into two groups. The firstis labeled demonstration tools to include vendors offeringstudents exposure to online surveys, access to design theirown survey, presenting their survey to an online sampleand graphically reviewing the results online. The secondgroup labeled project tools offers all the demonstrationtools benefits in addition to a facility to download onlinesurvey respondent data for further analysis by students.To assess the “look and feel” of each survey tool, a sampleof 22 undergraduate online marketing students respondedto a 5-item respondent impression survey. To determinesurvey software ease of use, students were invited to openan account, design a short survey (questions supplied bythe author) and launch the survey. From this, studentswere asked to evaluate their experience.


For “look and feel,” the best performing ProjectTools are Survey Methods (M = 1.80, SD = 0.35) andQuestion Pro (M = 1.8, SD = 0.57), whilst for Demonstra-tion Tools it is Survey Monkey (M = 1.63, SD = 0.36).

For ease of use, Question Pro (M = 1.58, SD = 0.9) isthe student’s most favored project tool, and Free OnlineSurveys (M = 1.38, SD = 0.6) is the most favored demon-stration tool. The average time students took to open anaccount, create a 3-item survey and create a link to asurvey was 10.8 minutes (SD = 6.1 minutes).

For in-class project use, Question Pro is the student’smost favored Project Tool. If more than two surveys areto be created, or more than 10 questions per survey, thenSurvey Gizmo would be the preferred tool. For Demon-stration Tools, Free Online Surveys is the most favored,tho ugh if over 50 and fewer than 100 responses areexpected, then Survey Monkey would be selected. Theaverage time students took to open an account, create a 3-item survey and create a link to a survey was 10.8 minutes,highlighting the user friendly nature of all the vendor’sservices and therefore allowing the instructor to confi-dently invite students to set-up their online vendor account,create the survey based on items supplied by the instructoror created by the students and create the link to the surveyfor respondent use.

If students represent the sample for an online survey,then the most appropriate surveys based on student atti-tudes to the look and feel of each survey is SurveyMethods is best as a project tool, though if student ease ofuse is a consideration, then Question Pro would be thepreferred best-compromise survey tool. As a demonstra-tion tool, Survey Monkey is ranked best and is the best-compromise when student ease of use is considered. It iscomforting to note that the best-compromise survey toolsSurvey Monkey (demonstration tools) and Question Pro(project tool) are ranked first and second in terms of in-links popularity and importance. References are availableupon request.


For further information contact:Matt Elbeck

College of BusinessTroy University

500 University AvenueDothan, AL 36303

Phone: 334.983.6556, Ext. 356E-Mail: [email protected]


MARKETING STUDENTS’ ATTITUDES TOWARD AND BELIEFS ABOUTCOMMERCIAL SIGNAGE

James J. Kellarism, University of Cincinnati

SUMMARY

On-premises signage is among the oldest and mostfundamental forms of marketing communication, yet it isoften overlooked as a topic in marketing courses andtextbooks. The studies reported here examine introduc-tory marketing students’ attitudes toward and beliefsabout commercial, on-premises signage (Study 1) andassess the impact of an educational intervention on suchattitudes and beliefs (Study 2).

Introduction

Signage is important to marketing because of thebenefits it confers as a communication medium. It playsmultiple roles in marketing communication, includingidentification, way-finding, and branding (Calori 2007).Despite the importance of signage as a communicationmedium and branding tool, the topic is largely ignored inintroductory marketing courses and textbooks.

As a first step toward addressing this informationgap, two studies examine students’ awareness of signageissues, as well as attitudes, beliefs, and the impact of aneducational intervention. These studies represent an ini-tial attempt to document and explore how marketingstudents think about an important but neglected form ofmarketing communication and visual branding. This topicis potentially important because it represents an economi-cally and strategically significant reality of marketingpractice that is underrepresented in marketing education.

Study 1

To assess students’ attitudes toward and beliefs aboutcommercial, on-premises signage, a questionnaire wasdesign and administered to undergraduate students enrolledin an Introduction to Marketing class at a large, urban,public university. The questionnaire contained 20 state-ments about signage, generated from multiple sources.Each statement was followed by a ten-point agreementscale. The analysis compiled descriptive statistics for eachattitude/belief statement and assessed the reliability oftwo composite scales representing positive and negativeattitudes.

Results show that although students have not devotedmuch thought to the topic of commercial signage, themajority tend to hold generally positive attitudes. Never-

theless, negative beliefs and misunderstandings are alsoevident. For example, 17 percent believe that signagecauses traffic accidents, 20.8 percent believe that signsalong the highway are “eyesores,” 25.8 percent believethat illuminated signs are an environmental threat, and32.1 percent believe that electric signs are a waste ofelectricity. Roughly a third of the students report that thereis no difference between signs and billboard ads.

Given that students’ awareness of signage issues isfairly low, attitudes and beliefs may be malleable andresponsive to educational intervention. A second studygathered additional evidence using a larger sample andassessed the impact of education on such attitudes/beliefs.

Study 2

The questionnaire used in Study 1 was administeredto students enrolled in an introductory marketing class ata large, urban, public university in the Midwestern regionof the United States. The class was split randomly into twogroups. Half were asked to fill out the questionnaireimmediately prior to hearing a “guest lecture” on signageas marketing communication; half were asked to fill outthe questionnaire shortly after hearing the lecture.

The lecture was delivered by a male, senior professor.Lecture content included basic definitions and types ofsigns, a brief history of commercial signage, purposes ofsignage, including the multiple roles of signage in market-ing communication, benefits of signage to customers,businesses, and society, including economic impact, andcontroversies regarding signage.

Results are general consistent with those of Study 1.Moreover they document the impact of hearing a lectureabout signage. The mean agreement with the statement“My attitude toward commercial signage is generallynegative” was 4.36 among those queried before the lec-ture versus 3.53 (i.e., stronger disagreement) among thosequeried after the lecture (t = 2.94, df = 155, p = .004 two-tailed). Other differences are also noted. The importanceof signage as a form of marketing communication becamemore salient after hearing a lecture and the benefits andeconomic impact of signage became more apparent. Posi-tive changes can also be seen regarding the protection ofsignage as freedom of speech and the impact of signage onproperty values.


Conclusion

These studies explored students’ attitudes and beliefsregarding an important, yet neglected marketing topic.They begin to document students’ views of signage, butmore importantly to point out a gap in marketing educa-tion and propose some topics for further exploration.

Given the pervasiveness of on-premises signage, themultiple roles it plays in marketing, its benefit to consum-ers and economic impact, this preliminary report is intendedto initiate a dialog among marketing educators regardingsignage as a marketing/branding topic. References areavailable upon request.

For further information contact:James J. Kellaris

Department of MarketingCollege of Business

University of CincinnatiCincinnati, OH 45221–0145



BRAND AND TECHNICAL INTEGRATION FOR RADICALINNOVATION: DO FIRMS THAT INTEGRATE DO

BETTER THAN FIRMS THAT DO NOT?

Raul Rivadeneyra, Judge Business School, United KingdomChirag Patel, Grenoble Ecole de Management, FranceAlina Sorescu, Mays Business School, College Station

Jaideep Prabhu, Judge Business School, United KingdomRajesh Chandy, London Business School, United Kingdom

SUMMARY

Firms that seek to introduce radical innovations arefaced with important questions regarding the manner inwhich the innovations are developed and marketed. Shouldthey use their existing brand to market the radical innova-tion, or should they develop a new brand? Should theydevelop the technology for the radical innovation in-house, or should they outsource the technology develop-ment? A common theme around these questions is thelevel of integration-in marketing and technical assets- thatis appropriate for the successful development and marketintroduction of radical innovations. While a sizable litera-ture has examined the antecedents and consequences ofradical innovation, there is little evidence on the driversand outcomes of the integration decisions associated withradical innovation. Our research seeks to fill this gap.

We seek to make three key contributions to research.First, we address the issue of branding in the context ofradical innovation. Branding decisions play a crucial rolein whether and how a radical innovation is marketed.Despite a significant stream of research on the determi-nants of the success of brand extensions, the brand inte-gration decision tends to be unexplored in the radicalinnovation literature. Second, we study technologyoutsourcing as it relates to radical innovation. In thiscontext, firms have to decide whether to develop the coretechnology needed for radical innovation in-house, orsource it from external entities. Despite many calls forresearch on technology outsourcing much of the empiri-cal research on radical innovation tends to ignore externalsources of technology, and as such implicitly assumes thattechnology development happens in-house. Third, weexamine the impact of technology and brand integrationon product superiority and marketplace performance inthe context of radical innovations. By linking decisionsregarding brand and technology integration to productand marketplace outcomes, we are able to make someinitial recommendations to managers about how best totackle these decisions.

We use the concepts of risk, resources, and control tobuild a theoretical framework and develop hypotheses

about technical and brand integration decisions in thecontext of radical innovation. Firms can develop thetechnical assets needed for radical innovation in-house oracquire them through market contracts. Similarly, firmscan use an existing brand identity or create a new one tomarket products using the new technology. We argue thatin the case of technical assets, integration requires moreresources, but leads to greater control. In the case of brandassets, integration requires fewer resources, but leads togreater risk and lower control. Finally, we examine whichaspects of long term performance are impacted by eachtype of integration decision.

First, we hypothesize that dominant firms are morelikely than non-dominant firms to integrate the technicalassets needed for radical innovation. While most firmshave the incentive to technically integrate – to gain controland accelerate learning – dominant firms are best posi-tioned to do so as they are likely to have greater slackresources and a better ability to absorb unsuccessfulattempts to develop a new technology, thus making iteasier for them to attempt it in the first place. Second, wehypothesize that firms that integrate the technical assetsneeded for radical innovation are likely to develop prod-ucts with greater superiority as compared to firms that donot integrate the technical assets. Indeed, firms that developthe technology in-house are in the best position to use –through time – existing internal knowledge as well asdetect and assimilate any external knowledge that isnecessary to optimally develop and subsequently updatethe radically new technology

Third, we hypothesize that dominant firms are lesslikely than non-dominant firms to integrate their brandassets for radical innovation. The uncertainty associatedwith the radical innovation creates the incentive to createa new brand which could mitigate the brand dilution thata potential failure of the radical innovation may entail.However, dominant firms are more likely to have theresources necessary to set up and successfully support anew brand. Finally, we hypothesize that firms that do notintegrate the brand assets needed for radical innovationare likely to develop products with greater market accep-tance as compared to firms that integrate the brand assets.


Indeed, a new brand increases the perception of newnessof the product, offers the firm substantial freedom inselecting the appropriate positioning and marketing strat-egy for the product, and provides greater control over theroutines necessary to maximizing the market success ofthe radical innovation.

We test our hypotheses on archival data from the U.S.retail banking industry. Specifically we examine theintroduction of Internet banking on a sample of 312 U.S.banks over a seven-year period. We collect data onvariables that measure the integration and performance ofInternet banking, as well as data on entry time, bankcharacteristics and on technologies that banks may haveused prior to the introduction of this radical innovation.Empirical tests provide supporting evidence for all fourhypotheses.

Our study offers prescriptive implications that canguide managers in making branding and technologicaldecisions associated with radical innovation. A few sampleimplications follow.

To differentiate, build your own technical assets.Outsourcing technical assets may appear to be the faster,easier and safer route to radical innovation. However,simply adopting an existing market solution is likely to

lead to the creation of a generic and standardized product.Building the technical assets needed for radical innova-tion in-house may potentially delay the launch of theinnovation, but following this route offers a solutioncustomized for its own market. This in turn will havepositive consequences in the long-term.

Build Brands, Don’t Extend Old One for I nnova-tion. There is little if any guidance in prior literature howbest to branding radical innovations. Insights from extantresearch on brand extensions may not easily translate toproducts that are so radically new that they may entail notonly new firm routines but also new consumer behaviors.Our results counter-intuitively suggest that greater marketacceptance is achieved with a new brand, rather than withan established one, due perhaps to the flexibility in posi-tioning that a new brand affords.

Keep Decisions on Technical and B rand AssetsSeparate. The decision to integrate technology for radicalinnovation might appear to go hand in hand with thedecision to integrate the brand associated the innovation.Results from our study would suggest that recognizingthat the two decisions are separate and that they havediffering effects has important consequences for firmsinvolved in the process of radical innovation.

For further information contact:Chirag Patel

Grenoble Ecole de Management12 rue Pierre Sémard

38000 Grenoble CedexFrance

Phone: 0033.4.76706531Fax: 0033.4.76706099



MANAGING SATISFYING SELLER-BUYER INTERACTIONS DURINGNEW PRODUCT DEVELOPMENT IN TECHNOLOGY-BASED,

INDUSTRIAL MARKETS

Jason Q. Zhang, Loyola University, BaltimoreGerard A. Athaide, Loyola University, Baltimore

SUMMARY

A growing body of literature notes that the newproduct development (NPD) process in technology-based,industrial settings is increasingly characterized by closeand frequent interactions between sellers and buyersthroughout the process (Gruner and Homburg 2000; Perks2000). Such relationships provide strategic benefits forboth sellers and buyers of technology-based innovations(Leonard-Barton 1995). For example, Hamel and Prahalad(1991) and Robertson (1993) note that co-developingproducts with knowledgeable buyers alerts sellers todesirable product characteristics and potential marketopportunities; buyers benefit by gaining access to newtechnologies ahead of competitors (Udwadia and Kumar1991) and by being able to influence performance attributes(von Hippel 2001). Despite these benefits, ineffectiverelationship management during product development isan important contributor to the high failure rates of tech-nology-based, industrial innovations (Nesse and Skjelnes1994). Unfortunately, empirical research on seller-buyerrelationships during NPD remains underdeveloped (Alam2002); therefore, to address this deficiency we examinethe following two research questions:

What are the behavioral dimensions of seller-buyerinteractions during NPD?

What are some important moderators of the linkagesbetween the behavioral dimensions and relationshipsatisfaction?

Our study addresses this deficiency by developing amodel of seller-buyer interactions during NPD. The modelis rooted in Transaction Cost Analysis (TCA) and wastested using data from 296 small to mid-sized firms in avariety of technology-based, industrial markets. It speci-fies education and product co-development as two keybehavioral dimensions that characterize seller-buyer inter-actions during NPD. Further, it proposes that a seller’ssatisfaction with undertaking these behaviors is moder-ated by buyer characteristics (i.e., perceived buyer knowl-edge and prior relationship history), innovation character-istics (i.e., product customization and innovation discon-tinuity) and an environmental characteristic (technologi-cal uncertainty in the seller’s industry). The overall sup-

port we find for the model can help managers optimizetheir relationships with buyers during NPD.

Results

We found that education and product co-developmentboth have statistically significant main effects on relation-ship satisfaction. The focus of our analyses, however,concerns the moderation effects. Our analysis showedthat as hypothesized, perceived buyer knowledge nega-tively interacts with education and positively interactswith product co-development to affect relationship satis-faction. But prior relationship history was not found tointeract with either education or product co-development.Therefore, neither of the hypothesized relationships wassupported. As expected, product customization positivelyinteracts with product co-development to affect relation-ship satisfaction but the hypothesized negative interactionbetween product customization and education was notsupported. Also, innovation discontinuity positively inter-acts with education to affect relationship satisfaction ashypothesized but the expected negative interaction withproduct co-development was not supported. Finally, tech-nological uncertainty negatively interacts with productco-development to affect relationship satisfaction, as pre-dicted but it did not positively moderate education asexpected.

Discussion

Drawing on TCA, our study makes several importanttheoretical and managerial contributions. Since empiricalresearch on seller-buyer NPD relationships remains under-developed, existing studies tend to be largely anecdotaland based on case studies. In addition, the relatively fewempirical studies are often restricted to single-industrysettings. While adequate at the initial stages of theorydevelopment, such single-industry studies limit thegeneralizability of the findings. Our study represents animportant next step in theory development in the areabecause we collect data from a variety of technology-based industries. Our study also has useful implicationsfor practitioners. Allocating resources to effectively man-age the nature and timing of buyer involvement duringproduct development remains a challenge for NPD man-agers. The normative framework inherent in our model


addresses this challenge. For example, early and bilateralinteractions via product co-development are appropriatewhen dealing with a buyer who is knowledgeable. Simi-larly, education, which occurs later in the NPD process

and entails a unilateral flow of information from a sellerto a buyer, is well suited for discontinuous innovations.References are available upon request.

For further information contact:Jason Q. Zhang

Loyola University Maryland4501 N. Charles StreetBaltimore, MD 21210Phone: 410.617.5837



THE ADOPTION OF TECHNOLOGY ORIENTATION INHEALTHCARE INDUSTRY

Olivia F. Lee, St. Cloud State UniversityMatthew L. Meuter, California State University, Chico

Wenzhi Wang, Nankai University Binhai College, China

SUMMARY

Introduction

Consumer demand for better health is a prime factordriving the adoption of technology orientation (TECHOR)within the healthcare industry. Today, healthcare deliveryoften involves the use of technological tools and equip-ment such as robotics, support devices and monitoringsystems. Indeed, the availability of technology is animportant consideration for many patients when theyevaluate and select health care treatment plans (Moseley2005). Despite the importance of technology in today’shealth service delivery, the study of TECHOR in healthcareorganization as a firm-level construct is scanty. Previousorientation research on technology narrowly focuses onlimited aspects of technological use such as sales technol-ogy (Hunter and Perreault 2006), and potentially benefi-cial technologies (Srinivasan, Lilien, and Rangaswamy2002). This paper attempts to fill this gap by establishingTECHOR as a firm-level construct to assess how organi-zations manage their technology initiatives and developstrategies accordingly to meet the growing demand oftechnological use.

Conceptual Domain

In accordance to the common practices in constructdevelopment (Cronbach and Meehl 1955), the universe ofTECHOR is articulated through a set of theoretical con-cepts by extensive literature review and field studies. Thispaper argues that the existence of TECHOR deals withphysical presence of technology and technology knowl-edge. The proposed TECHOR is a five-factor multidi-mensional construct consists of technology change, tech-nology allocation, technology deployment, technologymastery, and sense and respond to technology. Technol-ogy change is defined as any changes in clinical practicethat enhance the ability of providers to diagnose, treat, or

prevent health problems (Congressional Budget Office2008). Technology allocation refers to acquiring, holdingand locating technological resources into possession. Tech-nology deployment refers to arranging, moving and orga-nizing technological resources for a deliberate purpose.Technology mastery refers to the development of compre-hensive expertise, knowledge, tacit know-how, skills, andtechniques essential to master the use of specific technol-ogy. Sense and response to technology refers toorganization’s ability to response to technology opportu-nities. Technology sensing involves acquiring knowledgeabout new technological progresses, and technologyresponding reflects the willingness to respond to new,potentially beneficial technologies (Srinivasan, Lilien,and Rangaswamy 2002).

Methodology and Findings

The healthcare industry is selected because of its highuse of technologies in clinical practices and patient caremanagement. A total of 382 employees from a Midwest-ern hospital participated in this study. The item pools havea total of 31 questions designed to assess the multi-dimensional TECHOR construct. The analytical approachinvolved two major procedures: (1) a measurement assess-ment of TECHOR to explore the possible underlyingfactor structure without imposing a preconceived struc-ture on the outcome, and (2) a confirmatory analysis toverify the factor structure by constraining the number offactors (Jöreskog 2000).

A total of five factors were extracted for the TECHORconstruct. The explained variance in each TECHOR dimen-sion is as follows: 89 percent in technology change; 27percent in technology allocation; 82 percent in technol-ogy deployment, 67 percent in technology mastery, and89 percent in sense and respond to technology. Refer-ences are available upon request.


For further information contact:Olivia F. Lee

Department of Marketing and Business LawG.R. Herberger College of Business

St. Cloud State UniversitySt. Cloud, MN 56301–4498

Phone: 320.308.4256Fax: 320.308.4061



THE ROLE OF MARKETING RESOURCES IN RADICAL INNOVATION

Kyriakos Kyriakopoulos, ALBA Graduate Business School, GreeceMathew Hughes, Nottingham University Business School, United Kingdom

Paul Hughes, Loughborough University, United Kingdom

SUMMARY

Recent research in marketing has made an importantstep in acknowledging the role of marketing strengths inthe development of innovations (Dutta, Narasimhan, andRajiv 1999; Henard and Szymanski 2001; Sorescu et al.2003; Voss and Voss 2008). However, with the exceptionof study by Sorescu et al. (2003), marketing has not mademany inroads into understanding the role of marketingresources in radical innovation, despite an increasingempirical literature on marketing assets and capabilities(Bahadir, Bharadwaj, and Srivastava 2008; Dutta et al.1999; Krasnikov and Jayachandran 2008; Vorhies andMorgan 2005).

Our overall goal, thus, is to understand the role ofmarketing resources in generating and benefitting fromradical innovation. To this end, we distinguish three typesof marketing resources building on Srivastava, Shervani,and Fahey (1998, 2001). Intellectual resources character-ize the market knowledge resources of the firm, capturingthe stock of knowledge or firm’s memory about custom-ers, competitors, channel partners, and other externalentities affecting the firm environment (Moorman andMiner 1997; Srivastava et al. 1998, 2001). Relationalresources are the result of bonds between the firm andchannel partners in terms of duration, strength, closenessand quantity (Srivastava et al. 1998). Reputationalresources reflect the impact of a firm’s corporate name orpersonality in its market in terms of awareness, strength,and appeal relative to competitors. Such resources alsocapture brand equity and firm image and can operate as anendorser of brand or product extensions (Aaker 2004).

A radical product innovation is a new product thatincorporates a substantially different core technology andprovides substantially higher customer benefits relative toprevious products in the industry (Chandy and Tellis1998, 2000). The role of resource superiority in radicalinnovation is central though multi-faceted and debatable.A literature stream building on early work of Schumpeter(1934) portrays a bleak picture for resource rich firms:complacent, inert, and unwilling to cannibalize their prod-ucts and resources, they are displaced by much morenimble, ambitious, and lean new entrants (e.g., Hendersonand Clark 1990). There is other research, however, thatstresses the complexity of modern innovations requiresfirms with resources and market power (Chandy and Tellis

2000). We address these issues with a set of hypothesestreating the marketing resources as antecedents and mod-erators of radical innovation.

H1: There is inverted U-relationship between market

knowledge resources and radical innovation in whichintermediate levels of marketing knowledge pro-duces the highest level of radical innovation.

H2: There is a negative relationship between reputational

resources and radical innovations.

H3: There is a positive relationship between relational

resources and radical innovations.

H4: The greater the market knowledge resources, the

lower the likelihood that radical innovation willpromote the financial performance of the firm.

H5: The greater the reputational resources, the higher the

likelihood that radical innovation will promote thefinancial performance of the firm.

H6: The greater the relational resources, the higher the

likelihood that radical innovation will promote thefinancial performance of the firm.

We test our hypotheses in a sample of 110 techno-logically intensive Dutch firms utilizing a survey amonggeneral and marketing managers. We relied on existingscales to tap radical innovation and marketing resources(Chandy and Tellis 1998) while measures of the market-ing resources were derived from Slotegraaf, Vorhies, andMorgan (2007). After testing for common method biasand purifying the constructs, a preliminary analysis inLISREL 8.54 indicates market knowledge resources havenegative impact and reputational have no impact onradical innovation (no support for H

1 and H

2); while

relational resources benefit radical innovation in supportof and

H

3. We also find that market knowledge reduces

while reputational and relational resources strengthen thevalue of radical innovation in support of H

4, H

5, and H

6.

Despite some methodological limitations related toreliance on cross-sectional and primary data, our study isthe first to explore empirically marketing resources in anintegrated way and is the first to take a customer-marketingview toward marketing resources as opposed to a techni-


cal or aggregate explanation of resources. First, we mapspecific marketing resources that influence radical inno-vations. Our focus complements work that has beenconfined to examining marketing resources as proxies,either as investments in marketing activities or externalmarket outcomes (Moorman and Slotegraaf 1999; Sorecsuet al. 2003).

Second, although research has examined the finan-cial consequences of some marketing resources (Bahadiret al. 2008; Capron and Hulland 1999; Krasnikov andJayachandran 2008; Moorman and Slotegraaf 1999), it

has paid scant attention to their facilitating role in derivingfinancial benefits from radical innovations (Sorescsu et al.2003). Overall, we show that the value of marketingresources can vary depending on whether we examinetheir impact before or after radical innovation. Third, webuild on the intuition that firms benefit financially fromradical innovation in the presence of complementarymarketing resources. Although research has examined thefinancial consequences of some marketing resources in-dependently (Moorman and Miner 1997), it has paid scantattention to their facilitating role in deriving financialbenefits from radical innovations (Sorescsu et al. 2003).

For further information contact:Kyriakos Kyriakopoulos

ALBAAthinas Ave & 2A Areos Str.

166 71 VouliagmeniGreece

Phone: +30.210.8964531.8.221Fax: +30.210.8964737



COMMA N’ CENTS IN PRICING: THE EFFECTS OF AUDITORYREPRESENTATION ENCODING ON PRICE RECALL

AND MAGNITUDE PERCEPTIONS

Keith S. Coulter, Clark University, WorcesterPilsik Choi, Clark University, Worcester

SUMMARY

Research has demonstrated that consumers’ process-ing of price information may not always involve con-scious, rational processes (Monroe and Lee 1999). Rathernon-conscious processes can operate at encoding to dis-tort the numerical information that is retained in memory(Xia 2003). For example, non-numerical aspects of thestimulus such as color, spokesperson, layout, or contextcould impact both how the price information is laterrecalled, and the “magnitude” or “value” that is associatedwith the sale price (Babin, Hardesty, and Suter 2003;Nunes and Boatwright 2004; Grewal, Marmorstein, andSharma 1996).

In this paper, we suggest that the manner in which aprice “name” is verbally represented [e.g., $167 can be(either internally or externally) verbalized as “one hun-dred sixty-seven dollars” or “one sixty-seven” or “one sixseven”] can impact not only whether it is retained in short-term memory, but also how it is perceived in terms ofnumerical value or amount. Further, we demonstrate thatinclusion of commas (e.g., $1249 → $1,249) and cents(e.g., $1249 → $1249.99) in a price’s written form inArabic numerals (i.e., how it is perceived visually) canalter how that price is verbally encoded and recalled, andhence influence both recall as well as numerical magni-tude perceptions. We maintain that the effect occurs dueto variation in price-name syllabic length, and the positivecorrelation that consumers non-consciously attribute tothe relationship between syllabic length and numericvalue. Our theory is grounded in the numerical cognitionliterature, which suggests that the human cognitive sys-tem deals with numerical information (and hence prices)through a dedicated cognitive subsystem (Ashcraft 1992;Vanhuele, Laurent, and Dreze 2006). Dehaene (1992) andcolleagues developed a triple-code model of the cognitivenumerical subsystem, in which it is proposed that num-bers can be mentally represented and manipulated in threedifferent forms (or representations): visual, auditory, andanalog. Because each of the encoded representations areneurologically connected with one another, we posit that

if phonological representations can be manipulated byvarying the syllabic length of price-names (i.e., addingcommas or cents to the price), then analog representationsmay be affected.

We conduct an experiment in which subjects areexposed to a series of products (i.e., laptops and flat-screen televisions) and associated prices. The prices varyin terms of comma insertion, as well as the inclusion ofcents digits. Stimuli are presented visually; the visual cuesare hypothesized to determine the manner of phonologicalencoding, and hence impact recall and price magnitudeperceptions. In order to ensure phonological encoding,the experimental instructions direct subjects to read theprices to themselves “without speaking out loud” (thisreflects the way they read prices in a store environment),and subjects are also told to internally recite (i.e., mentallyrehearse) the observed prices in order to keep them inmemory.

Logistic regression results reveal that participants areunable to recall both prices in 62.5 percent of those trialswhere commas have been inserted; however inability torecall both prices drops to 33.3 percent when commas arenot present. Similarly, participants are unable to recallboth prices in 58.3 percent of those trials where centsdigits have been attached; however inability to recall bothprices drops to 37.5 percent when cents digits are notincluded. OLS regression reveals that comma insertionsignificantly increases magnitude perceptions involvingtwo different sets of prices [R2 = .115; F(1, 94) = 12.27, p =.001; beta = .34; R2 = .092; F(1,94) = 9.49, p = .003; beta =.30]. Similarly, cents-digit addition significantly increasesmagnitude perceptions for both our first [R2 = .091; F(1,44) = 4.40, p = .042; beta = .302] and second [R2 = .051;F(1,44) = 7.83, p = .008; beta = .389] set of prices. Theseresults support our hypotheses. Because marketing man-agers are typically interested in conveying the smallest(i.e., least expensive) price possible, from a practicalperspective our findings seem to argue against Comma N’Cents in pricing.


REFERENCES

Adaval, Rashmi and Kent B. Monroe (2002), “AutomaticConstruction and Use of Contextual Information forProduct and Price Evaluations,” Journal of Con-sumer Research, 28 (4), 572–88.

Ashcraft, Mark H. (1992), “Cognitive Arithmetic: AReview of Data and Theory,” Cognition, 44 (1/2),Special Issue, 75–106.

Coulter, Keith S. and Patricia A. Norberg (2009), “TheImpact of Physical Distance on Price Discount Per-ceptions,” Journal of Consumer Psychology, 19 (2),(April), 144–57.

Dehaene, Stanislas (1992), “Varieties of Numerical Abili-ties,” Cognition, 44 (1/2), Special Issue, 1–42.

____________ and L. Cohen (1995), “Toward an Ana-tomical and Functional Model of Number Process-ing,” Mathematical Cognition, 1 (1), 83–120.

McCloskey, M. (1992), “Cognitive Mechanisms in Nu-merical Processing: Evidence from AcquiredDyscalculia,” Cognition, 44, 107–57.

____________ and Paul Macaruso (1995), “Represent-ing and Using Numerical Information,” AmericanPsychologist, 50 (5), 351–63.

Noel, Marie-Pascale (2001), “Numerical Cognition,” inHandbook of Cognitive Neuropsychology: WhatDeficits Reveal About the Human Mind, Brenda Rapp,ed. New York: Psychology Press.

Vanhuele, Marc, Gilles Laurent, and Xavier Dreze (2006),“Consumers’ Immediate Memory for Prices,” Jour-nal of Consumer Research, 33 (2), 163–72.

Xia, L. (2003), “Consumers’ Judgments of Numerical andPrice Information,” Journal of Product and BrandManagement Featuring Pricing Strategy and Prac-tice, 12 (5), 275–92.

Complete list of references is available upon request.

For further information contact:Keith S. Coulter

Graduate School of ManagementClark University950 Main Street

Worcester, MA 01610–1477Phone: 860.429.1152



THE INFLUENCE OF ASSORTMENT SIZE ON CONSUMERS’INFORMATION-PROCESSING STRATEGY: RATIONALE

AND DESIGN OF AN EMPIRICAL APPROACH

Svenja Schmidt, University of St. Gallen, Switzerland

ABSTRACT

Studies on consumer choice often assume thatconsumers use a hierarchical two-step process in makingtheir decision on what to buy. The purpose of the researchoutlined in this manuscript is to renew the question ofbrand versus attribute processing in decision making.Choice processing is revisited in the light of recent evidenceindicating that assortment size induces choice overloadand heuristic processing. More important, this paperoutlines the rational and design for an empiricalexamination of the processing strategies of “attributesfirst” versus “brands first” consumers use when presentedwith assortments of increasing numbers of brands andproduct options.

INTRODUCTION

Research on consumer choice early on assumed thatconsumers administer a hierarchical two-step process inmaking their decision on what to buy (Bettman and Park1980; Gensch 1987; Lussier and Olshavsky 1979). Anumber of related studies have been concerned with thesequence and degree to which consumers process byattribute or by brand (Bettman et al. 1998; Payne 1976;Ursic and Helgeson 1990). Some studies suggest thatconsumers begin with a single-attribute, across-brandelimination of options, shifting to within-brand, multiple-attribute, and overall brand evaluation during the secondphase of the choice process (Bettman and Park 1980;Biehal and Chakravarti 1986). Other authors, however,have argued that when choosing to make a purchase,consumers will first consider the brands given (e.g.,Howard and Sheth 1969; for a summary see Shocker et al.1991). Thus, elaborations of the details of the two-stepdecision-making process seem to be leading to contradict-ing results.

However, one seemingly trivial, yet potentially im-portant, methodological characteristic has been neglectedso far and might help solve the contradiction. The conceptof brands being considered first is typically based on thenotion that consumers are faced with a large number ofbrands inducing them to use a simple brand heuristic(Hauser and Wernerfelt 1990) in eliminating choiceoptions. On the other hand, studies suggesting a choiceprocess beginning with single attribute evaluations haveusually worked with comparably small choice sets con-

taining no more than four to eight brands. It would appear,then, that what prior research has actually shown is thatchoice among relatively limited assortments is made byprocessing attributes first (“attributes first model,” AFM)whereas choice among relatively large assortments ismade by processing brands first (“brands first model,”BFM). Although the size of the choice set may seem atechnical detail of operationalization and design, it mayturn out to be a critical parameter as several studies havefound that the strategy by which individuals make adecision is influenced by the number of brands and theamount of information per brand in the choice set (Jacobyet al. 1977; Payne 1976).

Moreover, large assortments have recently been shownto influence consumer choice through effects of choiceoverload (Iyengar and Lepper 2000) and heuristic pro-cessing (Shah and Oppenheimer 2008). Yet, to date nospecific attempt has been made to conceptually and empiri-cally consider potential effects of assortment size in theconcept of two-stage decision making. Additional researchseems, thus, promising.

Hence, the purpose of this research is to renew thequestion of brand versus attribute processing in decisionmaking in the light of recent evidence indicating that anincreasing assortment induces choice overload and heu-ristic processing. More important, this paper outlines therational and design for an empirical examination of theinformation-processing strategies AFM versus BFM con-sumers use when presented with assortments of increas-ing numbers of brands and product options.

CONCEPTUAL BACKGROUND ANDPROPOSITIONS

To date, a fair number of studies have investigateddecision-making processes involving consumer choicesfrom a given assortment. Recently, however, the estab-lished premises of rational choice and utility maximiza-tion are being challenged by the concepts of choiceoverload and simplifying heuristics which seem to pro-vide psychologically more valid accounts of human deci-sion making. In contrary to rational choice theory, theconcept of choice overload indicates that consumers arenot necessarily better off having more options (Chernev2003b; Lehmann 1998). Instead, an inverted U-shapedrelation between assortment size and decision-making


outcomes such as, for example, choice satisfaction orlikelihood of product purchase, is found in empiricalstudies (Huffman and Kahn 1998; Iyengar and Lepper2000). Simplifying heuristics, on the other hand, counterthe notion of consumers as deriving a global utility valuefor each option by reviewing, weighing, and adding allavailable pieces of product information. Instead, deci-sion-makers have been shown to use shortcuts such asbrand information to ease their information processing(Maheswaran et al. 1992).

The concepts of both choice overload and simplify-ing heuristics are based on the notion that a human’scognitive capacity is limited and that, thus, one of con-sumers’ most important goals for decision making is“minimizing the cognitive effort required to make thechoice” (Bettman et al. 1998, p. 193; see also Shugan1980). As consumers have limited information process-ing capabilities, they “seek to conserve cognitive energywhen making perhaps dozens of purchases in lowerinvolvement supermarket shopping environments”(Andrews and Currim 2002, pp. 65–66). Decisions withmore options and attributes per option are typically evalu-ated by decision-makers as more difficult and morecognitively demanding (Malhotra 1982, 1984). Further-more, it has been found that information-processing strat-egies vary as a function of the number of choice options(Bettman et al. 1998). These findings indicate that anincreasing choice load will lead to more heuristic decisionstrategies, such as the usage of brands and a BFM choicestrategy in information processing.

Hence, this paper advances the extant research byaddressing the question of how the assortment size bymeans of choice overload and heuristic processing influ-ences consumers’ strategy about which information toselect for examination.

One of the principles underlying choice processes isthat choice options are made up of discrete pieces ofinformation and are evaluated on an information-by-information, or piecemeal, basis (Sujan 1985). Piecemealprocessing (Newell and Bröder 2008) is typically concep-tualized as sequential in nature, as “the mind operatessequentially, step by step and cue by cue” (Martignon andHoffrage 1999, p. 137; see also Payne et al. 1993).According to AFM, consumers begin with a single-at-tribute, across-brand elimination of options, shifting towithin-brand, multiple-attribute, and overall brand evalu-ation during the second phase of the choice process(Bettman and Park 1980; Biehal and Chakravarti 1986). Ifpurchasing jam, for example, the consumer would firstlook for the type of jam he/she desires (e.g., blueberry)and would then look for his/her desired brand (e.g.,Smucker’s). But as noted earlier, a BFM choice strategycan be proposed for large assortments as consumers areassumed to refrain from processing all different product

attributes when faced with a large assortment, but toinstead simplify the assortment by using brands as pro-cessing heuristics (Jacoby et al. 1974a). Hence, instead ofgoing through all types of jam offered, the consumerwould look for his/her desired brand (e.g., Smucker’s)and would then look for his/her favorite type of jam (e.g.,blueberry) within this diminished choice set. Thus, withregard to the present research context, it is proposed that:

P1: When choosing from a small assortment, consumerswill first process product attributes, then brands(“attributes first model,” AFM).

P2: When choosing from a large assortment, consumerswill first process brands, then product attributes(“brands first model,” BFM).

The notion of differences in choice strategies as afunction of assortment size is consistent with early find-ings on human information processing. Information pro-cessing is said to increase with growing input until anoptimal information-processing level is reached (Schroderet al. 1967). If input further increases, the information-processing level decreases, giving rise to simplifyingheuristics (Jacoby et al. 1974a; Malhotra 1982).

As for the critical number of information units cor-responding to this optimal information-processing level,empirical evidence is mixed. In his famous article, Miller(1956) proposes a processing limit of seven informationunits, other authors concluded that “information-process-ing level decreased beyond the information load of tenitems” (Malhotra 1982, p. 427), or found evidence that“six was expected to represent maximum comfortableload” (Wright 1975, p. 63). Several authors argue thatpurchase decisions are based on no more than about threeunits of information (Lussier and Olshavsky 1979), andBettman (1981) revealed that consumers were likely toadopt simplifying information-processing strategies whenthe number of choice options exceeded five. Not only hasthe absolute number of information units determining theoptimal information-processing level been subject to sci-entific debate, consumer researchers also dissent onwhether the number of attributes and the number of brandsin an assortment form independent dimensions of amountof information (Malhotra 1982) or ought to be seen asmultiplicative (Jacoby et al. 1974b).

Regardless of the absolute value of the optimal infor-mation-processing level and its calculation, the reportedfindings provide support for an optimal information-processing level or critical assortment size (CAS) whichtriggers a shift toward heuristic, brand-based processingonce it is being exceeded. Hence, the two-step process indecision making can be proposed to invert from AFM toBFM when the assortment size a consumer is presentedwith exceeds the CAS.


P3: There is a critical assortment size (CAS) where thehierarchical two-step decision-making process invertsfrom processing attributes first (AFM) to processingbrands first (BFM).

Brands make an appealing processing heuristic inthis context as consumers have been shown before to useless information for processing when brand names areavailable for a choice task (Jacoby et al. 1974a, 1977;Maheswaran et al. 1992). On the other hand, productcategories in supermarkets and stores are oftentimes pre-sorted by brands, and the organization of an assortmenthas shown to influence consumers’ perceptions and pro-cessing (Hoch et al. 1999; Kahn and Wansink 2004).Therefore, it could be argued that the proposed inversionfrom AFM to BFM is not necessarily due to an effectcaused by heuristic brand processing as a result of theactual assortment size. This effect, one might argue, couldrather be the result of any heuristic processing by what-ever kind of presorting given in the assortment or choiceset. If this assumption was true, BFM would apply only ifthe choice set was sorted by brands, but not if it was sortedby product attribute or not presorted at all. To furtherinvestigate the distinct role of brands within consumptionchoice, it is hence proposed:

P4: The model inversion from AFM to BFM occursregardless of the way the choice set is presorted.

The propositions articulated above provide a richresearch agenda where the next step is the operationalizationand development of an empirical design to put the propo-sitions to test.

RATIONALE OF EXPERIMENTAL DESIGN

Most of the current research approaches to consumerchoice focus on the predictive ability of different types ofchoice models tested, using the relative proportion ofchoices correctly predicted by each model as their crite-rion of success. Studies that are less concerned with thepredictive ability of concurrent models but more con-cerned with the main processes involved in decisionmaking are often hampered by methodological challengesimposed by the required costly process-tracing methodssuch as, for example, eye tracking (Russo and Leclerc1994). As a consequence, cognitive choice processeshave often been investigated by means of respondents’verbal reports (Bettman and Park 1980; Payne 1976;Ursic and Helgeson 1990). Such reports, however, onlyallow to “conjecture as to what actually happened duringthe temporal sequence leading to the purchase decision”(Jacoby et al. 1977, p. 209). This is particularly true ascognitive psychology teaches us that most of the factorsaffecting choice are unconscious. Glöckner (2009) there-fore recommends response-time measurement as a rela-

tively inexpensive experimental technique to determinehow individuals organize and process information.

The influence of time upon consumer choice is wellrecognized within the literature (Rose and Black 2006)and evidence has been provided that time has consider-able influence over the cognitive efforts of consumers andsignificant impact upon choice behavior (Dhar and Nowlis1999; Suri and Monroe 2004). In particular, the latency toreach a decision has been shown to be a function of thecomplexity of the choice, that is of the number of at-tributes and brands to be considered (Iyengar and Lepper2000; Jacoby et al. 1977; Lynch and Srull 1982). Accord-ingly, this empirical study is designed as an interactivecomputer simulation, individually run for each participant(Chernev 2003a), where decision time serves as an ex-tended method to classify decision strategies and to unveilconsumers’ underlying information processing. Respon-dents are presented with stimulus sets consisting of pic-tures of product assortments of different size and productcategories (e.g., noodles, jam, yogurt as suggested byprior research; e.g., Chernev 2003a; Iyengar and Lepper2000; Kahn and Wansink 2004). Participants are asked topick one of the displayed items as if they were doing theirregular grocery shopping at a store offering the givenassortment. Following the cover story of a marketingresearch study, participants are familiarized with thechoice task. They are given two sample sets (one beingrelatively small, one being relatively large) containingchoice sets similar to the ones used later in the experimentbut from a different product category. Next, respondentsare given the actual choice sets. Every stimulus set ischaracterized by a certain assortment size, number ofdifferent brands (N

B) and number of attribute dimensions

(NA), for example, number of different flavors. Assort-

ment sizes range from a small selection of three items tolarge sets of 18 items where participants have a reasonablylarge, but not unusual, number of options. Assortmentsizes and individual stimuli within a product category arerandomly assigned for each trial. All stimulus sets aredesigned as to avoid biases introduced by aspects ofattribute alignability (Gourville and Soman 2005;Herrmann et al. 2009), relative frequencies of items withinan assortment (Simonson and Winer 1992), differences inattribute importance (Chernev 1997) and undesired trade-offs between brands and attributes (Jacoby et al. 1974b).

Previous studies in choice research claim that “peopleoften approach decision problems with already formedproduct preferences and, provided that the preferred optionis available, make their selection without evaluating allalternatives” (Chernev 2003a, p. 151). Thus, to control forpossible confoundations by the absence or presence ofparticipants’ favorite products or brands within the stimu-lus set, these preferences are sampled. It is also controlledfor both purchase and consumption frequency of the


product categories used in this research, as previousstudies found these frequencies to be potential confoundvariables (Dechêne 2006). Time needed for decisionmaking on each given choice set is employed as the maindependent variable. It is complemented by self-reportmeasures to assess information load (Malhotra 1982),enjoyment, frustration and perceived difficulty of deci-sion making (Iyengar and Lepper 2000) as well as per-ceived variety (Kahn and Wansink 2004) and respon-dents’ degree of confusion (Huffman and Kahn 1998).These additional subjective measures are collected asMilford and Perry (1977, p. 133) and other authors main-tain that “the necessary condition for overload is whetheror not the individual feels overloaded rather than whetheror not he is required to process X bits of information.”Finally, the results of this computer simulation will alsodepend on a respondent’s computer expertise and cogni-tive processing skills where the latter will vary as afunction of respondents’ need for cognition (Gourvilleand Soman 2005) and socio-demographic factors such asage (Bettman et al. 1998). Hence, these variables need tobe accounted for.

Three different experiments are run to investigate thepropositions outlined. All trials are implemented as anexperimental within-subject-design. A first trial is run todetermine the decision time needed for assortments ofdifferent sizes and to identify the critical assortment size.A second trial is then set up to empirically investigate theoften assumed hierarchical two-step decision-making pro-cess within small and large assortments, where “small”and “large” are determined relative to the critical assort-ment size identified. Finally, to distinguish between re-sponse-time effects induced by consumers’ usage of brandsas heuristic choice strategies versus by sole assortmentpresorting, a third trial is run. The stimulus sets for thethree experimental trials are designed according to therespective requirements of each trial and are exemplifiedbelow.

Trial 1: Identification of the Critical Assortment Size

The aim of this first trial is to determine the decisiontime needed for assortments of different size and toidentify the corresponding maximum of time a consumeris willing to invest when choosing a product versus abrand. Two series of stimulus sets are designed. Series Asets contain N

A different attribute dimensions but only one

brand (NB = 1) whereas series B sets contain N

B different

brands but only one attribute dimension (NA = 1). Both

series are designed with different assortment sizes of threeto 18 randomly chosen stimuli that are randomly posi-tioned within the choice set. Using a range of three to 18items is in line with prior research and allows to reliablydetect the critical assortment size as well as to unveildifferences between the upper limit of time consumers are

willing to invest when choosing a product option from aseries A set (UL

A) respectively when choosing a brand

from a series B set (ULB). The assortment is manipulated

within subjects: All participants are asked to make twochoices for every product category used; one choice to bemade upon a series A set, and one choice upon a series Bset. The number and position of items displayed areassigned randomly by the computer software.

In the following, an example is given using noodlesas an exemplary product category, and, for clearnessreasons, pictograms instead of product images to illustratethe study’s rationale. Figure 1 gives an example for bothstimulus sets A and B where the assortment consists of anotional number of nine stimuli each. Example set Acontains noodles of nine different shapes (i.e., attributedimensions) but from the same brand, example set Bcontains noodles of the same shape but from nine differentbrands. With the author being German, the brand logosused in this exemplifying illustration are common Ger-man noodle brands any respondent to this study wouldencounter when doing his/her grocery shopping at aregular German supermarket. In addition to the pretest,recognition of brands and products used in the ultimatestudy are pretested, and product, respectively brand aware-ness are controlled for in every experimental trial.

Respondents’ decision time for choosing a productoption or attribute dimension given a certain brand (seriesA set) versus choosing a brand given a certain attributedimension (series B set) is measured as the dependentvariable. The optimal information-processing level isconsidered to be reached upon the assortment size wheredecision time saturates or decreases. No prior research isindicating any response-time differences for choosingamong attributes or brands per se, hence, UL

A and UL

B

can be expected to be equal.

Trial 2: The Two-Stage Decision-Making Process

Based on the results of Trial 1, this second trial ismeant to determine whether the often proposed hierarchi-cal two-stage decision-making process (Bettman and Park1980; Ursic and Helgeson 1990) is to be confirmed, andif so whether the assortment size influences it as assumedin propositions 1–3. As one condition of this experiment,two assortment sizes are chosen for all stimuli: a smallassortment size below the CAS and a large assortment sizeabove the CAS, as indicated by the results of Trial 1. Forboth assortment size conditions, pairs of stimulus sets aredesigned. Series 1 sets contain N

A (e.g., nine) different

attribute dimensions but NB (e.g., three) different brands;

the corresponding series 2 sets contain the same assort-ment size but in the vice versa allocation, in this example,that is, N

B being nine different brands and N

A being three

different attribute dimensions. Figure 2 illustrates an


3=NBA

1=NAB

example for two corresponding stimulus sets with anassortment size of nine items. The figure shows that forset 1, the number of brands to choose from after theattribute dimension has been chosen equals one ( ),whereas the number of products to choose from after thebrand has been chosen equals three ( ). The contraryis true for decisions made from set 2; here, the number ofbrands to choose from after the product has been chosenequals three ( ), whereas the number of productsto choose from after the brand has been chosen equals one( ).

If AFM was true and attributes were processed first,decision time for set 1 should be equal to the decision timefor set A given the same assortment size. In this paper’sexample, both sets contain nine different attribute dimen-sions. Set A contains only one brand for all attributedimensions, whereas set 1 contains three different brands.But if choice was first based on attribute dimensions, thesebrands should not be taken into account when making thechoice for set 1. Decision time should thus not be differentfrom the set containing only one brand; choosing fromeither set 1 or set A should take approximately the sameamount of time. This is expressed in Equation (1) and isfurther illustrated in Figure 3.

Similarly, if brands were processed first (BFM),decision time for set 2 should be equal to decision time forset B given the same assortment size. In this paper’sexample, both sets contain nine different brands. Set Bcontains only one attribute dimension for all brands,whereas set 2 contains three different attribute dimen-sions. But if choice was based on brands, these attributesshould not be taken into account when making the choicefor set 2. Decision time should thus not be different fromthe set containing only one attribute; choosing from eitherset 2 or set B should take approximately the same amountof time. This is expressed in Equation (2) and is furtherillustrated in Figure 4.

(2)

Furthermore, if AFM employed, decision time for set2 should be equal to the sum of decision time for a seriesA set, where the assortment size equaled the number ofattribute dimensions in set 2, plus the decision time for aseries B set, where the assortment size equaled the numberof brands per attribute dimension in set 2. This is expressedin Equation (3) and is further illustrated in Figure 5.

FIGURE 1Example of Stimulus Sets in Trial 1

3=NAB

1=NBA

1,Nt=N,Nt AABA(1)

1,Nt=N,Nt BBAB



FIGURE 3Illustration of Equation 1 (AFM, Set 1)


(3)

with v being the time needed for scanning one set (correc-tion factor).

And finally, if BFM employed, decision time for set1 should be equal to the sum of decision time for a seriesB set, where the assortment size equaled the number ofbrands in set 1, plus the decision time for a series A set,where the assortment size equaled the number of attributesper brand in set 1. This is expressed in Equation (4) and isfurther illustrated in Figure 6.

(4)

with v being the time needed for scanning one set (correc-tion factor).

Provided that trial 1 resulted in significant differ-ences of decision time for small and large assortment sizes(AS), where small and large are determined by the CASor UL, the following propositions can be made to substan-tiate the validity of Propositions 1 to 3:

P1a: For assortment sizes below the upper limit (AS <CAS), in the Set 1 condition there is

FIGURE 4Illustration of Equation 2 (BFM, Set 2)

νNt+,Nt=N,NtABAABA 1,1

νNt+,Nt=N,NtBABBAB 1,1

νNt+,Nt=N,NtABAABA 1,1

1,Nt=N,Nt AABA

1,Nt=N,Nt BBAB

νNt+,Nt=N,NtBABBAB 1,1

(AFM in small assortments, Equation 1).

P1b: For assortment sizes below the upper limit (AS <CAS), in the Set 2 condition there is

with v being the timeneeded for scanning one set (AFM in small assort-ments, Equation 3 ).

P2a: For assortment sizes above the upper limit (AS >CAS), in the Set 2 condition there is(BFM in large assortments, Equation 2).

P2b: For assortment sizes above the upper limit (AS >CAS), in the Set 1 condition there is

with v being thetime needed for scanning one set (BFM in largeassortments, Equation 4).

Trial 3: The Influence of Presorting

Following the trials outlined above, one may criticizethat the suggested inversion from AFM to BFM is not dueto consumers’ usage of brands as heuristics in largeassortments, but is caused by retailers’ habit to presortassortments by brand families. One may argue that onceassortments increase, consumers’ make use of any pre-


FIGURE 5Illustration of Equation 3 (AFM, Set 2)

FIGURE 6Illustration of Equation 5 (BFM, Set 1)


sorting as a matter of choice simplification instead ofexplicitly using brands as a choice heuristic. If this pleawas true, there should be a significant time difference forlarge assortment decisions made on presorted sets versusnon-sorted sets as well as no difference in decision timewhether the assortment was sorted by attribute or bybrand. Figure 7 illustrates the rationale of the correspond-ing stimulus sets used in this third trial designed todetermine the influence of assortment presorting.

EXPECTED CONTRIBUTION ANDIMPLICATIONS

For almost any common grocery product, consumersshopping in today’s typical supermarket are required tomake their choice from among numerous competing prod-ucts and brands (Mitchell et al. 2005). Research on con-sumer decision making and choice has long viewed suchvariety as beneficial for consumers, manufacturers, andretailers (Broniarczyk et al. 1998; Hoch et al. 1999).However, a burgeoning empirical literature calls this viewinto question, showing that increasing variety can lead tochoice overload and heuristic decision making (Iyengarand Lepper 2000; Shah and Oppenheimer 2008). Thepresent research extends these recent findings and inves-tigates how an increasing assortment will influence con-sumers’ choice processing. In particular, it is argued thatconsumers’ choice strategies will be systematically influ-enced by assortment size, with a large assortment leadingto a strategy where brands are processed first, but a smallassortment abetting attribute processing.

Empirical testing is the logical next step in establish-ing the validity of the decision-making model and proposi-tions outlined above. However, a traditional test strategy,where the null hypothesis would assume no difference

between means in reaction time, seems inappropriate(Mayer and Butler 1993); model confirmation woulddepend on a series of null effects for the outlined propo-sitions. Yet, failure to reject the null hypothesis could bedue to the model being acceptable, but it could also beinterpreted as a result of an insufficient sample size or ofa test with low power. Therefore, equivalence testing isrecommended for the analysis of the data collected (Rogerset al. 1993; Wellek 2003). To the best auf the author’sknowledge, equivalence testing has not been used inconsumer research so far; hence, the application of thistest strategy will contribute to the methodological discus-sions and the marketing literature.

Moreover, this study’s aim of investigating the effectsof assortment size on choice processes is of considerabletheoretical relevance (Dieckmann et al. 2009, p. 200):“Gaining insights into the processes people follow whilemaking purchase decisions will lead to better informeddecision theories.” As has been noted, consumer choiceresearch concerning the details of the hierarchical two-step decision-making process led to contradicting results.At the same time, related research has failed to account forthe moderating role of assortment size. Thus, in order toinvestigate the influence of assortment size on choiceprocessing, this research develops a novel procedure thatallows to examine the details of choice processing, utiliz-ing response-time measurement.

But the results of this research are of theoretical aswell as practical interest. Marketers are interested in morerealistic decision models for predicting market shares andfor optimizing marketing actions (Dieckmann et al. 2009),for example, by adapting their category management andassortment design to consumers’ choice processes. Fur-thermore, the results of this research will provide informa-



tion about the significance and optimal usage of brands aspurchase triggers depending on a given assortment. Thisresearch will thus allow for recommendations as of whatkind of assortment design and composition manufacturersand retailers should aim for. And it will answer some ofshopping psychologists’ most pressing questions: What isconsumers’ capacity for assortment processing and howdoes decision making for consumer products occur at the

PoS? Hence, the results of this research will help tounderstand the consumer and his/her behavior at the PoS.Such an understanding can then be made profitable by thedevelopment of guidelines for a consumer-optimizedassortment design. Consequently, marketing and retailmanagers will be able to offer optimal conditions at thePoS in an attempt to ease consumers’ buying decisionsand to increase shopping turnover.

REFERENCES

Andrews, R.L. and I.S. Currim (2002), “Identifying Seg-ments with Identical Choice Behaviors Across Prod-uct Categories: An Intercategory Logit MixtureModel,” International Journal of Research in Mar-keting, 19, 65–79.

Bettman, J. and W. Park (1980), “Effects of Prior Knowl-edge and Experience and Phase of the Choice Processon Consumer Decision Processes: A Protocol Analy-sis,” Journal of Consumer Research, 7, 234–48.

Bettman, J.R. (1981), An Information Processing Theoryof Consumer Choice. Reading, MA: Addison-Wesley.

____________, M.F. Luce, and J.W. Payne (1998), “Con-structive Consumer Choice Processes,” Journal ofConsumer Research, 25 (3), 187–217.

Biehal, G. and D. Chakravarti (1986), “Consumers’ Useof Memory and External Information in Choice:Macro and Micro Perspective,” Journal of ConsumerResearch, 12, 382–405.

Broniarczyk, S.M., W.D. Hoyer, and L. McAlister (1998),“Consumers’ Perceptions of the Assortment Offeredin a Grocery Category: The Impact of Item Reduc-tion,” Journal of Marketing Research, 35, 166–76.

Chernev, A. (1997), “The Effect of Common Features onBrand Choice: Moderating Role of Attribute Impor-tance,” Journal of Consumer Research, 23, 304–11.

____________ (2003a), “Product Assortment and Indi-vidual Decision Processes,” Journal of Personalityand Social Psychology, 85 (1), 151–62.

____________ (2003b), “When More Is Less and Less IsMore: The Role of Ideal Point Availability andAssortment in Consumer Choice,” Journal of Con-sumer Research, 30 (2), 170–83.

de Vries-van Ketel, E. (2005), “How Assortment VarietyAffects Assortment Attractiveness,” Ph.D. Thesis,Erasmus Research Institute of Management (ERIM).

Dechêne, C.F. (2006), Abwechslungsbedürfnis undWerbewirkung: Theoretische Überlegungen undExperimentelle Prüfung. Wiesbaden: Gabler.

Dhar, R. and S.M. Nowlis (1999), “The Effect of TimePressure on Consumer Choice Deferral,” Journal ofConsumer Research, 25 (4), 369–84.

Dieckmann, A., K. Dippold, and H. Dietrich (2009),

“Compensatory versus Noncompensatory Modelsfor Predicting Consumer Preferences,” Judgmentand Decision Making, 4 (3), 200–13.

Gensch, D. (1987), “A Two-Stage Disaggregate AttributeChoice Model,” Marketing Science, 6 (Summer),223–31.

Glöckner, A. (2009), “Investigating Intuitive and Delib-erate Processes Statistically: The Multiple-MeasureMaximum Likelihood Strategy ClassificationMethod,” Judgment and Decision Making, 4 (3),186–99.

Gourville, John T. and Dilip Soman (2005), “Overchoiceand Assortment Type: When and Why Variety Back-fires,” Marketing Science, 24 (3), 382–95.

Hauser, J.R. and B. Wernerfelt (1990), “An EvaluationCost Model of Consideration Sets,” Journal of Con-sumer Research, 22, 186–99.

Herrmann, A., M. Heitmann, R. Morgan, S.C. Henneberg,and J. Landwehr (2009), “Consumer Decision Mak-ing and Variety of Offerings: The Effect of AttributeAlignability,” Psychology and Marketing, 26 (4),333–58.

Hoch, S.J., E.T. Bradlow, and B. Wansink (1999), “TheVariety of an Assortment,” Marketing Science, 18(4), 527–46.

Howard, J.A. and J.N. Sheth (1969), The Theory of BuyerBehavior. New York: John Wiley.

Huffman, C. and B.E. Kahn (1998), “Variety for Sale:Mass Customization or Mass Confusion?” Journal ofRetailing, 74, 491–513.

Iyengar, S.S. and M.R. Lepper (2000), “When Choice isDemotivating: Can One Desire Too Much of a GoodThing?” Journal of Personality and Social Psychol-ogy, 79 (6), 995–1006.

Jacoby, J., D.E. Speller, and C.K. Berning (1974a), “BrandChoice Behavior as a Function of Information Load:Replication and Extension,” Journal of ConsumerResearch, 1, 33–42.

____________, D.E. Speller, and C.A. Kohn (1974b),“Brand Choice Behavior as a Function of Informa-tion Load,” Journal of Marketing, 11 (1), 63–39.

____________, G.J. Szybillo, and J. Busato-Schach(1977), “Information Acquisition Behavior in BrandChoice Situations,” Journal of Consumer Research,


3 (4), 209–16.Kahn, B.E. and B. Wansink (2004), “The Influence of

Assortment Structure on Perceived Variety and Con-sumption Quantities,” Journal of Consumer Research,30, 519–33.

Lehmann, D.R. (1998), “Customer Reactions to Variety:Too Much of a Good Thing?” Journal of Academy ofMarketing Science, 26 (1), 62–64.

Lussier, D.A. and R.W. Olshavsky (1979), “Task Com-plexity and Contingent Processing in Brand Choice,”Journal of Consumer Research, 6 (2), 154–65.

Lynch, Jr., J.G. and Th.K. Srull (1982), “Memory andAttentional Factors in Consumer Choice: Conceptsand Research Methods,” Journal of ConsumerResearch, 9 (1), 18–37.

Maheswaran, D., D.M. Mackie, and S. Chaiken (1992),“Brand Name as a Heuristic Cue: The Effects of TaskImportance and Expectancy Confirmation on Con-sumer Judgements,” Journal of Consumer Psychol-ogy, 1 (4), 317–36.

Malhotra, N.K. (1982), “Information Load and ConsumerDecision Making,” Journal of Consumer Research,8, 419–30.

____________ (1984), “Reflections on the InformationOverload Paradigm in Consumer Decision Making,”Journal of Consumer Research, 10, 436–40.

Martignon, L. and U. Hoffrage (1999), “Why Does One-Reason Decision Making Work? A Case Study inEcological Rationality,” in Simple Heuristics ThatMake Us Smart, G. Gigerenzer, P.M. Todd, and theABC Research Group, eds. New York: Oxford Uni-versity Press, 119–40.

Mayer, D.G. and D.G. Butler (1993), “Statistical Valida-tion,” Ecological Modeling, 21–32.

Miller, G.A. (1956), “The Magical Number Seven Plus orMinus Two: Some Limits on our Capacity for Pro-cessing Information,” Psychological Review, 63 (2),81–97.

Milord, J.T. and Ramond R. Perry (1977), “A Method-ological Study of Overload,” Journal of GeneralPsychology, 97, 131–37.

Mitchell, V.W., G. Walsh, and M. Yamin (2005), “Towarda Conceptual Model of Consumer Confusion,”Advanced Consumer Research, 32, 143–50.

Newell, B.R. and A. Bröder (2008), “Cognitive Pro-cesses, Models and Metaphors in Decision Research,”Judgment and Decision Making, 3 (3), 195–204.

Oppewal, H. and K. Koelemeijer (2005), “More Choice isBetter: Effects of Assortment Size and Compositionon Assortment Evaluation,” International Journal ofResearch in Marketing, 22 (1), 45–60.

Payne, J.W. (1976), “Task Complexity and Contingent

Processing in Decision Making: An InformationSearch and Protocol Analysis,” OrganizationalBehavior and Human Performance, 16, 366–87.

____________, J.R. Bettman, and E.J. Johnson (1993),The Adaptive Decision Maker. Cambridge, England:Cambridge University Press.

Rogers, J.L., K.I. Howard, and J.T. Vessey (1993), “UsingSignificance Tests to Evaluate Equivalence BetweenTwo Experimental Groups,” Psychological Bulletin,113 (3), 553–65.

Rose, J.M. and I.R. Black (2006), “Means Matter, ButVariance Matter Too: Decomposing ResponseLatency Influences on Variance Heterogeneity inStated Preference Experiments,” Marketing Letters,17, 295–310.

Russo, J.E. and F. Leclerc (1994), “An Eye-FixationAnalysis of Choice Processes for Consumer Nondu-rables,” Journal of Consumer Research, 21, 274–90.

Schroder, H.M., M.J. Driver, and S. Streufert (1967),Human Information Processing. New York: Holt,Rinehart and Winston.

Shah, A.K. and D.M. Oppenheimer (2008), “HeuristicsMade Easy: An Effort-Reduction Framework,” Psy-chological Bulletin, 134 (2), 207–22.

Shocker, A.D., M. Ben-Akiva, B. Bruno Boccara, and P.Nedungadi (1991), “Consideration Set Influences onConsumer Decision-Making and Choice: Issues,Models, and Suggestions,” Marketing Letters, 2 (3),181–97.

Shugan, S. (1980), “The Cost of Thinking,” Journal ofConsumer Research, 7, 99–111.

Simonson, I. and R.S. Winer (1992), “The Influence ofPurchase Quantity and Display Format on ConsumerPreference for Variety,” Journal of ConsumerResearch, 19, 133–38.

Sujan, M. (1985), “Consumer Knowledge: Effects onEvaluation Strategies Mediating Consumer Judge-ments,” Journal of Consumer Research, 12 (1), 31–46.

Suri, R. and K.B. Monroe (2004), “The Effects of TimeConstraints on Consumers’ Judgments of Prices andProducts,” Journal of Consumer Research, 30 (June),92–104.

Ursic, M.L. and J.G. Helgeson (1990), “The Impact ofChoice Phase and Task Complexity on ConsumerDecision Making,” Journal of Business Research,21, 69–90.

Wellek, S. (2003), Testing Statistical Hypotheses ofEquivalence. London: Chapman & Hall.

Wright, P. (1975), “Consumer Choice Strategies: Simpli-fying vs. Optimizing,” Journal of Marketing Research,12, 60–67.


For further information contact:Svenja Schmidt

Center for Customer InsightUniversity of St. Gallen

Rosenbergstr. 519000 St. Gallen

SwitzerlandE-Mail: [email protected]


PURELY INCIDENTAL NUMERIC PRIMING ON PRODUCTJUDGMENTS: THE EFFECTS OF USING HIGH VERSUS

LOW PRODUCT MODEL NUMBERS


SUMMARY

This research explores whether consumers’ productpreferences can be altered by a high versus a low productmodel number, whereby consumers’ increased accessi-bility of a product model number may exert a directinfluence on their judgments about a product. The prin-ciple of simple numeric priming (Wong and Kwong 2000)may be the key driver eliciting these consequences. How-ever, it has remained largely unexamined in the consumerlandscape. The current research addresses this gap from amotivational perspective in the new product evaluationsand consumption context by focusing on the identifica-tion of an incidental anchoring effect that has yet to bemuch examined in the anchoring (Critcher and Gilovich2008) and consumer psychology literature (Nunes andBoatwright 2004).

Previous research on basic anchoring has found thatsuch anchoring effects are remarkably robust, neitherdepending on anchor values’ informativeness (Tverskyand Kahneman 1974) nor the judge’s motivation (Joyceand Biddle 1981; Nunes and Boatwright 2004), expertise(Northcraf and Neale 1989), and careful attention to theanchor (Brewer and Champman 2002; Wilson et al. 1996).However, recent demonstrations of the impact of sublimi-nal anchors (Adaval and Monroe 2002; Mussweiler andEnglich 2005; Rooijen and Daamen 2006) imply that suchanchoring effects may affect a person’s numerical esti-mates by the peripheral, but not by the central route.Rooijen and Daamen (2006) found that a significantsubliminal anchor effect emerged only among those whowere under time pressure. Critcher and Gilovich (2008)showed that participants who could not recall the anchorvalue were more susceptible to the anchor’s pull thanthose who could correctly recall the anchor value.

These findings suggest the basic numeric anchoringeffect may not be emerged when participants are engagedin more effortful thinking process, because effortful andextensive information processing may wash out the impactof the anchor value (Critcher and Gilovich 2008). On theother hand, those who are not motivated to process targetinformation may still attribute their judgments to theanchor value, because the anchor value may be sufficientto leave a memory trace which is capable of influencingjudgments (Critcher and Gilovich 2008; Nunes andBoatwright 2004). Given this line of reasoning, we hypoth-

esized that the purely numeric anchoring effect on productevaluations will be observed only when people processproduct information heuristically, but not when they pro-cess target information systematically. Two experimentswere conducted to test this hypothesis. Experiment 1examined the influence of motivation on consumers’attitudes toward a product with a high versus a lowproduct model number. We extended this finding byshowing that the same effect persists when the dependentvariable was willingness to buy (experiment 2).

Forty-seven students participated in experiment 1 forpartial course credit. Participants were randomly assignedto one of the four conditions composing the 2 (motivation:high vs. low) × 2 (model number: high vs. low) between-subjects design. We gave participants the cover story thatthey are participating in a new product survey. Drawingon Wang and Lee (2006)’s involvement manipulationmethod, participants in the high involvement conditionwere told that the new product was targeted exclusively tocollege students and would soon be launched in the localmarket; furthermore, they were told that the survey wasconducted among a few selected college student groups toreceive important feedback on the new product before thelaunch. We told participants in the low involvementcondition that the new product was still in the develop-ment stage and that the manufacturer was conducting asurvey on a huge sample to receive preliminary feedback.Participants were then presented with the same descrip-tion of a Dell Studio laptop either identified as Dell Studio81 or Dell Studio 18. Attitudes toward the product wereassessed using three semantic items with the rage of oneto one hundred (1 = bad, negative, and unfavorable; 100 =good, positive, and favorable; α = .965).

The 2 (motivation) × 2 (model number) ANOVA onparticipants’ attitudes toward the product yielded a sig-nificant interaction effect (F(1, 46) = 10.64, p < .005). Aspredicted, follow-up contrasts showed that participants inthe low-motivation condition who saw Dell Studio 81Laptop expressed more favorable attitudes toward theproduct (M = 60.44) than did those who saw Dell Studio18 Laptop (M = 50.30; F(1, 21) = 7.09, p < .02). In thehigh-motivation conditions, however, the hint of using aproduct’s model number as an anchor value was disap-peared and was even reversed. Motivated participantswho saw Dell Studio 81 Laptop expressed less favorableattitudes toward the product (M = 62.96) than did those


who saw Dell Studio 18 Laptop (M = 78.89; F(1, 22) =3.98, p < .06).

Experiment 2 was conducted to test whether thedemonstrated effect of a product model number on unmo-tivated consumers’ product attitudes would generalize tothe more actionable dependent variable: consumers’ will-ingness to purchase. The design and procedure weresimilar to those of experiment 1, but with a differentproduct, a high-speed laser printer. The description of anew high-speed laser printer was accompanied by itsbrand name and model number, either HP-97D or HP11D.Having viewed the HP printer’s features, participantswere asked to indicate their willingness to buy the producton a 7-point scale (1= not at all likely, 7 = very likely).

The 2 (motivation) × 2 (model number) ANOVA onwillingness to buy yielded a marginally significant inter-

action effect (F(1, 92) = 2.84, p < .10). As hypothesized,follow-up contrasts showed that unmotivated partici-pants’ willingness to buy was higher when the modelnumber was 97 (M = 3.12) than was the model number 11(M = 2.34; F(1, 54) = 4.527, p < .04). Consistent with ourprediction, however, when participants were motivated toprocess the information of the target product, the effect ofa high versus a low product model number was notobserved (F(1, 34) < 1, ns).

The results of experiments 1 and 2 demonstrate thata product with a high model number appeals more tounmotivated consumers than a product with a low modelnumber. But consumers processing product informationsystematically in more effortful thinking, their brandpreferences or willingness to buy are not assimilatedtoward its model number. References are available uponrequest.



21 East Market St. S252 PBBIowa City, IA 52242–1000

Phone: 319.335.0967Fax: 319.335.3690



CONCEPTUALIZING CONSUMER ANIMOSITY

Stefan Hoffmann, Technische Universität Dresden, GermanyRobert Mai, Technische Universität Dresden, Germany

Maria Smirnova, Saint Petersburg State University, Russia

SUMMARY

Consumer animosity is the general antipathy of con-sumers toward a particular country (Klein, Ettenson, andMorris 1998). Consumers who display high levels ofanimosity toward a specific country are less likely to buyproducts made in this country. To date, the concept ofconsumer animosity is idiosyncratically depending on theinterplay of the consumers’ home country and a specifictarget country of animosity, however. This has negativeimpacts for both, marketing practice and research. Multi-national companies are interested in analyzing antipathiestoward different countries before entering a new market.Academics need a measurement instrument of consumeranimosity that can widely by applied to conduct cross-cultural research.

Therefore, the paper develops a concept of animositythat can be applied for different sets of home countries andtarget countries. The model relies on Riefler andDiamantopoulos (2007) who proposed to measure ani-mosity via several formatively specified sources of ani-mosity that are related to military-, political-, economicissues or personal experiences. However, in their concep-tion, the set of formative indicators needs to be adapted tothe target country. This study aims at developing a moreflexible model. It is suggested that the influence of numer-ous different causes (e.g., war, nuclear testing, economicexploitation) on animosity is mediated by a small set ofunderlying universal mechanism (e.g., perceived threat).The paper, therefore, suggests a conceptualization ofconsumer animosity that incorporates several universalmechanisms as antecedents.

A pre-test based on 32 participants reveals that themechanisms of animosity are threefold: (a) perceivedthreats (military, politically, economical) toward the re-spondents home country, (b) antithetical political atti-tudes, and (c) negative personal experiences. To validatethe suggested concept, the study investigates animosity of710 Western-European and Eastern-European consumerstoward three target countries. Convergence and discrimant

validity of the suggested construct is established. Theconstruct has explanatory power for product-specific CoO-image, boycotting behavior, and purchase intentions. It isshown that animosity is embedded in the nomologicalnetwork of the related constructs ethnocentrism, patrio-tism, and internationalism.

Thus, this study supports the suggested respecificationanimosity model addressing several key issues. First, toovercome weaknesses of previous research, the proposedconcept of animosity is based on three facets determiningthe general level of animosity. The proposed multi-dimensional structure of consumer animosity can beapplied for different sets of home countries and targetcountries. The paper demonstrates that for some consum-ers negative personal experiences have the strongest impacton animosity toward a certain country whereas otherconsumers rely their general level of animosity mainly onantithetic political attitudes. Third, empirical evidence isgiven that animosity causes a depreciation of the per-ceived quality of products made in the target country, itextenuates purchase intentions, and it fosters the tendencyto boycott products from a certain country. Further researchshould examine the robustness of the proposed model formore home countries and more target countries. Mostimportant, the model needs to be tested in different cul-tural settings.

The findings are of high relevance for practitioners.The suggested concept of animosity is not restricted tostrong salient feelings of animosity, but enables to iden-tify less obvious feelings of animosity toward generallyliked or ambivalent countries. Moreover, the concept caneasily be applied to every home country and every targetcountry. Therefore, multinational companies are welladvised to assess the level of animosity toward their homebase or the communicated country of brand before enter-ing a new country market by means of the suggestedoperationalization. Knowledge about the type and thedegree of animosity enables companies to choose the rightstrategies and tactics (e.g., public relations, foreign brand-ing, etc.) to prevent themselves from resentment.


REFERENCES

Klein, Jill, Richard Ettenson, and Marlene D. Morris(1998), “The Animosity Model of Foreign ProductPurchase: An Empirical Test in the People’s Repub-

lic of China,” Journal of Marketing, 62 (1), 89–100.Riefler, Petra and Adamantios Diamantopoulos (2007),

“Consumer Animosity: A Literature Review and aReconsideration of its Measurement,” InternationalMarketing Review, 24 (1), 87–119.

For further information contact:Stefan Hoffmann

Lehrstuhl für MarketingTechnische Universität Dresden

Fakultät Wirtschaftswissenschaften01062 Dresden

GermanyPhone: +49.351.463.32334



GLOBAL YOUTH AND THEIR RELATIONSHIPS WITH GLOBAL ANDLOCAL BRANDS IN EMERGING MARKETS

Yuliya Strizhakova, Rutgers University – CamdenRobin Coulter, University of Connecticut, Storrs

Linda Price, University of Arizona, Tucson

SUMMARY

Globalization signifies diffusion of market mecha-nisms on a world scale, and young consumers are fre-quently viewed as the most homogenous global consumersegment in this marketplace. For global corporations,global young adults are an attractive target because theyfacilitate entrance of global products and brands into theirlocal markets. For local firms in developed markets,global young adults facilitate globalization of some localproducts and brands by attaching the “coolness” ideologyto them (Kjeldgaard and Askegaard 2006). However,research has not explored the appeal of young consumersto local firms in the emerging markets, and has notexamined how young consumers in emerging marketsnegotiate frequently opposing discourses of the globalworld and their nation-states.

The goal of the present study is to gain a deeperunderstanding of young consumers’ consumption in East-ern Europe (Russia and Ukraine) and effects of globaliza-tion processes on their global identity. First, we evaluatethe global young adult market in terms of consumerethnocentrism, nationalism, and belief in global brands asa passport to global citizenship (BPGC). Our second goalis to assess any differences among young consumers intheir relationships with global and local brands. We focuson varying patterns of consumer involvement with brandsand consumer use of brands as expressions of quality andidentity in relation to both global and local brands.

We consider three belief structures – ethnocentrism,nationalism and BPGC – that are relevant to consumernegotiation of their global and local identities in a com-plex marketplace. As Anholt (2003) and Fong (2004)suggest, some consumers are expected to exhibit highlevels of all three beliefs. However, the complexity of themarketplace may result in other young consumer seg-ments, such as those who only support the nation-state andignore any global influences or those who do not see theiridentity at the global/national levels. Because branding isa very significant ideoscape of globalization (Askegaard2006), consumers who identify stronger with the globalworld are more likely to adopt brands as quality andidentity signals and be involved with brands at both globaland local levels (Strizhakova, Coulter, and Price 2008).

Undergraduate students from universities in far-eastern Russia (n = 250) and eastern Ukraine (n = 250)participated in our study for extra-credit (62% females,M

age = 18.87, SD = 1.46). About 73 percent of participants

reported having internet access either at work or home.Participants completed a pencil-and-paper survey wherethey were first presented with definitions and examples ofglobal and local brands. Next, participants were asked tolist examples of one global and one local brand for fiveproduct categories: TV sets, mineral water, beer, ice-cream and banks. By engaging participants’ recall ofglobal and local brands, we ensured their differentiationof global and local brands. Then, participants respondedto questions about their use of global brands as expres-sions of quality and identity and their involvement withglobal brands and a similar set of questions about localbrands. Finally, they answered questions that measuredtheir ethnocentrism, nationalism, BPGC, and demograph-ics. Because we confirmed configural and metric invari-ance of all measures (AMOS 17.0, Steenkamp andBaumgartner 1998), we were able to analyze our partici-pants as one sample and use three distinct belief measuresas a basis for clustering.

We first conducted hierarchical cluster analysis usingthe average linkage method. We proceeded by running aK-means cluster analysis with three clusters (Wilk’s λ =.19, F = 218.51, p < .001). Univariate ANOVA tests weresignificant (p < .001). The first cluster, the GloballyUninvolved, accounted for about 27 percent of participants(n = 133) and yielded the lowest levels of all three typesof beliefs. The second cluster, the Globally Involved,accounted for about 44 percent of participants (n = 218)and expressed the highest levels of BPGC andethnocentrism with moderate levels of nationalism. Thethird cluster, the Nationally Involved, accounted for theremaining 29 percent of participants (n = 149) andexpressed the highest levels of nationalism with moderatelevels of ethnocentrism and low levels of BPGC.

To identify differences in consumer use and attitudestoward global and local brands, we conducted a MANOVAtest. The overall test was significant (Wilk’s λ = .72, F =11.85, p < .001), and the six univariate tests were signifi-cant (p < .001). Examination of pairwise comparisonsindicated a consistent pattern of results. The Globally


Involved consumers were significantly more likely toexpress stronger relationships with both global and localbrands. Specifically, the Globally Involved were signifi-cantly (p < .001) more likely to use global and local brandsas expressions of quality and identity and were moreinvolved with both global and local brands than either theNationally Involved or the Globally Uninvolved youngadults. The Nationally Involved and the GloballyUninvolved consumers did not significantly differ in theirinvolvement and use of global brands (p > .05). However,the Nationally Involved consumers were significantly(p < .01) more likely to use local brands as expressions ofquality and were more involved with local brands than theGlobally Uninvolved consumers.

Consistent with the view of global corporations, thelargest segment of young consumers (about 44%) in-cluded the Globally Involved young consumers who holdthe strongest levels of consumer ethnocentrism and BPGC

and moderate levels of nationalism. In contrast, about aquarter of participants, the Globally Uninvolved, did notrelate to any of the beliefs, i.e., did not define theiridentities through global/national associations. Marketersand researchers do not seem to know much about theseyoung consumers in emerging markets who currently donot take part in the globalization processes. These indi-viduals reject global influences as much as they rejectnation-state discourses. Finally, slightly less than a thirdof participants, the Nationally Involved, expressed strongnationalistic beliefs while holding moderate levels ofconsumer ethnocentrism and low levels of BPGC. Theseconsumers seem to respond to nation-state discourses andreject any global influences. This latter segment appearsto be particularly appealing to local firms and brands andpresents a challenge to multinational firms entering emerg-ing markets and expecting initial acceptance young adultswho are presumed to have a global brand orientation.

For further information contact:Yuliya StrizhakovaSchool of Business

Rutgers University – Camden227 Penn Street

Camden, NJ 08102–6452Phone: 1.856.225.6922



WHY CONSUMERS BUY COUNTERFEIT SOFTWARE:A SEVEN COUNTRY STUDY

Elfriede Penz, Wirtschaftsuniversität Wien, AustriaBarbara Stöttinger, Wirtschaftsuniversität Wien, Austria

SUMMARY

The rapidly increasing sales of pirated software havebeen a constant headache to the software industry formany years (e.g., Nunes et al. 2004). The hard facts speakfor themselves: China as being among the countries withthe highest software piracy rates in the world at 98 percent(Wang et al. 2005); across Asia, sales of pirated softwareaverage around 50 percent; on a worldwide average, 37percent of all software in use is estimated to be pirated(Kini et al. 2004).

The academic literature has investigated the demanddrivers for pirated software from various angles (for acomprehensive review see Eisend and Schuchert-Güler2006): factors that were investigated as having an effecton the demand for pirated software are, for example, priceand pricing structure (Nunes et al. 2004), the consumers’cultural background (e.g., Husted 2000; Moores 2008;Wang et al. 2005; Yang and Sonmez 2007) or consumerquality perceptions of original vs. fake (e.g., Tom et al.1998; Wee et al. 1995). Some researchers investigatedwhether tolerating software piracy to a certain extent mayeven promote the diffusion of the original (e.g., Givonet al. 1995; Prasad and Mahajan 2003). In addition, a largenumber of studies on pirated software highlight the ethicalaspects of this illegal activity (e.g., Gupta et al. 2004; Kiniet al. 2003; Logsdon et al. 1994; Tan 2002).

Within this paper, we aim to consolidate existingfindings and to develop a comprehensive, yet parsimoni-ous model of the antecedents and drivers of the demandfor pirated software. Based on the Theory of PlannedBehavior, we develop a conceptual model and test it withnon-student samples in a seven-country study for itscross-national applicability to identify demand-orientedapproaches for the cross-national fight against softwarepiracy.

When looking at theoretical underpinnings to explainthe demand for pirated software, three streams of literatureappear helpful. First, counterfeit products would not existif it were not for branded products and what they promise(Bloch et al. 1993; Cordell et al. 1996). Second, prevailingliterature puts the intentional purchase pirated software inthe context of unethical consumer behavior (e.g., Mooresand Chang 2006; Tan 2002; Wagner and Sanders 2001).Often, misbehavior is provoked by certain characteristicsor situational factors such as price, penalty and situation-

specific elements (Dodge et al. 1996). This holdsparticularly true for pirated software which sells at muchlower prices than the original (e.g., Lau 2006; Moores andDhaliwal 2004). Thirdly, given the fragmented knowledgebase in the field, we use the theory of planned behavior(TPB) (Ajzen 1991) for guidance in systematizing existingfindings and adding additional variables, such aspsychographic and demographic determinants.

To increase its explanatory power, we tested ourtheoretical model in seven countries, namely Mexico,Thailand, Ukraine, Slovakia, the U.S., Austria, and Swe-den. Country selection is based on Husted’s (2000) workwho identified three factors as important drivers of thedemand for pirated software (GNP per capita, distributionof income, individualism).

The items were subjected to multi-group confirma-tory factor analyses using AMOS 16 (Arbuckle and Wothke1999). Maximum likelihood estimation was used to esti-mate the parameters. Two models were estimated, onethat included the intention to purchase software at asignificantly cheaper price (Model A) and one that esti-mated antecedents of the intention to purchase software ata slightly cheaper price (Model B) than the originalproduct.

Taken collectively, our analyses show that the Theoryof Planned Behavior is well suited to explain the demandfor pirated software. Both models at the different pricelevels have an acceptable fit across all countries involved.The strongest influence on the intention to buy fakeproducts comes from perceived behavioral control. Thefewer obstacles to purchase pirated software in terms oftime needed to find them, geographic barriers, etc. areperceived, the more likely consumers will intend to buythem. The actual access, however, is not decisive, as mostconsumers know how to get pirated software or knowsomeone who knows someone etc. Against most previousfindings, price consciousness did also not contributestrongly to the intention to buy pirated software. Thesubjective norm plays an important role in the demand forpirated software, particularly when measured through theimmediate social environment’s impact rather than on amore general level. As to the attitudes toward buyingpirated software and their antecedents, findings weremixed and point to national idiosyncrasies. Overall, theirinfluence on purchase intention was limited. Referencesare available upon request.


For further information contact:Elfriede Penz

Institute for International Marketing & ManagementWirtschaftsuniversität Wien

Augasse 2–61090 Vienna

AustriaPhone: +43.1.313.36.5102

Fax: +43.1.313.36.793E-Mail: [email protected]


COMPARING CONSUMER CULTURE PLOTS IN TELEVISIONADVERTISING FROM SOUTH AFRICA AND NIGERIA

Adesegun Oyedele, St Cloud State University

SUMMARY

Zhao and Belk (2008) contend that the findings fromprevious international marketing and advertising researchusing content analysis methodology are inadequate interms of understanding the interplay between the globaland the local in advertising. Zhao and Belk (2008) alsopointed out that the cultural complexity of globalization inadvertising has not been well researched and marketingand advertising researchers do not have a clear under-standing of this issue. They suggested that the explicationof global and local tension in advertising requires the useof rigorous interpretive qualitative technique such assemiotic analysis (Barthes 1977).

In response to the call by Zhao and Belk (2008) formore studies that will help explicate the cultural complexityof globalization in advertising, this current study employsa semiotic approach in addressing the following researchquestions: (1) What are the plots used in promotingconsumer culture in television advertising from SouthAfrica and Nigeria? (2) Are there signals of global andlocal contention in television advertising in South Africaand Nigeria? (3) Are the consumer culture plots in theSouth African television advertisements different fromthe plots revealed in the Nigerian television advertisements?

The role of advertising and global national media inthe promotion of global consumer culture has been high-lighted by researchers across different disciplines includ-ing marketing, advertising and communication disciplines(e.g., Zhao and Belk 2008; Cleveland and Laroche 2007;Tse, Belk, and Zhou 1989; Stearns 2001). Hence, theinvestigation of the plots used in promoting consumerculture in advertising and the investigation of global andlocal contention in advertising will contribute to thetheoretical debate about the cultural complexity of global-ization in advertising. Examples of studies linking adver-tising to global consumer culture include the work of Lin(1993), Tse, Belk, and Zhou’s (1989) and Mueller (1987)just to mention a few.

The methodology used in this study is based on thetheoretical frameworks of semiotics (Mick 1986), thetheories of dramatic performance in advertising and motionpicture narrative structure (Wright 1975) and visual rhetoric(Barthes 1977). These techniques are appropriate whenthe underlying objective is to evaluate the content (e.g.,

visual, sound imagery, settings, story lines, characters) ofthe advertisement beyond the obvious message impliedby the advertisement (Zhao and Belk 2008). A non-probability convenience sampling technique was used incollecting the sampled advertisements from South Africaand Nigeria. The sampled advertisements from SouthAfrica (N = 195) were recorded from the most widelyviewed network in South Africa (South Africa Broadcast-ing Corporation SABC1) in early February 2007 and thetwo weeks in mid-March. In the case of Nigeria, theadvertisements (N=196) were recorded from two popularnational channels (National Television Authority Chan-nel 7 and Channel 5).

The advertisements included in the final sample wereselected based on three criteria. First, the advertisementmust promote a consumer product; all non-consumerproduct ads were excluded from the final sample. Second,the ads must have human characters must be over 10seconds long. The final requirement was the conceptualrichness of the narrative plot. This form of selectivesampling is common in studies using semiotic analysis(e.g., Mick and Oswald 2006).

The initial analysis of the advertisements from SouthAfrica revealed four main plots. A closer look at theseplots and more iteration resulted in the identification oftwo unique plots: (1) Harmonizing loxion culture andWesternized middle-class identity (HLWM) and (2) Usinghumor to negotiate interracial interactions (UHNII). Theinitial analysis of the advertisements from Nigeria revealedseven main plots. However, a closer diagnosis of theseinitial plots and more iteration resulted in the identifica-tion of three unique plots: (1) Male breadwinner (MBW)(2) The Big man plot (TBM) and (3) Hedonism andMasqueradization of Western Ideals (HMWI).

A comparative evaluation of the plots from SouthAfrica and Nigeria revealed some interesting insights.First, unlike the Nigerian advertisements where the ele-ments and signals of the local promotes genderism ideol-ogy or inequality between genders, the signals and ele-ments of the local in the South African advertisementspromotes equality themes. For instance, the UHNII plotrevealed in the South African advertisements promotesracial equality while the MBW and BM plots revealed inthe Nigerian advertisements promote gender inequality.


Overall, the findings from this study suggest that thecontention between the local and the global is significantin emerging markets where the people have strong nation-alistic sentiments. Additionally, the study also revealed

that advertisements reflect a romanticized view of middleground politics promoted by political establishments bothat the national and the international level. References areavailable upon request.

For further information contact:Adesegun Oyedele

Department of Marketing and Business LawG.R. Herberger College of Business

St Cloud State University720 4th Ave South

St Cloud, MN 56301Phone: 320.308.3006



EXCITING RED AND COMPETENT BLUE: LINKING COLOR TOBRAND PERSONALITY

Lauren I. Labrecque, University of Massachusetts AmherstGeorge R. Milne, University of Massachusetts Amherst

SUMMARY

It has been five decades since marketers began toinvestigate the symbolic nature of brands (Levy 1959),thus giving rise to concept of brand personality andspawning extensive interest in this subject (e.g., Aaker1997; Aaker, Benet-Martinez, and Berrocal 2001; Plummer1984). Consumer researchers have devoted their efforts todeveloping reliable scales (Aaker 1997) investigatinghow brand personality encourages self-expression andassociation (e.g., BeIk 1988; Kleine, Kleine, and Kernan1993; Malhotra 1981) and have also empirically tested thebenefits and consequences of brand personality (Batraet al. 1993; Freling and Forbes 2005).

Despite the plethora of research on the topic of brandpersonality, there is little research that examines theantecendents of brand personality. While the literaturesuggests that brand personality can be shaped throughmany factors including brand name, product attributes,advertising, logo, and package design (Batra et al. 1993)empirical work on the matter is scarce (with the exceptionof Orth and Malkewitz 2008 and Grohmann 2009). Poten-tial factors that may influence brand personality such asnaming, distribution channels, and pricing strategies haveyet to be examined; additionally the influence of sensoryelements such as music, color, and scent remain virtuallyunexplored.

In this paper, we use experimental design to explorethe relationship between color and brand personality.Based on empirical research in the fields of psychology(Clarke and Costall 2007; Murray and Deabler 1957;Valdez and Mehrabian 1994; Wexner 1954) and market-ing (Bellizzi, Crowley, and Hite 1983; Crowley 1993;Gorn et al. 1997; Gorn et al. 2004), along with colorpsychology texts (Fraser and Banks 2004; Mahnke 1996;Wright 1988) hypotheses that link different hues to brandpersonality are developed. Our experiment examines theeleven hues (red, orange, yellow, green, blue, purple,pink, black, white, brown, and gray) that comprise the setof universal basic colors (Berlin and Kay 1969).

Variables and Procedure

A set of two fictitious logos adapted from previousresearch (Henderson and Cote 1998), were created withAdobe Illustrator CS4™ with the assistance of a profes-

sional designer. Eleven different color versions of eachlogo were created. Only hue was manipulated (e.g., redversus blue), keeping levels of saturation and value con-stant. Consistent with other research examining hue effectswith computers, levels of saturation were kept high [220out of 240 in the Hue Saturation Lightness (HSL)colorspace], while levels of lightness (value) were keptmidrange (120 out of 240 in the HSL colorspace) (Mehtaand Zhu 2009). Black, white, gray, and brown wereexceptions; white is a color with full value, black is a colorwith zero value, middle gray is a color with mid-level ofvalue, and brown is an orange with mid-saturation andlow value. We included white, black, brown, and gray inorder to cover the spectrum of Berlin and Kay’s (1969)universal colors.

Two hundred and seventy-nine undergraduatestudents participated for extra credit in an introductorymarketing course. The experiments were conducted in acomputer lab containing identical machines, lighting, andother atmospheric elements. Computer monitors wereprofessionally calibrated before the experiments to ensureequivalence among all screens and a uniform experienceacross the sample. Participants were told that we wereinterested in finding out which personality traits or humancharacteristics come to mind when you see a brand logoand read a paragraph describing brand personality. Thiswas the same description used by Aaker in her scaledevelopment study (1997). Participants were thenpresented with a logo and were asked to rate it usingAaker’s (1997) 42 item brand personality scale.Demographic information including age, gender, culturalbackground, and colorblindness were collected for theanalysis. Color preferences assessing participants’ threemost liked and three most disliked colors were alsocollected. In addition, participants provided ratings onlogo likability and familiarity.

Analysis and Results

Each logo was then analyzed separately. A multivari-ate regression was used to evaluate the five brand person-ality dimensions as the dependent variables, followed byindividual univariate regressions for each personalitydimension. Ten colors served as the predictor variables,with likability, familiarity, gender, and least/most favoritecolors as covariates.


Overall support was found for the majority of thehypotheses. In general, the empirical findings for thehypothesis tests at the dimension level and further explo-ration at the facet level provide evidence of a link betweenbrand personality and color; thus, further investigationinto this relationship is merited.

This work represents the first step in examining therelationship between color and brand personality. Thisstudy examined hue only and did not address the influenceof the other components of color, saturation and value, nordid it address other design elements such as logo shape.References are available upon request.

For further information contact:Lauren I. Labrecque

University of Massachusetts Amherst121 Presidents Dr.

Amherst, MA 01003Phone: 413.455.6783


George R. MilneMarketing

University of Massachusetts Amherst121 Presidents Dr.

Amherst, MA 01003Phone: 413.545.5669



THE IMPACT OF COLOR TRAITS ON CORPORATE BRANDING

Wei-Lun Chang, Tamkang University, TaiwanXia-Liang Lin, Tamkang University, Taiwan

SUMMARY

Like human beings who have specific characteristics,cultures, and philosophies, enterprises have symbols andsigns that deliver their corporate identities and representtheir corporate images. Napoles (1988) indicated thatcorporate image is how the mass population sees a com-pany and that a company aims for consumers to recognizeits corporate identity as a symbol for the firm (Dowing1994). Corporate identity and brand identity utilize asystematic approach to give meanings to symbols andvisualized concepts, although the starting points of corpo-rate identity and brand identity are different. Corporateidentity considers the enterprise perspective and empha-sizes the performance and management of the company,while brand identity takes into account the customerperspective in order to specify how customers perceivethe functions of the brand.

Ideally, brand identity adds value to a firm’s productsand services. It is the sum of all tangible and intangiblecharacteristics of the products and services offered by afirm and is identified by certain essential product labels,such as names, brand symbols, and colors. Color is aparticularly effective way to manifest brand identity. Anybrand can use a color as the focus of its brand identity,making it the major focus of brand identity by utilizing asingle distinctive color or color palette (Schmitt andSimonson 1997). If a color is consistently used across avariety of identity elements, it becomes part of a company’ssignature: Kodak’s yellow, IBM’s dark blue, Coca-Cola’sred, and Acer’s green.

More than the symbol of a brand, color is a source ofinformation. A study by the Institute for Color Researchrevealed that people make a subconscious judgment aboutan item within 90 seconds of initial viewing and that 62–90 percent of that assessment is based on color alone.Marc Gobé (2004) observed that color is powerful andrepresents emotions and so not only provides the skeletonof brand identity but also the traits of the brand personal-ity. In the perceptions of human beings, visualization isthe most common way to appeal to consumers’ senseorgans; in fact, Aristotle indicated that all recognition istriggered by vision, and Lindstrom (2005) specified thatvision is primary among all human sense organs. Thus, a

visual symbol of vision has the capability to draw con-sumers’ attention.

This research utilizes Birren’s color theory (1961) toidentify the characteristics of colors and interpret the traitsof brands and the colors they use in terms of perceptualstimulus and psychological recognition, and Aaker’sHoneycomb model to interpret the customers’ percep-tions of a firm’s brand identity. The purpose of the studyis to identify whether the traits of corporate identityindicated by the colors used match the key conceptsdelivered by the enterprise into customer perceptions. Thepositions, key concepts, and intents of the enterprises willbe manifested if they use appropriate colors.

This study showed that color traits can deliver themessages of brand identity and build an emotional con-nection between consumers and brands. In particular,consumers first react to a brand based on their perceptionof the traits associated with a color. Corporate identityresults in corporate images and associations, such as thecheerful and lively Ronald McDonald representingMcDonald’s. Appropriate colors can deliver accuratemessages and visions of the enterprise and its identity.While corporate identity represents the external imageand reputation that the company attempts to deliver, colordelivers its own traits and helps consumers to interpret themeaning of a brand by communicating the corporateidentity.

Corporate branding can also attract consumers byusing color traits that stimulate the memories of brands.Hence, this work provides preliminary evidence that colortraits affect brand identity and corporate identity and thatthe impact of color on corporate branding is both signifi-cant and essential to enterprises. The connection betweencorporate/brand identity and corporate branding is alsobuilt to demonstrate that color traits may affect corporatebranding. This study provides a preliminary analysisbased on color traits and clues to help companies obtain abetter understanding of corporate branding. Finally, wehope that other cases across different industries (insteadof eight brands) can be examined for future study in orderto build a comprehensive understanding of the impact ofcolors traits on corporate branding.


For further information contact:Wei-Lun Chang

Department of Business AdministrationTamkang University

No. 151, Ying-chuan RoadTamsui, Taipei County

Taiwan (251)Phone: 886.2.26215656, Ext. 3148



WHEN LESS IS MORE: THE EFFECTS OF PRODUCT QUANTITY ONCONSUMER BEHAVIOR

Stacey G. Robinson, Florida State University, Tallahassee

SUMMARY

Consumers often do not make final purchase deci-sions until they are physically in the store (Court et al.2009). As a result, in-store merchandising strategies, suchas product packaging, shelf placement, and inventorycontrol are increasingly influential and essential sellingfactors affecting consumer choice (Court et al. 2009;AMA 1982). Correspondingly, thorough understandingof the process by which consumers acquire prepurchaseinformation, ultimately affecting purchase decisions, hasbeen referred to as an enduring interest of academics andpractitioners alike (Schmidt and Spreng 1996).

While a number of prior studies focus on the effectsof product placement, high inventory levels, and productavailability (e.g., stockouts, soldouts) on consumer per-ceptions and purchase decisions (Balakrishnan et al. 2004;Campo et al. 2000; Ge, Messinger, and Li 2009; Koschat2008), research that specifically evaluates the effects oflow inventory levels on consumer prepurchase assess-ments and buying behavior is not available. This is acritical gap in the literature, because low inventory levelsare a reality for many retailers. Consumer purchases ofpopular products reduce available quantities and lowerinventory levels throughout the day.

In contrast to established beliefs, the present researchsuggests lower inventory levels may serve as a positivecue to potential consumers. We argue that under certainconditions, low inventory sends a signal that an item ispopular or recommended by other consumers, whichincreases expected quality perceptions and subsequentpurchase intentions. Two independent experimental stud-ies reveal that in the absence of other information, lowlevels of inventory, compared with higher levels of acompeting brand, can increase perceptions of quality andpurchase intentions.

It appears as though low inventory levels provideconsumers with behavioral evidence of choices made byother consumers, thus serving as a form of endorsement.This is an important finding because it provides an ironicupside to waning inventory levels that may occur as afunction of strong consumer demand for a particularproduct or a retailer’s limitations that may preclude quickinventory replenishment. Retailers faced with resourcelimitations, both human and inventory, may find consola-tion in this finding. The present research suggests theselimitations may not be detrimental and may even improvesales under certain conditions. For researchers, this workprovides additional insight into noninteractive social influ-ence and how seemingly unintentional observable envi-ronmental cues may be utilized in consumer decisionprocesses.

For further information contact:Stacey G. Robinson

The College of BusinessFlorida State University

Rovetta Business Annex, Room 307P.O. Box 3061110

Tallahassee, FL 32306–1110E-Mail: [email protected]


THE NEXT GENERATION OF BUSINESS LEADERS: INFLUENCES ONUNETHICAL STANDARDS ACROSS MULTIPLE CULTURES

Victoria L. Crittenden, Boston College, Chestnut HillRichard Hanna, Northeastern University, BostonRobert A. Peterson, University of Texas, Austin

William F. Crittenden, Northeastern University, Boston

SUMMARY

Reports of scandalous behavior have long dominatedthe popular press, with current business pages reportingeverything from an increase in earnings restatements tofraudulent business behavior. In an increasingly intercon-nected world, the fragility of the global marketplace hasbeen evidenced time and time again. From the Enron andWorldCom scandals to the Wall Street banking scandalsto the United States government bailout of American automanufacturers, today’s future business leaders are con-fronted early in their academic careers with history-making events that have had a profound impact on theircountry’s economic system. These business students arematuring in a digital age in which they do not have to readabout historical scandals; rather, today’s business schoolstudents are exposed to behaviors as they unfold and, assuch, may be developing professionally by modeling thebehaviors of both ethical and unethical business leaders.

The actions exhibited by too many managers todaymay not be those which we would like our future businessleaders to be mimicking. The results from a 2004 CMOMagazine poll to learn the opinions of marketing officerswith respect to the impact of ethics on the marketingfunction highlighted marketers’ specific ethical concerns.The top three ethical concerns facing U.S. businesses,according to the marketing executive surveyed, were:improper accounting practices (48%), conflicts of interest(42%), and deceptive sales and marketing practices (42%).Interestingly, 31 percent of these marketing executiveswere not confident that companies were taking appropri-ate actions to curtail scandalous behavior.

Future business leaders from across the globe willultimately be exposed to, and influenced by, the activitiesin their work environments. Hence, we propose SocialLearning Theory as a theoretical foundation for under-standing the ethical standards of future business leadersand to pursue explanatory constructs for unethical stan-dards worldwide. Social Learning Theory explains hu-man behavior in terms of continuous reciprocal interac-tion between cognitive, behavioral, and environmentalinfluences. Recognition of this reciprocal interdepen-dency is critical for understanding the potentially conflict-

ing signals that our future business leaders receive as theydevelop their own standards for unethical decision mak-ing. Additionally, moral philosophies and deviant behav-ior are included in the analysis.

Using data drawn from students from 115 four-yearundergraduate institutions in 36 different countries, therelationships among role models, capitalism, and lawswere examined using multiple linear regression analysis.The data suggest that future business leaders educated inenvironments espousing all moral philosophies exceptvirtue ethics are influenced by their role models. That is,students educated under cultural norms of egoism, for-malism, relativism, and utilitarianism who perceive thatmanagers engage in unethical acts are significantly morelikely to compromise their ethics. Also, students whoperceive that the ends justify the means in business aremore likely to compromise their ethics, and this is consis-tent across all moral philosophies. Thus, anomie theorydoes not necessarily hold true with respect to role model-ing as an influence since deviant behavior was foundacross all societies.

However, differences in deviant behavior are foundas related to the social influences of capitalism and laws.Capitalism as a social influence varied across the moralphilosophies. For example, students educated in the for-malism and virtue ethics societies who believe in therelationship between efficient management and capital-ism are more likely to compromise their ethics. However,within the capitalism construct, profit entitlement due torisk assumption is negatively related to the need to com-promise ethics in all moral philosophies (although onlysignificantly so within the egoist group). Additionally,students of egoist, formalist, and utilitarianism societiesare more likely to compromise their ethics if they perceivethat anything that is legal is ethical. Thus, the moralcompass for these societies appears to be the legal domain.

In conclusion, it appears that students’ likelihood tocompromise their ethics is consistently predicted by thebehavior of their role models. As well, especially in egoistsocieties (which includes the USA), using what is legal asa barometer for what is ethical plays a major role. Interest-ingly, beliefs about capitalism only have influence in


western societies and perhaps not in ways one wouldexpect. Thus, it would appear that capitalism by itself

cannot be blamed for future business leaders’ potential forunethical behavior.

For further information contact:Victoria L. Crittenden

Boston College140 Commonwealth Avenue

Chestnut Hill, MA 02467Phone: 617.552.0430



HOW DOES ONLINE WORD-OF-MOUTH INFLUENCE STUDENTS’COURSE SELECTIONS AND THEIR ATTITUDES

TOWARD PROFESSORS?

Cong Li, University of Miami, Coral GablesXiuli Wang, Peking University, China

SUMMARY

As an important communication venue, Word-of-Mouth (WOM) affects individuals’ decision-making inmany situations. With the development of Web 2.0, onlineWOM has become the focus of a growing body of research.Many prior studies have shown that valence and volumeare two important factors influencing the effects of onlineWOM, where valence represents the nature of informa-tion as either positive or negative and volume refers to thetotal amount of messages. However, prior research doesnot provide a clear and consistent conclusion on howWOM valence interacts with its volume. To fill thisknowledge gap, two studies are conducted in the currentresearch. A special industry, higher education, is selectedfor investigating online WOM effects.

Since more and more students go online and rate theirprofessors today, Study 1 is designed to be a field experi-ment. Randomly sampled students’ numerical ratings andwritten comments about their professors onratemyprofessor.com are recorded and content analyzed.The enrollment percentage for each of the sampled pro-fessors’ courses in the 2009 spring semester is alsorecorded. Through a mediation analysis, it is found thatthe relationship between perceived professor quality andactual course enrollment is fully mediated by onlineWOM valence. More specifically, “good-quality” profes-sors tend to have higher course enrollment percentagesbecause they receive more positive online student com-

ments. On the other hand, “bad-quality” professors receivemore negative online student comments, which lead tolower course enrollment percentages.

Built on the findings of Study 1, Study 2 is conductedto explore the interaction effects between WOM valenceand its volume. It is designed to be a full factorial 2x2 labexperiment. The two factors manipulated are WOMvalence (positive/negative) and WOM volume (high/low). A total of 80 undergraduate students participates inthe experiment. The study results replicate what is foundin Study 1. When the online comments about a professorare predominately positive, students tend to have morefavorite attitudes toward him and are more likely to takehis course. In contrast, when the online comments aboutthe professor are mostly negative, students’ attitudestoward him tend to be less favorable, and they are lesswilling to take his course too. Moreover, such effects ofWOM valence are found to be moderated by its volume.More specifically, when the volume is high, the WOMvalence effects get even stronger. However, when thevolume is low, the WOM valence effects become weaker.

Incorporating both findings in Study 1 and Study 2,an online WOM model is depicted in the current research.Both WOM valence and volume are illustrated to besignificant predictors of students’ attitudes toward theirprofessors and their course selection decisions. OnlineWOM is then argued to be an unneglectable force thathigher educators need to pay more attention to.

For further information contact:Cong Li

School of CommunicationUniversity of Miami

Coral Gables, FL 33124Phone: 305.284.2355



HOW DO INSTITUTIONAL PILLARS MODERATE THEEFFECTIVENESS OF FIRMS’ CUSTOMER-BASEDMARKETING CAPABILITY ACROSS COUNTRIES?

Fernando Angulo, Autonomous University of Barcelona, SpainDaniel Bello, Georgia State University, Atlanta

Diego Prior, Autonomous University of Barcelona, SpainJosep Rialp, Autonomous University of Barcelona, Spain

SUMMARY

Marketing accountability is a key area of concern formarketing academia and practitioners, as demonstratedby the Marketing Science Institute 2008–2010 researchpriorities (MSI 2008, p. 2–3). One of the key topics withinmarketing accountability is the financial contribution ofmarketing capabilities (e.g., Dutta, Narasimhan, and Rajiv1999). Interestingly, this topic has received much moreattention in domestic than in international research so far.Additionally, for managers to assess the impact ofmarketing in an international arena, they need measurementsystems that link intangible measures of marketing (suchas marketing capabilities) to the financial measures usedby CEOs and CFOs (MSI 2006). In an attempt to developsuch systems in an international arena, the role of theinternational context becomes critical in that there aremany country factors that can dilute the effect of managerialvariables (such as capabilities) on performance. Amongthose country factors, the literature underlines theimportance of countries’ institutional variables. To studythe role of institutional variables in, for example, thecapability-performance relationship and to developcomparative marketing studies across countries, someresearchers have suggested the use of the institutionaltheory (Iyer 1997). Though institutional theory has beenemployed to study different international phenomena,there is no evidence so far about to what extent institutionalvariables moderate the marketing capability-financialperformance relationship.

In this sense, the aim of this article is threefold: (1) tounderstand whether the impact of marketing capability onfinancial performance is significant beyond the effects offirms, industry, macroeconomic and institutional vari-ables; (2) to study the direct impact of countries’ institu-tional variables on firms’ financial performance; and (3)to assess whether the countries’ institutional variablessignificantly moderate the marketing capability-financialperformance link.

Methodologically, we have the following groups ofindependent variables. (a) Managerial: we employ mar-keting capability in terms of a customer oriented capabil-

ity and refer to customer-based marketing capability(CMC) as the superior and distinctive process to trans-form marketing investments into brand equity and cus-tomer satisfaction by developing, nurturing and leverag-ing relationships with and perceptions held by customers.We employ an intertemporal output-oriented data envel-opment analysis (DEA) Bootstrap for estimating CMC.(b) Firm: we use size which is the log of annual total assets.(c) Industry: dummy variables of telekom and bankingindustries are used. (d) Country: we employ macroeco-nomic variables such as GDP, inflation and unemploy-ment growth; and institutional variables such as norma-tive elements (the metric of country’s management prac-tices and attitudes and values), cognitive elements (thecountries’ scores of Hofstede masculinity, individualism,power distance, and uncertainty), and regulative elements(the tax over profits firms should pay in a country annu-ally, time of enforcement a contract at country level, andthe strength of legal rights in a country).

Regarding the dependent variable, we follow thefinancial literature (Damodaran 2002; Palepu, Healy, andBernard 2000) and employ the following performancemeasures: revenues, operating income, income beforetaxes, and net income. The model for estimation is asfollows: Performance

it as a function of Managerial

it,

Firmit, Industry

j, Macroeconomic

ct, and Institutional

c. As

it can be observed, we work with variables in annualvariation and not in levels

(to control for firm-specific

information not modeled). The variables regarding finan-cial, marketing, institutional, and other data cover eightconsecutive years, 2000–2007. To avoid potential prob-lems of serial autocorrelation and heteroskedasticity, weestimate the model using generalized least squares cor-recting estimations for autocorrelation and hetero-skedasticity (Mizik and Jacobson 2008).

We focus this study on firms from U.S., Denmark,Finland, Norway, and Sweden. These countries providean interesting picture about the differences in the institu-tional elements. For instance, Nordic countries exhibitsbetter scores in management practices and attitudes andvalues than U.S. Firms in Denmark should pay almost 33percent of taxes on profits, U.S. firms should pay on


average 44 percent, and Swedish firms should pay almost55 percent.

The key findings of this study are (1) the growth offirms’ CMC significantly and consistently creates growthof the different measures of firms’ operational perfor-mance beyond the effect of country variables, (2) norma-tive, cognitive and regulative institutional elements havea significant effect only on firms’ revenue growth, (3) theeffect of the firms’ CMC growth on revenue growth islarger in countries with more tax imposition, and finally(4) the effect of firms’ CMC growth on net income growthis superior in countries with less appropriate norms and inmore hierarchical and individualist cultures.

The findings interestingly reveal that more regulativeenvironments (e.g., tax on profits) provide benefits tofirms in terms of revenue growth and moderate the capa-bility-revenue relationship. We expected this result argu-ing that more regulation expands aggregate demand whichin turn influences revenue growth. It is important thatpractitioners see a more regulative environment as apotential for growing but they should also be cautious

with this idea because we developed this study on asample of developed economies (U.S., Denmark, Fin-land, Norway, and Sweden) and we do not know whetherthe results will be the same for developing economies.

Another interesting and surprising finding is thatmarketing capability-net income relationship is larger incountries with less normative orientation (e.g., less ethicalpractices and social responsibility). We expected thatcapability-net income link would be greater in countrieswith more normative orientation as the institutional theorypurports. It is important to be precise that data employedfor measuring management practices is an index thatrange from 0 to 100. However, we should caution that allcountries in our sample were above the median (index of50) and therefore had overall good ethical practices.Therefore, further research is needed in order to under-stand the marketing capability-performance relationshipin countries with index of ethical practices below themedian since it is possible that this relationship might bereversed as the index decreases. References are availableupon request.

ACKNOWLEDGMENT

The authors thank the financial support of the Commis-sioner for Research and Universities of the CatalanMinistry of Innovation, Universities and Enterpriseand the European Social Fund. The authors also

acknowledge the financial support of the SpanishMinistry of Science and Education (projects:SEJ2007-60995/ECON and SEJ2007-67895-C04-02). The authors also appreciate the constructivecomments of AMA reviewers.

For further information contact:Fernando Angulo

Departament d’Economia de l’EmpresaAutonomous University of BarcelonaEdifici B, Campus UAB, Bellaterra

Barcelona, 08193Spain

Phone: +34.93.581.4416Fax: +34.93.581.2555



IMPLEMENTATION OF SUBSIDIARY MARKETING STRATEGIES: ANEMPIRICAL INVESTIGATION OF THE ACTORS, MODALITIES,DRIVERS, PARAMETERS, AND PERFORMANCE OUTCOMES

Esra F. Gencturk, Ozyegin University, TurkeyDestan Kandemir, Bilkent University, Turkey

SUMMARY

Despite extensive research on international market-ing strategy, relatively little attention has been directed towhat is needed to implement the intended or adoptedstrategy. In fact, Birnik and Bowman (2007) refer toexecution as the missing link that should be explored toadvance our understanding. In an attempt to redress thissituation, this study develops and empirically tests agrounded model of subsidiary marketing strategy imple-mentation (SMSI) in the context of U.S. based multina-tional corporations (MNCs).

The following aspects of strategy implementation areconsidered to delineate its scope. Strategy Implementa-tion Actors: Notwithstanding subsidiary’s direct formalresponsibility for carrying out its marketing strategies, weadopt the upper-echelons framework (Hambrick andMason 1984) and consider the headquarter perspective toinvestigate HQ managers’ role in SMSI. For the StrategyImplementation Modality, we focus on deployment ofresources in terms of managerial time and effort expanded.The Strategy Implementation Conditions are identifiedbased on the resource dependence perspective, whichsuggests that the need to access knowledge from thesubsidiary would be the impetus for headquarter involve-ment in SMSI in order to protect its interest as well as tomaintain its supremacy (Ambos and Schlegelmilch 2007).Consequently, the conditions considered are: attitudinalinvolvement of HQ managers in SMSI, internationalexperience of HQ, customer as well as host-countrymarket diversity. For Strategy Implementation Param-eters, we focus on HQ managers’ resource deployment interms of the marketing tasks because the HQ managershave been observed to relate to the 4P framework incalibrating the managerial resources they allocate to theintended strategy. The Strategy Implementation Perfor-mance was assessed based on Ruekert, Walker, andRoering’s (1985) multidimensional performance con-struct where satisfaction with subsidiary market perfor-mance (MP) and profitability were used to determine theeffectiveness and efficiency of SMSI, respectively.

Data Collection and Psychometric Considerations

The hypothesized model (Figure 1) was tested withsurvey data obtained from 252 informants representing

227 divisions at 163 U.S. manufacturing firms that agreedto participate in this study, for a response rate of 61percent. The psychometric properties of the measureswere assessed with a confirmatory factor analysis (CFA)(Bentler 1995). Although the chi-square test was statisti-cally significant (χ2

(188) = 258.23, p < .05), the model

provided a satisfactory fit to the data (Bentler-Bonettnonnormed fit index [NNFI] = .96; comparative fit index[CFI] = .97; Bollen’s fit index [IFI] = .97; root meansquare error of approximation [RMSEA] = .053), indicat-ing the unidimensionality of the measures (Anderson andGerbing 1988). Furthermore, measures demonstrated ad-equate convergent validity and reliability with statisti-cally significant (p < .01) factor loadings, and the compos-ite reliabilities exceeding the threshold value of .70(Nunnally 1978). With average variance extracted by themeasure of each factor exceeding the squared correlationof that factor’s measure with all measures of other factors,strong discriminant validity was also observed for all thefactors considered.

Hypothesis Tests and Results

The hypothesized model was examined by SEM. Thechi-square test was statistically significant (χ2

(229) = 387.88,

p < .05), however, the fit indices [NNFI = .91; CFI = .92;IFI = .92; RMSEA = .073] indicated a good fit with thestructural model. The results reveal that HQ managers’attitudinal involvement has a significant positive effect onHQ resources allocated to product (β = .27; p ≤ .005),pricing (β = .19; p ≤ .05) and distribution activities (β =.13; p ≤ .10). In contrast, HQ’s international experiencewas found to be significantly and negatively associatedwith its resources allocated to product (β = -.26; p ≤ .005),pricing (β = -.26; p ≤ .005) and distribution activities (β =-.26; p ≤ .005). Higher degrees of customer diversity havesignificant positive effects on HQ resources allocated toproduct (β = .22; p ≤ .05), pricing (β = .21; p ≤ .05) anddistribution activities (β = .28; p ≤ .005). Host-countrydiversity was found to be positively and significantlyassociated with HQ resources allocated to both product(β = .16; p ≤ .05) and distribution activities (β = .20; p ≤.05). However, it was not found to be related to HQresources allocated to pricing activities (β = .06; p > .10).We found that HQ resources allocated to product activi-ties had significant positive effects on satisfaction withsubsidiary MP (β = .36; p ≤ .005) as well as its profitability


(β = .36; p ≤ .005). In contrast, HQ resources allocated topricing activities had significant negative effects on bothsatisfaction with MP (β = -.23; p ≤ .05) and profitability(β = -.17; p ≤ .05). Finally, HQ resources allocated todistribution activities were found to have a significantpositive impact on satisfaction with MP (β = .13; p ≤ .10) butnot significantly associated with profitability (β = -.06;p > .10).

Discussions and Conclusions

To bring strategy implementation to the forefront ofthe international marketing strategy research, we under-took a theoretically grounded empirical investigation ofSMSI where its nature, drivers, and consequences wereexamined within the context of U.S. based MNCs. Ingeneral, our empirical results support the upper-echelonstheory in the strategy literature (Hambrick and Mason1984) and reveals that managers at the strategic apex ofthe MNC are involved in SMSI and their involvementdoes influence the subsidiary performance. Furthermore,the overwhelming empirical support found for our hypoth-eses signify the relevance and importance of the strategyimplementation drivers considered in this study based on

Satisfaction with

Market Performance

Profitability

Distribution

Activities

Pricing Activities

Product Activities

Attitudinal Involvement

International Experience

Customer Diversity

Host Market Diversity

Subsidiary Marketing Strategy Implementation

Denotes significant paths.

Denotes non-significant paths.

FIGURE 1Antecedents and Performance Consequences of SMSI

the implications of resource dependency theory. In addi-tion, this study offers several recommendations to HQ andsubsidiary managers. In particular, the results reveal thatHQs interested in increasing their involvement in SMSIshould focus first on subsidiaries with diverse customerbase and those operating in diverse host countries becauseof their possession of country-specific knowledge. On theother hand, the depth and breadth of international experi-ence accumulated through appear to diminish the need forHQ involvement in SMSI. Our findings also validate thepresumed direct relationship between HQ involvement inSMSI and subsidiary performance. Specifically, HQ in-volvement in subsidiary product strategy appears to payoff in terms of both HQ satisfaction with subsidiary MP aswell as subsidiary’s financial performance. On the otherhand, HQ managers must be very cautious when consider-ing involvement in subsidiary pricing strategy since antici-pated performance payoffs may be not realized. Theseresults show that simplistic or ad hoc allocation of HQmanagerial time and effort to SMSI without careful exami-nation of the country-specific knowledge resources of thesubsidiary is not advised. References are available uponrequest.


For further information contact:Esra F. Gençtürk

Ozyegin UniversityUskudar, Istanbul 34662

TurkeyPhone: +90.216.559.2368



PERCEIVED BARRIERS FOR INTERNATIONALIZATION OF SMALLFIRMS: PERCEIVED LACK OF RESOURCES

Yu Henry Xie, College of CharlestonTaewon Suh, Texas State University, San Marcos

SUMMARY

Internationalization is a major strategic decision forsmall- and medium-sized enterprises (SMEs), because afirm must make an important strategic decision on whetherto internationalize or not in the global marketplace. How-ever, there are still a large number of SMEs that “do notdare cross national boundaries to sell their products andservices” (Leonidou 2004, p. 279). As Gilmore and Carson(1999) suggest, some small firms are constrained in finance,management, and market knowledge in their internation-alization, in comparison with larger multinational corpo-rations (MNCs). The behavioral modes of international-ization processes have stressed the important role ofperceived problems and chances in the firm’s internation-alization (Eriksson et al. 1997). The extant literature onsmall firms’ internationalization only provides a partialexamination of barriers to internationalization (Leonidou2004). Few studies have examined the perceived barriersto internationalization regarding firm resources. There-fore, the perceived lack of resources is the focal attentionof this current study.

This study adopts the resource-based view (Barney1991) to frame the resources that affect internationaliza-tion decisions. In addition, this study also integratestechnology acceptance model (Davis 1989) to examinethe impact of perception and behavioral intention ofowner-managers of small firms. Firm strategies are devel-

oped in accordance with firm-specific attributes to iden-tify, protect, and exploit their unique skills and propri-etary assets (Tallman 1991). In addition, the internation-alization models are also behaviorally-oriented. Percep-tions actually influence consequential behaviors. Someperceptions can become barriers to small firms’ interna-tionalization. Therefore, perception of firm resources byowner-management of small firms plays an important rolein identifying and deploying heterogeneous resources intheir internationalization decisions. Thus, we argue thatthe perceived lack of resources acts as a perceived barrier,thus negatively affecting internationalization of smallfirms.

This study tests a model to identify small firms’perceptual barriers to internationalization. The results ofthis study support the hypotheses that the perceived lackof general and specific experiential knowledge in interna-tional market hinders small firms’ intention to engage ininternationalization. Knowledge-based resources are per-ceived as a critical factor for both the internationalizedand non-internationalized smaller firms. On one hand,these resources may help some small firms to internation-alize in their early stages. On the other hand, some othersmall firms may be scared way from internationalizationdue to their perceptual barriers regarding the knowledge-based resources. We also discuss direction for futureresearch. References are available upon request.

For further information contact:Yu Henry Xie

School of BusinessCollege of Charleston

66 George StreetCharleston, SC 29424Phone: 843.953.6658



WHEN DOES INTERNATIONAL MARKETING STANDARDIZATIONMATTER TO FIRM PERFORMANCE?

Oliver Schilke, University of California, Los AngelesMartin Reimann, University of Southern California, Los Angeles

Jacquelyn Thomas, Southern Methodist University, Dallas

SUMMARY

In their fight for global competitive advantage, firmspick strategic options that enable them to save costs andeffort in marketing their goods and services on a globalscale. The cost benefits and ease of administering makethe strategy of standardizing international marketing pro-grams an attractive choice to numerous firms (Douglasand Wind 1987; Katsikeas, Samiee, and Theodosiou2006). Consequently, standardization is considered theperhaps most influential aspect of international marketingstrategy (Zou and Cavusgil 2002).

Most prior research on the topic primarily concernsthe antecedents to standardization, analyzing a range offactors that lead firms to adopt this strategy (e.g., Calantoneet al. 2006; Jain 1989; Powers and Loyka 2007). Perfor-mance implications have received less emphasis, andthus, the question about the impact of standardization onfirm performance remains an enduring research concern(Griffith, Cavusgil, and Xu 2008). Among the few studiesfocusing on this aspect, reported results are inconclusive(Özsomer and Prussia 2000; Theodosiou and Leonidou2003), limiting further development of theory and improve-ment of management practices. While prior researchpredominantly indicates overall beneficial effects of stan-dardization (e.g., O’Donnell and Jeong 2000; Szymanski,Bharadwaj, and Varadarajan 1993), some investigatorshave also argued that a standardization strategy can comewith disadvantages (Lages, Abrantes, and Lages 2008).Consequently, despite its demonstrated benefits, stan-dardization may not always improve performance out-comes.

Therefore, we agree with Ryans, Griffith, and White(2003) on the need to further substantiate “some of the keyunderlying assumptions regarding the value of standard-ization” (p. 589). Indeed, researchers are beginning torecognize that the relationship between standardizationand performance may be complicated and contingent onother factors. Katsikeas, Samiee, and Theodosiou (2006)argue that the effect of standardization on performancebecomes stronger if a fit or coalignment is present betweenoverall marketing program standardization and the mar-ket environment in which it is implemented. Their find-ings – as well as the inconclusive results in studies

investigating an unconditional direct link between stan-dardization and performance (Özsomer and Prussia 2000;Theodosiou and Leonidou 2003) – suggest that the perfor-mance effect of standardization may increase under cer-tain circumstances and decrease under others. However,researchers have paid little systematic attention to theconditions other than environmental fit that determinewhen and how standardization relates to firm success. Inmaking suggestions for future research, Zou and Cavusgil(2002) state that important moderators may include notonly the external industry environment but also internalorganizational attributes. Similarly, Samiee and Roth(1992) posit that standardization needs to be viewed inlight of other key firm policies and strategies; certainorganizational activities and characteristics may haveimportant implications for the effectiveness of globalstandardization.

This paper explores the moderating effect of severalorganizational factors on the relationship between stan-dardization and firm performance. More specifically, weinvestigate the role of competitive strategies, other aspectsof marketing strategy, product characteristics, and gen-eral firm characteristics. Thus, the study’s key contribu-tion is to improve our understanding of those internalorganization aspects that make standardization a particu-larly effective approach to international marketing. Froman academic viewpoint, our study helps to resolve some ofthe inconsistent results regarding to the link betweenstandardization and performance.

We conducted a large-scale survey among busi-nesses from various industries to test our research model.The usable returned responses totaled 489, representing aresponse rate of 19 percent. In our baseline evaluation, wefind that, ceteris paribus, standardization has a significantand positive effect on firm performance. Using multi-group analyses we then tested the contingent effect ofvarious organizational characteristics. We found supportfor five of the seven proposed moderating relationships,suggesting that substantial differences exists betweenfirms with regard to the extent to which standardizationeffects firm performance. More specifically, the factorscost leadership, coordination of marketing activities, glo-bal market participation, product homogeneity, and firmsize significantly moderated the relationship between


standardization and firm performance in such a way thatthe standardization-performance relationship was strength-ened at high levels of the moderating variables and weak-ened at low levels.

We conclude that managers evaluating the adequacyof a standardization strategy need to consider the list ofcontingencies advanced in this research. References areavailable upon request.

For further information contact:Oliver Schilke

University of California, Los Angeles264 Haines Hall

375 Portola PlazaLos Angeles, CA 90095–1551

Phone: +1.310.825.1313Fax: +1.310.206.9838



WHAT IS THE IDEAL CUSTOMIZATION APPROACH? THEMODERATING EFFECT OF CULTURAL MEANING

SYSTEM ACTIVATION ON CUSTOMIZATION

Cong Li, University of Miami, Coral Gables

SUMMARY

Customization is a widely adopted marketing strat-egy within online environments nowadays. The basic ideaof customization is to provide individualized messages tomessage recipients in response to their particular needs orpreferences. However, it remains a mystery in the litera-ture how a message should be customized in certaincircumstances. In other words, there seems to be noconsensus on what the ideal customization approach is, ifit exists. The current study examines the moderatingeffects of cultural meaning system (either individualismor collectivism) activation on two most frequently adoptedcustomization approaches: tailored customization andtargeted customization. Specifically, when a message iscreated for a single person based on his or her uniquecharacteristics, it is refereed as tailored customization. Onthe other hand, when a message is created for a group ofpeople based on their common characteristics, it is calledtargeted customization. Based on well established cul-tural psychology theories on individualism and collectiv-ism, it is predicted that when people’s individualisticmeaning systems are activated, they tend to have a morefavorable attitude toward tailored customization becauseindividuals’ uniqueness is prominent in tailored mes-sages. In contrast, when collectivistic meaning systems

are activated, people tend to generate a more favorableattitude toward targeted customization since the person-in-group relationship is emphasized in targeted messages.

The hypothesis is tested with 52 participants in a 2x2factorial experiment. Participants’ cultural meaning sys-tem activation is manipulated via a five-minute primingtask. After it, participants are requested to view a desig-nated website with either some tailored or targeted travelinformation. It is found that when participants are primedwith individualistic meanings, they hold more favorableattitudes toward tailored messages than targeted mes-sages. In contrast, when participants are primed withcollectivistic meanings, they tend to generate more favor-able attitudes toward targeted messages than tailoredones.

Based on the study findings, it is argued thatcustomization effects are indeed contingent upon culturalmeaning system activation. As tailored customization andtargeted customization tend to emphasize different cul-tural meanings, the contexts where they will generatemore favorable effects are not the same. Therefore, it isproblematic to equalize customization and tailoring. It isalso questionable to suggest that the more tailored amessage is, the more favorable effect it is going to gain.

For further information contact:Cong Li

School of CommunicationUniversity of Miami

Coral Gables, FL 33124Phone: 305.284.2355



ALTERNATIVE HEDONISM, DIGITAL VIRTUAL CONSUMPTION, ANDSTRUCTURES OF THE IMAGINATION

Rebecca Jenkins, Bournemouth University, United KingdomMike Molesworth, Bournemouth University, United Kingdom

Sue Eccles, Bournemouth University, United Kingdom

ABSTRACT

Using “alternative hedonism” and “digital virtualconsumption” as examples, we extend existing theories ofconsumer imagining in terms of temporal location, trig-gers and outcomes. We argue that understanding thiscomplexity will enable marketing practitioners to respondmore effectively to changes within the content of theconsumer imagination.

INTRODUCTION

Daydream, fantasy, anticipation, nostalgia, reminisce,regret, and worry are all terms that might relate to theimagination. Our ability to differentiate between suchterms indicates different functions and types of imaginings.Any of these terms may be used in place of imagination,but they may also help to differentiate between, albeitclosely related, concepts. The imagination is taken to beour overall ability to experience things in our mind in theabsence of their material presence (Casey 2000), it isresponsible for a variety of mental activities (Thomas2004), and unlike night dreams, the imagination is undervoluntary control (Currie and Ravenscroft 2002). Yet thiscomplexity of the imagination is not well covered inconsumer research.

Academics have noted that market goods and expe-riences are good to daydream with (e.g., see Campbell’s1987 “modern autonomous hedonism”). Yet our suspi-cion is that there is value in developing a more completemodel of the consumer imagination that makes the rangeand scope of imaginative spaces available – the “struc-tures of the consumer imagination” – more transparent;that recognizes changes in such structures over time, and;that therefore allows for new perspectives on markets andon consumer culture. That is our first aim here, but morethan this, we note a criticism that as the market hasstructured the imagination it has done so in a way that hasproduced excessive and damaging over-consumption thathas ultimately lead to a financial crisis based on consum-ers wanting to buy “too much trivial stuff” (see Barber2007) or alternatively, a sense of boredom and ennui(Shankar, Whittaker, and Fitchett 2006). The currentglobal recession is likely a time of reflection wherestructures of the imagination that define the relationshipbetween consumption and “the good life” are subject to

change, especially toward more sustainable living (e.g.,see Shi’s 1985 review of similar sentiment in the recessionof the 1970s). In this moment of collective reflection onour consumer society we note two possible trajectories:Soper’s (1998, 2007) analysis of “alternative hedonism”and Molesworth and Denegri-Knott’s (2006), (see alsoDenegri-Knott and Molesworth 2009 [forthcoming]) viewof “digital virtual consumption” (DVC). However we alsosee both of these as potentially limited in their focus onpositive, future orientated imagining and we therefore useour extended taxonomy of the imagination to furtherunpack the imaginative landscape of consumer matter.

THE EXPLICIT USE OF THE IMAGINATION INCONSUMER RESEARCH

Holbrook and Hirschman (1982) were amongst thefirst to acknowledge that consumers experience pleasurethrough fantasy. Subsequent research suggests a generalacceptance that individuals fantasize about things theywould like to own and experience (Belk, Ger, andAskergaard 2003; d’Astous and Deschênes 2005; Fournierand Guiry 1993). Research has also noted that consumersact on their dreams, they talk about them, search forinformation on desired commodities, and save towardtheir dream objects (d’Astous and Deschênes 2005).

An array of terms have been used to refer to pre-consumption imagining including; “day-dream”(Campbell 1987), “pre-purchase dreaming” (Fournier andGuiry 1993), “consumption visions” (Christensen 2002)and “consumption dreaming” (d’Astous and Deschênes2005). What these conceptualizations have in common isa future orientation and this leads to a dominant theory ofthe consumer imagination that is based on “desire.” Desiredrives the creation of consumption based daydreams,which in turn intensify desire and motivate actual con-sumption as consumers seek to actualize or partiallyactualize what they imagine (Belk 2001; Belk, Ger, andAskergaard 2003; Campbell 1987; d’Astous andDeschênes 2005; McCracken 1988).

According to Campbell (1987), McCracken (1988)and Belk, Ger, and Askergaard (2003) consumer desire iscreated through meaning that we attach to goods – theysymbolize a certain way of being: a better lifestyle; anenhanced identity, or; an idealized version of life, which


we seek to realize through purchasing material goods.One attraction of this approach is that it may explain theprocess of endless wants and desires as possession of thedesired good cannot live up to the embroidered pleasurescreated in the imagination (Campbell 1987; McCracken1988). Disappointment causes us to renew our desire fora new object and so the process of creating pleasurabledaydreams about desired goods starts over. Even if pos-sessed goods do not “fail” to live up to imaginings in thisway Belk, Ger, and Askergaard (2003) argue that once inpossession an object can no longer be desired, and asdesire is the sought after experience consumers mustattach their desire to something new such that wanting anddesiring rather than having becomes the consumer’s mainfocus.

The concept of desire highlights the important rolethat consumer goods take on in structuring the imagina-tion. We associate all sorts of life aspirations and eventswith goods and our consumer society stimulates suchfeelings, for example through: advertising, branding, retail-ing, packaging, window shopping, television, film, maga-zines, the internet, word of mouth conversing, and obser-vation of others (Belk 2001; Belk, Ger, and Askergaard2003; Falk and Campbell 1997; Featherstone 1991;Friedberg 1993; McCracken 1988; Stevens and Maclaran2005). In other words much of our leisure and media useaids the construction and actualisation of consumption-related imagining.

NEW CONSUMER DESIRES

A reflection on such positive, future orientated con-sumer imagination has prompted others to consider devel-opments in the ways in which the consumer imaginationmay be structured, and if we expect a relationship betweenthe imagination and economic activity, these “trajecto-ries” may be significant for areas of future economicgrowth. We consider two possibilities here.

Firstly, Soper (1998, 2007) explores “alternativehedonism.” Here the individual may come to imagine andthen seek to actualize pleasures that are generally deniedor even undermined by the market, possibly prompted bydisillusionment that the market is capable of delivering onits promises of the “good life.” These reflections maytherefore lead to resistance to the market and/or alignindividual desires with more civic and ethical possibili-ties. For example, a desire for more open, green spaces incities – for fresh air and peace – may lead to support forpublic parks and to a political will to prevent commercialdevelopments in green spaces. For Soper such sentimentin consumers may lead to the possibility for policy that ismore sympathetic to green issues in general and thereforechanges in economic structures. Such an analysis ofindividuals’ future-orientated, positive imagining alsoties this work to a much larger body on voluntary simplic-

ity (Doherty and Etzioni 2003; Elgin 1993[1981]; Schor1998) where thinking about the “good life” means imag-ining a society with less consumption.

Alternatively Molesworth and Denegri-Knott (2006)note that the failure of the material marketplace to sustainpleasurable cycles of desire may encourage “escape” intodigital virtual spaces where online and console videogamesand “game-like” websites offer new desires and newforms of actualisation, such as trips to other worlds,“ownership” of digital virtual luxury goods, and possibili-ties to live out exotic other lives. Here we see a warningthat individuals’ reflections on the failures of markets tosustain promises of the “good life” may lead to retreat intofurther individualized and episodic lives rather than somere-engagement with civic and collective narratives ofprogress (Molesworth 2009). This is an implicit referenceto a “market determinism” where the market is looked to,to solve social problems.

However both these discourses consider the con-sumer imagination in terms of future-orientated positiveemotions only – i.e., what individuals desire and how it iselaborated in their imaginations. Our task now is toexplore how the structures of imagining that are used heremay be further conceptualized such that a fuller under-standing of the complexities and nuances of consumerimaginings may be understood.

THE IMPLICIT USE OF THE IMAGINATION INCONSUMER RESEARCH

Other research has dealt with the consumer imagina-tion in more implicit ways - i.e., often not using the term“imagination,” but noting elaborated thoughts, or worriesabout consumption. For instance negative post-consump-tion thoughts have been explored as dissonance (Oshikawa1969) and more recently as “buyer’s remorse”; post-purchase feelings of regret, generally in relation to how aforegone alternative may have had a better outcome(Tsiros and Mittal 2000). Alternatively satisfaction researchfocuses on how consumers’ feel regarding the outcome ofa situation compared to the expected (read: imagined)outcome (Oliver 1980). These areas of study presentopportunities to compare imaginative and real experi-ences, and emphasize short-term reflective or past-orientated imagining.

Research has also been conducted on nostalgia as itrelates to consumption (Havlena and Holak 1991; Holbrook1993). Nostalgia is seen as positive reflection on pastexperiences, often from childhood, it is therefore alsopast-orientated imagining triggered by a variety of per-sonally relevant stimuli (Holbrook and Schindler 2003).This is significant because it suggests that consumptionexperiences may “live on” in our imaginations long afterinitial future-orientated desire has passed.


Together these studies suggest the possibility of bothpositive and negative imaginings and a focus on both thepast and the future. Thus, alongside Soper’s (1998, 2007)conceptualization of “alternative hedonism” with its callfor future orientated positive imaginings to be “nudged”away from the current focus on endless consumption, wemight consider other anti-consumerist calls. For instance,those of Greenpeace or Adbusters that may focus oninstilling a negative view of the future where the planet isdamaged (especially for others) or where the conse-quences of consumption produce a bleak personal outlook(in the case of health information about obesity, drinking,or smoking, for example). Similarly we might considerthat along with the desire for spectacular futures filledwith the exotic noted by Molesworth and Denegri-Knott(2006) regular users of videogames may become nostal-gic for “simple” pleasures, or even for material experi-ences, or may regret time “wasted” playing games (e.g.,consider Caru and Cova’s (2003) review of a loss ofcontemplative time). Market activity may therefore resultfrom the complex interactions of such positive and nega-tive, past and future structures of the imagination. Broad-ening the scope of the consumer imagination may uncoverother forms of imagining that take place throughout theconsumption experience.

A BROADER VIEW OF THE CONSUMERIMAGINATION

We now want to consider six key aspects of theimagination broadly identified by Singer (1966) whosework is drawn on by many researchers. These are theimportance of mental imagery and thought, bodily sensa-tions and emotion, temporal setting and the degree towhich imagined events are possible.

Mental Imagery and Thought

Mental imagery is a representation of something inthe mind and may incorporate sight, sound, smell, taste,and touch (MacInnis and Price 1987). For example, wecan picture what our living room looks like, the smell offresh bread, the taste of strong coffee, the sound of a sportscar, and we can imagine the warmth of a fire. A variety ofresearchers note that visual imagery is the most commonfeature of daydreaming, (Singer 1966; Klinger 1990) andperhaps it is this visual focus that allows the sorts of digitalvirtual experiences that Molesworth and Denegri-Knott(2006) note to be so compelling. In videogames playerscannot feel, smell, taste or touch, yet the largely visualillusion produces desire. On the other hand, Soper (1998,2007) notes that it is often the impoverishment of thebroader senses in the commercial city that may encouragea desire for alternative hedonism.

Daydreams also contain self-talk, or an “interiormonologue” (Klinger 1990; Singer 1966). Klinger (1990,

p. 68) asserts that we “silently talk to ourselves” themajority of the time, ranging from just a few words to fullrunning commentaries accompanying the mental imagerywe create. As a type of thought daydreaming may bedistinguished from other kinds of thought such as “prob-lem solving” or “decision making” on the basis that it doesnot seek to achieve a “mental result” (Klinger 1990, p. 18),rather the intention may be to escape reality and gainpleasure from our imagination (Campbell 1987). Thisleads to the consideration that different forms of imagin-ing may have different purposes and we consider thisbelow.

Emotion and Bodily Sensations

It is established that “imagery produces physiologi-cal effects that mirror perceptual processes” (MacInnisand Price 1987, p. 474). Put simply; visualizing an objector event, may arouse the same emotions and effect on thebody as actually seeing the object or experiencing theevent in material reality so that we react subjectively as ifit were “real” (Campbell 1987; Klinger 1990). Again, thisaspect of the imagination may allow for DVC and is thebasis for the imagination as a source of hedonism, orpleasure. We can also have subsequent emotional reac-tions that may conflict with how we felt during thedaydream. For instance, you may feel happy and excitedduring a daydream about a holiday, but afterwards realizethat it is unlikely to materialize due to various constraintsand feel negatively (depressed, upset) toward it as aconsequence (d’Astous and Deschênes 2005). These sortsof emotional reactions are very much classed as part of theimaginative experience (Christensen 2002). However thisaspect of reflecting on daydreams is under-considered inconsumer research and may have a significant bearing onactions taken by consumers to either modify their day-dreams so that such negative emotions are not experi-enced, or to change their behaviors.

Temporal Setting

As we have noted in the review of consumer research,we have the ability to imagine objects and experiencesfrom our past, present and future (see also Singer 1966;Klinger 1990; Giambra 2000). The temporal setting allowsus to distinguish forms of imagination, for example,nostalgia and reminiscing are a longing for the past(Holbrook 1993) and anticipation and wishing refer toimagining the future (Belk, Ger, and Askergaard 2003;Campbell 1987). Other terms are less easy to distinguishbased on these criteria. A daydream, for instance, can bebased in any temporal setting (Klinger 1990; Singer 1966)and can vary from mundane anticipation to elaboratewishes (Belk 2001; Cohen and Taylor 1976).

However, Klinger (1990) notes that daydreams aboutthe present actually make up the majority of daydreams,


occurring when we imagine how the situation we are inmay unfold in the next few moments. As such they arefleeting and unmemorable. This more mundane type ofimaginative anticipation requires little elaboration becauseour expectations are based on routine (Campbell 1987).On the other hand both past and future orientated day-dreams are considered to be more memorable becausethey involve more emotion and reflection (Klinger 1990).Here a greater temporal distance produces complex re-flection on life that may include the more “superficial”and routine imaginings such that it is possible that thedaily “grind” of negative thoughts produces larger day-dreams that lead to the sort of resistance that Soper (1998,2007) identifies.

Level of elaboration and temporal location of a day-dream denote different kinds of imaginative experiencebut whatever the temporal orientation, daydreams aregrounded in real events and this may differentiate day-dreams from fantasy (Campbell 1987).

Fantasy-Like Embellishment

The use of the terms “daydream” and “fantasy” areused inconsistently across the literature, sometimes assynonyms (Holbrook and Hirschman 1982), sometimeswith daydreaming as a form of fantasy (Klinger 1990;Rook 1988), and sometimes as significantly different(Giambra 2000) – which is our stance here. It may beuseful to distinguish between fantasy and daydreamaccording to the degree to which material reality isadhered to.

Campbell (1987) stipulates three imaginative phe-nomena, each of which hold a different position on thereality – fantasy continuum (Christensen 2002). Imagina-tive anticipation is our expectations about “the way inwhich an existing course of events might develop”(Campbell 1987, p. 83), guided by our current reality andexperience of the past. Fantasy on the other hand “by itsnature is not reality” (Christensen 2002, p. 55), to theextent that fantasies are considered to be “incompatiblewith paramount reality . . . an alternative world” (Cohenand Taylor 1976, p. 73), which may involve impossibili-ties, such as possessing magical powers. Daydreams,however, sit in the middle and with pleasure as the guidethey may be crafted to include events that are highlyunlikely to occur in real life, that is, they may be embel-lished with fantasy (Campbell 1987; Christensen 2002).Such fantasy-like embellishment or abstraction (Shields2003) refers to the stuff of fiction (e.g., films and bookswhich daydreams can come to represent) in that it is notprobable but may be hopeful. Pure fantasy represents nopossibility of actualisation, which is problematic for thepursuit of pleasure (Campbell 1987). However this ispossibly key to the idea of DVC, where videogames aid inthe negotiation of “real” experiences with fantasy embel-

lishments, allowing fantasy to be actualized within thegame (Molesworth and Denegri-Knott 2006).

The Triggers for Imaginative Experiences

We have hinted that daydreaming in a pre-consump-tion context is often activated on demand and returned toand modified to become more pleasurable and vivid(Campbell 1987; d’Astous and Deschênes 2005) how-ever, there are prompts and triggers that may cause anindividual to direct attention away from paramount real-ity, so we also experience spontaneous, involuntary day-dreams (Singer 1966). The assumption in psychology thatdaydreams are predominantly unintentional leads us toquestion how something spontaneous and involuntarycan be considered elaborate and involved. When we lookat how unintentional daydreams are triggered it becomesapparent that although they may not be indulged in delib-erately – individuals may not consciously “schedule”them – “their timing and substance . . . are definitely notrandom” (Klinger 1990, p. 76), because they are linked toour concerns and emotions.

Daydreams are based on our goals and desires, whethermajor or trivial (Klinger 1990), and consequently may betriggered by “goal reminders” (p. 35); something associ-ated with the goal, such as words, events or thoughts.When we come into contact with a goal reminder, at a timewhen we are unable to meet the goal perhaps becausesome kind of constraint prevents it, we react to it mentallyas a daydream. This view seems well suited to daydreamsand imaginings about current concerns (unmet goals andunfinished business) when daydreaming is said to helporganize and remind us (Klinger 1990). For example, yousee wine at the supermarket, which reminds you about adinner party you are to host, which then sends you into adaydream about dishes you could serve and so on, but thispattern can also be extrapolated to more distant andimportant goals and desires. Molesworth (2009) noteshow DVC may be a form of imagined escape to deal withdull experiences at work, problems at home, or eventhwarted career aspirations. For example, a frustratingevening commute reminds the player about their desire toown and drive a luxury sports car, which leads to a sessionon a driving videogame on arrival at home.

Stimuli may also be used intentionally by individualsto intensify their desire (Belk 2001; Belk, Ger, andAskergaard 2003). Hence we also start to see that in anenvironment dominated by commercial messages, theseform a significant amount of the triggers for our imagina-tion and as Soper (1998, 2007) implies, calls for alterna-tive hedonism are marginalised. We are surrounded andsurround ourselves with commercial messages about thingswe can, might and should really want, and embracingtechnology seems to provide an easier solution to manag-ing the imagination than resistance to material desires.


Unintentional past-orientated daydreams are trig-gered by stimuli that we have an emotional bond withbecause they were once related to our goals (Klinger1990). For example you may replay pleasurable experi-ences from a holiday for weeks or months after your returnbecause certain stimuli remind you of the wonderful timeyou had. As many of these past experiences are them-selves consumption experiences, the market is also wellplaced to remind us of these and this is a further “problem”for alternative hedonism because it may too often be ayearning for something we have never experienced, ordone so only fleetingly. We might see how both the Miniand Beetle brands have been created out of a sense ofnostalgia for their original 1960s versions – even thoughboth differ considerably from their originals – whereas therejection of the consumer society at the same time findslittle outlet today where every experience must be paid for.

Purposes and Outcomes of Imagining

It should be clear by now that as Klinger (1990)argues, the imagination has an important role and manyuses in our lives. For example anticipation may be indica-tive of planning or rehearsing and a daydream may bepleasurable (Campbell 1987), while fantasy is indicativeof escape (Cohen and Taylor 1976).

Pleasure is central to future orientated imagining tothe extent that Campbell (1987) recognizes that we mayexperience greater pleasure in our imagination than inreality. Yet it is also evident that we can attain pleasurefrom past orientated daydreams that offer the opportunityto replay enjoyable experiences (Klinger 1990). We havealso noted the potential for anticipatory consumption to beused for compensation (d’Astous and Deschênes 2005;Fournier and Guiry 1993). When constraints prevent usfrom achieving the stuff of our dreams, imagining pro-vides compensatory pleasure. Purposive imagining istherefore a coping mechanism (Christensen, Olson, andRoss 2004) or surrogate experience (MacInnis and Price1987); we turn to our imagination to escape (Cohen andTaylor 1976) and the development of videogames andother DVC is seen by Molesworth and Denegri-Knott(2006) as the multi-billion dollar market response to sucha need. Soper (1998, 2007), on the other hand, considersa retreat from the market as another possibility. During arecession constraints on consumption may become evenmore prevalent and as financial and employment worries,added to environmental concerns prevent consumers frommaking “unnecessary” purchases we may see more ofsuch imaginative “work” as necessary.

Daydreams about future events provide opportuni-ties to explore and rehearse, which aids decision-makingand planning (Christensen, Olson, and Ross 2004). Theimagination provides a valuable opportunity to play outvarious scenarios and imagine the consequences of sev-

eral alternatives, and this feeds into the decision makingprocess (Phillips, Olson, and Baumgartner 1995). Plan-ning, is also a benefit of imagining future situations andshares a likeness with Klinger’s (1990) organization func-tion of daydreaming; we can prepare, anticipate andorganize ourselves through the use of images (Christensen,Olson, and Ross 2004).

The rehearsal quality of imagining lends itself topositive and negative events. As well as helping us figureout goals and desires that we want to pursue, worried andunhappy daydreams help us to plan and rehearse based onthings that we fear and want to avoid (Klinger 1990). Soalong with alternative hedonism there may be usefulimaginings about the consequences of over-consumption,and along with the escapism pleasures of videogamesthere may be reflections on in-game experiences that leadto consideration of aspects of the material world (indeedthis is a common line in discussing the reflective, orpersuasive aspects of games, e.g., see Turkle 1995). Pastorientated daydreams in particular can provide lessons tolearn; negative experiences of past events can teach ushow to behave in the future or how we would behavedifferently if a situation arose again (Klinger 1990) and sowe might wonder at the lessons individuals will take fromthe current recession as they reflect.

A key point here is that unlike escapism and even theromantic pleasures of desire, the rehearsal, planning,learning and decision making processes of imaginingoffers motivation for action (Belk, Ger, and Askergaard2003; Klinger 1990); in this way imagining is essentiallypreparation for “real” life. The stuff of daydreams, and ofworries therefore constantly provides something for us toaim for, or aim to avoid.

A TAXONOMY OF CONSUMER IMAGINING

We have demonstrated that there are more aspects ofimagining than have been explicitly considered in con-sumer behavior research. The range of characteristics thathave been discussed exist on a dimensional scale andpresent a variety of imaginative experiences that may bemapped according to four main elements; the temporallocation, range of emotions experienced (positive or nega-tive), degree of abstraction and level of elaboration. Whenthese elements are experienced in different ways andcombinations the outcome is different imaginative expe-riences, which constitute different forms of imagination.For example “pre-consumption daydreaming” is set in thefuture, features positive emotions, is highly elaboratedand may involve abstraction. Nostalgia also features posi-tive emotions, is highly elaborated and may involveabstraction, but is set in the past. The temporal location isthe distinguishing feature here. In addition we can alsonote the range of things that might prompt imagining, thepurpose of different forms and the degree to which they


might set goals for behavior. Table 1 sets out a variety ofimaginative experiences that constitute different forms ofimagination related to consumer behavior, based on theelements explained above. An issue here is the degree towhich an understanding of structures of the imaginationmay lead to attempts by marketers and others (e.g., policymakers) to “manage” these structures.

We have tried to capture a range of types of imaginingthat illustrate the elements, purposes and outcomes thatwe have identified, but it is clear that there may be otherterms that could be added as synonyms or as discreetcategories. We note the degree to which prompts are theday-to-day stuff of everyday life, or some more persistentennui, or need to escape. We then consider the range ofemotions that allows us to distinguish between positiveand negative forms of imagining; temporal location con-siders a focus on the past or the future; level of abstractioncaptures the degree of fantasy-like embellishment pos-sible, and elaboration denotes the intensity of an imagina-tive experience in relation to the mental imagery, thought,emotion and bodily sensations. From this we note thepurpose and the outcome in terms of possible marketbehavior. Finally we include examples from the literatureto illustrate the types of goods and experiences that havebeen found to lend themselves to each form of imagining.However, it should be remembered that a broad range ofproducts, services and experiences could be associatedwith all forms of imagining and our invitation to marketersis to give more consideration to the ways products interactwith the imagination; how numerous forms of imaginingmay be experienced in relation to just one object.

This working taxonomy offers a broader, systematicoverview of different uses of the imagination than isacknowledged elsewhere, including by either Soper (1998,2007) or Molesworth and Denegri-Knott, (2006) and maybe used as a basis to guide research that should endeavorto understand the variety of imaginative experiencesrelated to consumption in greater depth. Although manyof these terms may be used synonymously, here we havetried to separate them according to the characteristicsreviewed. However, we note the complexity and the wayprompts, purposes, and outcomes may overlap or contra-dict. The imagination is highly situated, based on indi-vidual circumstance and past experiences. This has impli-cations for the sorts of studies that may capture details ofthe structures of the consumer imagination.

Implications for the Trajectory of Consumer Imagining

The consumer imagination is not just about pre-consumption desire as it is often presented in consumerresearch, although the reasons for the focus on positivefuture-orientated imagining is obvious (it is where ourdesire to buy and experience ever more in the marketcomes from and it is where the future may be glimpsed).

However, our call is for a fuller understanding of thecomplex and contradictory nature of consumer imagina-tion practices. In the current economic climate this is bothtimely and necessary in order to help marketers, byenabling them to respond more effectively to consumers’worries, regrets, hopes, and dreams, including perhapsdesires for more sustainable consumption that allowsalternative hedonism.

A way to understand the scope of the consumerimagination is to study a range of imaginative phenomenaas they occur in a consumption situation. To date researchoften focuses on categories of products in relation to aspecific (imaginative) concept (e.g., studies of buyer’sregret and satisfaction have looked at the purchase of cars(Keaveney, Huber, and Herrman 2007). While this givesresearch a context it may limit the scope of explorationpossible with regard to imaginative forms. Our call is forresearch to trace the ebbs and flows of imagination prac-tices, noting how they drift from one to another, forexample from pleasurable pre-consumption daydreamsthat get embellished over time, to more realistic plans, topost consumption worry and disappointment and then toenjoyable reminiscences and finally nostalgia. What seemsclear is that the same goods and experiences may movebetween these imaginative states such that a car that wasonce an exciting daydream may become that which isanticipated when the loan gets agreed, or that job is got.But this same desirable car may also then become the thingthat produces angst (did we buy the right one with the rightoptions?), and then worry (what if it gets damaged orstolen?) or even regret (as we consider its impact on theenvironment). Here disillusionment with the pleasure carsprovide or promise may lead to a desire for alternativehedonism, or even for the pleasures of DVC. There mayof course be many other trajectories and the issue iswhether these support market based desires, or result in arejection of the market as a source of pleasure altogether.Of course another possibility may be the Buddhist tradi-tion of training the mind not to desire, (e.g., see Fromm’s1976 call for a reflection on having and a focus on being).

We also need to consider prompts, purposes andoutcomes of imagining. The problem for a resistance lineof argument is that the prompts for a continued marketfocus may swamp calls for alternative hedonism. This isnot just advertising, but the collective experiences ofpleasure that consumers may draw from to construct newdaydreams or to produce pleasurable nostalgia. Alterna-tive hedonism may require not just new daydreams ofnon-market pleasures (a rejection of pleasurable consum-ing futures), but a rejection of the pleasures of a consum-ing past and inoculation against the ever-present calls ofadvertising, promotions and shop displays (e.g., considerthe calls in Klein (2002), or on the Adbusters website).Understanding the consumer imagination in such circum-stances may require sensitive phenomenological investi-


Form of imagination

Prompts

Emotion

Temporal location

Level of Abstraction

Elaboration

Purpose

Possible outcome

Example product category

Fantasy

Boredom/ unpleas ant exp erience

Positive

Not defined

Very High

Very high

Compensation/

Pleasure/ Escape

Escapism/

Digital vi rtual consumption

Videogames and websites, e.g. World of Warcraft (Molesworth and Denegri -Knott 2006). Picturing oneself as as someone else e.g. s omeone from his tory (Campbell 1987)

(Pre-

consumption) Daydream

Market ing material/ Boredom/ Current

concerns

Positive

Future

High

Very high

Compensation/

Pleasure/ Explore/

Plan

Cons umption/ Resistance to consumption

Cars, houses, holiday homes, travel experiences, expensive i tems of clothing (Belk, Ger and Askergaard 2003, d’Astous & Deschênes 2005) Relocat ion, greater sense of community, more t ime (Soper 1998; 2007)

Ant icipation/ Expectation

Immediate concerns

Positive/ Negative

Near future

Low

Low

Rehearse/

Decide

Consumption

Planning dinner, food s hopping, gett ing pet rol, going out to eat (Christensen 2002)

Reminisce

Familiar objects or

events

Positive

Near past

Low

Low

Pleasure/

Learn

Formation of

routine, or habit

As nostalgia but l ess intense

Nostalgia

Objects and events

from the past

Positive

Past

High

Very high

Compensation/

Pleasure/ / Learn

Escapism/

Consumption

Holidays, Souvenirs (Holak & Havlena 1998; Urry 2002) ‘Retro goods’, e.g. new VW Beetle (Rindfleisch, Freeman and Burroughs 2000).

Worry

Current concerns /

Unpleasant exp erience

Negative

Future/

Past

Low

High

Plan / Avoid


Cars, recreational possessions (Richins 1997)

Buyers Regret/

Disappointment/ Dissatis faction

Immediate purchase/

or non-purchas e

Negative

Near past

Low

Low

Learn

Cons umption/

Non-consumption

Cars, IT g oods , res taurant experiences (Keaveney, Huber and Herrmanl 2007) Branded Vs generic goods, engagement rings, gi fts (Tsiros & M ittal 2000)

Regret

Objects and events

from the past

Negative

Past

High

Very high

Learn/ Plan


As buyer’s regret but more long last ing/intense

TABLE 1Taxonomy of the Consumer Imagination

gations that are able to capture the life-worlds and livedexperiences of consumers (e.g., see Thompson 1997;Thompson, Locander, and Pollio 1989).

For marketing the implications of such scope in theimagination may already be implicitly understood. Desirethrough daydreaming and fantasy is sought in order todrive sales; reminiscing and nostalgia may result in loy-alty and more sales (especially of “retro” products); regretand worry may reduce sales and loyalty for some, butenergize insurance and other markets; the excitement ofanticipation may lead to sales, but angst may prevent

them. With consumer practices informed by elaborateimaginative activities we may also note the potential fornew business models that exploit in new ways the imagi-native spaces that we have described and speculate abouthow marketers could put this to use. For example, in orderto sustain desire whilst waiting for a holiday or fordelivery of a new car, businesses could point customers tovarious imaginative resources related to the purchase suchas film, magazine, book, or website recommendations – ineffect encouraging various “approaching behaviors”(d’Astous and Deschênes 2005). Post-purchase, positivereviews about the products could be sent in an effort to


ratify the purchase and reduce the likelihood of dissatis-faction, along with recommendations for complimentarypurchases. However there is also a social responsibility toconsider the ways in which the marketing process helps tostructure the imagination only in ways that encourageconsumption, even trivial, or over-consumption. If we areto hope for more alternative hedonism (and less individu-alized and escapist DVC) there may be difficult policy

issues relating to restrictions in advertising and promotionand in support of alternative daydreams (for exampleissues relating to planning of open and green spaces andof public goods in general) and as we noted at the start, thecurrent recession - when there is a collective moment ofreflection on our consumer society – may be a good timeto consider these ideas in detail.

REFERENCES

Barber, Benjamin R. (2007), Consumed: How MarketsCorrupt Children, Infantilize Adults, and SwallowCitizens Whole. New York: W.W. Norton.

Belk, Russell W. (2001), “Speciality Magazines and Flightsof Fancy: Feeding the Desire to Desire” in EuropeanAdvances in Consumer Research, Andrea Groeppel-Klien and Frank-Rudolf Esch, eds. Provo, UT: Asso-ciation for Consumer Research, 5, 197–202.

____________, Guliz Ger, and Soren Askergaard (2003),“The Fire of Desire: A Mulitsited Inquiry into Con-sumer Passion,” Journal of Consumer Research, 30(3), 326–51.

Campbell, Colin (1987), The Romantic Ethic and theSpirit of Modern Consumerism. Oxford: Blackwell.

Caru, Antonella and Bernard Cova (2003), “A CriticalApproach to Experiential Consumption: FightingAgainst the Disappearance of Contemplative Time,”Proceedings of the 3rd International Critical Man-agement Studies Conference, Lancaster University,July 7–9.

Casey, Edward S. (2000), Imagining: A Phenomenologi-cal Study, 2nd Edition. Bloomington: Indiana Univer-sity Press

Christensen, Glenn L. (2002), “Mental Simulations, An-ticipations, and Dreams: Toward a Theory of Con-sumption Visions in Consumer Behavior,” doctoraldissertation. The Smeal College of Business Admin-istration, The Pennsylvania State University.

____________, Jerry C. Olson, and William T. Ross(2004), “Why Consumption Vision? UnderstatingConsumer Value in Anticipatory Consumption Imag-ing,” Advances in Consumer Behavior, Barbara E.Kahn and Mary Frances Luce, eds. Valdosta, GA:Association for Consumer Research, 31, 130–31.

Cohen, Stanley and Laurie Taylor (1976), Escape Attempts:The Theory and Practice of Resistance to EverydayLife. Aylesbury: Pelican Books.

Currie, Gregory and Ian Ravenscroft (2002), RecreativeMinds. Cornwall: Oxford University Press.

d’Astous, Alain and Jonathan Deschênes (2005), “Con-suming in One’s Mind: An Exploration,” Psychologyand Marketing, 2 (1), 1–30.

Denegri-Knott, Janice and Mike Molesworth (2009 [forth-coming]), “Digital Virtual Consumption: Conceptsand Practices,” Consumption, Markets & Culture.

Doherty, Daniel and Amitai Etzioni (2003), VoluntarySimplicity: Responding to Consumer Culture. Oxford:Rowman and Littlefield.

Elgin, Duane (1993), Voluntary Simplicity: Toward aWay of Life That is Outwardly Simple, Inwardly Rich.New York: William Morrow.

Falk, Pasi and Colin Campbell (1997), The ShoppingExperience. London: Sage.

Featherstone, Mike (2007), Consumer Culture andPostmodernism, 2nd Edition. Trowbridge: Sage.

Fournier, Susan and Michael Guiry (1993), “An EmeraldGreen Jaguar, a House on Nantucket and an AfricanSafari: Wish Lists and Consumption Dreams in Mate-rialist Society,” Advances in Consumer Research,Leigh McAlister and Michael L. Rothschild, eds.Provo, UT: Association for Consumer Research, 20,352–58.

Friedberg, Anne (1993), Window Shopping: Cinema andthe Postmodern. Berkley: University of CaliforniaPress.

Fromm, Eric (1976), To Have or to Be? London: Abacus.Giambra, Leonard M. (2000), “Daydreaming Character-

istics Across the Life-Span: Age Differences andSeven to Twenty Year Longitudinal Changes,” inIndividual Differences in Conscious Experience,Rober G. Kunzendorf, ed. Philadelphia: JohnBenjamins Publishing Company, 147–206.

Havlena, William J. and Susan. L Holak (1991), “TheGood Old Days: Observations on Nostalgia and ItsRole in Consumer Behavior,” Advances in ConsumerResearch, Rebecca H. Holman and Michael R.Solomon, eds. Provo, UT: Association for ConsumerResearch, 18, 323–29.

Holak, Susan L. and William J. Havlena (1998), “Feel-ings, Fantasies, and Memories: An Examination ofthe Emotional Components of Nostalgia,” Journal ofBusiness Research, 42, 217–26.

Holbrook, Morris B. and Elizabeth C. Hirschman (1982),“The Experiential Aspects of Consumption: Con-sumer Fantasies, Feelings, and Fun,” Journal ofConsumer Research, 9 (2), 132–40.


____________ (1993) “Nostalgia and Consumption Pref-erences: Some Emerging Patterns of ConsumerTastes,” Journal of Consumer Research, 20 (Sep-tember), 245–56.

____________ and Robert M. Schindler (2003), “Nostal-gic Bonding: Exploring the Role of Nostalgia in theConsumption Experience,” Journal of ConsumerBehavior, 3 (2), 107–27.

Keaveney, Susan M., Frank Huber, and Andreas Herrman(2007), “A Model of Buyer Regret: SelectedPrepurchase and Postpurchase Antecedents withConsequences for the Brand and the Channel,” Jour-nal of Business Research, 60, 1207–15.

Klein, Naomi (2002), No Logo: Taking Aim at the BrandBullies. New York: MacMillan.

Klinger, Eric (1990), Daydreaming: Using Waking Fan-tasy and Imagery for Self-Knowledge and Creativity.Los Angeles: J.P. Tarcher.

MacInnis, Deborah J. and Linda L. Price (1987), “TheRole of Imagery in Information Processing: Reviewand Extensions,” Journal of Consumer Research, 13(4), 473–91.

McCracken, Grant (1988), Culture and Consumption:New Approaches to the Symbolic Character of Con-sumer Goods and Activities. Bloomington: IndianaUniversity Press

Molesworth, Mike and Janice Denegri-Knott (2006),“Desire for Commodities and Fantastic Consump-tion in Digital Games,” in the Players Realm: Studieson the Cultures of Videogames and Gaming, J.P.Williams and J.H. Smith, eds. Mc Farland: NorthCarolina.

____________ (2009), “Adults’ Consumption ofVideogames as Imaginative Escape from Routine,”in Advances in Consumer Research, Ann L. McGilland Sharon Shavitt, eds. Duluth, MN: Associationfor Consumer Research, 36, 378–83.

Oliver, Richard L. (1980), “A Cognitive Model of theAntecedents and Consequences of Satisfaction Deci-sions,” Journal of Marketing Research, 17 (Novem-ber), 460–69.

Oshikawa, Sadaomi (1969), “Can Cognitive DissonanceTheory Explain Consumer Behavior?” Journal ofMarketing, 30 (October), 44–49.

Phillips, Diane M., Jerry C. Olson, and Hans Baumgartner(1995), “Consumption Visions in Consumer Deci-sion Making,” Advances in Consumer Research,Frank Kardes and Mita Sujan, eds. Provo, UT: Asso-ciation for Consumer Research, 22, 280–84.

Richins, Marsha L. (1997), “Measuring Emotions in theConsumption Experience,” Journal of ConsumerResearch, 24, 127–46.

Rindfleisch, Aric, Dan Freeman, and James E. Burroughs(2000), “Nostalgia, Materialism, and Product Prefer-ence: An Initial Inquiry,” Advances in ConsumerResearch, Stephen J. Hoch and Robert J. Meyer, eds.Provo, UT: Association for Consumer Research, 27,36–41.

Rook, Dennis (1988), “Researching Consumer Fantasy,”Research in Consumer Behavior, 3, 247–70.

Schor, Juliet B. (1998), The Overspent American:Upscaling, Downshifting and the New Consumer.New York: Basic Books

Shankar, Avi, Julie Whittaker, and James Fitchett (2006),“Heaven Knows I’m Miserable Now,” MarketingTheory, 6 (4), 485–505.

Shi, David E. (1985), The Simple Life: Plain Living andHigh Thinking in American Culture. New York:Oxford University Press.

Shields, Rob (2002), The Virtual. London: Routledge.Singer, Jerome L. (1966), Daydreaming and Fantasy.

Suffolk: Oxford University Press.Soper, Kate (1998), “An Alternative Hedonism” (inter-

view by T. Benton), Radical Philosophy. 92, 28–38.____________ (2007), “Re-Thinking the ‘Good Life’:

The Citizenship Dimension of Consumer Disaffec-tion with Consumerism,” Journal of Consumer Cul-ture, 7 (2), 205–29.

Stevens, Lorna and Pauline Maclaran (2005) “Exploringthe ‘Shopping Imaginary,’ The Dreamworld ofWomen’s Magazines,” Journal of Consumer Behav-ior, 4 (4), 282–92.

Thomas, Nigel J.T. (2004), “Imagination,” Dictionary ofPhilosophy of Mind. Chris Eliasmith, ed. (Accessed10 October 2008) [available at: http://philosophy.uwaterloo.ca/MindDict/imagination.html].

Thompson, Craig J., William B. Locander, and Howard R.Pollio (1989), “Putting Consumer Experiences Backinto Consumer Research: The Philosophy and Methodof Existential-Phenomenology,” Journal of ConsumerResearch, 16 (September), 133–46.

____________ (1997), “Interpreting Consumers: AHermeneutical Framework for Deriving MarketingInsights from the Texts of Consumers’ ConsumptionStories,” Journal of Marketing Research, 34 (4),438–55.

Tsiros, Michael and Vikas Mittal (2000), “Regret: AModel of Its Antecedents and Consequences in Con-sumer Decision Making,” Journal of Consumer Re-search, 26 (March), 401–17.

Turkle, Sherry (1995), Life on the Screen: Identity in theAge of the Internet. New York: Touchstone.

Urry, John (2002), The Tourist Gaze, 2nd Edition. London:Sage.


For further information contact:Rebecca Jenkins

Bournemouth UniversityTalbot Campus

Fern BarrowPooleDorset

BH12 5BBUnited Kingdom

Phone: 01202.965362Fax: 01202.962736



THE COMPARISONS OF REAL VERSUS VIRTUAL PRODUCTPURCHASE: THE EFFECTS OF SELF-CONGRUENCY

AND PRODUCT TYPE

Chih-Ping Wang, National Chengchi University, Taiwan

ABSTRACT

This research explored the impacts of congruency ofself-image and product image on the purchase intention ofreal and virtual products via a 2 (congruency of self-image: actual self-image/ideal self-image) * 2 (real/vir-tual product) * 3 (product type: utilitarian/hedonic/sym-bolic) experimental design. It implicated that appropriateproduct image shaping enhances product identificationsand increases purchase intentions.

INTRODUCTION

Nowadays, consumers go to shops not only for thefunctional benefits brought by products but also for thepsychological utilities they entail. People seek sensationalpleasure, diversities, or self-image expressions to satisfytheir mental need.

Previous studies showed that self-image is not aunidimensional concept (Burns 1979; Rosenberg 1979).When purchasing products, people project different self-images to meet their needs. Marketing scholars provedthat when the congruency of self-image and products’image is high, it leads to a positive influence on people’spreferences toward the products, purchase intentions, andproduct loyalties (Sirgy 1982). However, when discuss-ing types of products, past research primary focused onthe public visibility of products, products’ characteristics,or consumer decision types (Ekinci and Riley 2003).There is less discussion of customers’ needs. This researchis based on the product categorization of Woods (1960),who divided goods into three types: utilitarian, hedonic,and symbolic products. Accordingly, our first purpose isto explore the influence of different self-image congruen-cies on the purchase intention of utilitarian, hedonic, andsymbolic products and to examine how this then affectspeople’s purchase intention.

Besides the real products, we also explore the con-sumer behavior of virtual products. Today, internet videogames are so splendid and novel that people are conse-quently often immersed in them. While consumers playvideo games, they can change several roles. These role-playing video games provide fantasies which go beyondpeople’s real lives. These people could use differentselves to bargain for products or manage business affairs,and thus these games indirectly meet different consumer

needs. The virtual identities might not be the same as thereal identities. Therefore, the second purpose of thisresearch is to analyze whether people consuming virtualproducts employ different selves to satisfy their utilitar-ian, hedonic, and symbolic needs and to examine how thisthen affects their purchase intention.

LITERATURE REVIEW

Self-image and Product Image

Rosenberg (1979) pointed out that self-concept is theoverall sum of personal thoughts and affections and couldbe a reference of self identity. In the cognitive domain, selfis a conceptual system of self information (Sirgy 1982).Burns (1979) noted that self-image is a multi-dimensionalconstruct. There are three most important componentsunder this concept (Malhotra 1981; Rogers 1951; Sirgy1982): (1) actual self: how one looks upon himself. It is theimpression of how one subjectively sees himself, (2) idealself: how one wants to be looked upon or ideally wants tobecome with regard to certain types of persons, and (3)social self: the self that one expresses to others or how oneimagines others evaluate him overall.

Levy (1959) noted that different products projecteddistinct images. Sirgy (1985a) stated that products andservices could also be analogized to have personalityimage. Product images are composed of a set of character-istics, such as friendly, traditional, modern . . . etc. Theseattributes are different from the functional characteris-tics – qualities, prices, and weights, for example. Undersome situations, product images are so vivid that theseproducts become the media conveying the messagebetween people’s self-images and the outer world. Therelationships and interaction between consumer self-images and product images should be explored further.

Roger (1951) proposed the Theory of IndividualSelf-Enhancement. He considered that consumers preferthe products that are fitted to their self-images. Dolich(1969) found that consumers showed greater similarity ofself concept and the image of the most preferred productsthan that and the image of the least preferred products.Grubb and Grathwohl (1967) noted that self-congruencyis the bonding between the individual’s mental constructof self-image and the symbolic value endowed in theproducts that he consumes.


Schenk and Holman (1980) brought up the concept ofsituational self-image. They thought that how the indi-vidual acts depend on the characteristics of the environ-ment in which he is situated. Once a person decides whatkind of self is the most suitable, he will find ways toexpress it. The usage of products is one of the ways torepresent his self-image (Martin and Bellizzi 1982).According to this concept, one might behave differently inthe real and virtual worlds if one employs different selfcognition, attitudes, and emotion between these two worlds.

The Relationships among the Self-Image, ProductTypes, and Purchase Behavior

Sirgy (1982) proved that the purchase motivationwould be affected by the interaction of self-images andproduct images. This result corresponds to the self-enhancing theory noted by Roger (1951). This impliesthat individual tends toward buying the products whichmatch his self-image, but avoids purchasing the productsincongruent to his self-image, which leads to post-con-sumption dissonance.

People use products to meet their needs. The con-sumer needs can be classified as three categories (Park,Jaworski, and Maclnnis 1986): functional needs, experi-ence needs, and symbolic needs. In line with this concept,Woods (1960) classified products based on consumerneeds. She categorized products as utilitarian products,hedonic products, and symbolic products. The first type ofproducts provides practical or necessary functions. Thesecond kind of products emphasizes attractiveness, whichbrings sensational delight, fantasy, and happy feeling.The last type of product can satisfy the need for self-enhancement, role-positioning, and strengthening grouprelationships or self-identification.

Hedonic real products specify attractively sensa-tional characteristics. They provide psychological enjoy-ment rather than being the instrument to achieve aspiredself-images. According to Johar and Sirgy (1991), thecongruency of actual self-image can satisfy self-consistency needs. The need of self-consistency is viewedas the motivational tendency which serves to insure con-sistency between one’s actual self-image and relatedcognition, attitudes, and behaviors. For example, if aconsumer buys a sports car with sensational image such assexy, outgoing, and youthful, the consumer may think ofhimself as sexy, outgoing and youthful. This leads toevoke his positive attitudes toward the products andhigher purchase intention. Thus, we infer the hypothesis:

Hypothesis 1a: The impact of the congruency ofactual self-image and hedonic real products on thepurchase intention is higher than that of congruencyof ideal-self image.

Johar and Sirgy (1991) proposed that the congruencyof ideal self could satisfy the needs for self-esteem.Shrauger and Lund (1975) pointed out that the self-esteemmotive is the tendency to enhance the self-image. Theself-esteem need is the process to surpass the self of thereal world and to attain the more perfect self (Sirgy 1982,1985a, 1985b). To attain the ideal self expectation, sym-bolic products could be the appropriate media to deliverythis message. Through the behavior of purchasing thesymbolic products, the individual can approach the idealgroups, which indirectly meet the need of enhancing self-esteem. This psychological mechanism would lead tohigher positive attitudes toward the symbolic products.

Thus, we infer the hypothesis:

Hypothesis 1b: The impact of the congruency of idealself-image and symbolic real products on the pur-chase intention is higher than that of the congruencyof actual self-image.

Utilitarian products don’t have cultural and socialmeaning. They are focused on practical and necessaryusage. People put more efforts into comparing and search-ing for the difference among various similar utilitariangoods. According to the Elaboration Likelihood Model(Petty, Cacioppo, and Schumann 1983), people are highlyinvolved before they purchase these goods and use thecentral route thought to deal with the product’s essentialinformation. Chang (2002) concluded that when the prod-uct belongs to the highly-involved category, it makespeople have high motivation to cope with the differentattributes of products. At this moment, the congruency ofself-image is less important. The influence of congruencyof self-image is inclined to peripheral route thinking, andthe strength is limited under the utilitarian products.Hence, we get the following hypothesis:

Hypothesis 1c: The impact of the congruency ofactual self-image and hedonic real products on thepurchase intention is higher than that of the congru-ency of utilitarian products.

Hypothesis 1d: The impact of the congruency of idealself-image and hedonic real products on the purchaseintention is higher than that of the congruency ofutilitarian real products.

Hypothesis 1e: The impact of the congruency ofactual self-image and symbolic real products on thepurchase intention is higher than that of the congru-ency of utilitarian real products.

Hypothesis 1f: The impact of the congruency of idealself-image and symbolic real products on the pur-chase intention is higher than that of the congruencyof utilitarian real products.


An on-line game player investigation (sample size =303) conducted by Bei (2004) disclosed that the imagina-tive roles are more attractive than are the prestige roles inthe real world (i.e., doctor, lawyer, president . . . etc.). Thefour most popular roles are a swordsman, magician, headof a gang, and FBI investigator. Regardless of whetherthese roles are positive or negative images, the partici-pants place more importance on the fantastic roles andproject their selves toward these roles. In the virtualworld, what people present may not be the actual self-image but rather the ideal self. A consumer could projecta more ideal self-image in a virtual world.

As mentioned before, hedonic products can satisfysensational needs. In a virtual world, hedonic goods canbring exciting experience and higher creative imagina-tion. Compared with hedonic real products, consumerswho buy virtual hedonic products expect more fantasticenjoyment. Therefore, the impact of the ideal self-imageis higher than that of the actual self-image.

Hypothesis 2a: The impact of the congruency of theideal self-image and hedonic virtual products on thepurchase intention is higher than that of the congru-ency of the actual self-image.

In a virtual world, the symbolic goods transmit themessage to others that the user has some specific statusdistinct from the common public. The display of one’susage of some symbolic products enhances the user’s self-esteem needs. Therefore, the symbolic virtual productscould indirectly heighten one’s ideal self-images. Weinfer that

Hypothesis 2b: The impact of the congruency of theideal self-image and symbolic virtual products on thepurchase intention is higher than that of the congru-ency of the actual self-image.

Under the situation of a virtual world, people stillfocus their emphasis on the essential characteristics ofutilitarian products. They select the utilitarian goods whichcan bring the most usefulness (e.g., solve problems). Thecongruency of self-image and product image is a minorfactor to be considered. The influence of self-image islimited. We conclude:

Hypothesis 2c: The impact of the congruency ofactual self-image and hedonic virtual products on thepurchase intention is higher than that of the congru-ency of utilitarian virtual products.

Hypothesis 2d: The impact of the congruency of idealself-image and hedonic virtual products on the pur-chase intention is higher than that of the congruencyof utilitarian virtual products.

Hypothesis 2e: The impact of the congruency ofactual self-image and symbolic virtual products onthe purchase intention is higher than that of thecongruency of the utilitarian virtual products.

Hypothesis 2f: The impact of the congruency of idealself-image and symbolic virtual products on the pur-chase intention is higher than that of the congruencyof the utilitarian virtual products.

METHOD

The Measurement of Congruency of Self-Image andProduct Image

The actual self is how the individual looks at oneself(Roger 1951; Burns and Farina 1979; Malhotra 1981;Sirgy 1982). This self perception is subjective. The idealself is how the individual treats himself, or what kind ofperson he is ideally. This is the cognitive result of one’sover-estimation of himself. In the past literature, the self-image has been measured in three methods. They are theQ-sort method, semantic differential method, and directmeasurement method. The first was developed by Sommers(1964). The semantic differential method is conducted byseveral sets of two opposite adjectives to evaluate self-image (Dolich 1969; Hoelter 2001; Malhotra 1981). Thesecond method estimates the self-image and product imagewith the same measurement items. The level of the con-gruency of self-image and product image is calculatedfrom the absolute value of these two different scores (i.e.,absolute difference model) or by squaring the differencescores and then radicalizing them (i.e., Euclidean Dis-tance Model).

Sirgy et al. (1997) commented that the semanticdifferential method is the mathematical result of two setsof evaluation scores, which is an indirect way. The directmethod could have higher predictive validity. However,the direct measurement method has a shortcoming. Wecan’t know which item affects the self-congruency. Fromthe result of the direct measurement method, we only getthe kind of conclusion such as the higher the self-congruency, the higher the dependent variable. We can’tget detailed reasoning such as “the higher the self-congruency, especially in the youth-mature item, thehigher the dependent variable.” This drawback reducesthe application to managerial practices. Besides, thisresearch included virtual products. It is hard for people toevaluate an item such as “I am like the person who usesthis product” because people use the Internet and can’tmake sure who buys the same virtual goods. Based on theabove two reasons, we adopt the semantic differentialmethod to measure the congruency of product image andself-image. Here, we use Malhotral’s (1981) measure-ment items, which comprise 15 sets of adjectives. This


version provides good validity and fits both in measuringindividual self-image and product image.

The dependent variable is purchase intention, whichis measured separately by the real products and virtualproducts. Fishbein (1975) used “consider” and “willing”to measure the purchase intentions. It is a five-pointLikert-type scale from “very impossible” to “very pos-sible.” Tauber (1973) addressed the fact that past mea-surement items couldn’t examine whether products finallysatisfied a consumer’s needs. It caused the gap betweenpurchase intentions and actual re-purchase behaviors. Hepointed out that it could be improved by adding the itemsof “product importance” to find out the agreement betweenproducts and individual inner needs. These are also goodpredictors of consuming behaviors. Thus, we added twoitems to measure the importance of products to consumers.

The moderators are the three types of products. Thedefinitions are introduced in the former sections. Weconducted two pretests to find out the three types of realproducts and virtual products. The purpose of pretest oneand two is to find the representative goods of each type(i.e., ask people to classify the types of products byassigning 100 percent to the three product types) and totest the Chinese version of Malhotra’s self-image scale.To find the relevant product candidate for the real andvirtual products, we adopted the expert opinion method.The questionnaire was distributed in two versions: one forreal products and the other for virtual products. Theeffective sample consisted of 32 people (i.e., 20 on-linegame players and 12 common people). The final realproducts which were selected as the experiment stimulusin the main study are toothbrush (utilitarian real product),KTV (hedonic real product), and ring (symbolic realproduct). The virtual representative products are detoxifiedpanacea (utilitarian virtual product), e-lover (hedonicvirtual product), and knight badge (symbolic virtualproduct).

The questionnaire of the main study was composed offour parts. The first is to test the level of the individual self-image. The second is to test the level of product image.This second part will have two versions designed for realproducts and virtual products participants. The third partis to measure the purchase intention, which includes fouritems: “considering to buy,” “the purchase intention,”“purchasing this product is an important consumption forme,” and “owning this product is important for me.” Thefourth part is the classification of product types andproduct familiarity measurement. We asked people toassign the weight of listed products to the three productsagain. If this weighting is inconsistent with the result ofpretests, then we discarded the whole questionnaire. Wealso add the items of product familiarity (i.e., 7 pointLikert-type scale). The purpose of the product familiarityscale is to screen out the people who are unfamiliar with

the listed products. An especially key factor is whether thevirtual product questionnaires filled by people are valid.Past research showed that product ownership influencesthe result of the self-congruency (Birdwell 1968). For thevirtual products, it is hard to define ownership. Besides,the hedonic products bring about almost a feeling orexperiential enjoyment rather than the ownership of cer-tain products. For example, the consumption of KTV isfor singing, not owning the KTV facilities. Thus, wereplace the concept of ownership with familiarity (Barone,Shimp, and Sprott 1999) in the questionnaire item.

The two versions of the questionnaire were distrib-uted at Warner Village in Taipei City. We randomlyselected people and asked them whether they played on-line games frequently. If they answered yes, then thevirtual product version questionnaire was given to them.If not, then the real product version questionnaire wasgiven.

RESULT

The numbers of real product and virtual productquestionnaires are 228 and 201. After deducting thecopies that were not filled out completely, we finally had226 real product questionnaires and 192 virtual productquestionnaires. The female respondents make up 75.66percent of the real product questionnaires and male respon-dents account for 70.31 percent of the virtual productquestionnaire, thus apparently displaying the actual phe-nomenon that male users still account for a great part of allon-line users. The ages are between 16–25 years old (74%and 80% for real and virtual products separately). In themanipulation check, we discarded the questionnaires thathad a familiarity score of below 3. The three product typesalso reached a high accuracy rate of classification in boththe real products and virtual products categories (at leastabove 92% in real products and above 85% in virtualproducts).

The Cronbach’s α of purchase intention items is .76in the real product category questionnaires and is .8 in thevirtual product category questionnaires. Due to the highreliability, we used the mean of these four purchaseintention items to represent the purchase intention.

By aggregating the difference scores of 15 self-imageitems and product images, we obtained the congruencyscores of self-images. Because some difference scores arenegative, we adopted the Euclidean-Distance method toresolve this problem. The sum of these 15 self-imagedifference scores represents the level of the self-congruency. The higher the sum is, the lower thecongruency is. When the actual self-image and productimages are compared, we called this figure the differencescores of actual self-image. When the ideal self-image andproduct images are compared, we called this the difference


scores of ideal self-image. A regression equation is builtto test the relationship of the self-congruency and purchaseintention.

Y = ß0 + ß

1X + ε

Y is the mean of the four purchase intentions items.X is the differences scores of self-congruency scores. ß

0 is

the intercept. ß1 is the coefficient of the self-congruency.

is the error term. The result of the real products is shownin Table 1.

About the hedonic real product – KTV, we found it isnot significant both in the congruency of actual and idealself-image. That is, the congruency of actual self-imageand ideal self-image would not affect the purchase inten-tion of the hedonic real product. Thus, hypothesis H1a isnot supported.

Regarding the regression result of the symbolic realproduct – ring, the coefficient of the difference scores ofactual self is -.02 (p < .05). Besides, the Pearson correla-tion coefficient between it and purchase intension is -.208(p = .015). It means that when the difference scores ofactual self-image is lower (i.e., the congruence of actualself-image is higher), the purchase intention is higher.This result shows that the purchase intention of the sym-bolic real product is more influenced by the congruency ofactual self-image. Thus, hypothesis H1b is not supported.To further analyze which self-image item has a higherinfluence on the ring purchase intention, we used the 15self-image items as the independent variable and foundthat the item “economical-extravagant” is significant (i.e.,ß = -.162, p < .05, R2 = .048) toward the ring purchase

intention. When the congruency of “economical-extravagant” product image and self-image is higher, thering purchase intention is higher.

From Table 1, we found that the coefficients of thedifference scores of actual self-image and the differencescores of ideal self-image are not significant toward thepurchase intention of the hedonic real product (i.e., theKTV). For this reason, we couldn’t corroborate whichkind of congruency of self-image influences the purchaseintention of utilitarian real products more. Therefore,hypothesis 1c and hypothesis 1d are not supported.

The coefficient of the congruency of ideal self-imagetoward the symbolic real product (i.e., the ring) purchaseintention is not significant, but the congruency of actualself-image is significant, compared with the former resultthat it is not significant toward the purchase intention oftoothbrush (utilitarian real products). Therefore, hypoth-esis 1e is supported but hypothesis 1f is not supported.

The result of the virtual products is showed in Table 2.The coefficient of purchase intention of the hedonicvirtual product – e-lover and the difference scores ofactual self-image is -.038 (p < .05). The Pearson correla-tion coefficient between these two variables is -.254 (p =.0165). That is, when the difference score of the actualself-image is lower (i.e., the congruency of the actual selfis higher), the purchase intention of the virtual hedonicproduct is higher. It is not the same as what we originallyinferred. The hypothesis 2a is not supported. We exploredthe 15 self-image items as independent variables to regressto the purchase intention of e-lover and found that “com-fortable-uncomfortable” (ß = -.019, p < .05) and “domi-

TABLE 1The Regression Analysis of Self -Congruency Toward Real

Product Purchase Intention

Depend Variable: IDV: EffectivePurchase Intention Congruency of Image Sample ß

1F R2

Toothbrush Actual Self 207 -.001 .02 .000Ideal Self 205 -.000 .01 .000

KTV Actual self 174 -.007 .82 .005Ideal Self 172 -.004 .17 .001

Ring Actual self 136 -.020* 6.05* .043Ideal Self 134 -.015 3.54 .026

* p < .05

Note: The negative value of ß1 means when the differences score between self-image and product image is lower

(i.e., the congruency of self-image is higher), the purchase intention is higher.


nating-submissive” (ß = -.0128, p < .01) were significant.That is, when the congruencies of these two items of actualself-images were high, it could increase the e-lover pur-chase intention.

In the symbolic virtual product – knight badge’saspect, the regression coefficient of the congruency of theideal self-image toward purchase intention is -.018 (p <.05). The Pearson correlation coefficient between thesetwo variables is -.197(p = .0359). That means the differ-ence of ideal self-image has significant negative influenceon the purchase intention. Thus, hypothesis 2b is sup-ported. We next examined the 15 self-image items andfound that “rugged-delicate” (ß = -.017, p < .01), “color-less-colorful” (ß= -.018, p < .01) and “modest-vain” (ß =-.098, p < .05) total differences of self-congruency havesignificant negative impact on purchase intention. That is,when the three items of self-congruency of self-image arehigh, it leads people to have higher purchase intentions tothe knight badge.

From the regression result of Table 2, the ideal self-congruency has no significant coefficient toward thepurchase intention of the hedonic virtual product –e-lover. The actual self-congruency has a significantcoefficient toward the purchase intention of virtual hedonicproduct. However, the coefficient of actual self-congruencytoward utilitarian product – detoxified panacea isinsignificant. This implies that the actual self congruencyhas higher impact on the purchase intention of hedonicproducts than on that of utilitarian products. Therefore,hypothesis 2c is supported but hypothesis 2d is notsupported.

Also, we found that the self-congruency has nosignificant influence on the purchase intention of theutilitarian product – detoxified panacea. However, theactual self-congruency has no significant influence on thepurchase intention of utilitarian products and symbolicproducts. Thus, hypothesis 2e is not supported but hypoth-esis 2f is supported.

DISCUSSION AND MANAGERIALIMPLICATION

From this research, we found that if we divide self-image into “actual self-image” and “ideal self-image,”they have different influences on different product types.Even the same types of products are affected by dissimilarself-images. The result of the congruency of the real worldcannot be generalized to the virtual world. Findings willbe discussed as follows.

To the common people, utilitarian products (for ex-ample: a toothbrush) are for daily use. They do not belongto any specific group. It is hard to form a vivid image ofwho is the product user. When people purchase utilitarianproducts, they focus more on the product attributes thanon the self-congruency. This result is consistent withChang’s (2002) finding.

Regardless of the actual self-congruency or idealself-image, there are no significant influences on thehedonic real products’ purchase intentions. We examinedthe 15 self-image items and the result showed that thestandard deviations of KTV product image are lower thanthat of other types of products. It appeared that people

TABLE 2The Regression Analysis of Self Congruency Toward Virtual

Product Purchase Intention

Depend Variable: IDV: EffectivePurchase Intention Congruency of Image Sample ß

1F R2

Detoxified Panacea Actual self 134 -.009 1.16 .009Ideal Self 134 -.013 2.64 .020

E-Lover Actual self 89 -.038* 5.96* .064Ideal Self 89 -.0032 3.85 .042

Knight Badge Actual self 114 -.010 1.05 .009Ideal Self 114 -.018* 4.51* .039

* p < .05

Note: The negative value of ß1 means when the difference scores of self-image and product image is lower (i.e.,

the congruency of self-image is higher), the purchase intention is higher.


have high agreement when recognizing KTV’s image. IfKTV has a certain specified image, the congruency ofimage is not the dominant factor before consuming it.Other factors that should be considered are the location,the expense, and the service, etc. This will weaken theeffect of product image toward the purchase intention.Another possible explanation is that the consumptionmode of KTV is different from other products used in thisresearch. The most distinct point is the number of peoplewho collectively consume it. We usually purchase atoothbrush or a ring alone or with 1–2 accompanyingfriends. However, the number of people who consumeKTV is usually over four. The individual congruency ofself-image could be suppressed by the group conformitybehaviors.

One of the findings is that the purchase intention of avirtual hedonic product is significantly affected by thecongruency of actual self-image. We infer that people arerepressed by the social norm and moral restraint in the realworld. They release the pressure when they are situated inthe virtual world. Thus, people could want to use moneyto purchase an e-lover which can comfort their dull lives.This is a psychological feedback mechanism which needsto be further explored.

In the aspect of the purchase intention of a symbolicproduct, the real product, the ring, is influenced signifi-cantly by the congruency of actual self, while the virtualproduct, the knight badge, is significantly affected by thecongruency of ideal self-image. The definition of sym-bolic products clearly expounds that these goods couldbring about the satisfaction of people’s symbolic needs.The symbolic needs are divided two categorizations:“Role Position” and “Self Enhancement.” The formerexpresses to others that you belong to some class or somegroup. The ring is the symbolic tool to declare to others

that there is a strong love relationship between two lovers.The latter is the ambition to promote the self to a higherclass or certain noble group (Sirgy 1982). The knightbadge reveals the message that people who own this arepromoted to a higher military official rank. These twokinds of symbolic meanings might give some reasons forour research result.

For managerial practitioners, they can shape theirproduct image according to the target consumers’ self-images. At the same time, they also need to examine theproduct type to which their products belong. The utilitar-ian product should focus on the appealing productattributes. If the product is a hedonic product, managersshould put more efforts on fitting the actual self-image oftarget consumers. However, if this hedonic product isconsumed by group of purchasers, managers should putmarketing resources on the product qualities or othersubstantial benefits. When selling symbolic products,especially luxury and fashionable goods, managers shouldnotice the symbolic meanings embodied in their merchan-dise. If the symbolic product is classified as a role posi-tioning good, the marketer should shape the product’simage by the actual self-image of their target consumers.If the symbolic product belongs to the self-enhancementcategorization, managers should develop marketing strat-egies to maintain their product image according to theideal self-image of target consumers.

Different product types project different self-images.The market practitioner should make clear the producttype to which their products belong. Moreover, they canexamine this in detail through the 15 specific self-imageitems employed in this research and then shape relevantproduct images. This will strengthen the identificationwith target consumers and will effectively raise theirpurchase intentions.

REFERENCES

Barone, Michael J., Terence A. Shimp, and David E.Sprott (1999), “Product Ownership as a Moderator ofSelf-Congruity Effect,” Marketing Letters, 10 (1),75–85.

Bei, Lien-Ti (2004), The Study of the Usage of MobileMultimedia. Sony Ericsson Project.

Birdwell, Al E. (1968), “A Study of Influence of ImageCongruence on Consumer Choice,” Journal of Busi-ness, 41 (January), 76–88.

Bracken, Bruce A. and M. Susan Lamprecht (2003),“Positive Self-Concept: An Equal Opportunity Con-struct,” School Psychology Quarterly, 18 (2) 103–21.

Burns, Robert B. (1979), The Self Concept. New York:

Longman Group Limited.Chang, Chingching (2002), “Self-Congruence as a Cue in

Different Advertising Processing Contexts,” Com-munication Research, 29 (October), 503–36.

Dolich, Ira J. (1969), “Congruence Relationship BetweenSelf-Image and Product Brands,” Journal of Market-ing Research, 6 (February), 201–14.

Ekinci, Yuksel and Michawl Riley (2003), “An Investiga-tion of Self-Concept: Actual and Ideal Self-Congru-ence Compared in the Context of Service Evalua-tion,” Journal of Retailing and Consumer Services,10 (July), 201–14.

Fishbein, M. and I’Ajzen (1975), Belief, Attitude, Inten-tion and Behavior. Reading, MA: Addison-Wesley.

French, Warren A. and Alan B. Glaschner (1971), “Levelsof Actualization as Matched Against Life Style Evalu-


ation of Products,” Proceedings of the AmericanMarketing Association, 30, 358–62.

Greeno, Daniel W., Montrose S. Sommers, and Jerome B.Kernan (1973), “Personality and Implicit BehaviorPatterns,” Journal of Marketing Research, 10 (Janu-ary), 63–69.

Grubb, Edward L. and Harrison L. Grathwhohl (1967),“Consumer Self-Concept, Symbolism, and MarketBehavior: A Theoretical Approach,” Journal of Mar-keting, 31 (October), 22–27

Hoelter, Jon W. (1985), “The Structure of Self-Concept:Conceptualization and Measurement,” Journal ofPersonality and Social Psychology, 49 (May), 1392–1407.

Johar, J.S. and M. Joseph Sirgy (1991), “Value-Expres-sive versus Utilitarian Advertising Appeals: Whenand Why to Use Which Appeal,” Journal of Advertis-ing, 20 (September), 23–33.

Kempf, DeAnna S. (1999), “Attitude Formation fromProduct Trial: Distinct Roles of Cognition and Affectfor Hedonic and Functional Products,” Psychologyand Marketing, 16 (January), 35–50.

Lavidge, Robert J. and Gary A. Steiner (1961), “A Modelfor Predictive Measurements of Advertising Effec-tiveness,” Journal of Marketing, 25 (October), 59–62.

Levy, Sidney J. (1959), “Symbols for Sales,” HarvardBusiness Review, 37 (4), 117–24.

Malhotra, Naresh K. (1981), “A Scale to Measure Self-Concepts, Person Concepts, and Product Concepts,”Journal of Marketing Research, 18 (November),456–64.

Martin, Warren S. and Joseph Bellizzi (1982), “An Analy-sis of Congruous Relationships Between Self-Imagesand Product-Images,” Journal of the Academy ofMarketing Science, 10 (September), 473–89.

O’Curry, Suzanne and Michal Strahilevitz (2001), “Prob-ability and Model of Acquisition Effects on ChoicesBetween Hedonic and Utilitarian Options,” Market-ing Letters, 12 (February), 37–49.

Park, C. Whan, Bernard J. Jaworski, and Deborah J.MacInnis (1986), “Strategic Brand Concept-ImageManagement,” Journal of Marketing, 50 (October),135–45.

Petty, Richard E., John T. Cacioppo, and David Schumann(1983), “Central and Peripheral Routes to Advertis-ing Effectiveness: The Moderating Role of Involve-ment,” Journal of Consumer Research, 10 (Septem-ber), 135–46.

Rogers, Carl (1951), Client-Centered Therapy: Its Cur-rent Practices, Implications and Theory. Boston:Hougton Mifflin.

Rosenberg, Morris (1979), Conceiving the Self. Malabar,FL: Robert E. Krieger.

Schenk, Carolyn T. and Rebecca H. Holman (1980), “ASociological Approach to Brand Choice: The Con-cept of Situational Self-Image,” in Advances in Con-sumer Research, Jerry Olson, ed. Ann Arbor, MI:Association for Consumer Research, 7, 610–14.

Schrauger, J. Sidney and Adrian K. Lund (1975), “Self-Evaluation and Reactions to Evaluations from Oth-ers,” Journal of Personality, 43 (March), 94–108.

Sirgy, M. Joseph (1982), “Self-Concept in ConsumerBehavior a Critical Review,” Journal of ConsumerResearch, 9 (December), 287–300.

____________ (1985a), “Self-Image/Product-Image Con-gruity and Consumer Decision-Making,” The Inter-national Journal of Management, 2 (December), 49–63.

____________ (1985b), “Using Self-Congruity and IdealCongruity to Predict Purchase Motivation,” Journalof Business Research, 13 (June), 195–206.

____________, Dhruv Grewal, Tamara F. Manglegurg,Jae-ok Park, Kye-Sung Chon, C.B. Claibirne, J.S.Johar, and Harold Berlman (1997), “Assessing thePredictive Validity of Two Methods of MeasuringSelf-Image Congruence,” Journal of the Academy ofMarketing Science, 25 (Summer), 229–41.

Sommers, Montrose S. (1964), “Product Symbolism andthe Perception of Social Strata,” Proceedings of theAmerican Marketing Association, 22, 200–16.

Tauber, Edward M. (1973), “Reduce New Product Fail-ures: Measure Needs as Well as Purchase Interest,”Journal of Marketing, 37 (July), 61–64.

Woods, A. Walter (1960), “Psychological Dimensions ofConsumer Decision,” Journal of Marketing, 24(April), 15–19.

For further information contact:Chih-Ping Wang

National Chengchi UniversityNo. 64, Sec. 2, ZhiNan Road

Wenshan DistrictTaipei, 11605

TaiwanPhone: 886.960.275295



THE EFFECTS OF PARTICIPATION IN ONLINE BRANDCOMMUNITIES: FINDINGS FROM XBOX

COMMUNITIES IN BRAZIL

Stefânia Ordovás de Almeida, Universidade de São Paulo, BrazilUtpal Dholakia, Rice University, Houston

José Afonso Mazzon, Universidade de São Paulo, BrazilHugo Müller Neto, Universidade de São Paulo, Brazil

ABSTRACT

This study of Brazilian consumers proposes andempirically tests a theoretical framework explaining theeffects of online brand communities on the company’sbrand marketing and future consumer behavior. Resultsconfirm positive effects of the antecedents tested onparticipation and community identification. Communityidentification impacts behavioral effects, which influ-ences marketing effects.

INTRODUCTION

Brand communities, defined as “specialized non-geographically bound communities, based on a structuredset of social relationships among admirers of a brand,”(Muñiz and O’Guinn 2001, p. 412) are increasingly im-portant, not only as conduits of information, places ofsocial support, and channels of self-expression for con-sumers, but also as marketing programs for firms (e.g.,Algesheimer and Dholakia 2006; McAlexander, Schouten,and Koenig 2002). One consequence of this increasedinterest is that popular brands often have many estab-lished online brand communities (OBCs), resident on thefirm’s website or organized by customer enthusiasts. Forexample, Microsoft XBOX 360, the leading video gamingconsole, has an established OBC (www.xbox.com/Com-munity) hosted on the firm’s web-site and managed byMicrosoft managers. Concurrently, there are dozens ofXBOX OBCs, founded and managed by fans such as theBrotherhood of the Box (www.bob.com.sg).

The objective of this paper is to develop and test atheoretical framework (see Figure 1) describing theconsumer’s engagement within online brand communi-ties, along with key antecedents that drive engagementand their marketing and behavioral consequences onconsumers. We developed the framework not only bydrawing upon prior research from marketing, social psy-chology, and sociology, but additionally by conducting anextensive netnography of Brazilian XBOX communitymembers over a period of more than two years. Surveydata gathered from active participants of the two largestBrazilian XBOX online brand communities is used to test

our proposed framework. The data collection site is espe-cially significant because few previous studies have exam-ined online consumer behavior in Brazil, even though it isa vibrant world economy with one of the highest per capitarates of overall internet use (Fusco 2008) and internet usefor social purposes (IBOPE/NetRatings 2008).

A BRIEF REVIEW OF ONLINE BRANDCOMMUNITY RESEARCH

We study online brand communities, which reside onweb-sites and are sustained through online interactions, inthe current research. OBCs serve many different func-tions for consumers and firms. For consumers, OBCs actas conduits of information, channels for solving product-related problems and learning how to use its features,places for finding new friends for social support and formeeting existing friends, and as a means for self-expres-sion through creation and sharing of symbolic content(e.g., Muniz and O’Guinn 2001; McAlexander et al.2002). For firms, OBCs are low-cost, high-efficacy mar-keting programs which can achieve a number of differentmarketing objectives simultaneously: the abilities to con-duct quick and low cost marketing research with the targetaudience, deliver prompt customer service at low expense,educate and socialize new customers, strengthen attach-ment to the firm’s brand for existing customers, andincrease the frequency and loyalty of customer purchasebehaviors (Algesheimer and Dholakia 2006; Muñiz andSchau 2005).

Several aspects of brand communities are worthnoting. First, a brand community is a social collectiveorganized around one particular brand, which means thatthe collective comprises of consumers who have at leastsome heightened enduring interest in that brand (Munizand O’Guinn 2001). Such an interest may stem from anattachment to the brand itself, from a more general interestin the product category to which the brand belongs, or(more likely) both. Second, regular social interactions andcommunication between members, accomplished throughonline channels such as bulletin boards, chat-rooms, andemail lists, is essential not only for the community’sbusiness to be conducted, but perhaps more importantly,


for the relationships between community members toform and strengthen.

Third, OBCs must possess three markers that soci-ologists stipulate as essential for any social collective to betruly considered a community: (1) a consciousness ofkind, (2) a sense of moral responsibility, and (3) theknowledge and acceptance of the collective’s rituals andtraditions. The first core marker of community, the so-called consciousness of kind, refers to the intrinsic con-nection that OBC members feel toward one another througha sense of belonging to the group, and a sense of differenceor separation from those who are non-members (Cova1997; Muniz and O’Guinn 2001). The second core com-munity marker is a feeling or sense of moral responsibilityor obligation toward the OBC itself and its other members,and may include a concern for their well-being as ex-pressed through acts of help or social support such as byteaching newer members how to use the product, andeducating them about the practices and norms of the OBC.The third core community marker is the knowledge andacceptance of the OBC’s shared rituals and traditions byits members. This discussion is helpful in providing thebackground necessary to develop our framework.

THEORETICAL FRAMEWORK ANDRESEARCH HYPOTHESES

Our proposed theoretical framework is graphicallysummarized in Figure 1. As can be seen the frameworkdescribes three key antecedents, three mediators, and anoutcome variable “marketing effects.” Hypotheses regard-ing the relations between these dimensions arenow presented.

Perceived Psychographic Homogeneity – PPH

Even though its participants are united in their inter-est in the brand, to say that an OBC is diverse or homoge-neous is to recognize other qualitative or categoricaldistinctions among them that are evident to its participants(DiTomaso, Post, and Parks-Yancy 2007). We defineperceived psychographic homogeneity as “participants’perceptions of the extent to which other members aresimilar to them in their values, interests, and hobbies.”

Perceived psychographic homogeneity is importantto study in OBCs for several reasons. First, psychographicvariables are distinct segmentation and targeting vari-ables used to choose potential community participants bymanagers (e.g., Li and Bernoff 2008). Community partici-pants are aware of these similarities and differences, asthey come up in conversations. They can also be ascer-tained through members’ profiles, and via the content andmethods of their online interactions Organizational psy-chology research shows that greater team diversity slowsdown decision making, reduces the amount of intra-groupcommunication, leading to more communication break-downs, and interpersonal conflict (Greening and Johnson1997; Hambrick and D’Aveni 1992; Hoffman 1985;O’Reilly, Snyder, and Boothe 1993). Based on thesefindings, we argue that homogeneity perceptions will bebeneficial to participants, leading us to hypothesize:

H1: Perceived psychographic homogeneity of OBC mem-bers will have positive impacts on (a) communityparticipation (H1a) and community identification(H1b).

Perceived Psychographic Homogeneity

Availability of Virtual Venues

Community Participation

Behavioral Effects

Marketing Effects

Community Identification

Brand Relationship

Quality

H1a

H1b

H2a

H2b

H3 H6

H4

H5

H7

FIGURE 1Proposed Theoretical Framework


Availability of Virtual Venues – AVV

Regular social interactions and communicationbetween OBC members, accomplished through onlinechannels such as bulletin boards, chat-rooms, and emaillists, is essential not only for the community’s business tobe conducted, but perhaps more importantly, for therelationships between OBC members to form andstrengthen. It is through timely and convenient communi-cation ability that the OBC is able to solve the problems ofindividual members, deliver technical service and supportto them, generate consumer feedback and new productideas, and deepen participants’ knowledge of one anotherand the strength of their relationships (Mathwick et al.2008). For individual participants, communicating withothers within the OBC serves specific functional purposessuch as solving a particular product-related problem orlearning how to use a product feature, or may simplyprovide the means to have a pleasant and enjoyablecommunal experience (Dholakia et al. 2008). Therefore,the more the number and type of venues available toconsumers, the greater will be the positive effects onparticipation and community identification. Thus, wehypothesize:

H2: The availability of virtual venues to OBC memberswill have positive impacts on community participa-tion (H2a) and community identification (H2b).

Brand Relationship Quality – BRQ

Algesheimer et al. (2005, p. 23) define brand rela-tionship quality as “the degree to which the consumer seesthe brand as a partner in a satisfactory relationship with it.”It involves cognitive aspects such as the degree to whichthe consumer believes that the brand’s image overlapswith his or her self-image, and emotional elements such asthe degree of the consumer’s emotional attachment to thebrand. We note that such a conceptualization is in linewith how Algesheimer and colleagues (2005) definedbrand relationship quality, as well as with priorconceptualizations of related constructs such asBhattacharya and Sen’s (2003) “consumer-brand identi-fication” construct and De Wulf, Oderkerken-Schröeder,and Iacobucci (2001)’s “relationship with the brand”construct. We argue that the consumer’s relationship withthe brand is something he or she brings to the communityand consequently to the extent that it is strong, it willinfluence the degree to which she/he identifies with thecommunity. In hypothesizing this effect, we underscorethe fact that most consumers already have an existingrelationship with the brand prior to joining and participat-ing in OBCs.

H3: Brand Relationship Quality will have a positive im-pact on community identification.

Community Participation – CP

Refers to the amount of user interactions in thecommunity. Burnett (2000) and Algesheimer and Dholakia(2006) distinguishes between members who are “non-interactive” – those who only read, but don’t activelyparticipate in the community (also called as “lurkers” byother researchers) – and those who are “interactive” –those that generate content. Participation is a crucialprocess in OBCs and a key mediator in our frameworkbecause it impacts other significant variables.

In brand communities, it is very common that itsorganizers confer status upon members based on theircontribution to the community (Dholakia et al. 2008;Mathwick et al. 2008). Participants use their status toinfluence the community’s functioning and the othermembers (Algesheimer et al. 2005). Recent research alsoshows that the identification is related to the interactionsof participants with each other and with the communityand to the value that they derive from these interactions(Dholakia et al. 2008). Therefore consumers who partici-pate more actively in the community are those that have atendency to continue participating, to continue being amember and recommending the community to others(Algesheimer et al. 2005), which creates a greater identi-fication with that community. These results suggest thatthe greater the participation, the greater the consumer’sidentification with their peers, which leads to closer tiesand relationships established, thereby impacting the extentof community influence on members (Nambisan andBaron 2007). Based on this discussion, we hypothesize:

H4: Community participation will have a positive impacton community identification.

H5: Community participation will have a positive impacton behavioral effects.

Community Identification – CI

The second mediator we included in our conceptualframework is community identification, which priorresearch has found to be an important driver of behavioraland marketing outcomes (e.g., Algesheimer et al. 2005).Theoretically, it is viewed as an essential condition, viaconsciousness of kind, for a social collective to be consid-ered a community (Muñiz and O’Guinn 2001). Commu-nity identification draws upon self-categorization theory(e.g., Turner 1987) which posits that people are hard-wired to categorize the social world into groups so as tosimplify their thinking and interactions with others, andview themselves as members or non-members of specificsocial groups. It also relies on social identity theory (e.g.,Tajfel and Turner 1986) which suggests that a categoriza-tion process is used by individuals to derive a valued self-


identity, which in turn satisfies the drive for positive self-esteem. Based on these ideas, and consistent withAlgesheimer and colleagues (2005), we define commu-nity identification as “the degree to which the participantcategorizes himself or herself as a member of the brandcommunity.” A higher degree of community identifica-tion implies that the participant places greater value on hisor her identity as a brand community member.

Many prior studies have shown that social identifica-tion influences the individual’s actions concerning thesocial category (Algesheimer et al. 2005; McAlexanderet al. 2002). Also consistent is research showing thatindividuals identifying with a particular group tend toimitate its members to reinforce the shared identity (e.g.,Mackie 1986). Consequently, we expect the degree of thecommunity’s influence on members, operationalized bybehavioral effects, to be positively affected by communityidentification.

H6: Community identification will have a positive impacton behavioral effects.

Behavioral Effects – BE

Is the third mediator and responsible for the influencethe OBC’s have on participants’ decisions regarding theproduct and the brand. This variable is defined as theimpact of OBC opinions on opinions of individual com-munity members and their decisions in the product cat-egory and brand (Dholakia et al. 2008). A number ofbrand communities studies have shown that communitiesexert substantial influence on their members, impactingtheir thoughts and actions (e.g., Muñiz and Hamer 2001;Algesheimer et al. 2005; Thompson and Sinha 2008).Bagozzi and Dholakia (2006a) concluded that greatershared intentions among community members lead tomore behaviors related to the product in question, asmoney spent on the product and product use.

Consumers who are members of brand communitiestend to buy only one brand and to provide positive wordof mouth. They tend to act as aggressive defenders of thebrand. We hypothesize that the extent to which the com-munity influences thoughts, beliefs, and decisions regard-ing the brand will dictate how strong these marketingeffects are for the consumer (Algesheimer et al. 2005;Mathwick et al. 2008; Muniz and O’Guinn 2001). Thecommunity’s influence will translate into marketing suc-cess. Thus, we posit:

H7: Behavioral effects will have a positive impact onmarketing effects.

Marketing Effects – ME

Is the terminal construct in our proposed framework.It captures outcomes that occur according to the commu-nity expressed through brand loyalty, repurchase andadvocacy intentions, given their consequence to the firm.The more integrated is the consumer into the brand com-munity, the more loyal she/he will be to the brand, becauseidentification is more common with the brand that con-sumers consume as part of the community (McAlexanderet al. 2002). The formation of relationships with consum-ers who share their interests is a reliable way to persuadeconsumers and connect with it, such a connection influ-ences causing them to have more behaviors associatedwith the repurchase and loyalty.

METHOD

The first stage of our research was exploratory, andemployed netnographic observation and analysis (Kozinets2002) and in-depth interviews. We chose netnographybecause this method is designed specifically to study “thelanguage, motivations, consumption linkages, and sym-bols of consumption-oriented online communities”(Kozinets 2002, p. 70). We conducted netnographic re-search in six different OBCs over a period of more thantwo years to better understand processes and bases ofsocial interactions within them. Additional details aboutour netnography are available upon request.

To complement this investigation with the percep-tions of individual consumers that interact in differentOBCs, we performed in-depth interviews with partici-pants of OBCs for high involvement goods and services.The measures employed in the study were mainly derivedfrom the literature review and development of items basedon the results of this exploratory study following Churchill(1979). The items developed were then submitted to threemarketing professors to check for face validity and con-sistency. The appendix provides the measures used in thestudy (English translations) along with sources.

To test our proposed conceptual model, we con-ducted an empirical study with active members of the twolargest XBOX OBCs in Brazil during early 2008. Thefirst, Portalxbox (http://www.portalxbox.com.br), is anOBC organized and managed by customer enthusiasts ofXBOX; the second, XBOX Brasil (http://www.xbox.com/pt-BR), is the sole XBOX community managed byMicrosoft in Brazil. In both cases, data were gathered incooperation with the community’s managers through alink in the communities’ webpage to the research ques-tionnaire. Data were collected between January and April2008. The final sample had 555 valid cases, 336 fromPortal XBOX and 219 from XBOX Brasil.


To test the research hypotheses, data were analyzedusing SPSS 15.0 and LISREL 8.5 (Jöreskog and Sörbom1998). First, exploratory data analyses were done to allowa better understanding of the sample and to verify adequacyof the psychometric properties. Exploratory factor analy-sis and reliability analysis – assessed via Cronbach’sAlpha and item-item and item-total correlations – wereperformed for each community individually and for thefull dataset including aggregated data. No major differ-ences in factor distribution or reliability were foundbetween the two community samples. All subsequentanalyses reported in this paper are done with the fullsample. Next, we evaluated construct validity, the pro-posed model dimensions were submitted as differentmeasurement models to confirmatory factor analysis (CFA)(Jöreskog and Sörbom 1996). The measurement modelsadjustment criterion, allied to unidimensionality, reliabil-ity, convergent, and discriminant validity (Jöreskog andSörbom 1996; Bagozzi et al. 1991; Steenkamp and Trijp1991; Dunn et al. 1994; Garver and Mentzer 1999) wereused to support construct validity. Content validity wasachieved through logical-theoretical analysis (Peter 1981;Devellis 1991) verified in the data collection measuresdevelopment.

The structural equation model (SEM) included allrespondents and was used to test the hypotheses. Good-ness-of-fit of the models was assessed with χ2-tests, theRoot Mean Square Error of Approximation (RMSEA),the Non-Normed Fit Index (NNFI), the Goodness-of-fitIndex (GFI); the Adjusted Goodness-of-fit Index (AGFI)and the Comparative Fit Index (CFI). Details regardingthese indices can be found in Bentler (1990) and Marsh,Balla, and Hau (1996). The Maximum Likelihood (ML)method was used. All analyses were performed usingcovariance matrices.

RESULTS

Demographics: Virtually all respondents are male:98.9 percent male, 1.1 percent female. Respondents rangedin age from under 18 to over 40. 17.5 percent were under18, 33.5 percent were between 18 and 25, 27 percent were26 to 30, 16 percent were 31 to 35, 4.3 percent werebetween 36 and 40, and the remaining, 1.6 percent wereover 40 years of age. By gross monthly household income,22.4 percent made less than 3,000 Reals1, 39.2 percentmade between 3,000 and 7,500 Reals, and 37.8 percentmade more than 7,500 Reals. Respondents also reporteda number of different professions.

Evaluation of the Measurement Model

Unidimensionability and Internal Consistency. Be-sides the adjustment of measurement models that provedto be in accordance with literature recommendations –

GFI, NNFI, and CFI > 0.90 and RSMEA < 0.08 (Hair et al.1998), we used three other measures to evaluate internalconsistency of constructs. Unidimensionality was pur-sued and achieved through examination of standardizedresiduals proposed by Garver and Mentzer (1999),Jöreskog and Sörbom (1988) and Steenkamp and Trijp(1991). According to the authors, in order to presentunidimensionality, a dimension should have all its stan-dardized residuals equal to or lower than |2.58|. Compos-ite reliability (ñ

å) is a measure analogous to coefficient

(Fornell and Larcker 1981; Bagozzi and Yi 1988), andaverage variance extracted (AVE; ñ

VC(å)) estimates the

amount of variance captured by a construct’s measuresrelative to random measurement error (Fornell and Larcker1981). Estimates of ñ

å above .60 and ñ

VC(å)) above .50 are

viewed as indicating good internal consistency (Bagozziand Yi 1988; Fornell and Larcker 1981). As Table 1shows, the values for all constructs for these three indexesare indicative of good unidimensionability and internalconsistency.

Convergent Validity. For Garver and Mentzer (1999)and Bagozzi et al. (1991), the convergent validity of ascale is given by the quality of its fit-indices. Besides allthe scale dimensions having been validated through theexamination of their fit-indices, the factor regressioncoefficients were also verified to certify the convergentvalidity. All the factor regression coefficients were statis-tically significant, which, according to Bagozzi et al.(1991), means that t-values are higher than |1.96| for p <0.05. As a strong condition for convergent validity,Steenkamp and Trijp (1991) suggest that all the factorregression coefficients should exceed 0.5. Both condi-tions were achieved and are provided in Table 1.

Discriminant Validity. We evaluated discriminantvalidity of the model using the procedure suggested byFornell and Larcker (1981) and widely used by otherresearchers (e.g., De Wulf et al. 2001; Ramani and Kumar2008). The AVE (ñ

VC(å)) for each of the seven factors was

compared to the highest variance that the factor sharedwith other factors in the model. These results are providedin Table 2.

As can be seen, the AVE extracted for each factor wasalways greater than the highest shared variance, indicat-ing the factors are discriminant.

Structural Model Estimation

Considering that constructs were validated we moveda step toward the structural model test. Only one strongmodification index was identified between latent vari-ables, suggesting a new path from Brand RelationshipQuality and Marketing Effects (BRQ ME – index 88,39).According to Bagozzi and Baumgartner (1994) one of the


biggest problems in structural models are related to weakspecification that can also occur by omission of relevantvariables, as for inclusion of irrelevant ones, as well as dueto bad specification for relations between latent variables.In this sense the authors recommend that besides a stronga priori theoretical base, alternative theoretical pathsshould be taken into consideration so relevant relationsbetween latent variables will not be undetected. The fitstatistics for this full model including the new path were:

(χ2): 1162.12; (Df): 312; χ2/Df: 3.725; p: .000; GFI: .87;NNFI: .91; CFI: .92; RMSEA: .07. The χ2 is significant(p < .05); however, this is usually the case for large samplesizes like the one we have. The NNFI, CFI, and RMSEAare within acceptable ranges, whereas the GFI is slightlylower than the threshold.

Next step we evaluated the proposed hypotheses.Considering the impact of perceived psychographic homo-

TABLE 1Unidimensionability, Internal Consistency, and Convergent Validity

Unidimens. Internal Consistency Convergent Validity

Standardized Composite Lowest Standardized LowestLatent Variables Residuals Reliability AVE Factor Loading t-value

Perceived Psychographic |2.58| 0.819 0.605 0.60 13.68Homogeneity

Av. of Virtual Venues |1.70| 0.856 0.749 0.75 19.63

Brand Relationship |2.48| 0.948 0.788 0.74 18.59Quality

Community Identification |1.95| 0.897 0.744 0.68 17.26

Community Participation |1.14| 0.941 0.763 0.75 19.63

Behavioral Effects |0.86| 0.916 0.735 0.71 18.28

Marketing Effects |2.24| 0.955 0.811 0.78 21.47

TABLE 2Discriminant Validity of the Constructs

PPH AVV CI BRQ CP BE ME

PPH .61

AVV .03 .75

CI .17 .15 .74

BRQ .08 .03 .14 .79

CP .09 .06 .50 .05 .76

BE .12 .08 .32 .11 .25 .74

ME .02 .02 .02 .21 .01 .07 .81


geneity on community participation, we found that, asexpected, it has a strong positive impact (γ = .31, p < .001)supporting H1a . Also as predicted, perceived psycho-graphic homogeneity impacts community identificationsignificantly (γ = .21, p < .001), supporting H1b. Hypoth-eses H2a and H2b checked the impact of availability ofvirtual venues on community participation (γ = .24, p <.001) and on community identification (γ = .24, p < .001).As one can see in both cases there is a positive directimpact of the antecedent variable on the other ones,providing support to both hypotheses. H3 examines theimpact of brand relationship quality on community iden-tification. Also as predicted there is a significant positiveimpact (γ = .17, p < .001), providing support to thishypothesis. H4 presupposed the impact of communityparticipation on community identification (β = .63, p <.001). This relation was strongly supported. The impact ofcommunity participation on behavioral effects was notsignificant (β = -.002, ns), not providing support to H5.Community identification proved to have a strong posi-tive impact on behavioral effects (β = .72, p < .001),supporting H6. The impact of behavioral effects on mar-keting effects proved to be significant and positive (β =.08, p < .05), supporting H7. The new proposed relationfrom brand relationship quality to marketing effects alsoproved positive significant effects (β = .46, p < .001). Thepercentages of variance in community identification, com-munity participation, behavioral effects and marketingeffects explained by their respective antecedents were 82percent, 18 percent, 49 percent and 24 percent respec-tively. Also four rival models were tested in order to betterassure the validity of the proposed model. All modelsprovided a poorer fit considering adjustment indexes andsignificant paths, providing evidence for the proposedmodel. Details for these procedure are available uponrequest.

GENERAL DISCUSSION ANDCONCLUSIONS

In the current research, we studied the effects ofperceived psychographic homogeneity and availability ofonline venues on community participation and commu-nity identification, as well as the effects of brand relation-ship quality on these two variables. Also the effects ofthese variables on firm-relevant outcomes such as behav-ioral effects and marketing effects were investigated.Information regarding how these variables affect thebrand community’s effectiveness is critical for its manag-ers in making decisions regarding which consumers totarget. Our analysis found support for our conceptualframework in a sample of 555 active members of the twolargest XBOX brand communities in Brazil.

Results show that community identification is ex-plained through community participation, but perceivedpsychographic homogeneity, availability of online ven-

ues and brand relationship quality also help to enhancesuch identification. Community participation, in turn,depends more strongly upon perceived psychographichomogeneity followed by availability of online venues.

Community identification had the highest impact onbehavioral effects. Marketing effects were explainedthrough effects from brand relationship quality (althoughthis relation was not in the hypothesized model) andbehavioral effects, as initially hypothesized. As market-ing effects are much related to loyalty to the parent brand,in this case Microsoft, on several dimensions (Zeithamlet al. 1996; Bennett and Rundle-Thiele 2005) as satisfac-tion, word of mouth and loyalty, it is desirable, from aconsumer behavior perspective, that relationship qualitythat consumer has with XBOX is an antecedent of suchmarketing effects, which theoretically justify the inclu-sion of such a previously unhypothesized relation in thetheoretical model.

Another interesting result to be noted, and havingsignificant academic relevance, refers to the corrobora-tion of the relation proposed by Algesheimer et al. (2005)that assumed a previous identification with the brand(brand relationship quality) enhances consumer integra-tion and identification with the brand community. Alsothis examination directly answers Bagozzi and Dholakia’s(2006b) call for future studies to investigate the power anddirection of the relation between brand relationship qual-ity and community identification.

To summarize, our framework provides a nice addi-tion to the corpus of studies examining psychologicalprocesses leading to social influence in brand communi-ties by examining relationships between key variables andtesting them in a Brazilian context.


Marketing professionals are very interested in theway communities will create value for their companies(Algesheimer et al. 2005). In this respect, one relevantmanagerial outcome of this study is related to the valueonline brand communities have as a research and informa-tion source for companies. Communities show their valueas a good marketing investment for organizations, notonly considering enthusiasts that take a part in such socialinteractions, but also the whole market segment that canbe influenced by segmentation and positioning strategiesthat were drawn based on subsides derived from thesecommunities (Ridings et al. 2002; Algesheimer andDholakia 2006).

Our results also shed light on an interesting paradoxabout marketing effects of online brand communitiesrelated to the fact that some of the most favorable resultsof brand communities are in fact related to brand relation-


ship quality and not to the community itself. In this sense,even there are evidences that community influences onbuying decisions affect marketing effects, as satisfaction,word of mouth and loyalty, apparently such effect aremore related to the consumer’s relationship with the brandthan with the community itself. These results, for example,can lead companies to evaluate when it is interesting forcompanies to develop official communities and whenthey can lead consumers take care of such a task.

From a manager’s perspective, we found two impor-tant drivers of positive processes and outcomes in com-munities that they have some control over: participants’perceptions of psychographic homogeneity and the avail-ability of virtual venues. Consider homogeneity percep-tions first, it is clear that consumers value similarity infellow participants regarding their values, interests andoutlooks, and to the extent that these issues are brought tothe forefront and discussed, the social influence of theOBC is likely to increase, according to our findings.Managers could do so by providing specific venues todiscuss these issues, organize events focusing on com-monalities, and do other creative things marketing-wise toreinforce the common interest. Likewise, our study findsthat greater availability of virtual venues has a positiveeffect on participation and identification, suggesting thatthis may be one cost-effective way that OBC managerscan produce positive outcomes.

In this sense, the company must have a strong rela-tionship with communities, helping them in a variety ofonline activities. However, our study also raises the impor-tant issue of the role of offline interactions that meritsfurther attention. In particular, if online venues are benefi-cial, does pushing participants to interact face-to-faceproduce further positive benefits? According to Lin (2007),although most of activities in online communities willhappen in online environments, the social bonds betweenconsumers cannot be sustained if there are not strongoffline interactions. Consequently, face to face interac-tions would help to create connections between members,

enhancing community value for companies (Shang et al.2006), besides collaborating for a sense of belongingbetween community members encouraging them to par-ticipate and share information (Lin 2007). A better under-standing of companies’ value in face to face online com-munities’ events, and also the role of face to face interac-tions in the development and strength of bonds betweencommunity members appears as a field camp for futurestudies in the area.

Limitations and Future Suggestions

Despite the study’s contributions, it does have somelimitations. The first limitation regarding study scope andgeneralization of results refers to the inclusion of only oneproduct and brand in this study. Despite the adherence ofsuch a brand and product to the study purposes and also tothe necessary characteristics for a brand to have a brandcommunity, we are aware the need of replication of theproposed model in another communities for other prod-ucts and brands, including consumers of a different nation-ality. Another limitation is that although the use of Brazil-ian respondent samples is novel and extends the study ofbrand communities to a new nationality, it does indicatecaution when extrapolating these findings to consumersof other countries. Still, we have no reason to expectsignificant cultural differences between Brazilian XBOXcustomers and those of other countries. Also anotherscope limitation refers to the profile of the customermanaged community studied as portalxbox has a structureand organization that is not easily found in consumer-managed communities.

In conclusion, our present findings provide new anduseful insights in answering key questions regarding theeffectiveness of OBCs. Online brand communities offer aunique approach for delivering low-cost service supportto the firm’s customers for virtually all products. As such,marketing researchers are offered a fertile domain to studyinteresting and practically important group-level pro-cesses and phenomena.

ENDNOTE

1 As of October 25, 2008, 1 Brazilian Real = .44 U.S.Dollars.

REFERENCES

Algesheimer, René, Utpal M. Dholakia, and AndreasHerrmann (2005), “The Social Influence of BrandCommunity: Evidence from European Car Clubs,”Journal of Marketing, 69 (3), 19–34.

____________, and ____________ (2006), “Do Cus-tomer Communities Pay Off?” Harvard BusinessReview, 84 (11), 26–30.

Bagozzi, Richard P. and Youjae Yi (1988), “On theEvaluation of Structural Equation Models,” Journalof the Academy of Marketing Science, 16 (1), 74–94.

____________ et al. (1991), “Assessing Construct Valid-ity in Organizational Research,” Administrative Sci-ence Quarterly, 36, 421–58.

____________ and Hans Baumgartner (1994), “TheEvaluation of Structural Equations Models and


Hypothesis Testing,” in Principles of MarketingResearch, Richard P. Bagozzi, ed. Oxford, England:Blackwell Publishers, 386–422.

____________ and Utpal M. Dholakia (2006), “Anteced-ents and Purchase Consequences of Customer Par-ticipation in Small Group Brand Communities,” Inter-national Journal of Research in Marketing, 23 (1),45–61.

Bentler, Peter M. (1990), “Comparative Fit Indexes inStructural Models,” Psychological Bulletin, 107 (2),238–46.

Bhattacharya, C.B. and Sankar Sen (2003), “Consumer-Company Identification: A Framework for Under-standing Consumers’ Relationships with Compa-nies,” Journal of Marketing, 67 (April), 76–88.

Churchill, Jr., Gilbert A. (1979), “A Paradigm for Devel-oping Better Measures of Marketing Constructs,”Journal of Marketing Research, 16 (February), 64–73.

De Wulf, Kristof, Gaby Odekerken-Schröder, and DawnIacobucci (2001), “Investments in Consumer Rela-tionships: A Cross-Country and Cross-Industry Explo-ration,” Journal of Marketing, 65 (4), 33–50.

Devellis, R.F. (1991), Scale Development: Theory andApplications. Newbury Park, CA: Sage.

Dholakia, Utpal M., Vera Blazevic, Caroline Wiertz, andRené Algesheimer (2008), “Communal Service Deliv-ery: How Customers Benefit From Participation inFirm-Hosted Virtual P3 Communities,” WorkingPaper, Rice University.

DiTomaso, Nancy, Corinne Post, and Rochelle Parks-Yancy (2007), “Workforce Diversity and Inequality:Power, Status, and Numbers,” Annual Review ofSociology, 33, 473–501.

Dunn, Steven C. et al. (1994), “Latent Variables in Busi-ness Logistics Research: Scale Development andValidation.” Journal of Business Logistics, 15 (2),45–172.

Fornell, Claes and David F. Larcker (1981), “EvaluatingStructural Equation Models with Unobservable Vari-ables and Measurement Error,” Journal of MarketingResearch, 18 (February), 39–50.

Garver, Michael S. and John T. Mentzer (1999), “Logis-tics Research Methods: Employing Structural Equa-tion Modeling to test for Construct Validity,” Journalof Business Logistics, 20 (1), 33–57.

Hambrick, Donald C. and Richard A. D’Aveni (1992),“Top Team Deterioration as Part of the DownwardSpiral of Large Corporate Bankruptcies,” Manage-ment Science, 38 (10), 1445–66.

Jöreskog, Karl and Dag Sörböm (1996), LISREL 8: User’sReference Guide. Chicago: Scientific Software Inter-national.

Kozinets, Robert V. (2006), “Click to Connect:Netnography and Tribal Advertising,” Journal ofAdvertising Research, 46 (3), 279–88.

Mackie, Diane M. (1986), “Social Identification Effectsin Group Polarization,” Journal of Personality andSocial Psychology, 50 (4), 720–28.

Marsh, Herbert W., John R. Balla, and Kit T. Hau (1996),“An Evaluation of Incremental Fit Indices: A Clari-fication of Mathematical and Empirical Properties,”in Advanced Structural Equation Modeling. Issuesand Techniques, Geroge A. Marcoulides and RandallE. Schumacker, eds. Mahwah, NJ: Erlbaum, 315–53.

Mathwick Charla, Caroline Wiertz, and Ko de Ruyter(2008), “Social Capital Production in a Virtual P3Community,” Journal of Consumer Research, 34 (6),832–49.

McAlexander, James H., John W. Schouten, and HaroldF. Koenig (2002), “Building Brand Community,”Journal of Marketing, 66 (1), 38–54.

Muñiz, Jr., Albert M. and Thomas C. O’Guinn (2001),“Brand Community,” Journal of Consumer Research,27 (4), 412–32.

____________ and Lawrence O. Hamer (2001), “UsVersus Them: Oppositional Brand Loyalty and theCola Wars,” Advances in Consumer Research, 28(1), 355–61.

____________ and Hope Jensen Schau (2005), “Religi-osity in the Abandoned Apple Newton Brand Com-munity,” Journal of Consumer Research, 31(4), 737–47.

____________ and ____________ (2007), “VigilanteMarketing and Consumer-Created Communications,”Journal of Advertising, 36 (3), 35–50.

O’Reilly, Charles A., Richard C. Snyder, and Joan N.Boothe (1993), “Effects of Executive Team Demog-raphy on Organizational Change,” in OrganizationalChange and Redesign: Ideas and Insights for Improv-ing Performance, George P. Huber and William H.Glick, eds. Oxford, U.K.: Oxford University Press,147–75.

Peter, J. Paul (1981), “Construct Validity: A Review ofBasic Issues and Marketing Practices,” Journal ofMarketing Research, 18, 133–45.

Ramani, Girish and V. Kumar (2008), “Interaction Orien-tation and Firm Performance,” Journal of Marketing,72 (1), 27–45.

Ridings, Catherine M., David Gefen, and Bay Arinze(2002), “Some Antecedents and Effects of Trust inVirtual Communities,” Journal of Strategic Informa-tion Systems, 11, 271–95.

Schouten, John W. and James H. McAlexander (1995),“Subcultures of Consumption: An Ethnography ofthe New Bikers,” Journal of Consumer Research, 22(1), 43–61.

Steenkamp, Jan-Benedict E.M. and Hans C.M. Trijp(1991), “The Use of LISREL in Validating Market-ing Constructs,” International Journal of Researchin Marketing, 8, 283–99.

Tajfel, Henri and John C. Turner (1986), “The Social


Identity Theory of Inter-Group Behavior,” in Psy-chology of Intergroup Relations, Stephen Worcheland William G. Austin, eds. Chicago: Nelson-Hall,7–24.

Thompson, Scott A. and Rajiv K. Sinha (2008), “BrandCommunities and New Product Adoption: The Influ-

ence and Limits of Oppositional Loyalty,” Journal ofMarketing, 72 (6), 65–80.

Turner, John C. (1987), Rediscovering the Social Group:A Self-Categorization Theory. New York: BasilBlackwell.

Missing references are available upon request.

APPENDIX 1Summary of Constructs and Their Measures

(English Version of Measures)

Perceived Psychographic Homogeneity (PPH)(v1) Other [Community] members and I share the same values. 1/9; (v2) I think the members of [Community]are very similar in their interests and hobbies; (v3) I have the same outlook on life as most other [Community]members do.

Availability of Online Venues (AVV)(v4) [Community] has many highly interactive venues such as chat rooms for its members to communicate withone another; (v5) [Community] has several options of online venues.

Community Identification (CI)(v6) I perceive myself to be part of [Community].2/7/9; (v7) I have meet wonderful people thanks to this brandcommunity2/7; (v8) The image that I have formed of myself fits perfectly with the [Community]’s identity7.

Brand Relationship Quality (BRQ)(v9) I feel emotionally attached to the XBOX brand 3; (v10) The XBOX brand has an important role in my life2/7; (v11) The XBOX brand’s image and my image are similar in many respects 2/7; (v12) I feel a strong senseof identification with the XBOX brand 3/7; (v13) This brand says a lot about the kind of person I am2/7.

Community Participation (CP)(v14) I help other [Community] members by answering their questions10; (v15) I spent a lot of time each weekcommunicating with others in [Community]10; (v16) I am motivated to participate in various [Community]community activities2; (v17) I like to provide support to other members in [Community]2; (v18) I usuallycontribute to generate an interesting communication flow between community members.

Behavioral Effects (BE)(v19) [Community] has a lot of influence in the way I take my buying decisions regarding video and onlinegames; (v20) This community has a lot of influence in my opinions about this brand; (v21) After participatingin the community, I buy more Microsoft products; (v22) I will continue to buy Microsoft products because ofmy attachment to this community.

Marketing Effects (ME)(v23) I’m very satisfied with Microsoft; (v24) I say positive things about Microsoft to my friends and relatives5;(v25) I will recommend Microsoft to anyone that asks for my advice2/4/5/6/8; (v26) I intend to continue buyingMicrosoft products in the future4/5/8; (v27) I will continue to be a loyal Microsoft customer5/6.

Items adapted 1Maxham III and Netemeyer (2003) by Porter and Donthu (2005); 2Algesheimer et al. (2005);3Allen and Meyer (1990) by Fullerton (2005); 4Mathwick (2002); 5Zeithaml et al. (1996) by Porter and Donthu(2005); 6Zeithaml et al. (1996) by Fullerton (2005); 7Algesheimer and Dholakia (2006); 8Floh and Treiblmaier(2006); 9Maxham III and Netemeyer (2003) by Porter and Donthu (2008); 10Bhattacharya et al. (1995) byDholakia et al. (2008). Non-numbered items resulted from exploratory research.


For further information contact:Stefânia Ordovás de Almeida

Universidade de São Paulo (USP)Pontifícia Universidade Católica do Rio Grande do Sul (PUCRS)

Afonso Taunay 180/605Porto Alegre/RS

Brazil – 90520540Phone: 005551.33203547

or005551.30266532

Fax: 005551.33203547E-Mail: [email protected]


CALLING OR BLATANT COMMERCE? CLIENTS’ VIEWS ONEFFORTS OF A FOR-PROFIT ORGANIZATION TO ENGAGE

IN A SOCIETAL ISSUE

Johan van Rekom, Erasmus University, The NetherlandsFrank Go, Erasmus University, The Netherlands

Dayenne Calter, Erasmus University, The Netherlands

SUMMARY

It is a great challenge for marketers to make theirbrand the most compelling symbol for a value societydeems important (Holt 2004). Johnson & Johnson’s“Campaign for Nursing’s Future” in 2002, intended toenhance the nursing profession, provides a telling exampleof how such an endeavor can be successful. Nursesreported feeling a surge of pride when they viewed thecommercials. The campaign appeared to foster theorganization’s reputation (Hatch and Schultz 2008).However, such efforts to enrich a brand with societalsignificance are not without risk. The claims anorganization makes should appear to be untainted by anypressure from the market economy – otherwise they maycome across as inauthentic window-dressing which canhave a negative effect on customers’ loyalty. Our corequestion is: What makes the difference between asuccessful endeavor to emphasize the societal meaning ofa corporate brand and a “fake” appeal to societal valuesthat makes clients wonder to what length an organizationis willing to go to enhance its profit?

We identify three key conditions that jointly underliethe success of such endeavors: perceived fit, perceivedauthenticity, and the perception that features of the soci-etal endeavor flow forth from the organization’s charac-teristics. An example of this last condition is that peoplemay believe that an organization’s intention to foster amore flexible way of working logically flows forth fromits innovative character. Following Tauber (1988, p. 28),we conceptualize “fit” as “when the customer acceptsengagement into the societal issue as logical and wouldexpect it from the organization.” Fit is a condition for theorganization to come across as authentic. Both perceivedfit and authenticity depend upon the degree to whichpeople perceive the societal contribution to flow forthlogically from the organization’s characteristics. We ar-gue that if these three conditions are fulfilled, the audience’sattitude toward the organization’s societal contributionwill be more positive and that they will be more loyal to theorganization.

This reasoning leads to five hypotheses. Firstly, themore people perceive the features of an organization’s

societal endeavor to flow forth from its own intrinsiccharacteristics the more (1) people will perceive a fitbetween the organization and its societal endeavor and (2)the more the organization will come across as authentic.As far as the societal endeavor is concerned: (3) morepeople perceive it to fit the organization, the more positivetheir attitude toward it will be. The perceived fit will (4)partially mediate the relation between the degree to whichthe societal endeavor flows forth from organization’scharacteristics and the perceived authenticity of the orga-nization. Finally, (5) the more the organization comesacross as authentic, the more customers will remain loyal.

A software house, labeled “SH” for reasons of confi-dentiality, provided the research setting. They engaged ina program to make people take more advantage from theopportunities of information and communication technol-ogy to improve their work life, “the Work Project.”Qualitative research delivered six attributes of SH andfive of “the Work Project.” These formed the input for anonline survey, which delivered 213 usable questionnairesfrom clients. The hypotheses were tested using structuralequation modeling.

The data supported all five hypotheses. Clients’ loy-alty intentions were fully mediated by perceptions ofauthenticity and their attitude toward the Work Projectwas fully mediated by the fit they perceived between theorganization and its Work Project. Perceived fit partiallymediated the relation between the degree to which fea-tures of “the Work Project” depended upon organizationcharacteristics and the perceived authenticity of the orga-nization. Surprisingly, if only the two characteristics ofSH that clients perceived to cause most of the societalendeavor’s features were included in the analysis, ourmodel fitted even slightly better. The results suggest thatthe positive effects of societal endeavors as undertaken bySH can be achieved with just two well chosen key orga-nization characteristics, which can be uncovered with theapproach designed in this study. Such a small numberfacilitates the task of marketing officers to communicatehow the organization’s societal ambitions flow forthlogically from the organization’s characteristics, and facili-tates its application in practice.


REFERENCES

Hatch, Mary-Jo and Majken Schultz (2008), Taking BrandInitiative: How Companies Can Align Strategy, Cul-ture and Identity Through Corporate Branding. SanFrancisco: Jossey Bass.

Holt, Douglas (2004), How Brands Become Icons: ThePrinciples of Cultural Branding. Boston, MA:Harvard Business School Press.

Tauber, Edward (1988), “Brand Leverage: Strategy forGrowth in a Cost-Control World,” Journal of Adver-tising Research, 28 (August/September), 25–30.

For further information contact:Johan van Rekom

Department of Marketing, RSMErasmus University

P.O. Box 17383000 DR Rotterdam

The NetherlandsPhone: +31104081967

Fax: +31104089011E-Mail: [email protected]


TO GIVE AND GET BACK: DO CAUSE PORTFOLIOCHARACTERISTICS INFLUENCE THE RETURN

ON CORPORATE SOCIAL RESPONSIBILITY?

A. Meike Eilert, University of South Carolina, ColumbiaStefanie Rosen, University of South Carolina, Columbia

SUMMARY

In 2008, corporate social responsibility (CSR) was ona steady increase, with company contributions to charitiestotaling $4.4 billion, an increase of 90 percent since 1990(Foundation Center 2009). A firm’s engagement in CSRactivities is associated with several benefits such as stron-ger customer-company relationships (Sen and Bhatta-charya 2001), increased customer satisfaction (Luo andBhattacharaya 2006), positive product evaluations (Brownand Dacin 1997), and equity preservation (Godfrey,Merrill, and Hansen 2009). Therefore, charitable givingcan be viewed as an investment that can help secure thecompany’s competitive advantage. However, researchhas also shown that the relationship between corporategiving and value creation is nonlinear (Wang, Choi, andLi 2008). Thus, it is relevant for managers to understandhow to manage their company’s charitable activities toachieve a high return on investment.

By engaging in a variety of activities that result incharitable giving (such as direct donations, cause-relatedmarketing, or sponsorships), companies, intentionally orunintentionally, create cause-portfolios through formingrelationships with these causes that are visible to theirstakeholders. These relationships can exist at the organi-zational, brand, or product level (Varadarajan and Menon1988) and can typically be categorized as communalobligation, goodwill building, or strategic giving (Porterand Kramer 2002). Thus, we conceptualize cause portfo-lios as consisting of all cause-related relationships that acompany forms as a consequence of engaging in sociallyresponsible practices. Consequently, cause portfolio man-agement refers to the strategic management of all thecompany’s cause-related relationships including selec-tion, maintenance, and exit of the causes. Here, we advancepropositions as to how cause portfolio characteristics caninfluence a company’s performance.

First, we distinguish cause portfolios through theirstructure, which refers to the number of causes to whicha company donates. While companies tend to supportcauses that its key stakeholders want it to support(Drumwright 1996), it has also been shown that agencycosts increase with charitable giving, resulting in a curvi-linear relationship between the magnitude of charitable

giving and company performance (Wang, Choi, and Li2008). We propose a similar effect for the number ofcauses to which a company donates. Depending on thestakeholders, companies can be expected to donate toseveral causes. However, giving to too many causes canincrease costs associated with charitable giving, forexample, through an unfocused cause portfolio and sub-sequently diluted philanthropic reputation. Hence, a com-pany has to find a balance between supporting too few ortoo many causes in order not to dilute its philanthropicreputation or alienate important stakeholders. Thus, weadvance the proposition that cause portfolios with a mod-erate number of causes perform better than ones with alow or high number of causes.

Our second proposition refers to the similarity andsubstitutability of causes within the portfolio. Specifi-cally, we propose that causes within a cause portfolio haveto be high in similarity and low in substitutability. Simi-larity has been used to describe brands in brand portfoliosand refers to the extent to which brands within the firm’sbrand portfolio are positioned close to one another andappeal to the same consumer segments (Morgan and Rego2009). Therefore, similarity within the portfolio increasesthe focus of the cause portfolio and should increase theefficiency of resource allocation among these causes.Causes that are substitutable, however, have to competefor company’s resources. By dedicating resources to twosubstitutable charities, the company forgoes the opportu-nity to dedicate these resources to a charity that canincrease the incremental value of the portfolio through,for example, emphasizing the company’s positioning ortargeting a different stakeholder group.

Last, we propose that cause portfolios have a strongerpositive influence on a company’s performance whenthey are unique in the marketplace. The rationale is thatsimilar portfolios from competitors erode the competitiveadvantage that a company can gain from charitable giv-ing. Bhattacharya and Sen (2004) show that positiveconsumer attitudes toward the company resulting fromCSR can only translate into greater loyalty when theattitudes toward key competitors of the focal company areless positive or different. Thus, a portfolio is expected totranslate into greater value for the company when it isdifferent from competitors’ giving practices.


In summary, as companies engage in charitable be-havior, they are presented with the challenge of managingthe different causes they support through a number ofphilanthropic activities. We propose that firms shouldconsider cause portfolios instead of assessing each cause

separately and manage their characteristics such as thenumber of causes, alignment within the portfolio, anduniqueness in the marketplace to increase the potentialreturn from charitable investments. References are avail-able upon request.

For further information contact:A. Meike Eilert

University of South Carolina1705 College St.

Columbia, SC 29208Phone: 803.777.6074



THE ODD COUPLE: MANAGING “IDEOLOGICAL DISTANCE” IN THECORPORATE-NONPROFIT RELATIONSHIP

Kathryn Lefroy, Monash University, AustraliaDayna Simpson, Oregon State University, Corvallis

Yelena Tsarenko, Monash University, Australia

SUMMARY

There has been a marked increase in the number ofrelationships forged between corporations and nonprofitorganizations (NPOs) in recent years. Despite this, previ-ous studies of these relationships have been limited, orfocused on issues more relevant to needs of the corporatepartner. This research focuses on the strategic marketingpractices of nonprofit organizations, specifically theirpartnerships with commercial enterprises, as they attemptto create a stable and advantageous financial environmentfor themselves. These partnerships form a unique bridgebetween the for-profit and nonprofit goals of value cre-ation and offer vast potential to produce new and moreinnovative ways of “doing business by doing good.”However, the potential for conflict within such relation-ships is significant and can lead to a disparity in thinkingthat we have termed “ideological distance.” This termrefers to the cognitive gap that often occurs betweencorporate financial goals and nonprofit social goals. Itmay cause conflict or create important boundaries withinthese relationships regardless, or because of, governancemechanisms.

The aim of this research was to develop an empiri-cally based framework of how nonprofit organizationscan use governance to manage ideological distance incorporate-nonprofit partnerships. Differences in approachof NPOs toward their corporate partners has been describedby scholars in terms of the NPO’s belief structure – e.g.,mistrust of business generally – or differences in financialideology. Corporate partners tend to bring to their rela-tionships experience in the economic traditions of fiscalaccountability and standard contracting, whilst NPO part-ners tend to bring the more social traditions of trust-building, engagement with stakeholders and political per-suasion. This can lead to either a mismatch of formal-informal expectations, leading to potential exploitation ofa less powerful partner, or an over-emphasis on contract-ing arising from low levels of trust. Social network pres-sure also impacts the corporate-nonprofit relationship andcan affect the NPOs perceived legitimacy. Thus, a degreeof “ideological distance” exists within each corporate-nonprofit relationship. If an alliance is effectively man-aged then it can give the NPO a heightened sense oflegitimacy, yet exploration of how these relationshipsshould be structured to do this is limited. We explored

both explicit (regulative institutions) and implicit (cogni-tive institutions and relational embeddedness) gover-nance in our assessment of these relationships.

We employed a qualitative method to establish aninherent logic in the management of these partnershipsmainly from the NPOs perspective. We used a strategicsampling procedure to identify experts or ‘decision mak-ers’ who had primary responsibility for one or morecorporate–NPO relationships amongst registered Austra-lian nonprofits. A total of 20 organizations with currentcorporate partnerships were interviewed. Data were col-lected through three sources at each organization: (a)extensive semi-structured interviews over multiple occa-sions with corporate relationship mangers; (b) documentsthat defined any examples of “contracting” for theserelationships (e.g., memos, policies, actual contracts); and(c) discussion with additional team members and generalobservation of the partnership “in action” whilst at theNPO. Respondents were asked to provide evidence ofboth a “successful” and an “unsuccessful” partnership.An interview protocol was developed from the literaturethat loosely guided discussion that included themes oforganizational legitimacy, relationship management andideological distance. The transcribed interviews and theother collected data were subjected to coding and analysisby the study authors. Final data analysis employed pat-tern-matching techniques and emphasis was placed onidentifying reoccurring themes and structures for facili-tating theory building.

We describe an important finding that advances ourunderstanding of these unique partnerships through therole of maintaining almost two levels of function (finan-cial focus and social focus). Embeddedness and structureare important contributors to relationship success andconflict negotiation. Our study found that the use ofcontractual agreements and embeddedness in corporate-NPO partnerships plays a significant role in the success ofthese relationships from the perspective of the NPO. Thismanagement of “ideological distance” can provide a pointof reference for the NPO through which they can makedecisions that balance their more social goals and searchfor equitable compromise. However, our findings alsosuggest that inter-organizational relationships with dia-metrically opposed organizational goals yet aligned rela-tionship goals are inherently risky for both partners and


are supported best where both parties – particularly theNPO – maintain their inherent identity within and outsideof the relationship. The results contribute to the small but

growing discourse around corporate-NPO partnerships.These relationships provide a vital addition to the study ofsocial marketing and social issue management.

For further information contact:Kathryn Lefroy

Department of MarketingMonash University

P.O. Box 197Caulfield, VicAustralia 3145

Phone: 61.3.9903.2365Fax: +61.3.9903.2900



CONSUMER’S PERCEIVED TRUST IN IT-ECOSYSTEMS:CONCEPTUALIZING SYSTEM-CENTRIC ANDUSER-CENTRIC DETERMINANTS OF TRUST

IN DIGITAL ENVIRONMENTS

Klaus-Peter Wiedmann, Leibniz University of Hanover, GermanyNadine Hennigs, Leibniz University of Hanover, Germany

Marc-Oliver Reeh, Leibniz University of Hanover, GermanyDieter Varelmann, Leibniz University of Hanover, Germany

SUMMARY

Introduction and Background

Digital ecosystems or IT-Ecosystems (ITE’s) arecomposed of multiple and independent entities such asindividuals, organizations, services, software, and appli-cations as a number of new independent systems thatoperate and communicate with their own infrastructure(man to machine; machine to machine; person to person)sharing one or several missions. As self-organizing envi-ronments and with re-combination or evolution of itsdigital components, ITE’s are highly complex to studyand to manage. A better understanding of how ITE’s andthe interconnected components create benefits and added-values in view of different types of consumers is ofparticular importance to support the establishment ofdigital environments and manage their resources. There-fore, considering and conceptualizing trust, acceptance,autonomy and controllability as crucial prerequisites atthe intersection of marketing and technology, ITE’s canturn to an information-based ecological system for con-sumers.

Encompassing different components of perceivedtrust in ITE’s, in our paper, we rely on a multi-dimensionalframework of trust effects referring to system-centric aswell as user-centric determinants of trust. Based on ourconceptual model, we develop two sets of propositions:The first ones dedicated to technological drivers of trust inITE’s, the second set of propositions referring to user’sindividual as well as social drivers of trust. The model andpropositions are discussed with reference to future re-search steps and business implications.

Conceptualization

In an attempt to structure existing definitions, struc-tural approaches, antecedents, and consequences of trust,we rely on six key constructs of trust as defined byMcKnight and Chervany (1996). These six trust-relatedconstructs named (1) system trust, (2) dispositional trust,(3) situational trust, (4) trusting beliefs, (5) trusting inten-tion, and (6) trusting behavior form a group of trust-

dimension that are conceptually distinguishable, but areinterrelated to each other in specified ways (McKnightand Chervany 1996). Especially in the context of interde-pendencies between trust and the adoption of ITE’s distin-guishing between (I) technological-driven and/or (II)user-driven components of trust as presented in Figure 1,these six constructs cover the most important elements ofperceived trust.

With reference to the technological characteristicsthat influence consumer’s perceived trust in ITE’s and itscomponents, we develop propositions focusing on threedifferent topics: (1) The security, privacy, and functional-ity of the technology, (2) the topic of usability referring toperceived usefulness, ease of use, and information qual-ity, and (3) the presentation against the background ofcomplexity reduction and orchestration of information.

Our set of user-centric trust relations relies on boththe theory of reasoned action (Fishbein and Ajzen 1980)and the theory of planned behavior (Fishbein and Ajzen1985). Adapted to our ITE scenario, trusting beliefs leadto an intention to trust that becomes manifest in a positivetrusting behavior at its best (McKnight and Chervany1996). This individual trust building process is influencedand determined by a lot of different individual and socialdimensions and characteristics (Grabner-Kraeuter andKaluscha 2003). The reason for this is that trusting beliefsare primarily formed “outside” the ITE components itself,but typically in the same way beliefs are formed in theadaptation process within product shopping or serviceacquiring (Gefen et al. 2003). This process is based on theinteraction of cognitive and emotional dimensions, whichinfluences many different kinds of individual factors(Johnson and Grayson 2005). Firstly, the (1) individualtrust forming factors knowledge and expertise,innovativeness, involvement, aspiration level, conve-nience, and risk tolerance are discussed in detail. Sec-ondly, (2) social influences on perceived trust and theadoption of innovative technologies and ITE componentswith regard to an individual’s need for uniqueness, pres-tige orientation, and susceptibility to social influence arepresented.


Conclusions

Focusing on the link between perceived trust andITE’s, we examine the crucial role of technological anduser-centric drivers in determining the success factors inorchestration and adoption of IT-ecosystems. In a numberof situational contexts and different ITE-scenarios, theperception of potential or actual users may lead to a betterunderstanding of the highly interrelated trust componentsand outcomes. Based on this, an empirically verifiedtypology of different segments of users who differ in their

Security P1aSecurity P1a

Privacy P1aPrivacy P1a

Technology

Functionality P1bFunctionality P1b

Trust in IT-Ecosystems

System TrustSystem Trust

DispositionalTrust

DispositionalTrust

SituationalTrust

SituationalTrust

TrustingBeliefs

TrustingBeliefs

TrustingIntention

TrustingIntention

TrustingBehaviour

TrustingBehaviour

Knowledge / Expertise P4 a

K nowledge / E xpertis e P4a

Innovativeness P4 bIn novativen ess P4 b

Individual Driven

Involvement P4 cInvolvement P4c

Aspiration Level P4 dA spiration Level P4d

Convenience P4eConvenience P4e

Risk Toleran ce P4 fR isk Tolerance P4 f

Social Driven

Need for Uniqueness P5 aNeed for Uni queness P5 a

Presti ge Orientat ion P5 bPrestige Orientat ion P5 b

Susceptibil ity to Socia l In fluenc e P5c

Sus ceptib il ity to Soc ia l Influence P5c

PerceivedUsefulness P2a

Pe rceive dUsefulness P2 a

In form ation Qual ity P2 cIn formation Qual ity P2c

Usability

Perceived Ease of Use P2bPercei ved Ease of Use P2b

Com plexity Reducti on P3aCom plexi ty Reduction P3a

Presentation

P rovisi on of Inform at ion P3bP rovi sion of Inform ati on P3b

Technology-Driven User-Centric

FIGURE 1The Conceptual Model

willingness to trust the ITE and its components are ofspecial importance with regard to assess possible marketopportunities. Based on our model, practitioners might beable to understand the multidimensional origins of trustand market adoption. Even if a product shows high-performance technological features, the lack of trustwor-thiness may be the highest barrier for market success.Therefore, we suggest including trust-related insights intothe process of research and design of technological inno-vations and the orchestration of ITE’s. References areavailable upon request.

For further information contact:Klaus-Peter Wiedmann

Institute of Marketing and ManagementLeibniz University of Hanover

Koenigsworther Platz 130167 Hanover




PROSPECTIVE CONFIDENCE IN THE KNOWLEDGE OF THE WEB

Kishore Gopalakrishna Pillai, University of Leeds, United KingdomConstantine Katsikeas, University of Leeds, United Kingdom

SUMMARY

Confidence in consumer knowledge has attracted theattention of researchers recently. In part, this owes itself tothe research stream of knowledge calibration (Alba andHutchinson 2000). Research in education has carefullydifferentiated between predictive and postdictive confi-dence in comprehension. We extend this distinction to thedomain of the Web and examine prospective confidencein knowledge of the web. Drawing from theory andcurrent research, we propose that greater levels of involve-ment with the web, general self-efficacy, domain specificself-efficacy, subjective knowledge of the web, and expe-

rience with the web lead to greater level of prospectiveconfidence in the web. We also propose that prospectiveand retrospective confidence in the web are correlated; inaddition, it is proposed that calibration is higher forretrospective confidence in the web than for prospectiveconfidence. Using data from a sample of adults, our studyfinds that prospective confidence is a function of subjec-tive knowledge and general self efficacy. We also foundthat prospective and retrospective confidence in knowl-edge of the Web are correlated. Moreover, the evidencesuggests that the calibration of prospective confidence islower than that of retrospective confidence.

REFERENCE

Alba, Joseph W. and J. Wesley Hutchinson (2000),

“Knowledge Calibration: What Consumers Knowand What They Think They Know,” Journal ofConsumer Research, 27 (September), 123–56.

For further information contact:Kishore Gopalakrishna Pillai

Leeds University Business SchoolUniversity of Leeds

United KingdomPhone: 44.113.3432636



TOWARD A BETTER UNDERSTANDING OF TRUST IN WEB 2.0SOCIAL NETWORKS

Sonja Grabner-Kräuter, University of Klagenfurt, Austria

ABSTRACT

Online social networks are becoming increasinglyimportant not only as venues for individuals to communi-cate and present themselves but also as marketing andadvertising vehicles. Millions of participants reveal stun-ning amounts of private information in an environmentthat is largely devoid of security standards and practices.This paper addresses the pivotal question why socialnetwork users are being so trusting. A conceptual frame-work that facilitates the identification of key issues relatedto trust in Web 2.0 networks is presented. Thus the papercontributes to a better understanding of factors that influ-ence adoption and usage of social network sites.

INTRODUCTION

The concepts of Web 2.0 facilitate a new type ofcommunication that becomes increasingly important. Web2.0 is the popular term for advanced Internet technologyand applications, including blogs, wikis, podcasts, RSS,and social networks (Lai and Turban 2008). The mostpopular Web 2.0 application in recent years is the emer-gence of online social networks (OSNs) or virtual com-munities, in which membership continues to grow expo-nentially (Lai and Turban 2008). OSNs such as Facebook,MySpace, Friendster, Xing, or studiVZ are a new form ofself-representation and communication, and imply a so-cial behavior that is different to the real world (Bonhardand Sasse 2006). Since their introduction, these OSNshave not only attracted millions of users, but have becomean essential part in their everyday activities – a paralleluniverse, satisfying in the virtual world the human needfor sociability (Ganley and Lampe 2009). Social network-ing sites generate billions of dollars in revenue and arebeing increasingly used in marketing and advertisingcampaigns. However, very little research has been done toinvestigate the factors that influence the usage of OSNs(Gangadharbatla 2008).

The amount and scope of information that OSNparticipants freely reveal are stunning and constitute ahighly attractive data mining and profiling source fordifferent interest groups, ranging from marketers to recruit-ers, private detectives, public authorities, and hackers.Information technology experts characterize Web 2.0social networks as “attractive targets for those with mali-cious intent,” because each site offers a huge user basesharing a common infrastructure, and the information

users willingly supply is highly valuable (Mansfield-Devine 2008). The average user’s profile contains infor-mation about her/his home address, her/his pet’s name,where she/he went to school, her/his mothers’ maidenname and other family details – just the kind of informa-tion used for security or “lost password” questions stated,e.g., by online banking services. Many participants alsoprovide more or less detailed information about theirinterests, sometimes including their political and sexualorientations or intimate portraits of their social or innerlife (Gross and Acquisti 2005). Every now and thenproblems related to privacy or security issues on socialnetwork sites are reported in the media. For instance, inMay 2008 the social networking website Bebo admittedthat a “bug” in its systems enabled users to view otherpeople’s private information. Phone numbers and addresseswere made available as some of Bebo’s 40 million usersfound themselves randomly switched to other people’saccounts (Eriksen 2008). Evidence from many other onlinesocial networks indicates that millions of social networkusers nevertheless do not hesitate to share their thoughts,experiences, images, files, videos, and links in an environ-ment that is largely devoid of security standards andpractices.

Disclosing personal information on the Internet pre-supposes trust because the consumer does not knowwhether her/his personal information may be used in waysthat could result in harm to the consumer, or lead tounwanted future solicitations, credit card theft, or even ahijacking of one’s online identity (Milne and Culnan2004). Obviously, social networking takes place within a(largely unwarranted) context of trust. Consequently, thequestion arises why social network users are being sotrusting. Little research has considered the interrelation-ships between trust, social networks, and the Web 2.0environment. In this paper, we analyze the role of trust inOSNs from a network governance perspective that inte-grates concepts of social network theory, social capitaland the role of value in relational exchanges. By placinggreater emphasis on trust this paper aims at affording abetter understanding of information revelation behaviorin Web 2.0 networks. Specifically, the following researchquestions are addressed:

• What are the basic goals of Internet users to engagein online social networks?

• What role does trust play in online social networks?


• What are the sources of trust in online social net-works?

The paper proceeds as follows. We begin by adoptinga working definition of online social networks, followedby an overview of goals and benefits of participation inOSNs. We then develop a conceptual framework thatfacilitates the identification of key issues related to trustand discuss, in turn, perspectives, types, and sources oftrust in OSNs. To evaluate and potentially extend thetheoretical argumentation and research propositions weconducted three focus group discussions with StudiVZ,Facebook, and XING participants.

DEFINING ONLINE SOCIAL NETWORKS

Broadly, a social network can be defined as a set ofactors and the set of ties representing some relationship –or lack of relationship – between the actors (Brass et al.1998). Actors in a social network (people, organizations,or other social entities) are connected by a set of relation-ships, such as friendship, affiliation, financial exchanges,trading relations, or information exchange. An OSN usescomputer support as the basis of communication amongits members (Andrews et al. 2002). OSNs are organizedaround users and provide a basis for maintaining socialrelationships, for finding users with similar interests, andfor locating content and knowledge that has been contrib-uted or endorsed by other users (Mislove et al. 2007).Web-based social networks provide different means forusers to communicate, such as e-mail, instant messagingservices, blogging, and photo/video-sharing. By thenhundreds of online social networks have been launched,with similar technological features that support a widerange of interests and practices (Ellison et al. 2007).Social network sites can be oriented to work- or business-related contexts (e.g., XING), romantic relationship ini-tiation (the original goal of Friendster), or they can aim atconnecting those with shared interests such as music (e.g.,MySpace) or the college student population (e.g., StudiVZ,or the original launch of Facebook). Most online socialnetworks support the maintenance of already existingsocial ties, but there are also networking services thatsupport the formation of new connections with strangers,based on shared interests, political views, or activities.Some online social networks are directed at diverse audi-ences, whereas others attract people based on commoninterests or shared racial, sexual, religious, or nationality-based identities (Boyd and Ellison 2007).

Drawing on Boyd and Ellison (2007) we defineonline social networks as web-based services that allowindividuals to (1) create a public or semi-public profile forthemselves within a bounded system, (2) indicate a list ofother users with whom they are connected, and (3) viewand traverse their list of connections and those made outby other users within the system. The type and specific

name of these connections may vary from network tonetwork.

In the marketing literature, the terms online socialnetwork and virtual community are often used synony-mously. Virtual communities are viewed as consumergroups of varying sizes that meet and interact online inorder to achieve personal as well as shared goals of theirmembers (Dholakia et al. 2004). When social networksites “hit the mainstream” after the launch of MySpace in2003, a shift in the organization of online communitiesbecame apparent. While websites dedicated to communi-ties of interest still exist and flourish (e.g., Dogster,CafeMom, Feierabend), online social networks cateringto a broader audience are primarily organized aroundpeople, and not interests. Early public online communitiessuch as Usenet and public discussion forums were struc-tured by topics or according to topical hierarchies, butprevailing social network sites are structured as personalor “egocentric” networks, with the individual at the centerof his/her own community (Boyd and Ellison 2007).These online social networks addressing a very broadtarget audience enable users to articulate and make visibletheir social networks, whereas the opportunity to comeinto contact with strangers usually is of minor importance.

MOTIVATION TO PARTICIPATE IN ONLINESOCIAL NETWORKS

For several years information exchange betweenconsumers on online social networking sites has grownexponentially. Thus, especially marketing researchershave been and still are challenged to provide insights whatmotivates consumers to participate in and contribute toonline social networks. Recent research into virtual com-munities has advanced our understanding of the reasonswhy people get involved in online social networks. How-ever, many knowledge gaps still exist (de Valck et al.2009; Pempek et al. 2009). Many people join online socialnetworks out of a desire to be part of a communitycomprised of people who share similar interest. However,participation in online social networks can meet a consid-erable number of needs. One important focus of onlinecommunities is on guidance and informational supportthat enhances decision-making (Macaulay et al. 2007).Other needs that can be met by online social networks areaffiliation and belonging, power and prestige, and enter-tainment (Andrews et al. 2002; Balasubramanian andMahajan 2001).

Using the perspective of goal and action identitytheories, value can be regarded as the most important goalof network users in relational exchanges. Concerningvalue typologies, the range and variety found in theliterature are very wide, although the hedonic versusutilitarian value difference transcends in most cases(Gallarza and Gil Saura 2006). Different types of Web 2.0


social networking groups focus on different kinds of valuefor their members. Social benefits have been shown to bethe influence factor that most strongly motivates consum-ers to participate in online communities and to articulatethemselves (de Valck et al. 2009; Ellison et al. 2007).Available research suggests that most online social net-works primarily support pre-existing offline relationshipsor solidify offline contacts, as opposed to meeting newpeople (Boyd and Ellison 2007). Recent studies reportthat students and alumni primarily use Facebook to com-municate, connect, and stay in contact with others (Ellisonet al. 2007; Fogel and Nehmad 2009). The particularimportance of social value also became apparent in threefocus group discussions the author conducted withStudiVZ, Facebook, and XING participants. For instance,when asked what they thought was the most importantbenefit of using Facebook, discussants responded thatthey could stay in touch and reconnect with friends allover the world. Dholakia et al. (2004) differentiate betweentwo kinds of social value. Maintaining interpersonal con-nectivity refers to the social benefits derived from estab-lishing and maintaining contact with other people, such asfriendship, and social support. Another type of socialvalue is social enhancement, the value that network par-ticipants derive from gaining acceptance and approval ofother members.

Utilitarian value refers to tangible or objective ben-efits and can be defined as the value derived from accom-plishing some pre-determined instrumental purpose(Chaudhuri and Holbrook 2001; Dholakia et al. 2004).For online social network participants either informa-tional value or instrumental value can be of special rel-evance. Informational value is derived from getting andsharing information in the online community (Dholakiaet al. 2004). Several XING participants responded thatone of the most important benefits of XING is the “auto-matic” updating of contact features – the contacts (thereare not friends but contacts in XING) of the participantsupgrade their profiles themselves so “I have always cur-rent data such as phone numbers, e-mail addresses, changesof job from my schoolmates, former fellow students, etc.”When social interactions in online communities helpparticipants to accomplish specific tasks, such as solvinga problem, validating a decision already reached, orbuying a product, online social networks provide instru-mental value. Hedonic value relates to the experientialaspects of human consumption in which emotions andfeelings of enjoyment or pleasure play a pivotal role(Chaudhuri and Holbrook 2001). Dholakia et al. (2004)speak of entertainment value that community membersderive from fun and relaxation through playing or other-wise interacting with others. As one Facebook participantarticulated, “for me, entertainment (on Facebook) is animportant factor – when I sit in front of the computer allafternoon and work on a project paper, I have about 10 to20 logins during the afternoon.”

In sum, connectivity and friendship oriented net-works such as StudiVZ and Facebook emphasize socialvalue and hedonic value (reflecting enjoyment, fun, andpleasure), whereas in online groups that operate on func-tional support and shared experience (e.g., eBay, XING)utilitarian values are more important. These assumptionslargely were affirmed in the focus group discussions withStudiVZ, Facebook, and XING participants. One specificmotive that was mentioned by several discussants in eachfocus group was curiosity – “curiosity to find out who hada look at my profile” (two XING participants), “curiositywhat’s the news” (a Facebook participant), or “curiositywho that person is” (one StudiVZ participant). Referringto the above value typology, the online social networksprovide instrumental value to their members in all thesecases.

TRUST IN ONLINE SOCIAL NETWORKS – AGOVERNANCE PERSPECTIVE

Exchange of valuable information in (online) socialnetworks would be limited without trust both in thenetwork infrastructure and the other network participants.The role of trust in OSNs can be investigated from agovernance perspective that allows to integrate conceptsof social network theory and exchange theory. Heregovernance relates to the different modes of co-ordinatingindividual actions. Trust can be seen as a powerful alter-native to formal governance mechanisms that allowexchange relationships to be formed and attempt to con-trol opportunism (Puranam and Vanneste 2009). I assumea complementary relationship between trust and gover-nance, as far as trust may enhance the effect of governanceon exchange performance. Conceptual debates and empiri-cal results suggest, in most cases, that trust has an impor-tant part to play in social networks and exchanges, but theprecise role is not completely clear (Gubbins andMacCurtain 2008). The relationships among concepts ofsocial network theory, exchange theory and trust are verycomplex and far from conclusive. As a more detailedconceptual discussion is beyond the scope of this paper, Ionly point out selected facets.

A better understanding of social processes in Web 2.0communities requires a finer-grained analysis of the qual-ity and configuration of network ties. Adopting a socialnetwork approach to the analysis of trust involves theassumption that individual actors are embedded within anetwork of relationships (Jones et al. 1997). The conceptof embeddedness refers to the influence of the network onits members’ behavior (Granovetter 1992). Granovetter(1992) distinguishes between two levels of embeddedness:Relational embeddedness and structural embeddedness.Relational embeddedness describes the kind of personalrelationships people have developed with each otherthrough a history of interactions. Structural embeddednessrefers to the network’s overall structure or architecture


and thus concerns the properties of the social system andthe network of relations as a whole. It provides the basisfor social mechanisms to adapt, coordinate, and safeguardexchanges and thus enhances the likelihood of networkgovernance (Jones et al. 1997). Structural embeddednessfocuses on social pressure that the network as a wholeexercises on the development of a single relationship. Theembedded perspective proposes an evolutionary concep-tualization of trust and argues that trust increases as aconsequence of the accumulation of positive experiences(Ganzaroli 2002). Granovetter (1992) suggests that beingembedded in cohesive networks accelerates the creationof trust. The cohesiveness of the network structure, wherea specific relationship is embedded, facilitates the circu-lation of information about parties’ reputation and thesocialization of common behavior. The prevalent charac-teristics of a network shape the behavior of its members(Ganzaroli 2002).

From a social network perspective, within OSNsthere are few strong relationships or ties that foster thedevelopment of trust, cooperation, and intimacy, andmany weak ties that frequently are characterized by occa-sional interaction and low emotional intensity (Granovetter1973). Levin, Cross, and Abrams (2002) introduced theconcept of trusted weak ties and empirically demonstratedthe structural benefits of weak ties, showing that theyprovide better access to non-redundant and innovativeinformation. Their findings suggest that there are trustingand non-trusting weak and strong ties, implying that trustand tie strength are related but not synonymous (see alsoGubbins and MacCurtain 2008). Weak ties in (online)social networks tend to link individuals with many otherpeople who might be more socially dissimilar, providingnew sources of useful information or new perspectives.Investigating the mediating role of trust in effective knowl-edge transfer (Levin and Cross 2004) found that trustedweak ties yielded the most useful knowledge. I assumethat trusted weak ties are the prevalent relationships inonline social networks that facilitate effective informa-tion exchange and knowledge transfer and thus essen-tially contribute to network performance.

By supporting both strong and loose social ties onlinesocial networks allow users to create and maintain largersets of relationships from which they could potentiallydraw resources, because the Web 2.0 technology is well-suited to maintaining such ties cheaply and easily (Ellisonet al. 2007). Thus, online social networks contribute to thegeneration of social capital, a term with numerous defini-tions in multiple fields (Adler and Kwon 2002; Burt2000). Nahapiet and Ghoshal (1998) define social capitalas “the sum of the actual and potential resources embed-ded within, available through and derived from the net-work of relationships possessed by an individual . . .”(p. 243). In this regard the concept of social capital can beconsidered as a way to describe the value that can be

accrued through a social network and from the socialresources of the actors embedded within that network(Gubbins and MacCurtain 2008). Or put it another way,the value of social networks manifests to participants associal capital (Ganley and Lampe 2009). In the socialcapital literature, the concept of trust has been frequentlydiscussed, but there is no decisive view on its particularrole in definitions of social capital (see the references in(Gubbins and MacCurtain 2008)). I posit that trust willaffect the intention to participate in online social networks(and the actual participation and the social capital that canbe accrued both from the network itself and the resourcesthat may be mobilized through the network) through itsinfluence on creating value (see also the arguments in(Sirdeshmukh et al. 2002) who propose a mediating roleof value in the trust-loyalty relationship).

Drawing on the logic of social exchange theoryonline consumer behavior is largely determined by indi-vidual users’ cost-benefit analyses (Blau 1964; Dwyeret al. 2007; Metzger 2006; Thibaut and Kelley 1969).Internet users weigh the benefits of engaging in OSNs(e.g., staying in touch with friends) against the costs (e.g.,the risk of privacy invasion). Only if the exchange orinteraction is perceived to be beneficial, the individual islikely to enter into an exchange relationship. Prior researchin the context of privacy concern found that consumerswho perceive that the benefits of information disclosureexceed the risks, both current and future, are more likelyto disclose (Milne and Culnan 2004). Consistent withsocial exchange theory tenets, trust can be conceived as acounteragent to risk in that higher trust allows individualsto determine that the benefits outweigh the risks. In thisperspective trust bears important functions for the con-sumer – it reduces information complexity and lowers theperceived risk of a transaction or an exchange (see alsoLuhmann 1989). Only where there are high levels of trust,people are more willing to take risk in information andother exchanges (Nahapiet and Ghoshal 1998).

TYPES OF TRUST IN ONLINE SOCIALNETWORKS

The relative importance of trust depends – amongother factors – upon the complexity and the context of anaction. To analyze trust in the Web 2.0 environmentdifferent types of trust have to be distinguished. First, trustcan be conceptualized on different levels of analysis,reflecting the array of entities, individuals, dyads, groups,networks, systems, firms, and inter-firm alliances in whichtrust and related processes play a role (Rousseau et al.1998). In Web 2.0 social networks trust is both a micro-and a macro-level phenomenon in which there is aninterplay between the macro-network created by the (cor-porate) actor who designed it and the micro-groups formedby the individual network users (Lai and Turban 2008).


In the organizational trust literature trust is mostlydefined as a belief or expectation about the other (trusted)party, or as a behavioral intention or willingness to dependor rely on another party, coupled with a sense of vulner-ability or risk if the trust is violated (e.g., Mayer et al.1995; Rousseau et al. 1998). Accordingly, trust in theWeb environment is most often defined as a belief orexpectation about the website, the web vendor and/or theInternet as the trusted party or object of trust or as abehavioral intention or willingness to depend or rely onthe trusted party (Grabner-Kräuter and Kaluscha 2003;McKnight and Chervany 1996; McKnight et al. 2002;McKnight and Chervany 2002). In the context of OSNsother network participants, the network provider and thesocial network site are essential objects of trust. Theanalysis of online trust has to consider impersonal formsof trust as well, because in computer-mediated environ-ments personal trust is a rather limited mechanism toreduce uncertainty (Grabner-Kräuter and Faullant 2008).The Web 2.0 technology itself has to be considered as anobject of trust. Trust in technical systems mainly is basedon the perceived functionality (e.g., reliability, capability,correctness, and availability) of a system (Lee and Turban2001; Thatcher et al. 2007).

To this point, trust has been treated as a situationalconstruct. But trust can also be conceptualized as a cross-situational, cross-personal construct, encompassing indi-vidual characteristics of the trustor. Following the trusttypology proposed by (McKnight and Chervany 1996)this type of trust is called dispositional trust. Mayer et al.(1995) include a very similar construct “propensity totrust” in their trust model. These constructs have theirroots in personality psychology (e.g., Rotter 1967) andrecognize that people develop, over the course of theirlives, generalized expectations about the trustworthinessof other people. Dispositional trust is a stable within-partyfactor that not only affects trust toward other individualsor groups of individuals but also toward technical systemsor social network sites (Grabner-Kräuter and Faullant2008). More specifically, it can be assumed that disposi-tional trust affects trusting beliefs about the trusted object(other participants, social network site, provider, Web 2.0technology) and the behavioral intention or willingness todepend on the trusted party. I also propose a direct effectof dispositional trust on both the kind and the amount ofinformation users provide in their profile.

Drawing on conceptualizations of trust in the socialpsychology literature (e.g., Lewis and Weigert 1985)many researchers differentiate cognitive and affective/emotional trust. Cognitive trust refers to “good rationalreasons why the object of trust merits trust” (Lewis andWeigert 1985, p. 972). Cognitive trust is based on evalu-ating the competence, reliability, and predictability of thetrusted object and reflects the economic understanding oftrust as a rational choice (Riegelsberger et al. 2005;

Johnson and Grayson 2003). Affective trust, on the otherhand, is the emotion-driven form of trust that is based onimmediate affective reactions, on attractiveness, aesthet-ics, and signals of benevolence. Frequently trust-basedbehavior results from a mix of affective and cognitivetrust (Riegelsberger et al. 2005; Corritore et al. 2003).Concerning the relationship between the value of partici-pation in OSNs and trust it can be assumed that cognitivetrust is greater when the perceived utilitarian value is high(Chaudhuri and Holbrook 2001). On the other hand,OSNs with a higher pleasure potential that provide non-tangible, symbolic benefits and are likely to hold a greaterpotential for evoking positive emotions and affect-basedtrust in a consumer. Assuming that trust in OSNs is basedboth on cognitive and affective reasons and reactions Ipropose that utilitarian value, hedonic value and socialvalue are positively related to trust in OSNs. FurthermoreI propose that the importance of cognitive and affectivetrust depends on the type(s) of value the participantsderive from participation in this specific OSN.

BASES OF TRUST IN ONLINE SOCIALNETWORKS

Trust may develop for a number of reasons, and oftenfor a variety of reasons working together. Drawing fromthe general trust literature this section briefly outlinessome important bases of trust in online social networks. Anumber of authors have identified different bases orsources of trust in relational exchanges (e.g., Zucker1986; Lewicki and Bunker 1995; Shapiro et al. 1992;Doney and Cannon 1997; Kramer 1999; McKnight et al.1998). In the context of online social networks it makessense to combine the discussion of different grounds orbases of trust with the perspective of trust as a dynamicconcept which can be divided into different developmen-tal stages or phases, each with specific characteristics(Lewicki et al. 2006). According to Rousseau and Sitkinet al. (1998) three different phases of trust can be distin-guished: the phase of trust building, where trust is formed;the phase of stabilizing trust, where trust already exists;and the phase of dissolution, where trust declines.

In the phase of trust building, trust in a specific socialnetwork site and/or in other network participants canemerge on different bases. Drawing on the trust-buildingprocess framework proposed by (Doney and Cannon1997), it can be argued that initial trust in online socialnetworks often develops through a transference process.The “extension pattern” of trust development suggeststhat trust can be transferred from one trusted “proofsource” to another person, group of persons or otherobject of trust with which the trustor has little or no directexperience (Doney and Cannon 1997). Preliminary re-sults of three focus group discussions the author hasconducted with StudiVZ, Facebook, and XING partici-pants indeed show that recommendations and invitations


of real world friends were by far the most importantreasons for joining these online social networks, generat-ing personal profiles, and thus revealing personal infor-mation.

In the first developmental stage online trust can alsobe based on rational calculation of potential costs andbenefits. Lewicki and Bunker (1995) named this firststage calculus-based trust. Similarly, Doney and Cannon(1997) speak of trust involving a calculative process,when the trustor calculates the costs and rewards ofanother party choosing between cheating or staying in therelationship. The majority of focus group discussantsindicated that they are not really concerned about apotential breach of trust by the social network provider,“. . . because there are millions of other users who revealpersonal information on social network sites and don’tworry at all” (one Facebook discussant) and, besides, “. . .for Facebook good reputation is very important.”Concerning the credibility of profile information of othernetwork participants several StudiVZ and Facebookdiscussants have doubted the tenability of wrong orflattering information, “. . . because other people whoreally know that person immediately would find out thathe is cheating.”

When an Internet user for the first time visits andexplores a less renowned social network site that has notbeen recommended by friends, her/his initial trust mightbe primarily based on cognition. Cognition-based trustrelies on rapid, cognitive cues, or first impressions, asopposed to personal interactions (McKnight et al. 1998).Cognitive perceptions of network characteristics such assize of the network, current number of participants online,discussed topics, privacy and security, usefulness andease of use of the network site can be considered asimportant bases or antecedents of network trust in thephase of initial trust formation and trust building on “self-researched” network sites.

After some time and continuous interactions on asocial network site the judgements of a participant aboutthis specific network become more a function of theinteractions themselves. The trust-relationship may enterthe second stage of trust development which is dominatedby trust based on the trustor’s knowledge and understand-ing about the trusted party resulting from past interactions(knowledge- or experience-based trust) (Lewicki andBunker 1995). Doney and Cannon (1997) describe theprediction process of developing trust that relies onrepeated interactions and the trustor’s assessment of theother party’s past behavior and evinced trustworthiness.Hence, in the phase of stabilizing trust, factors such asfamiliarity with the technological features and communi-cation tools of the social network site or satisfaction withpast interactions with other community members areimportant antecedents of online trust.

Explicit and tacit understandings regarding transac-tion norms, interactional routines, and exchange practicescan be considered as another source of trust in onlinesocial networks. Such explicit and tacit understandings,captured in both formal and informal rules, provide animportant basis for inferring that other participants in thesocial network are likely to behave in a trustworthymanner (Kramer 1999). Rule-based trust is predicated noton a conscious calculation of consequences, but rather ona shared understanding of the system of rules pertaining toappropriate behavior. The question of how explicit andimplicit rules function to create and sustain mutual trust inonline social networks was not discussed in the StudiVZ,Facebook, and XING focus groups, but it would warrantfurther investigation.

The most mature level of trust is restricted to interper-sonal trust and dominated by internalization of the other’spreferences, mutual empathy, and identification with eachother (identification-based trust). Identification-based trustrepresents the highest and most solid level of trust whichmay be reached by the parties to the trust relationship.Trust is mainly formed and influenced by joint values,tasks and goals, by creating a collective identity, and byphysical proximity or emotional closeness (cf., Shapiroet al. 1992; Lewicki and Bunker 1995; Ratnasingham1999). Using similar arguments Kramer (1999) has coinedthe term category-based trust, which refers to trust predi-cated on information regarding the trusted party’s mem-bership in a social or organizational category or network.Shared membership in a social network can serve as a rulefor defining the boundaries of low-risk interpersonal trustthat bypasses the need for closer personal knowledge(Kramer 1999). Because of the cognitive consequences ofcategory membership and ingroup bias, individuals tendto attribute positive characteristics such as honesty, benevo-lence, integrity, and cooperativeness to other networkmembers.

The assumption that trust in online social networksoften develops through a transference process was con-firmed in all focus group discussions. Beyond that, Iassume that the importance of different sources of trust iscontingent upon the type of Web 2.0 social networkinggroup. I posit for example that in online social networksthat operate on the principles of democratic participationand friendship (e.g., MySpace and Facebook; see also Laiand Turban 2008; who elaborate on the characteristics offive types of groups) cognition-based trust and calculus-based trust are of minor importance, whereas knowledge-based trust might be of more significance, e.g., for evalu-ating information exchange with former schoolmates. Ipropose that in “friendship” networking groups trust ismainly category- or identification-based and formed andinfluenced by joint values, tasks and goals, or by commonexperiences. Furthermore I suggest that in more utilityoriented networks, such as XING, calculus- and rule-


based trust are of major importance, whereas identifica-tion with other network participants and mutual empathyplay a minor role. Preliminary analysis and interpretationof the focus group discussions seem to affirm the abovepropositions.

DISCUSSION AND CONCLUSION

Online social networks are becoming more and moreimportant not just as places for individuals to communi-cate and present themselves but also as marketing/adver-tising vehicles (Gangadharbatla 2009). Despite the lim-ited scope of this research in progress, it marks a step in theright direction toward a better understanding of factorsthat influence adoption and usage of social network sitesby developing a conceptual framework that integratesconcepts of social network theory, social capital, and therole of value in relational exchanges. I have attempted tosystematically identify promising areas for upcomingresearch by discussing the role of trust in Web 2.0 social

networks, as both an antecedent and an outcome of net-work participation and information disclosure. Drawingfrom the general trust literature I distinguished betweendifferent types of trust and identified different bases oftrust in online social networks. Preliminary findings ofour qualitative focus group study affirm our propositionthat the importance of different sources of trust is contin-gent upon the type of the Web 2.0 social network. Build-ing on the conceptual framework developed in this papera number of specific questions can be addressed: What arethe antecedents (categories: cognition-based, experience-and knowledge-based, personality-based) of trust in OSNsand how do these antecedents influence trust develop-ment? How do different types of trust influence thewillingness to join OSNs and the willingness to providepersonal information? How does the interaction betweenthe characteristics of OSNs and different types of trustinfluence participation in OSNs and revelation of per-sonal information? Future research should analyze theseissues.

REFERENCES

Adler, Paul S. and Seok-Woo Kwon (2002), “SocialCapital: Prospects for a New Concept,” Academy ofManagement Review, 27, 17–40.

Andrews, Dorine, Jennifer Preece, and Murray Turoff(2002), “A Conceptual Framework for DemographicGroups Resistant to On-Line Community Interac-tion,” International Journal of Electronic Commerce,6, 9–24.

Balasubramanian, Sridhar and Vijay Mahajan (2001),“The Economic Leverage of the Virtual Commu-nity,” International Journal of Electronic Commerce,5, 103–38.

Blau, Peter M. (1964), Exchange and Power in SocialLife. New York: John Wiley & Sons, Inc.

Bonhard, P. and M.A. Sasse (2006), “‘Knowing Me,Knowing You’ – Using Profiles and Social Network-ing to Improve Recommender Systems,” BT Tech-nology Journal, 24 (3), 84–98.

Boyd, Danah M. and Nicole B. Ellison (2007), “SocialNetwork Sites: Definition, History, and Scholar-ship,” Journal of Computer-Mediated Communica-tion, 13 (1).

Brass, D., K. Butterfield, and B. Skaggs (1998), “Rela-tionships and Unethical Behavior: A Social NetworkPerspective,” Academy of Management Review, 23,14–31.

Burt, Ronald S. (2000), “The Network Structure of SocialCapital,” Research in Organizational Behavior, 22,345–423.

Chaudhuri, Arjun and Morris B. Holbrook (2001), “The

Chain of Effects from Brand Trust and Brand Affectto Brand Performance: The Role of Brand Loyalty,”Journal of Marketing, 65 (April), 81–93.

de Valck, Kristine, Gerrit H. van Bruggen, and BerendWierenga (2009), “Virtual Communities: A Market-ing Perspective,” Decision Support Systems, In Press,Corrected Proof.

Dholakia, Utpal M., Richard P. Bagozzi, and Lisa KleinPearo (2004), “A Social Influence Model of Con-sumer Participation in Network- and Small-Group-Based Virtual Communities,” International Journalof Research in Marketing, 21 (3), 241–63.

Doney, Patricia M. and Joseph P. Cannon (1997), “AnExamination of the Nature of Trust in Buyer-SellerRelationships,” Journal of Marketing, 61, 35–51.

Dwyer, Catherine, Starr Roxanne Hiltz, and Katia Passerini(2007), “Trust and Privacy Concern Within SocialNetworking Sites: A Comparison of Facebook andMySpace,” in Thirteenth Americas Conference onInformation Systems. Keystone, Colorado.

Ellison, Nicole B., Charles Steinfield, and Cliff Lampe(2007), “The Benefits of Facebook ‘Friends’: SocialCapital and College Students’ Use of Online SocialNetwork Sites,” Journal of Computer-Mediated Com-munication, 12 (4).

Eriksen, Alanah May (2008), “Glitch Opens Bebo Users’Private Details to Others,” in The New Zealand Her-ald.

Fogel, Joshua and Elham Nehmad (2009), “Internet SocialNetwork Communities: Risk Taking, Trust, and Pri-vacy Concerns,” Computers in Human Behavior, 25(1), 153–60.


Gallarza, Martina G. and Irene Gil Saura (2006), “ValueDimensions, Perceived Value, Satisfaction, and Loy-alty: An Investigation of University Students’ TravelBehavior,” Tourism Management, 27 (3), 437–52.

Gangadharbatla, Harsha (2008), “Facebook Me: Collec-tive Self-Esteem, Need to Belong, and Internet Self-Efficacy as Predictors of the iGeneration’s AttitudesToward Social Networking Sites,” Journal of Inter-active Advertising, 8 (2).

____________ (2009), “Individual Differences in SocialNetworking Site Adoption,” in Social NetworkingCommunities and E-Dating Services: Concepts andImplications, Celia Romm-Livermore and KristinaSetzekorn, eds. Hershey, New York: IGI Global.

Ganley, Dale and Cliff Lampe (2009), “The Ties ThatBind: Social Network Principles in Online Commu-nities,” In Press, Corrected Proof.

Ganzaroli, Andrea (2002), “Creating Trust Between Localand Global Systems,” PhD Thesis Erasmus Univer-sity Rotterdam.

Grabner-Kräuter, Sonja and Ewald A. Kaluscha (2003),“Empirical Research in On-Line Trust: A Reviewand Critical Assessment,” International Journal ofHuman-Computer-Studies, 58, 783–812.

____________ and Rita Faullant (2008), “ConsumerAcceptance of Internet Banking: The Influence ofInternet Trust,” International Journal of Bank Mar-keting, 26 (7), 483–504.

Granovetter, Mark S. (1973), “The Strength of WeakTies,” The American Journal of Sociology, 78 (6),1360–80.

____________ (1992), “Problems of Explanation in Eco-nomic Sociology,” in Networks and Organizations:Structure, Form, and Action, N. Nohria and R.G.Eccles, eds. Boston: Harvard Business School Press.

Gross, Ralph and Alessandro Acquisti (2005), “Informa-tion Revelation and Privacy in Online Social Net-works,” in WPES 05, ACM, ed. Alexandria, VA:ACM.

Gubbins, Claire and Sarah MacCurtain (2008), “Under-standing the Dynamics of Collective Learning: TheRole of Trust and Social Capital,” Advances in Devel-oping Human Resources, 10 (4), 578–99.

Jones, Candace, William S. Hesterly, and Stephen P.Borgatti (1997), “A General Theory of NetworkGovernance: Exchange Conditions and Social Mecha-nisms,” Academy of Management Review, 22 (4),911–45.

Kramer, Roderick M. (1999), “Trust and Distrust inOrganizations: Emerging Perspectives, EnduringQuestions,” Annual Review of Psychology, 50, 569–98.

Lai, Linda S.L. and Efraim Turban (2008), “GroupsFormation and Operations in the Web 2.0 Environ-ment and Social Networks,” Group Decision andNegotiation, 17 (5), 387–402.

Lee, Matthew K.O. and Efraim Turban (2001), “A Trust

Model for Consumer Internet Shopping,” Interna-tional Journal of Electronic Commerce, 6 (1), 75–91.

Levin, D.Z., R. Cross, and L.C. Abrams (2002), “TheStrenght of Weak Ties You Can Trust: The Mediat-ing Role of Trust in Effective Knowledge Transfer,”Academy of Management Proceedings 2002, 1–7.

____________ and ____________ (2004), “The Strengthof Weak Ties You Can Trust: The Mediating Role ofTrust in Effective Knowledge Transfer,” Manage-ment Science, 50 (11), 1477–90.

Lewicki, Roy J. and Barbara B. Bunker (1995), “Develop-ing and Maintaining Trust in Work Relationships,” inTrust in Organizations, Roderick M. Kramer andTom Tyler, eds. London: Sage Publications.

____________, Edward C. Tomlinson, and NicoleGillespie (2006), “Models of Interpersonal TrustDevelopment: Theoretical Approaches, EmpiricalEvidence, and Future Directions,” Journal of Man-agement, 32 (6), 991–1022.

Luhmann, Niklas (1989), Vertrauen. Ein Mechanismusder Reduktion sozialer Komplexität (3d ed.). Stuttgart.

Macaulay, Linda A., Kathy Keeling, Peter McGoldrick,George Dafoulas, Emmanouil Kalaitzakis, and DebbieKeeling (2007), “Co-Evolving E-Tail and On-LineCommunities: Conceptual Framework,” in Interna-tional Journal of Electronic Commerce, Vol. 11,M.E. Sharpe Inc.

Mansfield-Devine, Steve (2008), “Anti-Social Network-ing: Exploiting the Trusting Environment of Web2.0,” Network Security, (November), 4–7.

Mayer, Roger C., James H. Davis, and David F. Schoorman(1995), “An Integrative Model of OrganizationalTrust,” Academy of Management Review, 20 (3),709–34.

McKnight, D. Harrison and N.L. Chervany (1996), “TheMeanings of Trust,” MISRC Working Paper, Elec-tronic Publication URL: [http://www.misrc.umc.edu/wpaper/WorkingPapers/9604.pdf], (last access: Octo-ber 2002).

____________, Larry L. Cummings, and Norman L.Chervany (1998), “Initial Trust Formation in NewOrganizational Relationships,” Academy of Man-agement Review, 23 (3), 473–90.

____________, Vivek Choudhury, and Charles Kacmar(2002), “Developing and Validating Trust Measuresfor e-Commerce: An Integrative Typology,” Infor-mation System Research, 13 (3), 334–59.

____________ and Norman L. Chervany (2002), “WhatTrust Means in e-Commerce Customer Relation-ships: An Interdisciplinary Conceptual Typology,”International Journal of Electronic Commerce, 6 (2),35–59.

Metzger, Miriam J. (2006), “Effects of Site, Vendor, andConsumer Characteristics on Web Site Trust andDisclosure,” Communication Research, 33 (3), 155–79.

Milne, George R. and Mary J. Culnan (2004), “Strategies


for Reducing Online Privacy Risks: Why ConsumersRead [Or don’t Read] Online Privacy Notices,” Jour-nal of Interactive Marketing, 18 (3), 15–29.

Nahapiet, Janine and Sumantra Ghoshal (1998), “SocialCapital, Intellectual Capital, and the OrganizationalAdvantage,” Academy of Management Review, 23,242–66.

Pempek, Tiffany A., Yevdokiya A. Yermolayeva, andSandra L. Calvert (2009), “College Students’ SocialNetworking Experiences on Facebook,” Journal ofApplied Developmental Psychology, 30 (3), 227–38.

Puranam, Phanish and Bart S. Vanneste (2009), “Trustand Governance: Untangling a Tangled Web,” inAcademy of Management Review, Vol. 34, Academyof Management.

Ratnasingham, Pauline (1999), “Risks in Low Trust amongTrading Partners in Eletronic Commerce,” Comput-ers & Security, 18, 587–92.

Rotter, J.B. (1967), “A New Scale for the Measurement ofInterpersonal Trust,” Journal of Personality, 35, 651–65.

Rousseau, D.M., S.B. Sitkin, R.S. Butt, and C. Camerer

(1998), “Not So Different After All: A Cross-Disci-pline View of Trust,” Academy of ManagementReview, 23 (3), 393–404.

Shapiro, Debra L., Blair H. Sheppard, and Lisa Cheraskin(1992), “Business on a Handshake,” NegotiationJournal, 8, 365–77.

Sirdeshmukh, Deepak, Jagdip Singh, and Barry Sabol(2002), “Consumer Trust, Value, and Loyalty inRelational Exchanges,” Journal of Marketing, 66,15–37.

Thatcher, Jason Bennett, Misty L. Loughry, Jaejoo Lim,and D. Harrison McKnight (2007), “Internet Anxi-ety: An Empirical Study of the Effects of Personality,Beliefs, and Social Support,” Information & Man-agement, 44, 353–63.

Thibaut, John W. and Harold H. Kelley (1969), The SocialPsychology of Groups. New York: John Wiley &Sons, Inc.

Zucker, L.G. (1986), “Production of Trust: InstitutionalSource of Economic Structure, 1840–1920,” Researchin Organizational Behavior, 8, 53–111.

For further information contact:Sonja Grabner-Kräuter

University of KlagenfurtUniversitätsstrasse 65–67

9020 KlagenfurtAustria

Phone: ++43(0)463.2700.4042Fax: ++43(0)463.2700.994042

E-Mail: E [email protected]


DOES REWARDING REFERRALS AFFECT PERCEIVED SENDERCREDIBILITY? AN EMPIRICAL INVESTIGATION IN CELLULAR

TELECOMMUNICATION SERVICES

Sabrina Helm, The University of Arizona, TucsonAnne Willach, RWTH Aachen University, Germany

SUMMARY

Customers’ positive word of mouth (WOM) isregarded as a very effective marketing tool that enablescompanies to gain new customers (Haywood 1989;Wangenheim and Bayón 2007) as well as bond currentcustomers (Helm 2003). The “power of word of mouth”(Arndt 1967) is generally attributed to its high credibility.As customers communicate at no or low cost for the firm,WOM is also a highly efficient marketing tool (Wilson1994). Consequently, customer referral programs (CRP)have caught notable attention in scientific literature (e.g.,Godes and Mayzlin 2004; Ryu and Feick 2007; Wirtz andChew 2002). In such programs, the service provider offerscustomers a reward for persuading others to also becomecustomers. Ryu and Feick (2007) demonstrate that offer-ing rewards can increase customers’ referral intentions.However, research has neglected to investigate whetherrewarding referrals has an effect on the recipient’s percep-tion of the sender of the referral and, specifically, whethersender credibility is affected. A reduction in credibilitymight then affect downstream behavioral variables. There-fore, we address the following research questions: Doesrewarding referrals have a negative effect on sendercredibility as perceived by the recipient? Does a changein perceived credibility affect recipient attitudes andpurchase intentions toward the service provider?

To answer these questions, we draw on research onattribution theory, tie strength (Brown and Reingen 1987),and the theory of planned behavior. We understandcredibility as a two-dimensional concept for evaluating asource of information that consists of the componentscompetence (or expertness) and trustworthiness. In orderto test our set of hypotheses, we conduct an experiment.We chose a post-test-only control group between-subjectsdesign (Campbell and Stanley 1963) and manipulated theindependent variables “reward” (no reward; reward) and“tie strength” (weak tie; strong tie) using four differentscenarios. All groups were completely randomized.

All student subjects were told that they met anotherstudent during class and decided to have lunch together.While having their meal, they coincidently talked aboutcellular telecommunication providers. Subjects were alsotold that they already were customers of a service pro-vider, that their contract was about to expire, and that theywere uncertain about prolonging or switching. The fellow

student recommended switching to his provider“MobileStar” and to choose the student tariff“Student@Star.” The variable “reward” was manipulatedby telling subjects that, at the end of their conversation, thefellow student told them about the service provider’s CRPand that he asked the subject to name him as a referrer incase of choosing the provider. He indicated that he wouldreceive 50 Euros for a successful referral which is acommon reward for referrals in the industry. In the controlgroup, this part of the conversation was omitted, resem-bling a naturally occurring WOM situation. The variable“tie strength” was manipulated by describing the relation-ship that subjects had with the fellow student. In the weak-tie scenario, subjects were asked to think of one of theirdistant acquaintances. In the strong-tie scenario, theywere to think of one of their closest friends (Granovetter1973). Scenarios and questionnaire were pretested. Intotal, 184 (109 male, 74 female; 24 years on average)students from three German public universities wereincluded in the net sample (8 excluded).

Study results show that rewarded senders are ascribednegative motives by referral recipients more frequentlythan non-rewarded senders. This result holds for strongand weak ties. However, in case of rewarded referrals,recipients do not ascribe negative motives to weak-tiesmore often than to strong-ties.

Furthermore, we can show a significant main effectof tie strength on perceived sender credibility, indicatingthat, within strong-tie dyads, recipients regard senders asmore credible than in weak-tie dyads, irrespective ofreferral rewards. We find that none of the independentvariables have an influence on perceived competence ofthe sender, but that referral rewards reduce perceivedcredibility of the sender, irrespective of tie strength.

Referrals from friends lead to a more positive attitudetoward the service provider. Recipients of a referral froma strong-tie are more likely to switch to the referredservice provider. However, while we presumed thatrecipient attitude will be less favorable in case that therecipient believes that the sender will receive a reward,data did not support this assumption. Whether the referralis rewarded or not seems not to impact purchase intentions.

In summary, our results show that rewarding referralsmay decrease their overall effectiveness because rewarded


senders are attributed more negative motives for theirreferral and credibility decreases. Given that credibility isdeemed one of the main factors explaining the “power ofWOM,” this finding is important in gauging CRP. Thecredibility effect is based on the trustworthiness compo-nent while perceived expertise of the sender is not influ-enced. This highlights that the reward indeed is the causeof loss in credibility.

Unexpectedly, we did not find recipients’ attitudestoward the provider to be less favorable in case of re-warded referrals. Also, purchase intentions seem not to benegatively affected by referral rewards. These are rathersurprising findings, given that credibility is generallyregarded as a strong determinant of source persuasivenessand subsequent recipient behavior. In our study, persua-sive power of rewarded and natural referrals appears to beidentical; they have identical effects on attitudes andpurchase intentions. However, strong-ties have an overallstronger persuasive power, their referral – whether re-warded or not – has a stronger impact on recipient atti-tudes and purchase intentions.

To our knowledge, this study is the first to theoreti-cally analyze and empirically confirm the effect of rewardedreferrals on sender credibility as perceived by the recipi-ent of the referral. Unlike most approaches, our studyfocuses on the effects of WOM on the recipient’s view ofthe sender. Our findings complement prior studies thatconfirm superior effectiveness of targeting strong-ties inreferral programs (Wirtz and Chew 2002) insofar aspersuasiveness is stronger in strong-ties and likelihood toswitch to the referred service provider is higher.

However, managers should be aware of potential“social side effects” of using customers’ referrals forcorporate means: perceptions of a rewarded sender areworse than perceptions of a natural referrer, indicating athreat to the personal relationship between sender andrecipient. While we could not detect subsequent effects onattitudes toward the service provider and behavioral inten-tions, additional studies should replicate ours and evalu-ate whether the decrease in credibility has further conse-quences for a service provider and the evolution of WOMas a marketing tool.

REFERENCES

Arndt, Johan (1967), Word of Mouth Advertising. AReview of the Literature. New York: AdvertisingResearch Foundation.

Brown, Jaqueline and Peter Reingen (1987), “Social Tiesand Word-of-Mouth Referral Behavior,” Journal ofConsumer Research, 14 (3), 350–62.

Campbell, Donald and Julian Stanley (1963), Experimen-tal and Quasi-Experimental Designs for Research.Chicago: Rand McNally.

Godes, David and Dina Mayzlin (2004), “Firm-CreatedWord-of-Mouth Communication: A Field-BasedQuasi-Experiment,” HBS Marketing Research PaperNo. 04-03, Boston.

Granovetter, Mark (1973), “The Strength of Weak Ties,”American Journal of Sociology, 78 (6), 1360–80.

Haywood, Katerine (1989), “Managing Word-of-MouthCommunication,” Journal of Services Marketing, 3

(Spring), 55–67.Helm, Sabrina (2003), “Calculating the Value of Custom-

ers’ Referrals,” Managing Service Quality, 13, 124–33.

Ryu, Gangseog and Lawrence Feick (2007), “A Penny forYour Thoughts. Referral Reward Programs and Refer-ral Likelihood,” Journal of Marketing, 71 (1), 84–94.

Wangenheim, Florian and Tomas Bayón (2007), “TheChain from Customer Satisfaction Via Word-of-Mouth Referrals to New Customer Acquisition,”Journal of the Academy of Marketing Science, 35(Summer), 233–49.

Wilson, Aubrey (1994), “Stimulating Referrals,” Man-agement Decisions, 32, 13–15.

Wirtz, Jochen and Patricia Chew (2002), “The Effects ofIncentives, Deal Proneness, Satisfaction and TieStrength on Word-of-Mouth Behavior,” InternationalJournal of Service Industry Marketing, 13, (2), 141–62.

For further information contact:Sabrina V. Helm

John and Doris Norton SchoolThe University of Arizona

P.O. Box 210078Tucson, AZ 85721–0078Phone: 001.520.621.7130



EXPLORING THE EFFECTS OF REFERRAL REWARD PROGRAMS ONSATISFIED AND DISSATISFIED CUSTOMERS

Ina Garnefeld, University of Paderborn, GermanyEva Muenkhoff, University of Paderborn, GermanyJens Hogreve, University of Paderborn, Germany

Andreas Eggert, University of Paderborn, Germany

SUMMARY

Referral reward programs are a well accepted mar-keting instrument for service providers. In referral rewardprograms, a customer receives a reward for convincingothers to become a customer. To date, referral rewardprograms have been perceived as an adequate means forincreasing referral likelihood among satisfied customers.As a consequence of the proliferation of referral pro-grams, firms have recently increased the size of therewards.

In our research, we analyze consequences of anincrease in reward size for the effectiveness of referralreward programs. Large rewards may not only increasesatisfied but also dissatisfied customers’ recommendationbehavior. As satisfied customers are intrinsically moti-vated to give word-of-mouth (WOM), we assume thatlarge rewards may be even more effective in increasingdissatisfied customers’ recommendation behavior com-pared to satisfied customers.

As research so far has neglected potential referralbehavior of dissatisfied customers, we also examine posi-tive and negative consequences of recommendations bydissatisfied customers for service providers. A positiveeffect for service providers might result from a potentialcommitment effect of articulating WOM. Based on find-ings from public commitment literature, we assume acommitment effect of articulating positive WOM on dis-satisfied senders. However, there is also a potential nega-tive consequence for the service provider. In case thereceiver of WOM learns about the sender’s dissatisfactionwith the recommended service provider, this might nega-tively affect the relationship between receiver and pro-vider. To test our hypotheses, we conducted three experi-mental studies.

In Study 1, we research effects of reward size onsatisfied and dissatisfied customers. We conducted a 2 x2 between-subjects factorial design. The variables weremanipulated on two levels: customer satisfaction (dissat-isfied and satisfied) and reward size (10 reward and 50reward). The representative sample consists of 154 par-ticipants. Our results reveal a significant interaction effectbetween satisfaction and reward size on WOM. When

customers are satisfied, their intention to recommend ishigher in case a small reward is given compared to a largereward. In contrast, when customers are dissatisfied, theintention to recommend is significantly higher when largerewards are offered instead of small ones. Therefore,while increasing rewards positively affects dissatisfiedcustomers’ recommendation intentions, satisfied custom-ers’ recommendation intention decreases when a largereward is offered.

In Study 2, we focus on the effect of articulatingWOM on dissatisfied senders. We tested the proposedhypotheses by conducting an experiment. We manipu-lated the articulation of WOM (no WOM, WOM). In total,114 business students participated in this experiment. Theresults reveal a significant positive effect of articulatingWOM on senders’ loyalty. Dissatisfied customers’ loy-alty increases when a recommendation is voiced. There-fore, articulating WOM although being dissatisfied has apositive effect on the sender’s attitude toward the serviceprovider.

In Study 3, we research potential negative conse-quences of WOM in case the receiver learns about send-ers’ real attitude. In this experiment, we vary whether thesenders’ dissatisfaction with the service provider is detectedby the receiver or not (no detection, detection) and researchthe effect of the detection on the relationship betweenreceiver and service provider. Two hundred twenty-fivestudents took part in the study. We do not find anynegative effect of detecting senders’ dissatisfaction on thequality of the relationship between receiver and the ser-vice firm. Thus, the detection of a recommendation voicedby a dissatisfied sender does not have negative conse-quences on the relationship between service provider andnew customer.

Our results offer insight on the consequences ofreward size increases. Increasing the size of rewardsoffered within referral reward programs motivates dissat-isfied rather than satisfied customers to recommend theservice. Satisfied customers are even less motivated torecommend the service provider when a large reward isoffered compared to a small one. This has implications forservice providers’ decisions about reward sizes. Whenchoosing large rewards, firms must take into account that


some of their best customers, i.e., their most satisfied, willeven reduce their recommendation intention.

However, large rewards motivate dissatisfied cus-tomers to give WOM which in turn has several positiveeffects for service providers. The service provider doesnot only acquire a new customer but also profits from adissatisfied sender’s increased loyalty caused by thesender’s public commitment. According to this result,

customer referral programs are an additional option forservice providers to deal with dissatisfied customers. Wealso could not confirm a negative effect on the newcustomer-service provider relationship in case a newcustomer detects that the referral was given by a dissatis-fied customer. Hence, motivating dissatisfied customers0only leads to positive consequences for the service provider.References are available upon request ([email protected]).

For further information contact:Ina Garnefeld

University of PaderbornWarburger Str. 10033098 Paderborn




AN EMPIRICAL TEST OF THE DRIVERS OF CONSUMER TRUST IN ANE-RETAILER AND ITS OUTCOMES DIRECTED TOWARD

CUSTOMER RETENTION

Cuiping Chen, University of Ontario Institute of Technology, OshawaChuanyi Tang, The University of Arizona, Tucson

Matthew O’Brien, Bradley University, Peoria

SUMMARY

Trust is advocated as the single most important factorfor consumers choosing an online supplier (Bain andCompany/Mainspring 2000; Reichheld, Markey, andHopton 2000; Urban, Sultan, and Qualls 2000;Vatanasombut, Stylianou, and Igbaria 2004). To studytrust mechanisms underlying consumer-e-retailerexchange relationship, this research takes a new perspec-tive, a perspective directed toward customer retention. Bybringing together trust literature and global evaluationstheory, this study develops a conceptual model that iden-tifies a consumer’s attitude toward the four key compo-nents of his entire purchase experience with an onlinestore (i.e., website design, fulfillment/reliability, privacy/security, and customer service) as the key drivers ofconsumer trust in the online store and future intentionsand consumer loyalty as the direct consequences of trust.

Data for the study was collected using an onlinesurvey distributed via email to a national sample of 4,156online consumers who were randomly drawn from a panelcontaining about three million people. The response ratewas 23 percent. Hypotheses were tested using structuralequation modeling. Although the empirical test largelysupported the research hypotheses underlying the proposedrelationships in the conceptual model, some surprisingfindings were demonstrated by the empirical results.

Findings of the study imply that although trust wasstill central in online transactions, the degree of its central-ity was challenged by the empirical results that trustplayed a partially mediating role between its antecedentsand outcomes instead of the fully mediating role predictedby the study. This finding was discovered by testing theconceptual model using the overall sample, which com-bined the pure and multi-channel online store data points.Specifically, trust appears to partially mediate the effect offulfillment/reliability satisfaction on future intentions aswell as the effect of security/privacy attitudes on futureintentions; likewise, trust partially mediated the effect ofwebsite design attitudes on consumer loyalty as well as theeffect of customer service satisfaction on loyalty.

When controlling for offline experience satisfactionin predicting consumer trust in a multi-channel e-retailer,we found that trust mechanisms in the relationship betweenconsumer and a multi-channel online store were some-what different from what was predicted by the researcher.Specifically, the findings imply that customer servicesatisfaction and security/privacy attitudes still playedimportant roles in consumer trust in a multi-channelonline store when controlling for offline experience satis-faction. However, website design attitudes and fulfill-ment/reliability satisfaction no longer influenced trustafter controlling for offline experience satisfaction. Inter-estingly, trust was found to influence future intentionstoward a multi-channel online store but not loyalty. Instead,website design attitudes and offline experience satisfac-tion influenced consumer loyalty toward a multi-channelonline store directly. It was also found that trust played apartially mediating role between security/privacy atti-tudes and future intentions as well as between offlineexperience satisfaction and future intentions. Addition-ally, it was found that trust played a fully mediating rolebetween customer service satisfaction and future inten-tions. This finding is consistent with what was predictedby the researcher. Lastly, it was found that fulfillment/reliability satisfaction did not play any role when control-ling for offline experience satisfaction in the multi-channelcontext.

For both pure and multi-channel contexts, consumerloyalty was found to influence future intentions in thisstudy. This finding is not surprising given the fact thatboth attitudinal and behavioral dimensions were incorpo-rated into the definition of loyalty in this study.

As depicted above, some findings are contrary to theresearcher’s prediction. Those findings together implythat trust mechanisms in the relationship between con-sumer and a multi-channel online store may be differentfrom those in the relationship between consumer and apure online store. Future research into this area is war-ranted. References are available upon request.


For further information contact:Cuiping Chen

University of Ontario Institute of Technology2000 Simcoe Street North

Oshawa, ONCanada L1H 7K4



CONCENTRATION OF MARKETING CHANNELS AND BRAND-LEVELRETAILER MARGINS

Kenji Matsui, Kobe University, Japan

SUMMARY

Concentration in the retail industry has grown con-siderably during the last few decades in many advancedcountries. In the light of this trend, the shift in marketpower within marketing channels from manufacturersand intermediate wholesalers to dominant retailers, andthe resulting economic impact on retailers’ margins, havebecome increasingly important issues. Given such ongo-ing structural changes affecting retailers, the objective ofthis research is to investigate the impacts of marketingchannel concentration in each distributive stage on bothwholesale and retail prices, and on the difference betweenthese prices – namely, retail gross margins – at the level ofbrands. Many retailing studies have investigated retailmargins using either firm-level financial data or price dataon a small number of products supplied by a specificindustry. However, research measuring the impact of theconcentration of marketing channels on retailer marginsusing extensive wholesale price data, covering variousconsumer products, is missing from the existing literature,primarily owing to the scarcity of brand-level wholesaleprice datasets. Hence, the simultaneous investigation ofthe retail margins of various products undertaken in thisstudy provides new additional insights into the retailindustry. Specifically, this research is able to address thisissue empirically using a unique official Japanese dataset,the National Survey of Prices, which reports not only theretail prices but also the wholesale prices of famousnational brand items and does so not simply at the cat-egory or firm level but also at the brand level.

The most notable empirical finding from this researchis that retailers’ margins on a broad range of consumerproducts have an inverse relationship with the degree ofwholesaler concentration, which implies that interactionsbetween retailers and upstream wholesalers occur. Thisfinding is consistent with previous empirical studies thathave investigated the impacts of manufacturerconcentration in a specific industry on category-levelretailer margins (e.g., Steiner 1993; Ailawadi and Harlam2004). Although previous empirical studies have confirmedthis tendency (known as the Steiner effect) from price dataon a specific product category, this study is the first to findthe inverse relationship between supplier concentrationand retailer margins by employing a comprehensive brand-level wholesale price dataset that covers many nationalbrands over a broad range of consumer product categories,ranging from convenience goods to shopping goods. A

further unique contribution of the present study is that itinvestigates the impact of wholesaler concentration ratherthan manufacturer concentration.

In contrast to the case of wholesaler concentration,this study finds that the degree of retailer concentrationitself has no influence on the retail margin, consistent withempirical findings in Betancourt and Gautschi (1992,1993); the one exception, however, is the case wheremanufacturers’ branches sell products directly to retail-ers. In this case only, the degree of retailer concentrationpositively affects retailers’ margins. As at least one whole-sale stage becomes disintermediated when one of themanufacturer’s branches sells a product directly to aretailer, “double marginalization” by both a manufacturerand a wholesaler does not occur. Therefore, the supplyquantity is not reduced below the desired level, enablingretailers to obtain a higher margin more easily. Therefore,we may interpret the result as indicating that thedisintermediation of wholesalers prevents the retailers’purchase prices from being driven up and the quantity ofsupply being reduced, thus enabling retailers to earnhigher margins, ceteris paribus.

Other empirical findings in this research relate towholesale prices and retail prices. As generally expected,wholesale prices depend positively on the degree ofwholesaler concentration. In relation to retailers, whenthese are classified by retail format, supermarkets face asmaller increase in their purchasing prices than do ordi-nary retail stores as a result of supplier concentration. Thissuggests that a large retail format exerts a stronger bar-gaining power over suppliers compared with the smallretail format of ordinary retail stores. As expected, retailprices are positively associated with the degree of retailconcentration. However, a closer examination of retailformat classifications reveals that supermarkets do notraise retail prices according to their degree of concentra-tion. This result implies that a significant part of theirbenefits from concentration is passed on to consumers inthe form of lower prices, which is consistent with thetheoretical framework presented by Aalto-Setälä (2002).Additionally, when sales companies established by manu-facturers or manufacturers’ branches – both of which maybe regarded as distributors substantially integrated by aspecific manufacturer – sell products directly to retailers,a greater wholesale concentration reduces retail prices.We may interpret this result to mean that vertical integra-tion of the distribution function by a manufacturer leads to


lower retail prices by avoiding double marginalizationand thus enhances consumer welfare.

A further supplementary result shows that the brandholding the highest share in each item category tends to

exhibit a low retailer margin. This result may reflect thefact that a retailer is willing to accept a lower gross marginto deal in a popular brand item when the brand holdsconsumer loyalty (Steiner 1993). References are availableupon request.

For further information contact:Kenji Matsui

Graduate School of Business AdministrationKobe University

2-1, Rokkodai-cho, Nada-kuKobe, Japan, 657–8501Phone: +81.78.803.6908



EFFECTS OF THE EXTANT ENVIRONMENT ON INFLUENCESTRATEGY USAGE IN THE FRANCHISE

SYSTEM RELATIONSHIP

Kyle A. Huggins, James Madison University, HarrisonburgDarin W. White, Samford University, Birmingham

SUMMARY

Franchising is literally sweeping the world. There arebelieved to be more than 750,000 franchise businessesoperating in the U.S. employing over eighteen millionpeople, accounting for over $1.5 trillion in economicoutput (Redstone Advertising 2009). However, this ex-plosive growth in franchising has not been without diffi-culty. Corporations are finding it increasingly challeng-ing to deal with more diverse and uncertain externalenvironments as they grow all across the globe. Priorresearch has found the franchise failure rate to be 15percent to 34 percent (Olson 2008; Bates 1998;Castrogiovanni, Justis, and Julian 1993). Thus, studyingfactors that influence the success of franchises in varyingenvironments is a promising field for academic research.

Given the contractual nature of franchise systemrelationships, the macro environment presents specialchallenges for decision makers in these organizations.Difficulty in gathering and processing information on theenvironment can negatively affect a channel relationshipin numerous ways (Dwyer and Welsh 1985). Therefore,when a franchiser desires for a franchisee to undertakecertain actions it will seek to exercise its position of powerin the system by communicating in different ways thatinclude information exchange, requests, recommenda-tions, promises, threats, and legalistic pleas (Frazier andSummers 1984, 1986; Boyle et al. 1992). Thus, influencestrategies are “the communicated portion of influenceattempts that one channel member uses to gain the compli-ance of another channel member” (Payan and McFarland2005).

Previous research has found that influence strategiesfall into one of three categories: non-coercive, soft coer-cive, or hard coercive (Venkatesh, Kohli, and Zaltman1995). Non coercive strategies (which include informa-tion exchange and recommendations) operate by chang-ing the attitude of the target about the attractiveness of theintended behavior (Payan and McFarland 2005). Whenutilizing a soft coercive strategy (which include requestsand promises), a source explicitly identifies a particularcourse of action that the source would like the target totake. These forms of communication intrude into thefranchisee’s decision making realm to a degree and entailan implied element of coercion (Frazier and Summers

1984). Hard coercive influence strategies (which includethreats and legalistic pleas) stimulate conformity on thebasis of the influence mechanism of source-controlledpunishments and rewards (Payan and McFarland 2005).Therefore, because environmentally uncertainty affectsdependence levels of franchisees on franchisers, we con-tend that the franchisers’ choice of influence strategy(hard, soft, or non-coercive) is driven, at least to somedegree, by perceived conditions of environmental uncer-tainty. Additionally, Achrol and Stern (1988) identifiedthree dimensions that are useful in studying environmen-tal effects on franchise system processes. In this tradition,researchers have studied the impact of environmentalheterogeneity, variability, and munificence on channelstructures and processes (Grewal and Dharwadkar 2002).It is proposed that three dimensions of the environment(heterogeneity, variability, and munificence) will inde-pendently exert varying degrees of influence on thefranchiser’s selection of communication strategies withina franchise channel dyad.

Our data collection consisted of a single franchisesector, which was selected to minimize the effect ofpossible cross industry managerial differences. For anindustry to be a good candidate, its participants shouldface differing environmental conditions across the sampleframe. After considering a variety of possibilities, the fastfood industry in the United States was selected for thestudy. According to the National Restaurant Associationresearch and figures from the U.S. Bureau of EconomicAnalysis and the Bureau of Labor Statistics, vast regionaland local environmental differences exist within the res-taurant industry, which fits the criteria for candidacy. Totest our hypotheses, we measured franchisees’ percep-tions of the franchise system relationship. We targetedfranchisee owners that control five or less total units in aneffort to locate owners that are intimately involved in theday-to-day operations of their business. These owner/managers typically deal directly with the franchiser, so thekey informant data collect method was implemented(Quinn 2009).

A national sample of one thousand one hundred andforty-five (1,145) franchisee-owned restaurants was cho-sen to participate in the study. Franchisees of 10 differentfast food companies were mailed survey questionnaires.Ninety-three percent of the names and addresses, repre-


senting eight of the companies, were purchased from amarketing research company in the Midwest. Companiessurveyed included McDonalds, Burger King, Taco Johns,Subway, Blimpies, Pizza Inn, International House ofPancakes, and Arby’s. The remaining names and addresseswere received from two companies (Sonic and Steak-Out)that voluntarily agreed to participate in the study. A totalnumber of 180 usable responses were obtained, resultingin a net response rate of 16.1 percent. Although theresponse rate is low, it is comparable to those of severalother empirical studies in the channel arena (cf., Etgar1977; Brown and Day 1981; Weiss and Anderson 1992).

Generally, our findings differentiate from previousresearch which indicates channel members commonly

prefer to use non coercive influence to coordinate theirchannel partners. In fact, our study demonstrates that theuse of hard coercive strategies significantly increase dur-ing uncertain and tenuous environmental conditions. Inthese conditions, franchisers refuse to avoid dysfunc-tional conflict when coordinating franchisees as previousresearch suggests. However, we do see that as uncertaintygives way to more stable environments, the use of hardcoercive strategies likewise decrease. In more stable envi-ronments, and especially in rich munificent environ-ments, franchisers move toward softer coercion tactics, ifany are needed at all. Ultimately, our research demon-strates that franchiser use of influence strategies are cer-tainly determined by environmental influences.

For further information contact:Kyle A. Huggins

College of BusinessJames Madison University

800 Main Street, MSC 0205Harrisonburg, VA 22801



EFFECTIVE BRAND STRATEGY IMPLEMENTATION: A REVIEW OFTHE LITERATURE AND FRAMEWORK FOR FUTURE RESEARCH

Alexander Haas, University of Bern, SwitzerlandAdele J. Huber, University of Bern, Switzerland

SUMMARY

Understanding brand strategy implementation (BSI)has been a long-standing goal of researchers and manag-ers alike and scholars widely agree on the importance ofBSI for enhancing firm performance (e.g., Aaker 1996;Keller and Lehmann 2006). While the focus of extantresearch has been on marketing instruments, organiza-tional BSI capabilities and their link to firm performancehave attracted only limited attention and still are not wellunderstood (Chimhanzi and Morgan 2005; Hickson et al.2003; Menon et al. 1999; Piercy 1998). Accordingly,although organizations invest huge amounts of resourcesin BSI efforts, many strategy implementation initiativesfall far behind expectations (Bigler 2001; Hickson et al.2003; Ind 2007; Wong and Merrilees 2007).

Against this background, the objective of this paperis to conceptually investigate effective BSI and clarifyingthe link between strategy development and implementa-tion. To this end, we review the extant conceptual andempirical literature on BSI and, based on the literaturereview, develop a conceptual framework of effective BSI.In line with extant research we define brand strategyimplementation (BSI) as the communication, interpreta-tion, adoption, and enactment of a brand strategy or abrand strategy initiative (Noble and Mokwa 1999).

The literature review shows that there is a variety ofconceptual work and a limited but growing body ofempirical evidence, which provide various insights intoBSI. However, these insights fail to provide a clear pictureof effective BSI. This is mainly due to researchers’ limiteduse of theories and the variety of theories used in thosestudies with a theoretical foundation. Given the frag-mented empirical evidence and the lack of consistentefforts to integrate the evidence systematically, researchon effective BSI requires conceptual research that inte-grates previous BSI-related research, organizes andexplains existing evidence of this research (e.g., by usingwell-established bodies of theory), and provides guide-lines for theory-driven, systematic future research on BSI.Further, the literature review shows that various issuesimportant to understanding effective BSI (i.e., theconceptualization of organizational BSI capabilities, pos-sible implementation effects of brand strategy develop-ment, and the relationship between organizational BSIcapabilities and firm performance) remain unclear. Hence,

these issues should be addressed in conceptual workaimed at understanding BSI effectiveness.

We developed a conceptual framework for investi-gating effective BSI with four basic elements: (1) BSIeffectiveness, (2) implementation-related brand strategydevelopment capabilities as determinants of BSI effec-tiveness, (3) firm performance as outcome of BSI effec-tiveness, and (4) organizational learning capabilities asmoderators of the relationship between brand strategydevelopment capabilities and BSI effectiveness. Basi-cally, the proposed framework links organizational BSIcapabilities to firm performance and suggests implemen-tation-related brand strategy development capabilities,organizational learning capabilities, and BSI effective-ness as core elements of organizational BSI capabilities.Our framework organizes both important variables relatedto, and extant evidence on, effective BSI. It builds onscholars’ conclusion that a better integration of brandstrategy development and brand strategy implementationwould result in higher BSI effectiveness (Noble 1999;Thorpe and Morgan 2007). More specifically, its basicrationale is that brand strategist, strategy developmentprocess, and brand strategy, all usually considered aselements of strategy development, possess particular imple-mentation-related characteristics that strongly affect BSIeffectiveness.

We use organizational learning theory (Argyris 1989;Huber 1991; Kim 1993) and the resource-based theory ofcompetitive advantage (Barney 1991; Barney 2001) tolink organizational BSI capabilities to firm performance,and based on these theories, identify key variables ofeffective BSI and explain how these variable relate to eachother. While the framework focuses on a set of relation-ships that the theories and previous research suggest askey to the investigation of effective BSI, based on thetheoretical rationale, additional relationships may easilybe identified and integrated into the framework.

Theoretically, we identify the core elements of orga-nizational BSI capabilities (i.e., implementation-relatedbrand strategy development capabilities, execution- andimprovement-oriented Organizational learning capabili-ties, and BSI effectiveness). Drawing on the organiza-tional learning theory and the resource-based theory ofcompetitive advantage we link organizational BSI capa-bilities to firm performance, and based on these theories,


identify key variables of effective BSI and explain howthese variables relate to each other. Furthermore, weclarify the link between brand strategy development andBSI, which the literature review has identified as impor-tant, but largely neglected by previous research. Finally,we provide an answer to the question of how organiza-tional BSI capabilities affect firm performance, adding to

the literature on performance-relevant organizational skills(Day 1994). Managerially, the paper suggests key vari-ables and mechanisms managers should focus their atten-tion on in their efforts to enhance BSI effectiveness andBSI’s contribution to firm performance. References areavailable upon request.

For further information contact:Alexander Haas

Marketing DepartmentInstitute of Marketing and Management

University of BernEngehaldenstrasse 4

CH-3012 BerneSwitzerland

Phone: +41(0)31.631.4541Fax: +41(0)31.631.8032



EXPLORING THE BRAND PRODUCTIVITY GAP

Maik Hammerschmidt, University of Mannheim, GermanyTomas Falk, University of Mannheim, Germany

Frank Germann, Pennsylvania State University, University ParkWilliam T. Ross, Jr., Pennsylvania State University, University Park

Hans H. Bauer, University of Mannheim, Germany

SUMMARY

Brands are generally considered among the mostvaluable intangible assets that a firm can have whichmeans that investment to manage those brands is a highpriority for top management (Keller and Lehmann 2006).As a response to the worldwide economic crisis, manyfirms are cutting back on their brand building invest-ments. The important question is whether or not thisdecision is best for the firm. Reducing brand investmentswill reduce costs, but it could reduce brand outcomes, too.Taken together, these two countervailing effects mayresult in either higher or lower returns on branding, andthe questions of which will result and why are difficult toanswer.

Drawing on the production economics literature (e.g.,Seiford and Zhu 1999; Luo and Donthu 2006), in thispaper, we introduce a two-step brand productivity modelthat measures and investigates how firms convert multiplebrand inputs into multiple outputs during the brand man-agement process. The first step incorporates conversionof brand investments into customer-related outcomes andthus captures brand efficiency, how well brand invest-ments do what they are meant to do, influence customers.The second step addresses the translation of customer-level outcomes into financial outcomes and captures brandeffectiveness, the degree to which customer outcomesresult in better firm performance. Brand productivitycombines both perspectives into an overall assessment ofbranding capabilities. This two-step approach fits with therecent notion that value creation at the customer-levelalone is insufficient to ensure the ultimate success of thefirm (Grewal et al. 2009). Firms must do two transforma-tions well to obtain an appropriate level of financialreturns from their brand investments. Distinguishing be-tween the two steps opens the brand productivity “blackbox” so that where, in the brand productivity chain, agiven firm is not as efficient and/or effective as its com-petitors can be determined.

We also expand on existing studies linking brandinvestments to financial outcomes by applying data envel-opment analysis (DEA) instead of parametric (regression-

based) methods. Using DEA is a response to recent callsfor methods that identify above-average brands that out-perform their industries (Muhanna et al. 2004) and allowsbenchmarking the actual investment (i.e., inputs) of eachbrand against the level invested by best-performing, fron-tier brands operating under the same conditions. We testour brand productivity model with comprehensive paneldata for 244 brands in 12 product categories, allowinggeneralization beyond existing studies which have focusedon a few top brands in a single category. Moreover, weexpand the brand-specific inputs we investigate to includeproduct quality and distribution as well as communica-tion, i.e., advertising, the sole focus of previous work.This use of more of the important brand investments,combined with the DEA methodology, allows us to opti-mize resource allocation across the different brandinginstruments. In addition, we include multiple outcomemeasures from different sources. We use panel data fromthe large scale Young & Rubicam Brand Asset Valuatorto capture customer-level brand outcomes (brand aware-ness and brand image), and we employ product-marketmetrics (brand revenue premium and EBITDA) and stock-market related information to capture financial brandoutcomes.

Our results reveal several interesting patterns. Thereare considerable brand efficiency and effectiveness dif-ferences across brands and across product categories.Specifically, unproductive brands exhibit two typicalinput-output transformation patterns resulting in brandproductivity gaps: inefficient-effective (e.g., financialservice brands) and efficient-ineffective (e.g., desktopcomputer brands). Additionally, differentiating betweenthe two steps provides insights into the specific sources ofbrand productivity gaps, and these insights enhance thefirms’ sense and ability to respond as it aims to improve itsbrand management process. By contrasting a brand’scurrent productivity level with a best-performing bench-mark brand in its respective industry we provide sugges-tions on how a brand can reduce its brand investmentswithout threatening either its customer or financial out-comes. Finally, we find that our two-step model bearssignificant advantages over a less sophisticated one-stepmodel.


REFERENCES

Grewal, Dhruv, Gopalkrishnan R. Iyer, Wagner A.Kamakura, Anuj Mehrotra, and Arun Sharma (2009),“Evaluation of Subsidiary Marketing Performance:Combining Process and Outcome PerformanceMetrics,” Journal of the Academy of Marketing Sci-ence, 37 (2), 117–29.

Keller, Kevin Lane and Donald R. Lehmann (2006),“Brands and Branding: Research Findings and FuturePriorities,” Marketing Science, 25 (6), 740–59.

Luo, Xueming and Naveen Donthu (2006), “Marketing’sCredibility: A Longitudinal Investigation of Market-ing Communication Productivity and ShareholderValue,” Journal of Marketing, 70 (4), 70–91.

Muhanna, Waleed A., Ray Gautam, and Jay B. Barney(2004), “Capabilities, Business Processes, and Com-petitive Advantage,” Strategic Management Jour-nal, 25 (1), 23–38.

Seiford, Lawrence M. and Joe Zhu (1999), “Profitabilityand Marketability of the Top 55 U.S. CommercialBanks,” Management Science, 45 (9), 1270–88.

For further information contact:Maik Hammerschmidt

University of MannheimL 5, 1

D-68131 MannheimGermany

Phone: +49.621.181.1569Fax: +49.621.181.1571



SPONSORSHIP DEAL CHARACTERISTICS, FAN PERCEPTIONS, ANDSPONSORSHIP FIT: AN EMPIRICAL EXAMINATION

David M. Woisetschläger, TU Dortmund University, GermanyChristof Backhaus, TU Dortmund University, GermanyManuel Michaelis, TU Dortmund University, Germany

SUMMARY

Sponsorship fit is frequently mentioned and empiri-cally examined as a success factor of sponsorship (e.g.,Becker-Olsen and Hill 2006; Simmons and Becker-Olsen2006; Speed and Thompson 2000). Sponsors that areperceived as congruent with the sponsored object havebeen found to benefit more from their sponsorship en-gagements in terms of increasing brand equity (Becker-Olsen and Hill 2006; Gwinner and Eaton 1999; Simmonsand Becker-Olsen 2006) and purchase intentions (Speedand Thompson 2000) than incongruent sponsors. Spon-sorship fit can be defined as perceived match of attributesbetween sponsoring firms and sponsored objects (Becker-Olsen and Hill 2006). While sponsorship fit has beenconsidered as a determinant of sponsorship success, littleknowledge exists about the antecedents of sponsorship fit.However, from a managerial point of view and especiallyfor sponsors that are in a particular situation of low fit,identifying antecedents of (in)congruence is highly rel-evant for improving the effectiveness of sponsorships. Onthe one hand, sponsorship fit can be seen as a result of acognitive process including different antecedents whichis different for each individual. On the other hand, spon-sorship fit is likely to be different between brands, i.e., thegroup level. From a managerial perspective, it is highlyrelevant to determine if firm actions are reflected in fanevaluations of sponsorship fit, especially if fans are arelevant target group. The size of this target group can beas large as 50.3 (CF Barcelona), 32.8 (Manchester United)or 19.9 (Bayern Munich) Million fans in Europe alone(figures: Sport+Markt 2008).

Existing literature on group level determinants ofsponsorship effectiveness is predominantly based on self-reported data (e.g., Cornwell, Roy, and Steinard 2001)rather than examinations of objective differences betweensponsorship deals and their impact on consumerperceptions.

Hence, the present paper contributes to the literatureby examining the individual and firm-level antecedents ofsponsorship fit. We draw on (1) a sample of 1,799 soccerfans from 31 clubs of the first and second German soccerleague in order to obtain individual drivers of sponsorshipfit. In addition, we gathered (2) objective data, i.e., char-

acteristics of shirt sponsorship deals of clubs analyzed inorder to explain between-group differences. Given thenested structure of the data, hierarchical linear modeling(HLM) is an adequate method for analysis and henceapplied in this study (cf., Bryk and Raudenbush 2002 foran overview on the method).

Several factors have been mentioned in previousstudies that are proposed to contribute positively to spon-sorship fit or sponsorship related outcomes. Functionalsimilarity (Gwinner and Eaton 1999), relatedness to thesponsored activity, perceived benefits of the partnership(Woisetschläger and Haselhoff 2009), regional identifi-cation of the sponsor (Woisetschläger and Haselhoff2009), the sponsors sincerity (Speed and Thompson 2000),dominance in the partnership, and the ubiquity of thesponsorship (Speed and Thompson 2000) are potentialdeterminants of sponsorship fit.

Besides determinants of sponsorship fit that are per-ceived differently by individual fans, characteristics ofthe sponsorship partnership are proposed to affect spon-sorship fit. Typically, sponsorship contracts can be char-acterized by the contract value (i.e., the amount paid by thesponsor per year), the duration of the contract, and theregionality of the sponsor. These objective measuresindicate the firm’s commitment in terms of budget, timeline,and shared regional identity. Differences in contract value,contract length, and regionality are expected to partiallyexplain differences of the level of sponsorship fit overdifferent sponsor-partnerships.

Results indicate that at the individual fan-level, spon-sorship fit is influenced by relatedness to the sponsoredactivity, dominance, sincerity, regional identification, andperceived benefits. Functional similarity and ubiquity didnot show an impact on sponsorship fit. The model explains47.35 percent of the variance in sponsorship fit.

At the group level, results show that the intercept ofthe regression is significantly influenced by past contractlength, future contract length, and regionality. However,there is no significant influence of contract value. About31 percent of the variance in the intercept of sponsorshipfit can be explained by past contract length, future contractlength, and regionality of the sponsor.


Our findings reveal that, on the individual fan level,perceived benefits are the strongest determinant of spon-sorship fit. Sponsors should therefore actively communi-cate the benefits of their engagement. Regional identifica-tion is also an important antecedent of sponsorship fit.Firms that show their belongingness to the region of theclub are perceived as in-group members, and therefore ascongruent to the club. Perceived sincerity of the sponsoris also a positive determinant of sponsorship fit. There-fore, firms should communicate that they are reliablepartners of the particular club but should avoid to beperceived as a dominator of the club. Relatedness to othersports activities is also positively contributing to sponsor-ship fit. Contrary to our propositions, functional similarityand perceived ubiquity are not significantly related tosponsorship fit. As a consequence, functionally dissimilarfirms such as McDonald’s in their current partnershipwith the FIFA football federation could be perceived as

congruent partners as a result of other determinants ofcongruence. Moreover, sponsorship engagements withseveral clubs are not negatively perceived by the fans.

On the group level, results show that past and futurecontract lengths contribute positively to sponsorship fit.Sponsors that strive for short-term impact are evaluatedless favorable, while sponsors that are situated in theregion of the club are evaluated significantly better interms of sponsorship fit. If companies are willing to takeadvantage of this fact is likely to depend on the attractive-ness of the regional market and the fans as target custom-ers. Contrary to our expectations, contract value is notsignificantly related to sponsorship fit. Hence, firms shouldinvest in long-term sponsorship relations instead of spend-ing more of their budget over a shorter time period.References are available upon request.

For further information contact:David Woisetschläger

Business and Social SciencesTU Dortmund University

Otto-Hahn-Str. 6Dortmund, 44227




THE SPILLOVER OF BRAND EQUITY WITHIN A CORPORATESPONSORSHIP PORTFOLIO

Joe Cobbs, Northern Kentucky University, Highland HeightsMark Groza, University of Massachusetts, Amherst

SUMMARY

The image transfer between a sports enterprise orcelebrity endorser and a single corporate sponsor’s brandis well documented in the literature (e.g., McCracken1989; Gwinner and Eaton 1999; Speed and Thompson2000; Till and Busler 1998, 2000). However, these sponsoror endorsement relationships seldom exist in isolation(Smith 2004). Numerous firms often associate their brandwith the same iconic enterprise or celebrity simultaneously.While the research to date on brand associations in thecontext of sponsorship has focused on consumers’association of a sponsored enterprise with a sponsoringfirm (Gwinner and Eaton 1999; Meenaghan 1991, 2001;Meenaghan and Shipley 1999), this study investigatesbrand spillover effects between sponsors partnered with acommon sports enterprise. An experimental designutilizing actual consumer brands is employed to ascertainthe influence of a sponsorship portfolio on consumers’perceptions of brand equity and purchase intentions of aparticular sponsoring brand.

Although empirical research on the impact of a port-folio of brand images in a sponsorship environment islacking, early brand alliance research has shown thephysical or symbolic combination of two or more indi-vidual brands can result in spillover effects (Fang andMishra 2002; Rao, Qu, and Ruekert 1999). These effectsoccur when consumers’ perceptions of a single brand areinfluenced by the other brands in an alliance or jointbranding situation (Lebar et al. 2005; Samu, Kirshnan,and Smith 1999; Simonin and Ruth 1998). Drawing on thetheories of brand associations (Keller 1993, 2003) andbrand equity (Aaker 1991), and considering empiricalevidence in brand alliance situations, the following twohypotheses were tested:

H1: Within a sponsorship portfolio composed of multiplebrands, consumers’ perceptions of the brand equityof a sponsor are positively influenced by the brandequity of the other brands contained within theportfolio.

H2: Within a sponsorship portfolio composed of multiplebrands, consumers’ purchase intentions for a particularbrand are positively influenced by the brand equity ofthe other brands contained within the portfolio.

In the primary study, 160 undergraduate studentsfrom two Northeastern universities were randomlyassigned into one of two sponsorship portfolio conditions.Each condition employed three distinct brands and onecommon focal brand within the sponsorship portfolio.Brands were chosen based on their brand equity scores inBrandweek’s annual Superbrands issue (2008) and afteran initial pretest and focus group. Marriott Hotels wasselected to serve as the common brand in both sponsorshipportfolios. Dodge, Kmart, and Discover Card comprisedthe LOW portfolio condition; while the HIGH portfoliocondition consisted of Toyota, Target, and VISA.

Subjects were subsequently asked (as a distractertask) to assess the clarity and effectiveness of three differ-ent advertisements, each of which featured a professionalhockey league thanking the same four corporate sponsorsfor their support. Then, respondents’ perceptions of brandequity and purchase intentions were ascertained throughpreviously validated scales (Aaker 1996; MacKenzie,Lutz, and Belch 1986; Voo, Donthu, and Lee 2000).

An analysis of variance (ANOVA) was employed toanalyze the difference between subjects’ perceptions ofMarriott in the low and high portfolio conditions. Thebrand equity impressions of Marriott (H1) were signifi-cantly higher when Marriott was presented within a spon-sorship portfolio consisting of higher equity brands (p =.017). Though respondents exposed to the portfolio ofsponsors consisting of higher equity brands demonstrateda greater intention of choosing Marriott for their next hotelstay, the difference between portfolio groups (H2) wasnot statistically distinct (p = .155).

The study undertaken in this paper reveals empiricalevidence of a spillover effect in brand equity betweencorporate sponsors aligned with a common sports enter-prise. This finding extends the literature on commercialsponsorship, which has primarily been concerned with thedyadic relationship between the sponsored enterprise anda single sponsoring firm (e.g., Gwinner 1997; Meenaghan2001; Speed and Thompson 2000). Future research shouldtake a broader perspective of the sponsorship environ-ment by focusing on the full context of the commercialsponsorship portfolio instead of a dyadic alliance betweena single corporate partner and a sponsored enterprise.


This research also advances the brand alliance litera-ture by extending the evidence of spillover effects tosecondary associations (Keller 1993; Lederer and Hill2001). While the sponsoring brands in this experiment didnot directly align with each other to produce a co-brandedproduct or engage in an intentional co-marketing initia-tive (Bucklin and Sengupta 1993), their connection to a

common promotional enterprise led respondents to seem-ingly form associations between brands that impactedassessments of brand equity. As a result, firms enteringalliances to achieve promotional objectives should care-fully map the secondary associations that accompanysuch an alliance. References are available upon request.

For further information contact:Joe Cobbs

Haile/US Bank College of BusinessNorthern Kentucky UniversityHighland Heights, KY 41099



EXPLORING THE POWER OF QUALITATIVE COMPARATIVEANALYSIS FOR MARKETING RESEARCH: REPLICATION

AND EXTENSION OF A NEW PRODUCTDEVELOPMENT STUDY

Andrea Ordanini, Bocconi University, ItalyEdwin Nijssen, Eindhoven Technical University, The NetherlandsFred Langerak, Eindhoven Technical University, The Netherlands

SUMMARY

Many marketing problems exist of severalinterconnected traits simultaneously driving an outcome.Examples may include brand value decisions, productlaunch decisions or firms’ foreign market entry. In suchsituations, seeking for the individual effect of anindependent variable or simple interaction effects may beof little interest. However, marketing researchers rarelylook for configurations of decisions or consider multiplesets of solutions leading to the same outcome. This lack ofattention for additivity and equifinality of solutionsprobably is an artifact of prevalent research methods inmarketing.

This study introduces qualitative comparative analy-sis (QCA) as a useful complementary method to addressthese issues in marketing research. Based on Booleanalgebra rules, QCA can identify complex combinations oftraits that have the necessary and/or sufficient conditionsfor an outcome. QCA is a mixed qualitative – quantitativetechnique originally developed for sociology and politicalsciences that models relations between variables in termsof their set membership rather than using traditionalcorrelation- or regression-based methods. Specifically,instead of disaggregating observations into independent,analytically separate elements, and then testing their aver-age net effects on an outcome using statistical theory,QCA conceptualizes cases as any possible combinationsof elements or traits and uses Boolean algebra to identifycombinations that offer necessary or sufficient conditionsto lead to specific outcomes. This makes QCA capable todescribe complex sets of relationships overcoming theassumptions of additivity and equifinality. Moreover, theuse of a mathematical/descriptive logic also makes QCAsuitable to deal with configuration analyses even in verysmall populations, thus being a powerful complementarytechnique to existing research methods in marketing.

To demonstrate the potential of QCA, we replicateand extend an existing study on new product development(NPD) that focuses on the effectiveness of a hierarchicalapproach as a means to accelerate NPD. Apart from thecentrality of the NPD in marketing, our choice was

informed by access to the original data and the fact thatsuch study is one of the few already published works thatsimultaneously contain the three interesting elements forapplying QCA: additivity, equifinality, and small samplesize as an extra complicating factor.

The results of the QCA replication show some simi-larities with the original study, in that the effects ofacceleration approaches on market opportunities and firmperformance are highly congruent. A hierarchical imple-mentation approach improves firm performance, and ran-dom approaches are not effective for increasing firmperformance. Moreover, the implementation of NPD accel-eration approaches in a manner consistent with the MRWhierarchy does not enhance market opportunities, butother, nonhierarchical approaches can do so. The otherQCA results appear only partially consistent with theoriginal findings and suggest that a more hierarchicalimplementation pattern relates to neither improved newproduct profitability nor higher development speed. Fur-thermore, we only find partial support for the originalclaim that a fully random approach is associated withhigher development speed.

Our replication also helps to understand the impact ofindividual NPD acceleration approaches, something thatwas not investigated in the original study. Along this line,we find that the simplify approach is most important forincreasing development speed and improving perfor-mance. Conversely, eliminating steps apparently is a poorapproach that should be never implemented except toenhance firm performance. The other approaches in theMRW hierarchy have contingent effects, depending onthe other acceleration approaches implemented. In gen-eral, the role of doing things in parallel seems less impor-tant than that of speeding up or eliminating delays. Fur-thermore, our QCA replication results reveal that theapproaches that increase development speed also tend toimprove new product profitability and market opportuni-ties. The main difference pertains to doing things inparallel.

To further show the potential of QCA, we extend thereplication by including another possible contingent fac-


tor for which we had information in the original dataset:the degree of sophistication of the NPD organization. Theresults of our additional QCA analyses support the notionthat the organizational structure of the NPD processinfluences the effect that specific acceleration approacheshave on development speed and/or performance. Firmswith simple NPD organizations have multiple pathwaysthey might follow to accelerate their development pro-cess. In particular, they may adopt less sophisticatedacceleration approaches in isolation, such as speeding up,or pursue combinations, such as speeding up versus theelimination of delays or the simplification of activities.Firms with sophisticated, team-based organizations ben-efit most from combining the simplification approachwith parallel processing while not engaging in task elimi-nation and removing delays.

We make a threefold contribution to marketing litera-ture. First, we provide initial evidence of the power ofQCA in conditions where additivity, equifinality, and asmall sample size affect the research context. We do notadvocate QCA as the only or a superior technique per sebut promote it as an alternative and complementaryapproach for circumstances in which multiple and com-

plex interaction effects are possible, yet too complex toinvestigate using traditional analytical techniques. Sec-ond, our paper adds to the existing stream of methodsresearch by highlighting the generally overlooked rel-evance of method triangulation. Researchers widelyacknowledge the importance of using multiple method-ologies to increase the robustness of findings, yet mostmarketing research continues to rely on single analyticalmethods. Moreover, studies based on methodologicalreplications and/or extensions are negligible in market-ing. Our study proposes QCA as a useful complementarytool to increase the robustness of findings in the marketingfield. Third, our study adds to the NPD literature, espe-cially on the stream that cites the effectiveness of differentacceleration approaches to reduce NPD time. By applyingQCA to an existing study, we enhance our understandingof the effect of simultaneous use of several accelerationapproaches on development speed and performance.Because additional data on NPD organization was avail-able in the original data set (not used in the originalanalysis) we also were able to make an interesting exten-sion, exploring how the use of acceleration approaches iscontingent on the NPD organization within a firm.

For further information contact:Andrea Ordanini

Management DepartmentBocconi University

Via Rontgen, 120136 – Milano

ItalyPhone: +39.0258363623



IMPROVING THE DIRECT ESTIMATION OF DEMAND BY ADJUSTINGFOR INCORRECT PRICE-STATEMENTS

David Blatter, University of Bern, SwitzerlandReto Hofstetter, University of Bern, Switzerland

Klaus Miller, University of Bern, Switzerland

SUMMARY

The key to a profitable pricing of product innovationsis to estimate consumer demand accurately. To gaugedemand for new and existing products, market research-ers often measure consumers’ maximum willingness topay (WTP) using direct WTP elicitation methods such asthe open-ended (OE) question format. The OE questionformat offers several benefits to the market researchersuch as cost and time savings and is often used in marketresearch. However, under the hypothetical OE questionformat survey respondents have little incentive to performthe necessary cognitive effort in order to accurately statetheir WTP, as their WTP statement has no economicconsequences. Therefore, direct measurement of con-sumers’ WTP using the OE question format may yieldbiased results. In this research, we argue that respondentsmay reduce their cognitive effort by reporting some othertype of price than their current maximum WTP, as otherprice concepts such as internal reference prices might bemore readily available to respondents. Based on thisargument, we present a new approach to adjust for suchincorrect price statements and hence, reduce the bias ofthe OE question format.

In our first study, we ask 99 subjects to state theirWTP for a newly introduced electronic bike, using the OEquestion format. In addition, the participants have to self-select a price-type among six possible price concepts thatwould best reflect the type of price they had stated duringthe previous WTP elicitation task. Such price-types include,e.g., the price the product would presumably cost in astore, a fair price, or a price the respondent would pay inthe future. Our data reveals that most respondents (56.57%)indeed do not state their current maximum WTP but reportother price concepts instead.

Based on this finding, we develop a second study inwhich we ask 781 consumers to state their WTP for anewly developed consumer good, a scarf with the imprintof the university’s logo. The study design comprises three

experimental groups (between-subjects design). In thefirst group (OE group), consumers only have to state theircurrent maximum WTP using the OE question format. Inthe second group (OE-Cut group), the OE question isfollowed by a price-type self-selection task (similar toStudy 1). In the third group (BDM group) we determineWTP directly by the incentive compatible BDM mecha-nism proposed by Becker, DeGroot, and Marschak (1964).We first test the existence of a bias of the WTP statementsin the OE group. We define the bias as the differencebetween the hypothetically (i.e., OE group) and actually(i.e., BDM group) stated mean WTP. A t-test and aKolmogorov-Smirnov (KS) test show WTP statements inthe BDM group are significant lower compared to the OEgroup (9.11 Swiss Francs versus 14.31 Swiss Francs).Similar to the first study, the data reveals that a majority(55.72%) of the respondents in the OE-Cut group state aprice that differs form maximum WTP. In order to adjustfor these incorrect price statements, we exclude thoserespondents from the sample who did not correctly answer(i.e., they claim to have stated their true maximum WTPin the price-type self-selection task) the WTP questionex-post. On this reduced sample, we run a separate analy-sis. Both, a t-test and a KS-test show that mean WTP doesno longer significantly differs between the BDM bench-mark and the reduced OE-Cut group (9.11 Swiss Francsversus 9.71 Swiss Francs). This finding indicates thathypothetical bias might be mitigated by only consideringstatements of those respondents that are able to correctlystate their WTP.

To conclude, we find that the respondents do notalways answer in a way that is intended by the marketresearcher and that our proposed approach to improveWTP measurement with the OE question format is capableto significantly improve the external validity of the mea-surement. By means of adding a price-type self-selectiontask after the OE question and only considering therespondents who give the correct answer, we show that itis possible to improve the WTP results in such a way thatthe hypothetical bias is removed to an insignificant level.


For further information contact:David Blatter

Marketing DepartmentInstitute of Marketing and Management

University of BernEngehaldenstrasse 4

3012 BernSwitzerland

Phone: +41.31.631.80.24Fax: +41.31.631.80.32



QUANTIFYING QUALITATIVE DATA: A NOVEL APPROACH TOMEASURE BRAND ASSOCIATIONS

Kaleel Rahman, American University in Dubai, United Arab EmiratesCharles S. Areni, University of Sydney, Australia

SUMMARY

The branding literature suggests that not all brandassociations are equal and the strength, uniqueness andvalence of brand associations need to be considered whenmeasuring brand associations (Keller 1993). We providean approach to address all these dimensions simulta-neously. Using a pilot study, uniqueness of brand associa-tions is determined by coding associations into severalmutually exclusive meaning components. The serial orderof free association elicitation is used to assess associationstrength. This serial order, combined with a measure ofvalence, is used to create a quantification of open-endedbrand associations called a “Weighted Valence Index”(WVI). Then, the WVI is used to predict brand attitudes.

Aaker (1991) asserted that the underlying value of abrand name is often the set of associations – its meaningto people – anything connected in memory to the brandname. Associations are the “the heart of brand-building”(Aaker and Joachimsthaler 2000a, p. 263). Brand associa-tion in marketing is typically elicited through free associa-tion method (e.g., Chen 2001). Free association, alsoknown as word association, allows consumers from vari-ous backgrounds to respond freely in their own terms withwhatever verbal form they feel most appropriate. Freeassociation data in marketing is typically used in a quali-tative fashion where the semantic meaning of associationsis considered and interpreted together with frequenciesand or percentages of such semantic meaning. In thisresearch, in a series of stages, we turn free associationsinto measures equivalent to that of interval scale items,such as semantic differential and Likert scales, which lendthemselves to rigorous assessment of reliability and valid-ity. We used quantified free associations as independentvariables and typical scale items as dependent variables inour study.

First, using pilot study data, free associations inresponse to two sub-brands (e.g., American Express Plati-num) from nine different services sectors were elicited.The subjects for this pilot study were 58 male and femaleundergraduate students enrolled in a marketing course ata major university in Australia. The pilot study datacollected using the free association method was analyzedthrough a technique called associative group analysis,advocated by Szalay and his colleagues (see Szalay et al.1993; Szalay and Bryson 1974; Szalay and Deese 1978;

Szalay and Maday 1973) in the field of social psychologyand later adopted by researchers in marketing (Friedmannand Zimmer 1988; Marsden 2002; Philips 1996). Accord-ing to associative group analysis, one way of identifyingmeaning from free association is to place the responsesobtained in categories based on their semantic relation-ship to the stimulus.

Using the pilot study data, the research team devel-oped the following five domains of brand meaning com-ponents that were equally applicable to all of the brands:

1. The SUB-BRAND meaning component captures thenotion of genuine brand equity at the sub-brand levelas it included anything judged to be mentioned aboutthe sub-brand.

2. The PARENT company meaning component wasconceptualized as anything judged to be mentionedabout the parent company.

3. Generic associations with the specific service CAT-EGORY included reference to the service categoryand its attributes, characteristics and features in anyform.

4. The SEMANTIC meaning component includes asso-ciations referring to the general semantics of the sub-brand’s brand name, font, logo, color, visual style andslogan that do not indicate any specific knowledge ofthe underlying sub-brand.

5. Finally, there was evidence of reference to COM-PETITORS in response to the sub-brand stimulus.

In the main study, word associations in response to 14actual sub-brands were generated from 447 respondents.As opposed to the typical procedure where the frequen-cies of associations are simply counted, here the frequen-cies were adjusted based on the order of elicitation, whichserved as a proxy for strength. Based on Szalay and hiscolleagues (see Szalay and Bryson 1974; Szalay andDeese 1978; Szalay and Maday 1973; Szalay et al. 1993),the first mentioned association is weighted by six points,the second 4.8 points, the third 4.2 points, the fourth 3.4points, the fifth 3.2 points, the sixth 3 points, the seventh2.5 points, the eighth 2 points, the ninth 1.5 points, thetenth 1.1 points, and one point each for any further


elicitations. This research also employed a rating schemeto account for differences in valence. The respondentswere asked to rate each of their brand associations from –3 to +3, where -3 was very unfavorable and +3 veryfavorable. The resultant strength and valence indices wereused to develop a brand association index WVI, where the

strength score is multiplied by the valence score for eachassociation. A WVI association index was calculated foreach meaning component for every subject for every sub-brand. Content, convergent and discriminant validity arerigorously discussed and the reliability of this technique isestablished using Cronbach’s alpha analysis.

For further information contact:Kaleel Rahman

School of Business AdministrationAmerican University in Dubai

P.O. Box 28282 DubaiUnited Arab Emirates

Phone: +971.4.3183327Fax: +971.4.3998899



A CONCEPTUAL FRAMEWORK OF HOW PRODUCT DESIGN IMPACTSCONSUMER PERCEPTIONS AND MARKET-BASED PERFORMANCE

Janell D. Townsend, Oakland University, RochesterMitzi M. Montoya, North Carolina State University, Raleigh

SUMMARY

Product design is a priority for firms as they competein an increasingly global marketplace, where cost andeven quality are no longer differentiators. While signifi-cant resources are often invested to improve productperformance based on design, foundations for under-standing of the controllable factors that influence con-sumer perceptions of design remain largely conceptual,even though important frameworks have been developedfrom a number of perspectives. Bloch (1995) considersthe interface between firm actions, consumer perceptionsand expected behavioral responses, and builds on earlierwork from Bloch and Richins (1983) which develops acomprehensive means to understanding the role of prod-uct importance. In the design literature, Crilly et al. (2004)integrate existing models based on theories and conceptsfrom a variety of fields including design research, psy-chology, consumer behavior and sociology. Drawingfrom psychology, Rindova and Petkova (2007) suggestantecedents and value perceptions of new technologicallyadvanced products based on cognitive an affective re-sponses. Similarly, in a book widely renowned amongdesign professionals, Norman suggests emotions regard-ing products arise from the visceral, behavioral, andreflective levels of brain processing activities (Norman2004). Expectations are that design influences consumerperceptions, which in turn engenders improved marketperformance.

Components of theorized relationships between vari-ables in frameworks have also begun to be reported in theliterature. Hertenstein et al. (2005) present a conceptualmapping of the role of industrial design, establishing anassociation between design effectiveness and firm levelfinancial performance. Further, robust design capabilitiesare also shown to have a relationship with firm perfor-mance and speed to market by Swan et al. (2005). Withrespect to aesthetic components of product design, Veryzerfinds that universal principles of design illicit significantaesthetic responses, and uses this as a foundation for abroader conceptualization of the role of design aestheticsin consumer research (1995a). However, to our knowl-edge, no single study has yet established the relationship

between product design, consumer perceptions and mar-ket-based performance.

The lack of an empirically tested longitudinal frame-work has both practical and academic relevance sincedefining measurable components of design would givemanagers an important tool in the development process,and researchers a broader foundation for further study.Understanding consumer’s perceptual response to designprovides managers in the strategic and tactical domain ofproduct development with knowledge that can be employedto influence consumer behavior, and shed insight intoexpected market outcomes. Although the literature hasbegun to address components of this phenomenon, majorgaps in our knowledge related to how design factorsinfluence consumer perceptions still exist.

In order to form a general and coherent perspectiveon the subject, this study develops a conceptual frame-work based on the existing literature, and further contrib-utes to knowledge by empirically testing the framework,and deriving new insights. We develop a conceptualframework illustrating how factors such as form, func-tion, and aesthetics impact market performance whenmediated by perceptions of design, and propose a longi-tudinal model based on objective measures of designcoupled with consumer perceptions and market perfor-mance. The broad purpose of this study is to betterunderstand how form, function, and aesthetic features ofa product impact market performance, as mediated byconsumer perceptions. The three main contributions ofthis research are: (1) To develop a conceptual frameworkthat integrates concrete and abstract measures of profes-sional product design to explain the impact of theseconstructs on consumer perceptions of design and marketperformance; (2) to set up a means to empirically test theconceptual model on a longitudinal basis; and (3) to usethe results to describe, explain, and understand the roledesign in achieving a competitive advantage in the mar-ketplace, providing guidance to product development,marketing, and executive management in the new productdecision making process with respect to future productdesign.


For further information contact:Janell D. Townsend

Marketing and International BusinessOakland University

348 Elliott HallRochester, MI 48309Phone: 248.370.2544



PRODUCT PLACEMENT IN THE MOVIES: FROM DAWN TILL DUSK?

Grigori Erenburg, Chapman University, OrangeEkaterina V. Karniouchina, Chapman University, Orange

Can Uslay, Chapman University, Orange

SUMMARY

For better (e.g., the Nokia 7110 in The Matrix 1999)or worse (e.g., over four dozen brands in The Departed2006), product placement in the movies has become a partof today’s marketing arsenal lending its power to offer-ings ranging from pregnancy tests to stock exchanges toluxury cars (Grover 2009). Industry sources boast thatproduct placements in the movies can significantly boostsales of individual brands as well as entire product catego-ries (Harlow 2006). For example, wineries and grocerystores saw an explosion of Pinot Noir sales after it wasfeatured in the movie Sideways (2004). Nevertheless,evidence for the tangible benefits of product placement ismostly anecdotal, and studies that empirically demon-strate its economic worth are scant at best.

This article examines the effectiveness of productplacements in the movies over a time span of almost fortyyears (1968–2007). During this time period, this advertis-ing practice has been transformed from a relatively rarecovert marketing technique to a part of the mainstreamintegrated marketing communications portfolio. The studyuses event study and multi-level mixed coefficient mod-eling to link branded product placements to firm value,and examine the applicability of “wear-out” theory in thecontext of product placement effectiveness. We identifysignificant cumulative abnormal returns in the (-10, 16)event window (the stocks gain .75% on average). Thisfinding indicates a favorable, yet not immediate, marketresponse to product placements in movies. We also ob-serve an inverted U-shaped relationship between the yearof the movie release and the returns associated withproduct placements. We estimate that the effectiveness ofproduct placement peaked in the early nineties (i.e., 1993)and has been declining since then. The dusk of thisadvertising medium may already have arrived for genericplacements. The market response to placements from2005–2007 time-frame is slightly negative and signifi-cantly different from the response associated with the restof the sample. These recent declines could be potentiallyexplained by soaring costs of product placement. There-fore, a detailed examination of recent product placementeffectiveness is warranted to justify continued adoption ofthis marketing practice. This finding is consistent withhabituation-tedium theory which is traditionally used toexplain the inverted U-shaped pattern in response to novelads. This article suggests that similar habituation-tedium

mechanism could be moderating response to an entireadvertising form. The results of this study reinforce thenotion that marketers find it increasingly difficult to gettheir message across using traditional display methods.They underscore the need for the advertising industry toreinvent itself when new tactics lose their luster.

Our results also indicate that more overt productplacements tend to be associated with negative marketresponse. This is consistent with previous literature thatfinds that overt placements may lead to decreased con-sumer response. Counter-intuitively, selective and lessexpensive placements (e.g., brief fleeting appearance)may be more effective than marquee, repeat-exposureplacements that are very costly. Thus, the value of repeatexposure in placements and what represents a marqueeplacement should be qualified carefully.

The cumulative knowledge from this study and thatof Wiles and Danielova (2009) suggest that blockbusterfilms may be associated with more efficient marketresponses and higher initial CAR to product placements infilms, but also with a strong downward adjustment thattakes place upon the movie’s opening. This new findingsuggests an interesting area of future research: an analysisof the relative success of product placements acrossdifferent types of movies (e.g., blockbusters vs. nichefilms). Placements in blockbuster films may generatemore hype but not necessarily a sustainable increase in thefirm’s economic value. The lack of long-term value couldbe tied to several factors, including the high costs associ-ated with these placements. Therefore, the ultimate eco-nomic worth of placement in very successful movies maybe questionable more so than placement in moderatelysuccessful movies (suggesting an inverted-U relation-ship). Furthermore, faster market response for top gross-ing movies also implies the need for investors to pay moreattention to pre-release box office receipts expectations(e.g., Hollywood Stock Exchange), and pay attention tocritical reviews that may mention placement and/or signalquality of placements in big budget movies, which influ-ence amateur ratings, and ultimately movie performance(Moon, Bergey, and Iacobucci 2010).

In summary, we adopt a longitudinal perspective, andexamine the evolution of the effectiveness of productplacement in movies over time. First, we provide a literaturereview that focuses on the history and the suggested


efficacy of this advertising approach. Second, we introduceour conceptual framework, and present our hypotheses.Third, we discuss the data and our methodologies (eventstudy and multi-level mixed coefficient model). Fourth,

we present the results and discuss our findings. Weconclude with implications, limitations, and future researchdirections. References are available upon request.

For further information contact:Ekaterina V. Karniouchina

Chapman UniversityOne University Drive

Orange, CA 92866Phone: 714.289.2068



CO-MARKETING CAPABILITY AND ITS IMPACT ON MARKETINGALLIANCE PERFORMANCE

Marcus Schögel, University of St. Gallen, SwitzerlandDennis Herhausen, University of St. Gallen, Switzerland

SUMMARY

Nike is performing well in an area where othercorporations, from Coca-Cola to General Motors, haveunsuccessfully tried to enter: Building brand loyalty viaonline social networking. The key to bringing their cus-tomers into the Web was the introduction of the Nike+Sport Kit sensor, developed and marketed together withApple. The partnership between Nike and Apple is a verysuccessful example of what might be termed a co-market-ing alliance (Bucklin and Sengupta 1993).

Unlike buyer-seller or manufacturer-distributorpartnerships, co-marketing alliances are lateral relation-ships between firms at the same level in the value chain.Firms with pronounced competences in co-marketing,referring to superior skills how to create and sustainfruitful cooperative arrangements with partners, have asignificant edge over their competitors (Lambe, Spekman,and Hunt 2002). Despite their potential contribution, co-marketing alliances pose significant managementchallenges. As a consequence, 60 to 70 percent of alliancescan be considered as failures (Das, Sen, and Sengupta2003). Although Sivadas and Dwyer (2000) as well asLambe et al. (2002) conducted early examinations ofalliance competence, it still lacks an integrated under-standing of what capabilities allow firms to successfullyfind, develop, and manage co-marketing alliances. Hence,we identify and empirically examine a specific co-marketing capability (CMC) that enable firms to effectivelyengage in marketing alliances. Our findings provide newinsights regarding the process of forming and conductingsuccessful alliances in marketing.

We conceptualize a CMC as the essential capabilityof an organization to achieve superior co-marketing alli-ance performance and a sustained competitive advantage.Drawing on an extensive literature review, in-depth per-sonal interviews with marketing alliance managers, andconsulting experience concerning co-marketing alliancesat various firms in the consumer good and business tobusiness market, we built a preliminary model of CMCand confronted 15 knowledgeable managers and ninemarketing experts with it. Based on the obtained insightsand suggestions from practitioners and academics, weextended our model and identified five distinct dimen-sions of CMC: Partner identification, alliance configura-tion, interfirm coordination, effective communication,

and knowledge management. Furthermore, the literaturereview as well as our interviews highlight certain charac-teristics of alliances that may increase or reduce theinfluence of co-marketing capabilities. In particular, threefactors have been shown to be important and are assumedto moderate the relationship between CMC and co-market-ing alliance performance: Age of a co-marketing alliance,complexity of tasks within the alliance and power imbal-ance.

Primary data for testing our hypotheses were col-lected via a mail survey of firms operating in consumerand business markets. We selected firms with more than100 employees to avoid an interference of small firmsbeing dependent on alliances. Such firms may cooperatewithout actively developing and employing CMC due toa lack of own assets. The results of our study suggest fourconclusions: First, CMC has a high impact on co-market-ing alliance success. Second, CMC is especially impor-tant during the early honeymoon stage of co-marketingalliances (Levinthal and Fichman 1988). Third, firmsneed high CMC to successfully handle alliances withcomplex tasks, e.g., co-development of new products(Sivadas and Dwyer 2000). Fourth, power imbalancedoes not moderate the CMC and co-marketing perfor-mance relationship. Contrary to what we expected andprior research suggested, CMC is equally important foralliance partners having similar resources and power andthose within an imbalanced alliance, e.g., a co-marketingalliance between a multinational and a small or medium-sized company.

High failure rates indicate that managing alliances isextremely challenging. Drawing on our research, a lack ofCMC may be a primary reason for unsuccessful co-marketing alliances. Our results demonstrate that sophis-ticated CMC leads to an effective work environmentbetween the partner firms in which alliance objectives willbe achieved. Hence, CMC needs to be anchored within thefirm. Firms should therefore aim to establish CMC at allpositions touched by the co-marketing alliance. The mod-erating role of co-marketing alliance age confirms priorfindings that alliance experience plays a significant rolefor alliance success (Lambe et al. 2002). Firms having along-lasting co-marketing alliance learn about each otherand internalize roles and responsibilities. Hence, CMCwill become more tacit and embedded within the partner-ship. As ambitious alliances with complex tasks require


extensive CMC, we empirically prove Day’s (1995) sug-gestion that firm’s initial attempts to establish co-marketing alliances should be made with relativelynoncomplex and modest objectives. The lack of relation-ship between power imbalance and CMC was a surprisebecause prior research has found that co-marketing alli-ances dominated by a single partner require higher mana-

gerial skills to avoid detrimental effects from the partner-ship (Bucklin and Sengupta 1993; Sivadas and Dwyer2000). We conclude that the identified CMC serves as ageneral foundation for achieving alliance objectives andis mandatory for all kinds of co-marketing alliances.References are available upon request.

For further information contact:Dennis Herhausen

Institute of MarketingUniversity of St. GallenDufour Street No. 40a

St. Gallen, 9000Switzerland

Phone: +41(0)71.224.28.59Fax: +41(0)71.224.28.57



CONCEPTUALIZING TRUST IN THE RELATIONSHIP BETWEENSALES AND MARKETING AND THE CUSTOMER

Kenneth Le Meunier-FitzHugh, University of East Anglia, United KingdomJasmin Baumann, University of East Anglia, United KingdomLeslie Le Meunier-FitzHugh, City College, United Kingdom

SUMMARY

There has been considerable research into the effectof trust on the intra-functional relationship between vari-ous internal groups (e.g., Das and Teng 1998; Beech andHuxham 2003; Vangen and Huxham 2003) and into theimportance of trust in the buyer-seller relationship (e.g.,Cannon and Perreault 1999; Jap, Manolis, and Weitz1999), but to our knowledge no one has considered howthe intra-functional relationship between sales and mar-keting affects the trust in the inter-organizational relation-ship between the company and customers in a business-to-business (B2B) setting. This research will considerhow far conflict/trust in the relationship between thesetwo customer-facing elements of the organization, salesand marketing, affects the trust in the organization’srelationship with its customers. The proposition is that ifthere is conflict/lack of trust between sales and marketing,collaboration between these two functions will be dam-aged and this will adversely affect the inter-organizationaltrust between B2B organizations and their customers. Ifthe sales and marketing relationship is collaborative, trustwithin the organization-customer relationship will bestronger and business performance will increase.

Trust between organizations is a complex constructas it arises initially between individuals and is influencedby many factors (Zaheer, McEvily, and Perrone 1998).However, according to Sako (2002) three types of trusthave been identified in B2B relationships: contractual,competence-based, and goodwill trust, but it is proposedthat specifically the interpersonal elements of trust, i.e.,goodwill and competence-based trust, may affect collabo-ration between sales and marketing.

If there is dysfunctional conflict between sales andmarketing, the sales department may not support market-ing events and promotions. Conversely, marketing depart-ments may not help sales to meet targets or support salesefforts in certain areas. There are examples of mistimeddirect mail events and the failure to upgrade product/service offers in line with sales requests (damaging good-will trust among the customers). Essentially, the absenceof trust and presence of dysfunctional conflict in this

relationship is expected to have an adverse effect oncollaboration between sales and marketing. If sales andmarketing are not collaborating, their efficiency is likelyto be impaired. Under these circumstances customers maystart to question the organization’s ability to deliver accu-rately and on time (competence-based trust). Trust is afragile commodity (Lane 2002) and even trust built upover an extended period may be permanently damaged bysuch actions. Customers may stop relying on the organi-zation and may even doubt its reputation. Sales may fail ascustomers replace the organization’s offers with those ofits competitors when building new allegiances. Custom-ers may also choose to switch to competitors for promo-tional support and solutions to new problems. Businessperformance between the customer and of the organiza-tion as a whole would begin to suffer.

Establishing interpersonal trust, consisting of a cog-nition-based and an affect-based dimension, in the salesand marketing interface is one way of improving thecollaboration between sales and marketing. Collaboratingto provide the best solutions for customers will reinforcecognition-based trust between sales and marketing, andbuild ties (affect-based trust) between the two groups(McAllister 1995). It is suggested that trust and dysfunc-tional conflict cannot coexist within the relationship. Asinterpersonal trust is developed, dysfunctional conflictwill be reduced, thereby improving collaboration betweensales and marketing. With greater collaboration the solu-tions bundles offered and promoted to customers will alsoimprove, and this will positively affect business perfor-mance. This work makes a number of contributions toliterature. First, the paper adds to the scant literature on theimportance of the relationship between sales and market-ing and its impact on business performance. Second, theinterdependence of trust and dysfunctional conflict isconsidered in relation to collaboration between sales andmarketing. Third, the effect of collaboration betweensales and marketing on trust in customer relationships isexplored and, finally, the effect of trust between thecustomer and organization is considered in terms ofbusiness performance. References are available from thelead author upon request.


For further information contact:Kenneth Le Meunier-FitzHugh

Norwich Business SchoolUniversity of East Anglia

Norwich, NR4 7TJUnited Kingdom

Phone: +44(0)1603.456161Fax: +44(0)1603.458553



ASSESSING THE IMPACT OF SALESPEOPLE’S FIELD-BASEDCOMPETITIVE INTELLIGENCE ON PRODUCT

COMPETITIVENESS AND BRAND PREFERENCE

Joel Le Bon, ESSEC Business School, SingaporeAdam Rapp, Clemson University

SUMMARY

Over the years, a number of theoretical and empiricalresearch studies have postulated conceptual models ofcompetitive intelligence (hereafter, CI) and offered vari-ous perspectives on CI practices. Despite diversity interms of approach and method, one common theme isvisible among previous studies. The CI literature agreeson the importance of the role that firm personnel play inthe CI process. Among a firm’s employees, the sales forceis recognized as the single best internal source of informa-tion concerning what is occurring in the market withcustomers and competitors (Porter 1980). This recogni-tion relies on the fact that salespeople are the boundaryspanners between the firm and the outside world and havea vital role in bringing information to the firm (Le Bon andMerunka 2006; Rapp et al. 2006).

With this in mind, this research examines CI at theindividual-level of analysis. We review the relevant lit-eratures and develop a model of CI that is grounded ininformation theory as well as social network and environ-mental scanning theories. Notably, we treat CI as anoutcome of the intelligence process and therefore a prod-uct or instrument that can be used by salespeople duringsales activities. Therefore, the purpose of this research isto: (a) determine the extent to which competitive intelli-gence gathered through the sales force impacts productcompetitiveness and brand preference, and (b) determinewhich aspect of competitive intelligence (i.e., quantityand quality of collected CI) impacts product competitive-ness. We also explore other factors that may strengthen orweaken these relationships. Importantly, we include mea-sures as assessed by customers of the sales representativessampled in this research to ensure an accurate representa-tion of the customer side of the buyer-seller exchangeprocess.

We present a framework that integrates the impact ofsalespeople’s field intelligence collection on customers’preferences. This represents a theoretical sequence thathas yet to be empirically verified. The proposed modelpresents product competitiveness as a consequence of thequantity and the quality of CI that customers share withsalespeople, and as an antecedent of customer satisfactionand brand preference. We also posit that the relationshipsbetween shared CI and product competitiveness may be

moderated by salespeople’s social network quality as wellas their perceived customer orientation.

The first set of hypotheses investigates the factorsrelated to the quantity (i.e., volume) and the quality (i.e.,value) of information shared by the customer with thesalesperson. Differentiating between CI quantity and qual-ity is important, because gathering proper marketplaceinformation certainly makes a difference when address-ing customers’ needs and expectations (Cleland and King1975; Montgomery and Weinberg 1979). Our next set ofhypotheses relates to the effects of CI quantity and qualityon product competitiveness. Business success and salesperformance depends on product differentiation and com-petitiveness (Holbrook 1994; Porter 1980). Product com-petitiveness, defined as a product’s superior utility andproposed value relative to competitors (Zeithaml 1988),incorporates firms’ ability to fulfill customer needs andpurchasing constraints (Phillips, Chang, and Buzzell 1983).Previous research indicates that high-performing sales-people effectively plan and maintain a high level ofknowledge concerning their products, customers, andcompetitors (Sujan et al. 1994). During a sales encounter,an exhaustive and independent assessment of customerperceptions about other sources of supply can enable thesalesperson to position the product in a better competitiveposition. The last set of hypotheses then examines theimpact of product competitiveness on customer satisfac-tion and brand preference, a causal chain of great impor-tance at the firm level.

The research setting for this study was a logistics firmwhose primary activities included assisting customers intransportation management and product shipment. Thefirm was engaged in business-to-business selling andworked with firms across a wide range of industries.Survey data was collected from both salespeople and theircustomers. All 67 of the firm’s salespeople were surveyedand 1440 customers were contacted as well. Due to themulti-level nature of the data, the most appropriate ana-lytical technique to empirically examine this model ishierarchical linear modeling.

Seven of the nine linear hypotheses were supportedand one was significant but opposite the anticipateddirection. Interestingly, two of the hypothesized interac-tions were supported. Namely, a salesperson’s network


quality interacted with quantity of information shared andsalesperson customer orientation interacted with cus-tomer quantity of CI shared.

At a broad level, the purpose of considering CI at anindividual level has been to bring attention to this widelyneglected determinant of product competitiveness. Despitethe abundance of frameworks linking different salesper-son related variables and performance, superior productperformance only arises when salespeople identify newmarket opportunities and react to them faster and more

proficiently than their competitors. The proposed theo-retical framework is intended to encourage academiciansto extend this debate and to focus their interest on theindividual level CI use. This study offers that researchersshould not only concentrate on the link between thesalesperson and customer, but also address the question ofhow competitors’ information can weaken or strengthenthis link. Moreover, salesperson’s utilization of CI mayprovide the opportunity to bring in the much needed angleof “competitor” to the salesperson-customer-competitortriangle. References are available upon request.

For further information contact:Joël Le Bon

ESSEC Business School100 Victoria Street

National Library #13–02188064, Singapore

Phone: +65.68.35.77.69E-Mail: [email protected]


A SALESPERSON’S CROSS-FUNCTIONAL ORIENTATION:ANTECEDENTS AND EFFECTS ON

CROSS-SELLING SUCCESS

Oliver Malms, University of St.Gallen, SwitzerlandChristian Schmitz, University of St.Gallen, Switzerland

SUMMARY

Introduction

Instead of following high capital intensive expansionstrategies research has revealed that leveraging currentcustomer relations provides many growth opportunitiesfor manufacturers containing very low initial investmentsand low levels of risks (Hartline, Maxham, and McKee2000). One strategy to leverage current customer poten-tial is best known as cross-selling. In a few words, cross-selling can best be described as a sales strategy thatleverages sales potential among existing customers bycombining offerings from different product areas or busi-ness units of an organization.

Research has shown that realizing a cross-sellingstrategy is considered to be a highly challenging taskrequiring strong intraorganizational ties between singlebusiness units (Akcura and Srinivasan 2005; Gulati 2007;Luo, Slotegraaf, and Pan 2006). Research by Maltz andKohli (1996) and Gronroos (1994) implies that incum-bents need to establish very well functioning internalprocesses that enable the flow of (sales) information anddata between participating business units in order forcross-selling strategy to work. Hence, the state of mind, orthe cross-functional orientation (CFO) of the sales forceseems to be a necessary ingredient for cross-selling suc-cess. Thus, the question this research aims to answers is:

What effect does a cross-functional orientation haveon cross-selling success and which prerequisitesneed to be in place to establish a strong CFO?

Theory and Hypotheses

Since, to the best of our knowledge no study yet hasresearched the antecedents of a CFO and the effects oncross-selling success we conducted a primary qualitativestudy to explore the phenomenon of CFO in its real lifecontext and to gain general insights on factors influencingCFO. These are discussed following.

Organizational Commitment. Organizational com-mitment is considered to be the strength of an individual’sidentification with and involvement in a particular orga-nization (Porter et al. 1974). Individuals highly commit-

ted to an organization’s goals and willing to devote a greatdeal of energy toward those goals are assumed toproactively exert CFO enhancing behavior compared toindividuals who are less committed to the organization.

Corporate Brand Identity. As Castro, Armario, andDel Rio (2005) state and verify, there is a significantpositive correlation between the level of an employee’sorganizational belongingness and his or her cross-functional activities. Following that, drawing on theresearch of Ashforth and Mael (1989) it can be stated thatfrom the moment on a person’s self-concept is tied to asocial category (e.g., a group, a community, or anorganization) the person is likely to choose activities thatare congruent with his or her salient role-identity (and thevalues tied to it).

Cross-Selling Motivation. We argue that salespeoplehaving a strong motivation for cross-selling activitiesengage in many different adjacent sales opportunities(e.g., cross-divisional opportunities, multiple productopportunities, long-term relationship opportunities fortheir own division, service opportunities and solutionopportunities) and thereby get a broader and profoundcross-functional involvement and belongingness.

Cross-Selling Readiness. Cross-Selling Readiness isdefined as the combination of knowledge about productsof different business units and the ability of flexiblymaking use of this knowledge in a sales presentation witha given customer. To be cross-selling ready means to beable to offer a wide range of different products andservices not only from one but from a multitude ofbusiness units.

Cross-Selling Success. Cross-selling success mea-sures the performance orientation of successful cross-selling activities in the organization. This is done byconsidering the amount of additional products sold to agiven customer in relation to its future cross-sellingpotential.

Measures and Results

From the qualitative study a set of hypotheses werederived to be tested via structural equation modeling(Figure 1).


All together the fit of the measurement model isa c c e p t a b l e : { 0 0 0 2 C E 0 3 - 0 0 0 0 - 0 0 0 0 - C 0 0 0 -000000000046} = 331,3, degrees of freedom [df] = 280, p =0.02, root mean square error of approximation [RMSEA] =0.03, comparative fit index [CFI] = 0.97, tucker-lewisindex [TLI] = 0.97.

Taken together, results indicate that the impact onCFO is significantly influenced by four different indi-vidual traits (brand identity, cross-selling motivation,cross-selling readiness, and organizational commitment).Only two of those individual traits show a direct effect on

FIGURE 1Stuctural Equation Model

cross-selling success (commitment and cross-selling moti-vation). Surprisingly, there was no significant effect testedfor the relationship of brand identity, cross-selling readi-ness, and cross-selling success. Moreover, we found anegative relationship for the effect of organizational com-mitment on cross-selling success and for the effect ofcross-selling motivation on CFO. A very strong correla-tion can be found between the level of CFO and cross-selling success. Finally, the individual traits in the pro-posed order did all show a positive significant effect oneach other. References are available upon request.

For further information contact:Oliver Malms

Institute of MarketingUniversity of St. Gallen

Dufour Street 40a9000 St. Gallen

SwitzerlandPhone: +41(0)71.224.7155


H5(+)

BrandIdentity

OrganizationalCommitment

Cross-SellingMotivation

Cross-SellingReadiness

CFO CS-Success

H2a(+)

H5(+)

H1a(+)H4a(+)

H2b(+) H1c(+) H3b(+)

H2c(+)

H3c(+)H4b(+)

H3a(+)

H1b(+)


COMBINING VISUAL AND TEXTUAL RHETORICAL FIGURESIN ADVERTISING

Robert O. Fabrize, University of North Texas, Denton

SUMMARY

Advertisers today rely so frequently on indirect claimssuch as rhetorical figures in headlines, text, and pictorialcontent that it is now normal to see the abnormal: readerscan see lemonade turn into butterflies (Crystal Light).These indirect claims tend to be more persuasive becausethe intent to persuade is subtle (Walster and Festinger1962). Rhetorical figures divert subjects from conventionand force them to make sense of the ad. They let viewersself-generate multiple positive inferences making adsmore believable.

Metaphors are the most powerful of rhetorical figuresbecause they include both nonrational and symbolic com-ponents. Their complexity and deviation from the literalmake them effective (McQuarrie and Mick 1996). Forexample, in the metaphor “business is war,” the readertransfers onto business the qualities of war: chaos, danger,death, triumph, and loss. Metaphors effectively developtop of mind recall in low involvement readers (Toncar andMunch 2001).

The effective use of metaphor is usually assessed bymeasuring the resulting changes in attitude toward the ad(Aad) and ad recall. While ads with visual indirect claimsare more effective than those with textual indirect claims(McQuarrie and Phillips 2005), all plausible combina-tions of visual and textual metaphors have not beenexamined. The combination that yields the greater Aadand ad recall is unknown. Based on this line of reasoningthe following propositions are offered:

P1a

Ads with both visual metaphors and textual meta-phors will have a greater effect than ads with onlyvisual metaphors. (VM + TM > VM).

P1b

Ads with only visual metaphors and no text will havea greater effect than ads with visual metaphors andliteral text. (VM > VM + LT).

P1c

Ads with visual metaphors and literal text will have agreater effect than ads with literal visual and textualmetaphors. (VM + LT > LV + MT).

P1d

Ads with literal visual and textual metaphors willhave a greater effect than ads with literal visual andliteral text (VM + TM > LV + LT).

P1e

Ads with literal visual and literal text will have agreater effect than ads with only literal images (LI +LT > LI).

The existence of constructs that may moderate theimpact of metaphor on Aad and ad recall has also receivedlittle attention. Theoretical considerations suggest thatsubjects’ inclination to decode may be affected by toler-ance of ambiguity (TA) and need for cognition (NFC).

TA affects subjects’ emotions and perceptions ofambiguous situations and information (Frenkel-Brunswik1949). Subjects with low levels of TA tend to reduceambiguous thoughts into certainties and rigidly dichoto-mized situations. Though McQuarrie and Mick have usedTA as a measure of ad liking, TA has not been used as amoderating variable (1992). Thus, the following proposi-tion is offered:

P2a

Subjects with high levels of TA will exhibit higherlevels of Aad and ad recall.

P2b

Subjects with low levels of TA will exhibit lowerlevels of Aad and ad recall.

NFC measures the need to make sense of experienceand structure it meaningfully in an integrated manner(Cohen et al. 1955). Ambiguous information creates atension in people with high NFC that increases their effortto understand. Subjects who exhibit high NFC also havea higher correlation between the message evaluation andpost-message attitudes than subjects with lower NFC(Caccioppo and Petty 1982, 1983, 1984). Though NFChas been used to screen subjects, it has not been examinedas a potential moderator. Based on this line of reasoningthe following propositions are offered:

P3a

Subject with high levels of NFC will exhibit higherlevels of Aad and ad recall.

P3b

Subject with low levels of NFC will exhibit lowerlevels of Aad and ad recall.

The proposed experimental design is adapted fromMcQuarrie and Mick (1992, 1999) for use on the Internet.Alternative metaphorical combinations will be manipu-lated via a series of test ads based on actual magazine adscontaining the desired visual and/or textual metaphors.


The dependent variables in this study are Aad and adrecall. Aad will be measured with Coulter’s semanticdifferential scale (1998). The ten point scale achieved aCronbach’s alpha of .90. Ad recall, defined as the approxi-mate reproduction of the graphic, the headline, and theproduct, will be patterned after McQuarrie and Mick(1992). Working independently, two judges who do notknow the purpose of the experiment will code recall. Thedifferences will be resolved through conversation.

Data will be collected from subjects using web sur-vey software. Subjects will be told they are participatingin a study on magazine reading habits. They will answerquestions about print magazine readership, editorial in-fluence, desired topics, etc. They will subsequently beinstructed to examine the “virtual” magazine for as long asthey wish before answering questions on the editorialstyle and value, as well as items assessing Aad. Lastly,they will answer ad recall questions. References areavailable upon request.

For further information contact:Robert O. Fabrize

University of North Texas1155 Union Circle #311396

Denton, TX 76203–5017Phone: 940.565.3120



DESIGNING ADVERTISEMENTS FOR CONSUMERS OLDERTHAN 50 YEARS

Stefan Hoffmann, Technische Universität Dresden, GermanySusanne C. Liebermann, Technische Universität Dresden, Germany

Uta Schwarz, Technische Universität Dresden, Germany

SUMMARY

The average life expectancy of inhabitants in indus-trialized nations has been rising continuously. In conse-quence, marketers are more and more interested in thesegment of individuals older than 50 years as potentialcustomers for their products and services. Marketingresearch, however, so far does not provide sufficientfindings on how to address these consumers effectively.This article aims at filling this void.

The basic assumption of the paper is that advertise-ment campaigns should provide platforms for self-verification. Therefore, marketers have to take into accountthe heterogeneity of life styles and values in the olderstratum of the population. To date, several typologiesshow that the concepts activity and modesty are of specialconcern for German consumers older than 50 years. Thus,the present article focuses on these concepts. Moreover,values and lifestyles of consumers older than 50 yearsmight vary due to cohort effects and value changes.Therefore, the paper surveys consumers who are between50 and 59 years (age group 50+ in the following) andconsumers who are 60 years or older (age group 60+ in thefollowing). Modesty as a traditional virtue in Germany isbecoming less relevant for younger consumers, while it isstill very important for older cohorts. On the contrary, theconcept activity is becoming more and more relevant forthe generation 50+. Hence, the paper aims at answeringthe following questions: Which advertisement effects willbe achieved by advertisement models personifying con-cepts of activity or modesty within the segments of con-sumers in the age groups 50+ and 60+? To what extentdoes the interplay of the self-view of the observer and hisperception of the advertisement model explain the effec-tiveness of an advertisement?

A 2x2 factorial experiment (active vs. modest adver-tisement model; age group 50+ vs. 60+ as respondents) is

conducted. Two fictive ads are developed as experimentaltreatment showing a modest or an active ad model, respec-tively. Respondents exposed to the active model state thatthey are more similar to the advertisement model than theones exposed to the modest model. The active ad model isperceived as more attractive than the modest one. Theattitude toward the ad is significantly higher after beingexposed to the active ad model than after having seen themodest one. As suggested, the age groups interact signifi-cantly with the concept of the advertisement model. Whilefor the age group 60+ the active model does not have asignificantly better effect than the modest one on theperceived similarity, credibility and attitude toward thead, the age group 50+ reacts significantly better to theactive model when it comes to perceived similarity, attrac-tiveness, credibility and attitude to the ad. Thus, the valuechange seems to have influenced the age group 50+stronger than the age group 60+.

Furthermore, the paper confirms a model stating thatthe perceived similarity between the respondent and theadvertisement model plays a key role to explain differ-ences between the effect of the active and the modestadvertisement model. Perceived similarity depends on thedifference between the self-concept and the perception ofthe advertisement model on the dimensions modesty andactivity. Perceived similarity, in turn, exerts a positiveeffect on perceived attractiveness and perceived credibil-ity, which foster the attitude toward the ad.

The analysis was conducted for the dimensions activ-ity and modesty. Further research should examine themodel for other dimensions that might be important forgeneration 50+. Moreover, cross-cultural studies shouldanalyze if the importance of the analyzed concepts dependson the cultural profile of a country.

For further information contact:Stefan Hoffmann

Lehrstuhl für MarketingTechnische Universität Dresden

Fakultät Wirtschaftswissenschaften01062 Dresden, Germany

Phone: +49.351.463.32334 ♦ Fax: +49.351.463.37176E-Mail: [email protected]


EXAMINING THE EFFECTIVENESS OF LANGUAGE CHOICE FORDIFFERENT AD APPEALS IN ADVERTISING TO

BILINGUAL CONSUMERS

Fernando R. Jiménez, The University of Texas at El PasoXiang Fang, Oklahoma State University, Stillwater

Darrell Bartholomew, Oklahoma State University, Stillwater

SUMMARY

Prior research favors the use of bilingual people’snative language (L1 henceforth) over the second language(L2) to attain better ad effectiveness, which we call “thesuperiority effect of native language” (e.g., Puntoni, DeLanghe, and Van Osselaer 2009; Noriega and Blair 2008;Luna and Peracchio 2001). Recent theory emphasizes thesuperiority effect especially for ads with emotional con-tent (e.g., Puntoni et al. 2009). Puntoni and colleagues(2009, p. 1022) recommend that “it is generally preferableto communicate with consumers using their own nativelanguage, as doing so should result in more emotionalmessages.” However, would the superiority of L1 holdunder different emotional appeals? This is an open gap inthe literature that we attempt to fill in this article. This gapis relevant because distinct emotions have been found toinfluence consumer responses to advertising (Bagozzi,Gopinath, and Nyer 1999; Holbrook and Batra 1987).

Emotions are individually distinctive and are usuallygrouped in valence terms as positive or negative (Roseman1991). We investigate the effect of language choice on adevaluations for ads eliciting hope (positive valence emo-tion), a functional ad (no emotional content), and adseliciting disgust (negative valence emotion). Hope andfunctional appeals (no emotional text, only product fea-tures) are common intentional appeals (McInnis and deMello 2005). Disgust is a common but unintended emo-tion in different product categories, e.g., acne, athlete’sfoot, etc. that has been related to negative ad and productevaluations (Shimp and Stuart 2004; Argo, Dahl, andMorales 2006; Morales and Fitzsimons 2007).

Previous theories of language emotionality (Puntoniet al. 2009; Pavlenko 2008) suggest that communicationsin the native language should lead to better messageevaluations because they are perceived as more emo-tional. In our study, we hypothesized that ads elicitinghope and disgust would be evaluated as higher in emo-tional intensity in L1 rather than in L2. In turn, emotionselicited by the ads can also influence ad evaluations viaaffect transfer (Pham 2004, p. 361). Thus, we expectedthat increased emotionality of the hope appeal should leadto better attitudes toward the ad in L1 vs. L2. Also, we

expected more positive evaluations for the disgust ad inL2 rather than L1 since L1 would make the negativeemotion more salient. Finally, we expected no differencein ad evaluations of a functional ad when presented in L1or L2.

Finally, consistent with prior research on emotions(Holbrook and Batra 1987; Shimp and Stuart 2004), weexpected that the emotion felt would mediate the effects oflanguage on ad evaluations for both the hope and thedisgust ad.

In order to test our hypotheses, we conducted a 2(Language: native vs. second) × 3 (Type of Ad Appeal:Functional, Hope, Disgust) mixed design lab experimentwith language as a between-subjects factor and type of adas a within-subjects factor. Forty-nine pre-screenedEnglish-Spanish bilinguals were included in the analysis.Ads were pretested and translated to Spanish using backtranslation procedures (Marin and Marin 1991). Partici-pants were randomly assigned to ads in L1 or L2, and eachviewed the three ads on a computer screen with a distract-ing task between ads in order to avoid carryover effects(Gross and Levenson 1995).

Our findings show that for a functional ad, languagechoice (L1 vs. L2) made no difference on attitudes towardthe ad. Consistent with previous studies, for an ad thatelicits hope (positive emotion), the results show thatparticipants reported more positive evaluations towardthe ad when it is presented in L1 than in L2. In contrast, foran ad that elicits disgust (negative emotion), the resultsshow that L2 led to more positive ad evaluations than L1.Furthermore, we found that only the disgust emotionelicited by the ad partially mediated the effect of languagechoice on ad evaluation.

Next, we investigated the effect of disgust on adevaluation by empirically testing the “distancing effect”of L2 (Bond and Tat-Ming 1986) via viewing time of thead. The analyses reveal that respondents spent more timeviewing the disgust ad in L2 than in L1. The results holdacross gender, and there was no difference on comprehen-sion level between ads in L1 and L2.


In sum, our research supports the superiority effect ofnative language for ads eliciting hope, with higher atti-tudes toward the ad in L1 vs. L2. However, language madeno difference on attitudes toward functional ads. For adseliciting disgust, attitudes toward the ad were higher in L2rather than in L1. Finally, we provide empirical support

for the notion that increased liking of the negative-emotionad in L2 is due to its distancing effect. This effect wasshown as individuals spent more time viewing a disgust adin L2 compared to L1. References are available uponrequest.

For further information contact:Fernando R. Jiménez

212 College of BusinessThe University of Texas at El Paso

El Paso, TX, 79968–0539Phone: 915.747.7724



PREDICTING THE EFFECTIVENESS OF CELEBRITYENDORSEMENTS USING BALANCE THEORY

Subhadip Roy, ICFAI University, Hyderabad, IndiaBashar S. Gammoh, University of Toledo

Anthony A. Koh, University of Toledo

SUMMARY

The practice of celebrity endorsements for advertis-ing and promoting products dates back to more than ahundred years and continues today. Research findingssuggest that the percentage of celebrity advertisementsout of the total number of ads aired is as high as 25percent – 30 percent in Western countries and around 60percent in India. In such a scenario both academia andindustry look for the issues in selection, use and effective-ness of celebrities as product endorsers.

Literature Review

Research in celebrity endorsements in the last thirtyyears have tried to address various issues in celebrityendorsements such as the effects of credibility of endorseron the consumers (Lafferty and Goldsmith 1999, 2004;Ohanian 1990; Silvera and Austad 2004) issues in celeb-rity product “match-up” (Friedman and Friedman 1979;Kamins 1990; Misra and Beatty 1990; Mittelstaedt et al.2000), meaning transfer in celebrity endorsements(McCracken 1989) and a host of other issues, such aseffect of negative celebrity information, cross-countrycomparison of celebrity advertising, gender and celebrityendorsement perceptions, and so on. However, the celeb-rity endorsement literature has quite a few contradictoryfindings and still the researchers are working on to find theideal model that can explain the effects of celebrityendorsements. Mowen (1980) proposed the use of Bal-ance Theory to explain the phenomenon of celebrityendorsements. The model proposed by Mowen (1980) hasnot been comprehensively tested and thus allowed anopportunity to test the model in the present scenario. Thepresent research study developed and validated a theoreti-cal model on similar thoughts given by Mowen (1980).

Objectives

The study tested the effects of a celebrity and aproduct brand endorsed by the celebrity on consumerattitudes. The hypothesized model also incorporated theeffects of the source credibility dimensions as covariates.In this regard, the Source Credibility Scale developed byOhanian was included in the study. Three main hypothesisand four supporting hypotheses were developed to test

various aspects of the conceptual model. For example,Hypothesis 1 was stated as “Given the relationship betweenthe consumer and the endorser (C – E) and the endorserand the product (E – P), the relationship between theconsumer and the product (C – P) (determined by thepurchase intent) will be such that the C–E–P triad willhave a balanced state.”

Methodology

In the beginning, a series of pretests were conductedto identify (1) one liked and one disliked celebrity, (2) amatching and a mismatching product with each celebrity,and (3) a matching and a mismatching message with eachcelebrity product combinations. The final study was con-ducted using a 2 (Liked/Disliked Celebrity) X 2 (Match-ing/Mismatching Product) X 2 (Matching/MismatchingMessage) full factorial design. The dependent variableswere Attitude toward the Advertisement (AAD), Attitudetoward the Brand (AB) and Purchase Intention (PI). Themethodological tool used was Multivariate Analysis ofVariance (MANOVA) and MANOVA with Covariates.The main experiment was conducted on 448 postgraduatestudents in MBA programs spread across five BusinessSchools in India.

Results and Implications

The results indicated that a celebrity has a significantimpact on the consumer attitude which may be favorableor unfavorable depending on whether the celebrity is likedor disliked. The study also found that the match of thecelebrity with the product and advertisement messagewould have a favorable impact on the consumer attitudes.The study also provided validation of the Source Credibil-ity Scale developed by Ohanian (1990) in the Indiancontext. The source credibility dimensions, namelyAttractiveness, Trustworthiness and Expertise were foundto have differential effects on the dependent variables,i.e., consumer attitudes in the model. Thus, the studyintegrated the two main streams of celebrity endorsementresearch, namely the Source Credibility concept and theCelebrity-Product Match up concept. The study answeredsome of the gaps found in celebrity endorsement literatureand opened up some new areas for future research. Refer-ences are available on request.


For further information contact:Subhadip Roy

ICFAI University, HyderabadSurvey No. 156/57

Dontanapally VillageShankerpally Mandal, R.R. District

Pin: 501203, Andhra PradeshIndia

Phone: 91.9989864694E-Mail: [email protected]

TABLE 1Key Results

Model Specification MIWC FFIC

Source Dependent Variable F Sig. F Sig.

Celebrity AAD 39.798 .000 6.751 .010AB 21.384 .000 4.429 .036PI 31.526 .000 2.702 .101

Product AAD 92.699 .000 70.248 .000AB 42.204 .000 31.819 .000PI 19.091 .000 14.023 .000

Message AAD 50.572 .000 48.512 .000AB 19.030 .000 17.603 .000PI 10.578 .001 9.070 .003

Attractiveness AAD 17.706 .000AB 8.324 .004PI 14.241 .000

Trustworthiness AAD .356 .551AB .280 .597PI 1.109 .293

Expertise AAD 6.463 .011AB 1.859 .173PI 1.619 .204


DESCRIPTIONS OF A RADICAL INNOVATION CAN IMPACTCONSUMERS’ RISK AND VALUE PERCEPTIONS

Arjun Chaudhuri, Fairfield University

SUMMARY

A radical innovation presents potential consumerswith both uncertainty and opportunity. On the one hand,there is risk for the consumer if the innovation has thepotential for loss. At the same time, there is the potentialfor gain if the innovation can benefit the consumer. Thus,it behooves consumer researchers to try to understandhow perceptions of risk and value can simultaneouslyinfluence consumers’ willingness to try (WTT) a radicalinnovation. I suggest a model in which hedonic descrip-tions can, initially, increase perceived value and decreaseperceived risk in a radically new product. However,subsequently, more utilitarian descriptions may be morepowerful in generating greater perceived value leading togreater willingness to try the innovation. I attribute theinitial impact to the effects of arousal and pleasure that aregenerated by hedonic descriptions of the innovation.Overall, the model depicts that arousal is indirectly relatedto WTT via the mediation of perceived value and per-ceived risk and that the type of description will moderatethe pathways from arousal to perceived value and per-ceived risk and, in turn, to WTT.

I chose the hydrogen fuel cell (HFC) car as theinnovation to use in the study. I used various sources(websites, advertisements, news reports) to develop sepa-rate hedonic and utilitarian versions of the descriptions. Ichanged the headline and the blurb in the description toreflect the hedonic and utilitarian definitions of eachconstruct. Similarly, I changed the beginning and the endof each description.

Undergraduates from a small university in northeast-ern U.S. participated in the study for extra credit. Half thesubjects saw the utilitarian description and half saw thehedonic description. In both conditions, subjects weregiven a self-administered questionnaire. The question-naire contained instructions, the innovation descriptionmanipulation, and the measures for the latent constructs.After reading through the instructions on the first page,

participants were asked to indicate their current arousallevels. Next, depending on the experimental condition,participants read an excerpt from a news article thatoffered either a hedonic or utilitarian description of theinnovation. Immediately after participants read this descrip-tion, they were asked to once again indicate their level ofarousal, followed by a series of measures for the otherlatent constructs.

I found significant differences between hedonic andutilitarian types of description in all four paths in themodel. The findings show that (a) the path from arousal toperceived value is positive and significant (p < .01) for thehedonic group but not significant for the utilitarian group.(b) The path from arousal to perceived risk is negative andsignificant (p < .10) for the hedonic group but not signifi-cant for the utilitarian group. (c) The path from perceivedrisk to WTT is positive and significant (p < .10) for thehedonic group but not significant for the utilitarian group.(d) The path from perceived value to WTT differs signifi-cantly across the utilitarian and hedonic group (p < .05).

Future research needs to consider whether the rela-tionship between perceived risk and WTT is sufficientlydifferent for hedonic and utilitarian descriptions to be ofmanagerial significance. Future research could also exam-ine the interaction of verbal and visual elements in thedescription of an innovation. What happens, for instance,if the risk producing message elements in the verbaldescription (say, the hydrogen fuel tanks in the HFC car)are accompanied by visual stimuli that heighten the dan-ger (pictorial depiction of the tanks)? Will perceivedvalue still predominate over perceived risk as in thepresent study? If not, the findings would be important toboth journalists and public relations managers in deter-mining the full content of press releases of an innovation.As stated at the outset, radical innovations usually presentsome element of risk for consumers. To what extent canthis risk be depicted without harming the potential successof the innovation?

For further information contact:Arjun Chaudhuri

Charles F. Dolan School of BusinessFairfield University

Fairfield, CT 06824–5171Phone: 203.254.4000 ♦ Fax: 203.254.4105



HOW DO CONSUMERS FORM PERCEPTIONS OF ECONOMIC VALUE:AN EXPLORATORY STUDY

Gerald Smith, Boston College, Chestnut HillSajeev Varki, University of South Florida, Tampa

SUMMARY

The concept of value is central to marketing as mostmarketing activity involves the exchange of value (e.g.,Hunt 1976; Kotler and Keller 2005). Yet, the notion ofvalue itself has been the subject of considerable debateboth in terms of what it represents and how it is formed.While value (economic) is often conceptualized as a tradeoff between benefits and sacrifices (e.g., Monroe 2002;Zeithaml 1988), with quality representing the benefit sideand price representing the sacrifice, the exact process (ortrade-off) by which value is formed is the subject of somecontroversy and remains unclear.

Different researchers have proposed different meansby which consumers arrive at value perceptions. Authorssuch as Monroe (2002) and Gale (1994) argue that valueis the ratio of quality to price (ratio rule), whereas authorssuch as Dodds et al. (1991) and Rust and Oliver (1994)argue that value is the utility of quality minus the disutilityof price (linear difference rule). This issue of how con-sumers arrive at perceptions of economic value is ofimportance to managers when it comes to “value-pricing”;a concern addressed by practitioners and academiciansalike (Nagle and Hogan 2006; Kotler and Keller 2005;Marn, Roegner, and Zwada 2004; Monroe 2002).

Consider the following example: Alpha plans tointroduce a new “internal feed paint brush” that automati-cally supplies paint to the brush while painting, saving the

trouble of constantly stopping to dip the brush. This newinnovation in painting, patterned after automatic paintroller systems, allows painters to paint the exterior of ahouse in half the time it takes a traditional 3-inch paintbrush. Assuming that the average house takes 40 hours topaint, the new brush results in a saving of about 20 hourson average. Painters, such as college students and part-time helpers, usually receive about $15 per hour of paint-ing. The cost for a traditional paint brush is $15. Whatwould be a reasonable price for Alpha to charge for thenew paint brush?

In our paper, we examine several different ways inwhich consumers could combine quality and price infor-mation to arrive at different valuations of the new brush.We do this examination in the form of a series of simula-tions with different levels of quality and price one ordi-narily encounters in real life. We propose that consumersuse one of four classes of value processing strategies toarrive at a value judgment in a trade up situation. Thesestrategies are: (a) Proportional Value Processing; (b)Proportional Difference Value Processing; (c) LinearDifference Value Processing; and (d) Holistic ValueProcessing. We present these strategies via simulation toillustrate the different processes, calculations, and likelydecision choices. We are especially interested in identify-ing when these different strategies for arriving at valua-tion lead to different likely decision choices. Referencesare available upon request.

For further information contact:Gerald Smith

450A Fulton HallCarroll School of Management

Boston CollegeChestnut Hill, MA 02467



SHIFTING CONSUMER PREFERENCE: THE EFFECTS OF COGNITIVESTYLES AND TEMPORAL PERSPECTIVE


SUMMARY

How are consumers’ cognitive styles altered over thecourse of time, and how does this alteration shift consumers’product preferences? The present paper addresses thisquestion by examining the impact of temporal perspectiveeliciting different types of cognitive modes on consumers’perceptions of products. The results reveal that when apurchase is temporally proximal, eliciting analytic (vs.holistic) processing modes enhances consumers’ productpreferences. When a purchase is temporally distant,however, eliciting holistic (vs. analytic) processing modesbetter drives consumers’ product preferences.

Keywords: Temporal distance; Mental construal;Cognitive styles; Attitude change; Product preference;Construal level theory.

Marketers often emphasize time pressure in theiradvertising or promotional messages. For example, whenVictoria’s Secret sends out promotional email messages,it frequently sets up a specific time frame through whichan exclusive online special offer is valid (e.g., “Freeshopping with any purchases! Ends Saturday! Use offercode BRASVS!”). IKEA, an international home productsand self-assembled furniture retailer, is often referencingto a limited-time frame for its solid wood furniture sales(e.g., “Seize the days! Grab these spectacular limited-timeoffers before they’re gone!”). In these ubiquitous mes-sages guiding consumers’ attention to temporal time,consumers may differently react to them, because theperceived proximity of an event in time alters their mentalrepresentations of the event (Förster, Friedman, andLiberman 2004). For example, a person buying a car ayear from now might think it in terms of more abstract andgeneral goals, such as “getting a safe and convenienttransportation for family,” whereas a person thinkingabout a car buying now might construe it in terms of moresubordinate and concrete goals, such as “finding a dealer-ship close to my house.”

We wonder whether this mental representation sys-tematically changed by the temporal distance might encour-age people to involve different types of cognitive process-ing modes, such as analytic thinking modes focusing on“detachment of the object from its context,” or holisticthinking modes taking more attention to “the context orfield as a whole” (Nisbett et al. 2001). Identifying thespecific thinking mode relative to the temporally proxi-

mal versus distant future purchases can be important ininfluencing consumers’ judgments and decisions about aproduct, because it may help marketing managers andadvertisers to create efficient and persuasive advertisingmessages eliciting consumer preferences toward theirproducts.

Therefore, in this research, we highlight the role ofdifferent types of thinking modes in consumer prefer-ences, when the purchase is proximal- versus distant. Wepropose and demonstrate that eliciting analytic versusholistic thinking modes influences consumer preferencestoward the temporally proximal versus distant futurepurchases, and thus this relative preference subsequentlyaffects the type of products they prefer. We argue that, dueto the heuristic tendency to construe near future eventsconcretely and distant future events abstractly (Förster,Friedman, and Liberman 2004), individuals elicited foranalytical thinking modes prefer the products framed bytemporally proximal versus distant appeal. When a pur-chase is temporally distant, however, individuals elicitedfor holistic thinking modes prefer products framed bytemporally distant appeal.

In the vacation trip experiment adopted from Mogilner,Aaker, and Pennington (2008), we examined whethereliciting analytic versus holistic thinking modes wouldexert differential effects on consumers’ attitudes towardthe vacation airline ticket messages with the temporalframe of purchase event. The findings of the experimentsuggest that analytic thinkers prefer a product messageframed by temporally proximal (vs. distant) appeal(Mproximal = 4.42 vs. Mdistant = 3.48; F(1,21) = 3.94,p < .07), but holistic thinkers prefer a product messageframed by temporally distant (vs. proximal) appeal(Mproximal = 3.55 vs. Mdistant = 5.03; F(1,22) = 15.59,p = .001). The results also showed that analytic thinkersexpressed higher purchase intention for a product framedby temporally proximal appeal than for a product framedby temporally distant appeal (Mproximal = 4.17 vs.Mdistant = 2.64; F(1,21) = 6.23, p < .03). Conversely,holistic thinkers expressed higher purchase intention fora product framed by temporally distant appeal than for aproduct framed by temporally proximal appeal(Mproximal = 2.91 vs. Mdistant = 3.92; F(1,22) = 6.12,p < .03).

These results suggest that different types of cognitiveprocessing modes interact with the temporal perspective


of a purchase. Our findings contribute to prior researchexamining the relationship between temporal perspectiveand cognitive processing (Förster, Friedman, and Liberman2004; Schooler 2002; Schooler, Fiore, and Brandimonte

1997). We believe that this is the first research to identifyanalytic versus holistic thinking modes as a determinantfor the temporal frame of the future purchase event.References are available upon request.



21 East Market St. S252 PBBIowa City, IA 52242Phone: 319.335.0967



ON-THE-GO CONSUMPTION: BECAUSE CONSUMERS WANT TO ANDNOT BECAUSE THEY HAVE TO

Sabine Moeller, European Business School (EBS), GermanyTobias Schaefers, European Business School (EBS), Germany

SUMMARY

An ongoing increase in on-the-go consumption offood and beverages can be observed especially inindustrialized countries, thus being perceived to be one ofthe major trends in consumer behavior (e.g., Figee andVringer 2007; Morganosky 1986). Regarding the factorsinfluencing consumers’ decision to consume on the go,common understanding as well as most prior researchrefer to pragmatic or utilitarian reasons like time pressure,price sensitivity or health orientation (Berry 1979;Botonaki, Natos, and Mattas 2009; Buckley et al. 2005;Darian and Cohen 1995; Warde 1999). We certainly agreethat these factors – which are determined by consumers’living conditions such as income, workload or healthcondition – are of high importance. However, based onprior studies (e.g., Gofton 1995; Harris and Shiptsova2007) as well as qualitative research we assume thathedonic factors, related to consumers’ enjoyment of on-the-go consumption, are highly relevant as well. Thus, wepresent and empirically test a model investigating utilitarianas well as hedonic determinants of on-the-go consumptionof food and beverages. The contribution of our researchresults is to advance the understanding of consumerbehavior regarding on-the-go consumption of food andbeverages – especially with regard to the fact that noknown research has investigated, empirically tested, andcompared the effects of utilitarian versus hedonicdeterminants.

Model and hypotheses development were based onan extensive literature review and a preliminary qualita-tive study. A total of 16 semi-standardized interviewswith international experts as well as six focus groupinterviews with consumers were conducted. All of thesequalitative research steps revealed that beyond the utili-tarian reasons prior research has focused on, enjoyment ofconsuming on-the-go offerings is at least equally relevant.These insights were utilized to deduce our hypotheses andspecify our research model.

Survey data was collected from 805 German respon-dents via a nationwide online access panel. Respondentswere qualified so that the sample was ensured to berepresentative to the German population by gender, age,and employments status. The measurement model evalu-ation using confirmatory factor analysis indicated that all

measures used exhibited sufficient degrees of reliability,convergent and discriminant validity. Regarding the struc-tural model all goodness-of-fit indicators satisfied gener-ally accepted criteria. Except for health orientation allstructural paths toward on-the-go consumption are sig-nificant at the .01 level.

The model estimation supported the hypothesizednegative effect of price sensitivity on the inclination toconsume on-the-go. This finding is in line with the deci-sion of more and more discounters and other retailers withaggressive pricing strategies to enter the market for on-the-go products. Secondly, the assumption that consum-ers with a higher health orientation show a lower inclina-tion to consume on-the-go needed to be rejected. Onepossible explanation is that the assumption of on-the-goproducts being generally perceived as unhealthy onlypartly reflects reality. Providers of on-the-go food anddrinks might have reacted to the ongoing trend of anincreasing health orientation by offering products withbetter nutrition values. Third, our model estimation showsthat time pressure and enjoyment positively influenceconsumers’ inclination towards on-the-go consumption.This suggests that on-the-go offers are indeed perceivedas time saving and that for a large part consumers do enjoyconsuming on the go. However, the most remarkablefinding in this regard is that the influence of enjoymentexceeds that of time pressure as a determinant for on-the-go consumption by far. This implies that consumersconsume on-the-go not because they have to but ratherbecause they want to. Subsequently, our findings chal-lenge the common view of time pressure as the maindeterminant of consumers’ inclination to consume on thego. Instead of the utilitarian reasons – time pressure,health orientation, and price sensitivity – it is rather ahedonic shopping motive that has the highest influencethe demand for food and beverage that can be consumedon the go.

The implication for service providers and retailersoffering on-the-go food and drinks is to focus more on thehedonic motive of on-the-go consumption. The impor-tance of enjoyment should not only be considered regard-ing the offerings, but also regarding service, store design,and personnel. Although we still think that time pressurein general plays an important role for many consumers itdoes not seem to lead to a competitive edge anymore.


Some limitations have to be taken into account wheninterpreting our research findings. First, our study focusedon only one European country. Our model is therefore notable to account for cultural influences regarding on-the-go consumption. Further research should therefore try tovalidate and compare the model across several countries.

Second, previous research has shown that other aspectslike demographic variables influence on-the-goconsumption (Binkley 2006). These variables should beutilized in further research to compare for groupdifferences.

For further information contact:Sabine Moeller

Lekkerland Endowed Chair for Convenience and MarketingEuropean Business School (EBS)

International University, Schloss Reichartshausen65375 Oestrich-Winkel




THE ROLE OF SELF REGULATORY FOCUS ONCOMPLIANCE BEHAVIOR

Stephanie Dellande, University of New OrleansPrashanth U. Nyer, Chapman University, Orange

SUMMARY

The purpose of this study is to better understand thefactors that influence compliance in compliance depen-dent services (CDS); services that are long-term in nature.Compliance involves behaviors the customer must per-form when away from the service provider in order toachieve the desired outcome (Dellande and Gilly 1998).We empirically examine the impact of self regulatoryfocus (SRF) on long term customer compliance withprescribed behaviors in a weight loss center in south India.

Theoretical Construct and Hypotheses

Self-Regulatory Focus (SRF). Research on SRF theory(Higgins 1997, 2002) has indicated that promotion focusedindividuals have a greater persistence when pursuinggoals, are more motivated to continue with effortful goals(Roney, Higgins, and Shah 1995), have a better ability tosucceed after experiencing a failure (Crowe and Higgins1997), and exhibit higher standards of achievement thatresult in better performance (Higgins 2002) in compari-son to prevention focused individuals. This leads to thefollowing hypothesis.

H1: Promotion focused subjects are more likely to engagein compliance behavior compared to preventionfocused subjects.

Keller and Lehman (2008) indicate that individualcharacteristics such as SRF moderate health intentions.Since promotion focused subjects are motivated by gainswhile prevention focused subjects are motivated by theurge to prevent losses, these two types of subjects shouldrespond differently to persuasive messages which prom-ise gains versus those which promise the prevention oflosses.

H2: The compliance behavior of promotion focused sub-jects will be more influenced by messages promisinggains while the behavior of prevention focused sub-jects will be more influenced by messages promisingthe prevention of losses, thus resulting in a two-wayinteraction of the regulatory focus of the subjects andthe regulatory focus of the message.

In recent years there have been a large number ofresearch projects exploring SRF and there is some uncer-tainty involving the validity and reliability of the mea-sures of global regulatory focus (cf., Summerville andRoese 2008; Haaga et al. in press), e.g., the RFQ (Higginset al. 2001), the six item regulatory focus scale (Higginset al. 1994), Lockwood, Jordon, and Kunda’s (2002)general regulatory focus measure, etc. It seems reasonablethat these generalized measures of regulatory focus wouldcompare poorly with situation specific measures of regu-latory focus based on the subjects’ stated reasons for theirgoals (e.g., weight loss).

H3: Hypotheses H1 and H2 will show more significantresults when SRF is measured using situation spe-cific measures of SRF rather than general measuresof SRF.

Methodology

The subjects were 243 female clients (aged 20 to 45)of a fitness/weight loss center in south India who signedup for an eight-week long program to lose modest amountsof excess weight (e.g., 10 to 20 pounds). Upon joining,respondents completed a questionnaire that included sev-eral scales of regulatory focus, and a question elicitingreasons for wanting to lose weight. Subjects were weighedupon enrollment and at the conclusion of the eight-weekprogram. Subjects were randomly assigned to one of threeconditions: gain-focused message, loss focused message,or gain and loss-focused message.

Measures

Four different measures of SRF were used: the RFQ(Higgins et al. 1994), a modified version of the GeneralRegulatory Focus Measure (GRFM; Lockwood, Jordon,and Kunda 2002), a single item measure (GLOM) adaptedfrom Galinsky, Leonardelli, Okhuysen, and Mussweiler(2005), and subjects’ list of reasons (Reasons) why theywanted to lose weight.

In testing H3, we treat RFQ and GRFM as the generalmeasures of regulatory focus (only 4 of the 18 items in themodified GRFM inventory are specific to weight loss),


and GLOM and Reasons as the situation specific mea-sures of regulatory focus. The dependent variable (WLA)was the percentage of the weight loss goal that wasachieved by the end of the eight-week period.

Findings

We found support for H1 and H2. Further, whensubjects are grouped as promotion or prevention focusedon the basis of situation specific measures of regulatoryfocus a significant effect of regulatory focus on WLA isfound, with promotion focused individuals achievinggreater weight loss compared to prevention focused indi-viduals. However, a significant effect for SRF was notfound when using generalized measures of SRF.


This research examines consumer compliancebehavior in the context of weight loss activities and hasimplications for other services requiring consumers toengage in prescribed behaviors over the long term. Forexample, success of debt counseling services andretirement savings programs require clients to engage incertain behaviors over the long run. Noncompliance couldresult in consumers not achieving the desired outcome,customer dissatisfaction, drop in customer retention, andnegative word-of-mouth. Marketers of CDS programswill be able to use the findings of this research project tofind new ways to increase long term customer compliance.

REFERENCES

Crowe, Ellen and E. Tory Higgins (1997), “RegulatoryFocus and Strategic Inclinations: Promotion and Pre-vention in Decision Making,” Organizational Behav-ior and Human Decision Processes, 69, 117–32.

Dellande, Stephanie and Mary Gilly (1998), “GainingCustomer Compliance in Services,” Advances inServices Marketing and Management, 7, 265–92.

Galinsky, Adam D., Geoffrey J. Leonardelli, Gerardo A.Okhuysen, and Thomas Mussweiler (2005), “Regu-latory Focus at the Bargaining Table: PromotingDistributive and Integrative Success,” Personalityand Social Psychology Bulletin, 31 (8), (August),1087–98.

Haaga, David A.F., Dara G. Friedman-Wheeler, Eliza-beth McIntosh, and Anthony H. Ahrens (in press),“Assessment of Individual Differences in RegulatoryFocus Among Cigarette Smokers,” Journal of Psy-chopathology and Behavioral Assessment.

Higgins, E. Tory, Christopher J.R. Roney, Ellen Crowe,and Charles Hymes (1994), “Ideal Versus OughtPredilections for Approach and Avoidance – DistinctSelf-Regulatory Systems,” Journal of Personalityand Social Psychology, 66 (2), 276–86.

____________ (1997), “Beyond Pleasure and Pain,”American Psychologist, 52, 1280–1300.

____________, Ronald S. Friedman, Robert E. Harlow,Lorraine C. Idson, Ozlem N. Ayduk, and Amy Taylor(2001), “Achievement Orientations from SubjectiveHistories of Success: Promotion Pride Versus Pre-vention Pride,” European Journal of Social Psychol-ogy, 31, 3–23.

____________ (2002), “How Self-Regulation CreatesDistinct Values: The Case of Promotion and Preven-tion Decision Making,” Journal of Consumer Psy-chology, 12, 177–91.

Keller, Punam A. and Donald R. Lehman (2008), “De-signing Effective Health Communications: A MetaAnalysis,” Journal of Public Policy & Marketing, 27(2), 117–30.

Lockwood, Penelope, Christian H. Jordan, and Ziva Kunda(2002), “Motivation by Positive and Negative RoleModels: Regulatory Focus Determines Who WillBest Inspire Us,” Journal of Personality and SocialPsychology, 83, 854–64.

Roney, Christopher J., E. Tory Higgins, and James Shah(1995), “Goals and Framing: How Outcome FocusInfluences Motivation and Emotion,” Personalityand Social Psychology Bulletin, 21, 1151–60.

Summerville, Amy and Neal J. Roese (2008), “Self-Report Measures of Individual Differences in Regu-latory Focus: A Cautionary Note,” Journal of Re-search in Personality, 42 (1), 247–54.

For further information contact:Stephanie Dellande

Department of Marketing and LogisticsCollege of Business Administration

University of New OrleansNew Orleans, LA 70148

Phone: 504.280.7313Fax: 504.280.5443



ANTECEDENTS OF NEW PRODUCT DEVELOPMENT: A COMPARISONBETWEEN GERMANY AND CHINA

Malte Brettel, RWTH Aachen University, GermanyDominik Austermann, RWTH Aachen University, Germany

Andreas Engelen, RWTH Aachen University, Germany

SUMMARY

There is agreement in the literature that new productsuccess is a major driver of organizational performance(Henard and Szymanski 2001). Therefore, it is not sur-prising that the literature provides extant insights on howto positively impact new product development (NPD)activities. These insights have almost exclusively beenderived from Western cultural environments, most nota-bly Western Europe and North America (Garrett et al.2006). Only a few fragmented studies on Asian particu-larities in NPD processes exist (e.g., Im et al. 2003). Thereality, however, shows that NPD processes are no longerlimited to one single national or cultural setting (Nakataand Sivakumar 1996; Lee et al. 2008).

Building upon literature on the nation- and culture-dependency of management practices, it is at least disput-able whether insights for best practice for NPD processesgained in the Western context are transferable to othercultural contexts. Literature on NPD to date, however, hasnot treated the interface between NPD and national cul-ture (Im et al. 2003). In a review on major studies on NPD,Guo, 2008 claims that “in a broader environmental con-text, different national cultures and economic and socialsystems largely affect firm’s new product development,”while he derives from his literature review that “NPD hasnot been studied extensively across a range of nationalboundaries” (p. 257).

The present research addresses this dearth andinvestigates major organization-level strategic andorganizational antecedents of NPD stages with surveydata from Germany and China as two culturally highlydistinct settings (Hofstede 1980). Building upon a literaturereview, five major antecedents are selected and examinedregarding their effect on NPD initiation and implementation.

Our Chinese sample consists of 164 companies ofsmall and medium size. Consistent with Gao et al. 2007,we used face-to-face and telephone interviews to increasethe response rate. This method is better than traditional

mail surveys in emerging economies because it can gen-erate more valid information. Furthermore, cooperationwith local researchers offers a key means to obtain reliableinformation. Therefore, we hired a major market researchcompany headquartered in Shanghai. The data was col-lected during a two-month period from August to Septem-ber 2008. For the German sample, we contacted about2,500 SMEs by mail between June and July 2008, drawingon the German Chamber for Industry and Commercedatabase. In total, 208 companies participated in thesurvey

Managers learn that a customer orientation is a stron-ger driver of NPD activities in the individualistic and lowuncertainty avoidance German culture. This finding runscounter to our expectations. The higher importance ofcustomer orientation in Germany may be justified by thefollowing: Collectivist cultures define themselves by theiradherence to groups. However, they make clear distinc-tions between in-groups and out-groups (Triandis 1994).One may argue that this is not necessarily the case. Itfollows that even in individualistic cultures as in Ger-many, relationships with customers – who are not mem-bers of their own organization – are valued more highlythan in collectivist cultures. In individualistic cultures, thedifferentiation between in-groups and out-groups doesnot exist. In line with our expectations, competitor orien-tation is more important in high uncertainty avoidanceChina. In terms of organizational antecedents, the differ-ences are largely confirmed. Whereas partially theseantecedents have positive effects across national cultures,the strength of these effects is clearly subject to culturalcircumstances. In low power distance Germany, the effectof participation is stronger. Consideration, however, linkedto the cultural dimensions of power distance and feminin-ity, turns out to be stronger in China. Top managementemphasis positively impacts initiation and implementa-tion in both settings. Due to obedience being one majorfacet of higher power distance, the effect of managementemphasis is stronger in China where power distance ishigh.


For further information contact:Malte Brettel

Centre for EntrepreneurshipRWTH Aachen University

Templergraben 6452062 Aachen

GermanyPhone: +49.241.8096177



DISTRIBUTION CHANNELS FOR CONSUMER PRODUCTS IN INDIA:PAST, PRESENT, AND FUTURE DIRECTIONS

Debi P. Mishra, State University of New York at Binghamton

SUMMARY

With a population in excess of one billion and currentannual GDP growth of 9 percent (Vietor and Thompson2007), India is a major player in the world economy. Notsurprisingly, by 2050 the country is projected to becomethe third largest economy after China and the UnitedStates (Hawksworth 2006). India’s economic prowess isbeing driven by the purchasing power of a burgeoningmiddle class as wealth steadily trickles down to the bottomof the economic pyramid. Given this brisk growth, domesticindustries are in a race against time to ramp up capacity,increase production, and achieve market access via chan-nels of distribution. One sector that is expected to bear thebrunt of this demand is the consumer goods industry withretail sales expected to top $40 billion by 2015 (IndiaBrand Equity Foundation 2008).

Despite its potential, the consumer goods industryfaces several significant marketing constraints. First, manu-facturers and retailers have to grapple with fragmentedmarkets and a plethora of channel forms in a constant stateof flux. In particular, numerous street-side vendors, hawk-ers, and roughly 12 million unregulated neighborhoodmom-and-pop or kirana stores create strong institutionalforces that cannot be ignored. Second, frequent regulatorychanges affect channel structure and exacerbate adapta-tion challenges. For example, in 2006 the governmentallowed direct foreign entry by single brand retailers(Lakshman 2007). Consequently, firms scampered forupscale retail space in a hypercompetitive real estatemarket while domestic manufacturers faced a multitudeof challenges in the areas of new product introduction,line stretching, and branding.

Given the importance of distribution channels to theemerging Indian economy, one would expect a consider-able body of relevant academic research to be readilyavailable. However, a careful appraisal of extant researchbelies this expectation. While India has garnered muchattention, the focus has primarily been on general topicspertaining to the socio-economic, political, and businessenvironments (Basu 2008; Vietor and Thompson 2007).In recent years, the emphasis has shifted to include researchon other topics like entry modes (Johnson and Tellis2008), and outsourcing. However, there remains a paucityof systematic work on the impact of distribution on the

Indian economy in general and the FMCG industry inparticular.

This study attempts to bridge the gap in our under-standing of distribution channels for consumer goods inIndia. More specifically, the objectives of this researchare: (a) to appraise distribution channel structure andmanagement challenges for consumer products, (b) todelineate variations in channel forms across markets, and(c) to outline the strategic role of distribution channels ingaining market access and achieving competitiveadvantage.

We develop several channel archetypes by consideringa 2X2 classificatory scheme with market access (easy,difficult) and per capita demand (high, low) as the relevantdimensions. In each resulting cell, the implications forchannel structure and management are discussed. Themain challenges and opportunities for firms lies in theirability to carefully appraise channel practices in each cell.For example, when market access is easy and per-capitademand is high, big-box retailers are in a better position toovercome significant entry barriers relative to other cells.The main challenge for modern retailers is to position theiroffering as a superior alternative to the neighborhoodkirana store that has historically offered a familiar andconvenient shopping option to customers. Customers inIndia are often skeptical of large retail stores and feel thatthey have to pay a high price at modern stores. To over-come such negative perceptions, modern stores such asBig Bazaar have developed creative ways of signalingvalue to customers (Raman and Winig 2006). For example,to create a familiar shopping environment, Big Bazaar’sstore layout mimicked the chaos of a traditional bazaar.The store also prominently advertised trade-ins wherebycustomers could bring their old merchandise and exchangethem for store coupons. Finally, the company successfullyorganized “Big Day” sales around the time of local religiousfestivals when customers typically go on a shoppingspree. All these efforts have paid off handsomely andtoday Big Bazaar is one of the fastest growing retail chainsin India.

In conclusion, we offer several guidelines to practic-ing managers for considering distribution channel deci-sions as a strategic tool for market entry and growth.


REFERENCES

Basu, Kaushik (2008), “The Enigma of India,” Journal ofEconomic Literature, 46 (2), June, 396–406.

Hawksworth, John (2006), The World in 2050: How BigWill the Major Emerging Market Economies Get andHow Can the OECD Compete? New York:PricewaterhouseCoopers.

India Brand Equity Foundation (2008), “Consumer Markets,”[http://www.ibef.org/economy/consumermarket.aspx].

Johnson, Joseph and Gerald J. Tellis (2008), “Drivers ofSuccess for Market Entry into China and India,”

Journal of Marketing, 72 (May), 1–13.Lakshman, Nandini (2007), “Protesters Tell Wal-Mart to

Quit India,” Business Week, (October 12). Accessedonline at: [http://www.businessweek.com/print/g loba lb i z / con t en t / oc t2007 /gb20071011_524356.htm].

Raman, Ananth and Laura Winig (2006), “Big Bazaar,”Harvard Business School Case No 9-606-099. Bos-ton: Harvard Business School Publishing.

Vietor, Richard H. and Emily J. Thompson (2007), “Indiaon the Move,” Harvard Business School Case No: 9-703-050. Boston: Harvard Business School Publishing.

For further information contact:Debi P. Mishra

School of ManagementSUNY Binghamton

Binghamton, NY 13850Phone: 607.777.2893



PRODUCT AND ENVIRONMENTAL INFLUENCES ON MARKETING-MIX STANDARDIZATION: A CENTRAL EASTERN

EUROPEAN PERSPECTIVE

Roxana Codita, Technische Universität München, GermanyFrank-Martin Belz, Technische Universität München, Germany

SUMMARY

In spite of the high attractiveness of Central EasternEuropean (CEE) countries from a theoretical and manage-rial perspective, empirical studies of international market-ing standardization analyzing this region as a host-regionare to a large extent lacking. Therefore, this work intendsto contribute to international marketing theory by testingthe standardization-adaptation framework within a lessexplored, yet highly attractive region. The present papertakes an approach of theory development and refinement.Concerning theory development, product characteristics,often mentioned as one of the most important determi-nants of marketing-mix standardization, are aggregatedinto an index describing a “product’s standardizationpotential,” whose influence on the adopted strategy isempirically tested. By introducing and testing the“product’s standardization potential” construct as a for-mative measure, this study makes a theoretical contribu-tion to the contingency perspective of international mar-keting standardization. From a theory refinement point ofview, the use of formative measurement models within thestandardization-adaptation framework, though theoreti-cally sensible, is still in the starting blocks. Thus, thisstudy reports first empirical experiences concerning theapplication and validation of formative measurementmodels using the PLS path modeling approach. More-over, the influence of the external environment andproduct’s characteristics upon a firm’s marketing-mixstrategy is empirically tested within the CEE context.

Conceptual Framework

Drawing from industrial organization theory, thisstudy investigates the influence of the external environ-ment and product’s characteristics upon a firm’s market-ing-mix strategy. Also known as the structure-conduct-performance paradigm, this theory sees the source of afirm’s competitive advantage in the developing of a cor-porate strategy (conduct) aligned with market and indus-try conditions (structure) (Porter 1980). Implementing anappropriate marketing strategy is considered the mecha-nism by which companies respond successfully to theexternal environment, thus achieving competitive advan-tage and superior performance (Katsikeas, Samiee, and

Theodosiou 2006; Zou and Cavusgil 2002; Zou and Stan1998).

As Figure 1 illustrates, the model we propose in thispaper investigates the influence of five contingency vari-ables, similarity of macro-environment, similarity of con-sumer characteristics, similarity of marketing infrastruc-ture, competition intensity, and product’s standardizationpotential, on the degree of standardization of the fourmarketing-mix elements. The impact of marketing-mixstandardization on performance is tested.

Measures, Sample, and Data Collection

Following a review of the relevant literature, scalesused in previous international marketing research wereidentified. Most scales use the pair-wise comparisonmethod developed by Sorenson and Wiechmann (1975),asking the respondent to draw a comparison between thehome- and host-market when evaluating the items. Theproduct characteristics were merged into an index repre-senting the “product’s standardization potential,” con-taining a set of seven semantic differential 5-point scaleslabeled: low-tech/high-tech, simple/complex, unique/stan-dard, innovative/traditional, rational/emotional, culture-free/culture-bound, symbolic/functional. Based on theo-retical and empirical considerations, four constructs, i.e.,product’s standardization potential, competition inten-sity, similarity of marketing infrastructure, and perfor-mance were modeled in a formative way.

The population of this study consisted of interna-tional marketing managers within German consumer goodsmanufacturers with marketing responsibilities for a prod-uct or product group in at least one of the ten CEEcountries, members of the European Union, i.e., Poland,Slovenia, Hungary, Slovakia, Czech Republic, Latvia,Lithuania, Estonia, Romania, and Bulgaria. The respon-dent was asked at the beginning of the questionnaire toselect among the 10 CEE markets listed, one market,where his/her company markets consumer products ANDwith which he/she is personally most familiar with. Datawas collected between March and April 2009, using onefollow-up mailing. A number of 132 questionnaires werereturned.


Results

The hypotheses were tested using PLS path model-ing. Similarity of consumer characteristics has a signifi-cant positive effect on all marketing-mix elements (prod-uct β: 0.16 t: 1.69*; promotion β: 0.24 t: 2.46**; pricingβ: 0.33 t: 2.99**) except for distribution standardization.H2 is thus supported. The hypothesized positive influenceof similarity of marketing infrastructure (H3) could besupported for the promotion (β: 0.31, t: 3.27***) anddistribution standardization (β: 0.53, t: 5.13***). Theproduct’s standardization potential has been shown tohave a significant positive impact on pricing (β: 0.25,t: 1.96*) and product standardization (ß: 0.37, t: 2.67**),in support of hypothesis H5. No significant influence onany of the marketing-mix elements has been detected incase of the variables similarity of the macro-environmentand competition intensity. Consequently Hypotheses H1and H4 must be rejected. Though all path coefficientsconnecting the marketing-mix variables to performanceare positive, only distribution standardization has also asignificant impact (β: 0.38, t: 3.33***), providing partialsupport for H6.

The variance explained, R2, amounts to 20 percent forproduct standardization, 24 percent for promotion stan-dardization, 23 percent for pricing standardization, and 32percent for place standardization. The model explains 24percent of the variance in the performance construct. Thisindicates that, especially in the case of product standard-ization, there are other factors that play a major role in thedecision to standardize or adapt. These may be related tofirm specific contingencies such as management’s inter-national orientation, global marketing structure, globalmarketing processes, market entry mode.

Though four out of six hypotheses could find partialsupport in this study, the results indicate that, in practice,product characteristics and the external environment havea relatively minor influence on the configuration of themarketing-mix strategy in CEE. The high degree of mar-keting-mix standardization, especially product standard-ization (4.32) exhibited by the companies in our samplesuggests that CEE markets may still be considered testmarkets, where investments are still kept to a minimum.

FIGURE 1A Contingency Framework of Marketing-Mix Standardization and

Performance Outcome

Degree of Marketing-Mix Standardization• Product• Pricing• Promotion• Distribution

Performance

Similarity of Macro-Environment

Similarity of Consumer Characteristics

Similarity of Marketing Infrastructure

Competition Intensity

Product’s Standardization Potential

H1 (+)

H2 (+)

H3 (+)

H4 (-)

H5 (+)

H6 (+)

Degree of Marketing-Mix Standardization• Product• Pricing• Promotion• Distribution

Performance

Similarity of Macro-Environment

Similarity of Consumer Characteristics

Similarity of Marketing Infrastructure

Competition Intensity

Product’s Standardization Potential

H1 (+)

H2 (+)

H3 (+)

H4 (-)

H5 (+)

H6 (+)

REFERENCES

Katsikeas, C.S., S. Samiee, and M. Theodosiou (2006),“Strategy Fit and Performance Consequences of Inter-national Marketing Standardization,” Strategic Man-agement Journal, 27 (9), 867–90.

Porter, M. (1980), Competitive Strategy. New York: Free

Press.Sorenson, R.Z. and U.E. Wiechmann (1975), “How Mul-

tinationals View Marketing Standardization,”Harvard Business Review, 53 (3), 38–44.

Zou, S. and S. Stan (1998), “The Determinants of ExportPerformance: A Review of the Empirical LiteratureBetween 1987 and 1997,” International Marketing


Review, 15 (5), 333–56.____________ and S.T. Cavusgil (2002), “The GMS: A

Broad Conceptualization of Global Marketing Strat-

egy and its Effect on Firm Performance,” Journal ofMarketing, 66 (4), 40–56.

For further information contact:Roxana Codita

Frank-Martin BelzBrewey and Food Industry Management

Technische Universität MünchenAlte Akademie 14

85354 FreisingGermany

Phone: +49.8161.71.3098Fax: +49.8161.71.3279



THE EFFECTS OF SPONSORSHIP ON BRAND IMAGE: A CROSS-NATIONAL ANALYSIS

Christian Lucas, TU Dortmund University, GermanyDavid M. Woisetschläger, TU Dortmund University, Germany

Alexander Eiting, TU Dortmund University, Germany

SUMMARY

Even though global growth rate dropped from 13.7percent in 2008 to 3.1 percent for 2009, sponsorshipspending is still increasing compared to investments inadvertising or sales promotions. Especially for globalplayers, such as Lenovo, Credit Suisse, BMW, or Panasonicthere are not many alternative sponsorship opportunitiesthat aim at a global audience. For instance, Formula Oneoffered 18 races with cumulated 1.8 million race dayattendances in 2006 (F1 Black Book 2007). According toa Financial Times special report, about 597 million peopleare estimated to have watched the 17 races of FormulaOne in 2007 (Sylt and Reid 2008).

Global sponsorship is believed to depend on specificfactors on a country-level, meaning that sponsorshipmight work differently from country to country (Ruth andSimonin 2003). Such variations in sponsorship effectswould be unintentional from a sponsor’s perspective.From a management perspective, it is highly relevant to(1) assess if country-differences in sponsorship effectscan be observed, and (2) if such differences can beexplained by potentially manageable variables on a coun-try-level. Prior research has predominantly focused onvariation in sponsorship outcomes that result from indi-vidual factors such as someone’s personal history with anevent (Gwinner 1997). On a country-level, it can beexpected that in countries with a long history of motorsportevents and many personal experiences with motorsports,people will have a different event image than TV viewersin countries without the ability to visit motorsport racespersonally. Moreover, broadcast times differ from coun-try to country. Hence, consumers have more or less timeto learn sponsor names and to attribute the image ofFormula One to the sponsor. Existing sponsorship researchthat considers country-variables such as nationality of thesponsor is conducted in a single country (e.g., Ruth andSimonin 2003) and does not consider variation on acountry-level.

Against this background, the present article’s contri-bution to the literature is to conceptualize and empiricallyidentify individual and country specific group-level fac-tors that explain differences in brand image, an importantsponsorship outcome for marketing management. Usingmarket research data obtained in a multinational field

survey by a major automotive company in 2007, a theory-based model is tested using field data from 14 countriesand objective data from secondary sources.

One of the key objectives of sponsorship is to evokepositive feelings and attitudes toward the sponsor. Thisstudy attempts to replicate findings on the individual-level. More precisely, the influence of event image (e.g.,Grohs et al. 2004), consumption of Formula One (e.g.,Lacey et al. 2007), sponsor recognition (Smith 2004), andfandom (Gwinner and Swanson 2003) on brand imagewill be assessed. Because this article is a cross-nationalanalysis and not comparing different brands, sponsorshipfit is not considered as a potential influencing variable.

Prior research has predominantly focused on varia-tion in sponsorship outcomes that result from individualfactors and largely ignored variation from group-levelfactors on a national basis. But as these variations insponsorship effects are unintentional from a sponsor’sperspective it is highly relevant to sponsors to know (1)what kind of effects these are and (2) how to manage theseeffects. Effects that vary from country to country, whichare believed to be influential are (1) whether there is aFormula One race in the country and (2) whether a team’sdriver comes from that country.

Data was collected in 14 different countries simulta-neously in August of 2007. The countries were selectedaccording to their economic relevance to the manufac-turer and capture the most relevant automobile markets inthe world (Australia, Brazil, Canada, China, France, Ger-many, India, Italy, Japan, Malaysia, Russia, Spain, U.K.,and the USA). Due to target market restrictions onlyrespondents aged between 18 and 60 years old, holding adriving license and living in a household that owns at leastone car were interviewed. The sample size for eachcountry was in the range of 622 to 677, with a total of9,520 respondents.

Results indicate that at the individual consumer level,brand image is influenced by event image, fandom, andpossession. Contrary to our expectations, consumption ofFormula One races and recognition of the sponsor branddo not influence brand image. At level-two, results showthat the intercept of the regression is significantly lower ifthere is a race in the country of the respondent and the


driver of the team comes from that country. Thesecounterintuitive findings can be attributed to the facts that(1) brand image is worse in the countries were FormulaOne races take place and that (2) the team has drivers fromtwo countries in which the market position of the brand isrelatively weak. As proposed, the effect of fandom onbrand image is increased, if there is a race in the countryof the respondent and the driver of the team is from thatcountry.

The analysis presented here extends current researchby controlling for heterogeneity between countries usingHLM and offers results from which managerial implica-tions can be derived. On the individual-level, we showthat event image as well as fandom influences the attitudetoward a sponsor positively. These results confirm previ-ously reported effects of fan identification (e.g., Madrigal2001) and image transfer in sport sponsorship contexts(e.g., Grohs et al. 2004). The results are stable across

countries, although the influence of fan identification isfound to differ significantly. The positive effect of fandomis increased, if races take place in the country of therespondent and drivers are from the fan’s country. Thisimplies that sponsors should influence the decision-makingprocess of having Formula One races in new and for thesponsor’s strategically important countries. Additionally,in order to boost a new entry of the sponsor in a certaincountry or to improve the brand’s image in a country,sponsors should consider employing a national driver.Consequently, marketing management should not onlydecide driver selection according to sportive achieve-ments and capabilities but also according to relevantmarkets. Generally, fandom is found to have a stronginfluence on brand image. Therefore, sponsorship man-agers as well as sport series representatives should investin building a strong fan base. References are availableupon request.

For further information contact:David Woisetschläger

Business and Social SciencesTU Dortmund University

Otto-Hahn-Str. 6Dortmund, 44227




PROMOTING BUYER-SELLER COLLABORATIONS: THEROLE OF COOPERATIVE CAPACITY

Maggie Chuoyan Dong, City University of Hong Kong

SUMMARY

Besides the critical resources invested, does a certaincapability is needed to achieve superior inter-firm col-laborative outcomes? To address this question, our studyextends the existing resource-based view to investigatewhat capabilities embedded in the specific relationships –in particular, the dyad’s cooperative capacity – facilitatethe successful cooperation between buyers and sellers.The relation-specific “cooperative capacity” depicts thecritical capabilities gained in a certain inter-firm collabo-ration, through the integration of the partners’ valuableresources and knowledge. Drawing from the features ofcooperation, we conceptulize three major dimensions ofcooperative capacity: learning, changing, and creatingcapabilities. These three dimensions of cooperative capac-ity reflect the major benefits that can be acquired throughthe buyer-seller collaboration rather than pure markettransactions or hierarchies. These three dimensions arehighly related with each other and together constitute theconcept of cooperative capacity.

We also delineate the beneficial roles of some eco-nomic, social, and contractual factors between the dyadon the dyad’s cooperative capacity. At the economic level,resource complementarity is examined as an antecedentof cooperative capacity. We study social fitness as anothersource of the dyad’s cooperative capacity, and examineinstitutional proximity as an indicator of the social fitnessof partners. In addition, we test role clarity as a contractualfactor to the dyad’s cooperative capacity. We then inves-tigate the influence of cooperative capacity on the jointstrategic outcomes of the dyad and the contingency factorof the role of cooperative capacity.

The results of a dyadic survey reveal the importanceof cooperative capacity together with relation-specificinvestments. The effects of cooperative capacity andrelation-specific investments on joint strategic outcomes

are oppositely moderated by the expectation of relation-ship continuity. Expectation of an enduring relationshipcan activate the utilization of cooperative capacity andaccordingly help the dyad to achieve much greater com-petitive advantages. On the contrary, the longer the part-ners expect the relationship to last, the less useful will bethe relation-specific investments to joint performance.The empirical test also shows that resource comple-mentarity and institutional proximity both enhance thecooperative capacity. Role clarity has a positive but mar-ginal significant effect on cooperative capacity. Collec-tively, inter-organizational fitness (economic, social, andcontractual) can be an important source of the dyad’scooperative capacity.

The findings provide several implications for re-searchers and managers regarding how to strengthen thecollaborations with channel members. First, the conceptof cooperative capacity enriches resource-based view ininter-organization studies by suggesting the importanceof the dynamic capabilities which are embedded in thespecific relationship. The three components of coopera-tive capacity provide holistic insights into the complexityof inter-firm collaborative processes, as reflected in themultiple purposes of cooperation. The joint learning,changing, and creating capabilities are needed to achievethese objectives in their cooperation. This finding canhelp managers develop a set of co-working systems orroutines for the complex inter-firm collaboration. Second,our study reveals that economic, social, and contractualfitness between the dyad determines the cooperative ca-pacity of the dyad. Third, the extent to which the coopera-tive capacity and idiosyncratic investments can be servedto achieve strategic outcomes depends on each partner’sexpectation of relation continuity.

Keywords: buyer-seller relationship, cooperative ca-pacity, relation-specific investments, resourcecomplementarity, institutional distance, role ambiguity.

For further information contact:Maggie Chuoyan Dong

Department of MarketingCity University of Hong Kong

Tat Chee AvenueKowloon

Hong KongPhone: 852.3442.5754



SEEKING ENHANCED UNDERSTANDING OF MULTINATIONALSUPPLIERS’ CHOICE OF GLOBAL ACCOUNTS: AN

INTEGRATIVE FRAMEWORK ANDEMPIRICAL STUDY

Tao (Tony) Gao, Northeastern University, BostonLinda Hui Shi, University of Victoria

SUMMARY

Despite the growing popularity of global accountmanagement (GAM) practices among multinational com-panies and the recognition that GAM represents the “newfrontier of relationship management” (Yip and Madsen1996, p. 24), empirical GAM research is still limited andthe existing literature is also lacking in several importantways. In this study, we intend to rectify a key weakness inthe GAM literature by combining four previously isolatedperspectives of account selection into one integrativeframework. Specifically, we study how a host of factorsrelated to intra-organizational support, inter-organizational fit, inter-organizational power dependency,and customer characteristics jointly affect multinationalsuppliers’ choice of global accounts. We also seek torectify another limitation in the GAM literature by distin-guishing between two forms of coordination (inter-organizational and inter-country coordination) in the glo-bal account management context. We further advancenew-to-the-literature hypotheses on the curvilinear natureof eight antecedent-to-coordination relationships. Theconceptual model is tested using data collected from across-national sample of global account managers.

Inter-country coordination (ICC) is defined as theextent to which a global supplier coordinates its market-ing-mix and supply chain management activities acrossnations within its organizational boundary in order to best

serve a global account customer (Birkinshaw et al. 2001;Homburg et al. 2002). Inter-organizational coordination(IOC) refers to the extent to which a global supplierpurposively organizes joint activities at each level (i.e.,senior executive, global account manager, and opera-tional-level employee) with a global account customer(Shi et al. 2009).

Our findings show that ICC and IOC decisions havedifferent sets of key antecedents and should therefore betreated separately in practice. While intra-organizationalsupport (i.e., strategic priority and vertical involvement),goal congruency, complementarity, and customer de-mand lead to more inter-organizational coordination, it isthe combination of inter-organizational fit (i.e., goal con-gruency and resource complementarity), customer depen-dency, customer demand, and revenue share that give riseto more inter-country coordination within the boundary ofthe supplier firm. These findings highlight the divergentneeds of IOC and ICC for internal resources needs. Thefact that revenue share does not significantly influenceIOC indicates existence of motivations other than accountsize for investing in hierarchical GAM coordination. Ourfindings also show that the relationship between goalcongruency and IOC and the influence of conflict on ICCare inverted U-shaped, cautioning about the duality ofimpacts of goal congruency and conflict on GAM accountselection decisions.

REFERENCES

Birkinshaw, Julian, Omar Toulan, and David Arnold(2001), “Global Account Management in Multina-tional Corporations: Theory and Evidence,” Journalof International Business Studies, 32 (2), 231–48.

Homburg, Christian, John P. Workman, Jr., and OveJensen (2002), “A Configurational Perspective onKey Account Management,” Journal of Marketing,

66 (2), 38–60.Shi, Linda Hui, Chris J. White, Shaoming Zou, and S.

Tamer Cavusgil (2009), “Global Account Manage-ment Strategies: Drivers and Ouctomes,” Journal ofInternational Business Studies, Forthcoming.

Yip, George S. and Tammy L. Madsen (1996), “GlobalAccount Management: The New Frontier in Rela-tionship Management,” International MarketingReview, 13 (3), 24–42.


For further information contact:Tao (Tony) Gao

College of Business AdministrationNortheastern University

Boston, MA 02115Phone: 617.373.5744



THE CONTRASTING IMPACT OF RELATIONAL AND BRANDINGCAPABILITIES ON MARKETING CONTROL RIGHTS

DECISIONS WITHIN TECHNOLOGY ALLIANCES

D. Eric Boyd, James Madison University, HarrisonburgBrian P. Brown, University of Massachusetts Amherst

SUMMARY

Technology alliances represent an inter-firm strategyin which firms collaborate in developing new producttechnology (Rindfleisch and Moorman 2001). Of greatestinterest to marketers are the marketing control rights thatinvolve the management of marketing activities related toco-developed product technology. Alliance partners oftenmanage marketing control rights through a collaborativeagreement in which one partner (hereafter referred to asthe licensor) licenses a share of marketing control rights tothe other partner (hereafter referred to as the licensee). Akey decision in licensing marketing control rights involvesdeciding how restrictive or lenient the licensor should bein distributing these rights to the licensee partner. Researchto-date has investigated the distribution of marketingcontrol rights from a bargaining power and resource-based view (RBV). We argue that prior research has takentoo limited a view of the impact capabilities can have ondistribution decisions. Instead of focusing on how capa-bilities determine what a partner can do, this paper focuseson how capabilities determine what a partner will do inregard to investing in an alliance. Analysis of technologyalliance agreements in the pharmaceutical industry sup-ports the hypothesized opposing effects of relational andbranding capabilities on the allocation of marketing con-trol rights between technology alliance partners.

Technology Alliances and Marketing Control Rights

Through each partner’s contribution of resources andperformance of alliance-related activities, each partneraccrues control rights granting them with the right tomarket co-developed technology which results in market-ing decision-making authority in a technology alliancebeing shared between the partners (e.g., Hart and Moore1988). The empirical evidence to-date suggests the bar-gaining power of each partner determines the allocation ofcontrol rights between technology alliance partners whilesupport for the role of firm capabilities in determiningcontrol right distribution has failed to be supported.

This paper also explores the role of firm capabilitiesbut offers an alternative perspective regarding the role ofcapabilities in determining the distribution of marketingcontrol rights between technology alliance partners. Focus-

ing on the partner’s relational and branding capabilities,the discussion focuses on how these two capabilities canhave opposing effects on a partner’s likelihood of invest-ing in carrying out the marketing activities associated witha technology alliance. In line with prior research (e.g.,Khanna, Gulati, and Nohria 1998), high relational capa-bilities are argued to promote greater likelihood of alli-ance investment while branding capabilities are argued tohave the opposite effect.

Methodology

A sample frame including 85 technology allianceswas examined in testing the role of relational and brandingcapabilities on the distribution of marketing control rightsbetween technology alliance partners. Results from a logitmodel supported the hypothesized effects. Specifically,the possession of high relational capability led to a partnerbeing allocated a greater level of marketing control rightswhile the possession of high branding capability led to alower level of marketing control rights being allocated toa partner.

Theoretical Implications

This research draws attention to the view that market-ing capabilities influence the distribution of marketingcontrol rights within technology alliances because of theirinfluence on expected investment behavior. Thus, thispaper finds evidence supporting an alternative conceptionof capabilities and how they affect the distribution ofcontrol rights than that found in the extant literature whichhas focused on how capabilities affect what a partner “cando” versus this paper’s focus on how capabilities influ-ence what a partner “will do.” Importantly, this researchshows that marketing capabilities influence contract struc-ture but in notably distinct ways. Thus, academic research-ers must continue to conceptualize and study the variety ofmarketing capabilities in addition to their interdependen-cies.


As technology firms come to recognize the value ofmarketing as a means of differentiating their products, itis becoming more important for firms within technology


alliances to understand the implications of expanding orrestricting marketing control. This effort suggests thatmanagers inventory and evaluate their firm’s marketing

capabilities prior to establishing a technology alliance ordetermining the distribution of control rights.

REFERENCES

Hart, Oliver and John Moore (1999), “Foundations ofIncomplete Contracts,” Review of Economic Studies,66 (226), 115–38.

Khanna, Tarun, Ranjay Gulati, and Nitin Nohria (1998),“The Dynamics of Learning Alliances: Competition,

Cooperation, and Relative Scope,” Strategic Man-agement Journal, 19 (3), 193–210.

Rindfleisch, Aric and Christine Moorman (2001), “TheAcquisition and Utilization of Information in NewProduct Alliances: A Strength-of-Ties Perspective,”Journal of Marketing, 65 (2), 1–18.

For further information contact:D. Eric Boyd

College of BusinessJames Madison UniversityHarrisonburg, VA 22807

Phone: 540.568.2721Fax: 540.568.3587



REAL OPTIONS: A GOLD MINE, OR FOOL’S GOLD?

Abhijit Guha, Wayne State University, DetroitWilliam Boulding, Duke University, DurhamRichard Staelin, Duke University, Durham

SUMMARY

Real options are now an important part of marketingstrategy, and are used in cases related to new productintroductions, setting up marketing alliances etc. Forexample, a focal firm may decide to buy an option to buya new drug, developed by another firm, in the future. In thefuture, if it looks like FDA approvals will be forthcoming,the focal firm may exercise the option and buy the drug.If it looks like FDA approvals will not be forthcoming, thefocal firm may let the option lapse. Research on realoptions has hitherto focused on how best to set up realoptions, and how then to value real options. However,there is little research on how to manage real options afterthe firm has bought a real option. Specifically, in thispaper, we focus on the influence of two factors on firms’propensity to make a suitable option-exercise decision.

Real Options and Option-Structure

The first factor we considered was option-structure.Based on differences in option-structure, there are twotypes of options. Options may be structured as call options,whereby the firm does not own the asset (underlying theoption), but has the option to buy such asset in the future.Options may also be structured as put options, wherebythe firm owns the asset, but the firm has the option to sellsuch asset in the future. The literature on the endowmenteffect suggests that individuals have relatively greatervaluation for owned objects (vis. other objects), and byextension managers should be more likely to overvaluethe asset in the case of put options (vis. call options),leading to managers being more likely to make a subop-timal option-exercise decision (specifically, if the assetvalue is low, managers may overvalue the asset, and maymake the suboptimal decision of not exercising the optionto sell the asset). Further, building off the literature on theendowment effect, differences in the propensity to makea suboptimal option-exercise decision (between calloptions and put options) should be mediated by differ-ences in psychological ownership toward the asset.

We ran a series of controlled experiments, wherein(1) in a business outsourcing context, we considered thecase of a firm using real options to acquire an outsourcingunit, and (2) we used post-graduate business students asexperiment participants. First, we contrasted the case of acall option with that of a put option, and care was taken to

ensure that the cash flows in both cases were mathemati-cally equivalent. As predicted, we found that participantswere more likely to make a suboptimal exercise decisionin the case of a put option than in the case of a call option.Further, differences in the propensity to make a subopti-mal option-exercise decision were fully mediated bydifferences in levels of psychological ownership. In afollow-on experiment, which fully supported our earlierfindings, we showed that by specifically reducing psycho-logical ownership, we reduced the propensity to make asuboptimal option-exercise decision.

Our results imply that at the time of structuring realoptions, firms must trade-off the benefits of using putoptions with the potential drawback that (down the road)the firm is more likely to make suboptimal option-exercisedecisions. If firms still choose to use put options, ourresults imply that the follow-on tendency to make subop-timal option-exercise decisions is due to the endowmenteffect. Consequently, firms can devise and use suitabledebiasing mechanisms.

Apart from contributing to the literature on realoptions, this research also contributed to the work on theendowment effect. First, it illustrated how psychologicalownership mediates/explains the endowment effect. Spe-cifically, it showed a measure that fully mediates theendowment effect, an important advance for the field.Second, it extended endowment effect. This researchshowed that endowment effect (and the psychologicalownership construct) relates not just to objects owned bythe focal person, but also to objects owned by the organi-zation with which the focal person is affiliated. In theory,this suggests fertile research opportunities to investigatehow endowment effect plays out in settings like in-groupvs. out-group, collectivist vs. individualist cultures etc.

Real Options and Involvement with the Option-Pur-chase Decision

The second factor we considered was involvement inthe option-purchase decision. The real options process isa two-stage process. In the first stage, the firm makes thedecision about whether to purchase the real option. In thesecond stage, the firm makes the decision about whetherto exercise the real option. The literature on escalation ofcommitment indicates that managers are more likely topersist with a failing project (i.e., more likely to make a


suboptimal decision) if such managers were also involvedwith initiating such project. By extension, this wouldsuggest that if managers were involved with the first stageoption-purchase decision, then such managers shouldalso be more likely to make a suboptimal option-exercisedecision. Across a series of experiments, we found theexact opposite result. That is, we found that if managerswere involved with the first stage option-purchase deci-sion, then such managers were less likely (and not morelikely) to make a suboptimal option purchase decision.Further, we found that this was such managers were lesslikely to bias second stage negative information.

This research has implications for how firms shouldmanage real options after the firm has bought the option.Specifically, the literature on escalation of commitmentsuggests that the best way to decide whether to persist witha prior-launched new product is to bring in a new manager

to relook at the project. This research suggests otherwise,and suggests that there are benefits to ensuring that thesame manager makes both the first stage option-purchasedecision and the second stage option-exercise decision.

Conclusion

Real options are an integral part of marketing strat-egy. As such, real options could well be a gold mine.However, if real options are to deliver on their promise,research is needed on reducing the propensity to makesuboptimal option-exercise decisions. This yet unresolved,under-researched problem may well result in real optionsturning out to be fool’s gold. For real options to occupy itsrightful role in marketing strategy, firms need to betterunderstand how to manage a real option after havingbought the option. This work is an early effort toward thatend. References are available upon request.

For further information contact:Abhijit Guha

School of Business AdministrationWayne State University, Detroit

Prentis Building5201 Cass

Detroit, MI 48202Phone: 313.577.4501



STRATEGIC DISAGREEMENT PARADOX IN MARKETINGSTRATEGY MAKING

Kwaku Atuahene-Gima, China Europe International Business School, ChinaNamwoon Kim, Hong Kong Polytechnic University, Hong Kong

Deborah Andrus, University of Calgary, Alberta

SUMMARY

The value of strategic disagreement (also known ascognitive, task conflict) is implicitly enshrined in thediscourse in marketing strategy making, but also inmarketing’s interactions with other functions (e.g., Grif-fin and Hauser 1996; Kohli and Jaworski 1990; Ruekertand Walker 1987). For example, in the new productdevelopment process the benefits of strategic disagree-ment has led to increased emphasis on the use of cross-functional teams, involving marketing, R&D, manufac-turing and other functions (Atuahene-Gima andEvangelista 2000; Griffin and Hauser 1996; Madhavanand Grover 1998; Xie, Song, and Stringfellow 1998). Thekey insight from this literature is that strategic disagree-ment unearths the inherent value of the diversity of astrategy group to allow effective discussions and qualitydecisions and that without it the diversity of decision-making group will remain an untapped resource.

In contrast to this positive view, a less sanguine viewof strategic disagreement that has received little attentionin the marketing literature is that it may also lead tointerpersonal conflict which is injurious to effective mar-keting strategy making. Interpersonal conflict refers to thepersonality clashes, tension, anger, and hostility towardother group members in strategy making (Jehn 1997; Xie,Song, and Stringfellow 1998). The open debate over taskideas, goals, and processes that occur during the strategydevelopment could be misconstrued as personal criticismand attack that leaves people with hurt feelings.

There-in lies the paradox of strategic disagreement;the very properties of strategic disagreement that make ituseful in marketing strategy making may also lead tointerpersonal conflict that hinders the effectiveness ofmarketing strategy making. If indeed strategicdisagreement is linked with interpersonal conflict, thenthe ability of marketing managers to perform effectivelyin the marketing strategy making process is endangered.

For this reason, several scholars have emphasized that thequality of the marketing strategy process should beexpected only to the extent that the disagreements amonggroup members are successfully managed (Maltz andKohli 2000; Menon, Bharadwaj, and Howell 1996). Thisbegs some important questions: under what conditions isstrategic disagreement likely to spiral into interpersonalconflict; and what managerial intervention strategies areopen to marketing managers to prevent such an occurrence.

We address these questions in this study to makethree new contributions to the marketing literature. First,we draw attention to the paradox of strategic disagree-ment that it has a positive link with marketing strategycomprehensiveness as well as interpersonal conflict. Fur-ther, we posit that the positive relationship between stra-tegic disagreement and marketing strategy comprehen-siveness is buffered by interpersonal conflict. We providethe novel insight that a discourse of strategic disagreementdevoid of a concurrent examination of the attendantinterpersonal conflict implications may be an oversimpli-fication of the dynamics of marketing strategy making andcould lead to an overly optimistic, and perhaps evenerroneous practical implications.

As a second contribution, we suggest that strategicdisagreement is more likely to lead to interpersonal con-flict (which is detrimental to marketing strategy compre-hensiveness) when the market and technological environ-ments are uncertain. This means that previous studieswhich have extolled the importance of comprehensivemarketing strategies under uncertain environmental con-ditions may have overlooked the onerous internal man-agement dynamics that made them possible. As our thirdcontribution, we suggest that marketing managers usetrust-building and collaborative conflict resolution asintervention strategies to simultaneously achieve the ben-efits of strategic disagreement and at the same timeovercome its potential interpersonal conflict downside.References are available upon request.


For further information contact:Namwoon Kim

Department of Management and MarketingHong Kong Polytechnic University

Hung Hom, KowloonHong Kong

Phone: 852.2766.7141Fax: 852.2765.0611



THE IMPLICATIONS OF KNOWLEDGE GRAFTING FORMARKETING STRATEGY

Sreedhar Madhavaram, Cleveland State UniversityVishag Badrinarayanan, Texas State University, San Marcos

Robert E. McDonald, Texas Tech University, Lubbock

SUMMARY

The concept of grafting has been around for centuriesand has benefitted the fields of horticulture, medicine, andengineering. In fact, grafting has been credited withenormous successes in these three fields. Eventually, theconcept of grafting, under the rubric of knowledge graft-ing, made its way to the organizational learning literature.However, the potential of knowledge grafting for busi-ness and marketing strategy remains unexplored. In fact,since the introduction of the grafting concept into thestrategy literature in the 1990s, there is no systematicresearch into knowledge grafting. Before we go anyfurther into the specifics of knowledge grafting, it isimportant to clarify what one means by knowledge graft-ing. Huber’s (1991) conceptualization of grafting in-volves organizations increasing their store of knowledgeby formally acquiring employees, firms, and/or by devel-oping long-term alliances with other organizations thatpossess information not previously available within theorganization. Furthermore, Huber (1991) predicts thatgrafting will become a more frequently used approach fororganizations to acquire quickly knowledge that is new tothem. This acquisition of knowledge is often faster thanlearning by experience and more complete than learningthrough imitation (Huber 1991; Osland and Yaprak 1995).

Although Huber’s (1991) concept of grafting is quitesignificant, it is somewhat limiting in terms of itsconceptualization. Huber’s (1991) concept of graftingconcerns the acquisition of knowledge not previouslypossessed by firms. However, the concept of grafting inhorticulture concerns a process that, in addition toacquisition, focuses on union, synergistic growth, andsuperior outcomes. That is, grafting in horticulture isoften done in the quest for superior outcomes andacquisition is only the beginning. Traditionally, inhorticulture, grafting involves preparing and placingtogether plant parts so they may grow together (Garner1988). Lee and Sears (1994) suggest that grafting is atechnique that makes it possible to add one variety of a treeto another. Spurgeon (1994) discusses grafting as a craft

that helps in propagating superior trees. The graft must becarefully prepared, joined, and nurtured to be effective.Hence, the acquisition of the entities necessary for thegraft, the union of the grafted entities, and the subsequentsynergistic growth that, in turn, results in superior outcomesfall under the rubric of grafting. Paralleling this graftingprocess, knowledge acquisition, knowledge integration,and synergistic knowledge growth that, in turn, results insuperior knowledge outcomes fall under the rubric ofknowledge grafting.

Drawing from the concept of grafting from horticul-ture, we define knowledge grafting as “bringing one ormore individuals, groups, and/or organizations togetherfor the purpose of integrating the new and existing knowl-edge, in order to produce superior values or outcomesthrough synergistic knowledge growth.” We conceptual-ize knowledge grafting beyond mere acquisition of exter-nally developed knowledge. We propose that graftingalso involves what one does before, during, and after theacquisition of externally developed knowledge, to ensurethat the organization becomes more effective or efficient,and perhaps, gains competitive advantages. We furthercontend that the organization’s overall emphasis on, andregular/continuing efforts toward, knowledge graftingcould potentially provide the basis for sustaining competi-tive advantages. Consequently, knowledge grafting canbe useful for achieving strategic changes in firms (Nag,Corley, and Gioia 2007).

In this paper, we specifically explore the implicationsof knowledge grafting for marketing strategy. First, weprovide a brief overview of knowledge grafting. In thissection, we introduce the different types of knowledgegrafting. Second, we discuss the specific types of knowl-edge grafting in how they relate to marketing strategy.Here, we also discuss the importance of marketing func-tion for the firm and its role in relevant strategic changesthat the firms can bring about. Finally, we conclude witha discussion of our paper’s contributions to theory andpractice. References are available upon request.


For further information contact:Sreedhar Madhavaram

Department of MarketingNance College of Business Administration

Cleveland State University2121 Euclid Avenue

Cleveland, OH 44115Phone: 216.687.3797



WHAT’S IN A NAME? AN ANALYSIS OF THE STRATEGIC BEHAVIOROF FAMILY FIRMS

Saim Kashmiri, University of Texas at AustinVijay Mahajan, University of Texas at Austin

SUMMARY

Family firms play a significant role in the U.S.economy making up about one-third of the firms listed onthe S&P 500 and Fortune 500 indices (Anderson and Reeb2003) and employing about 60 percent of the U.S.workforce (Astrachan and Shanker 2003). Surprisingly,in spite of the importance of family firms to the U.S.economy, marketing literature on family firms is almostnon-existent. The scant literature that does exist, prima-rily in the fields of management and finance assumesfamily firms to be a homogenous group displaying similarkinds of behavior; it does not explore differences instrategic behavior between different types of family firms.Do all family firms behave similarly or are there differenttypes of family firms that behave differently? The authorstry to address this question by focusing on the linkbetween family firms’ names and firm behavior. Somefamily firms such as Dell, Ford, Walgreen, and Marriottuse their founding family’s name as part of their firmname; others like Nike, FedEx, Progressive, and Gap, donot. The authors refer to the former group as family-named family firms (FN family firms) and the latter asnon-family-named family firms (NFN family firms). Thisresearch explores differences in strategic behavior be-tween these two types of family firms. It also investigatesif FN family firms perform better than NFN family firms,and if their superior performance is mediated by differ-ences in their strategic behavior.

Family-named family firms are expected to behavedifferently than non-family-named family firms becauseof the greater visibility of the family-firm linkage in theeyes of the family-named firms’ consumers. A firm namelike Dell Inc. that includes the family name explicitlycommunicates that the firm is a family firm and puts thefounding family in the spotlight. FN family firms maytherefore find it harder separating the firm’s reputationfrom the family’s reputation. One would expect FN fam-ily firms, then, to be relatively more concerned aboutprotecting their firm’s reputation. This greater emphasisby FN family firms on protecting their reputation may leadthem to behave differently from NFN family firms acrossa number of strategic factors such as the level of custom-ers’ voice in the boardroom (i.e., the choice of includingor not including a chief marketing officer in the topmanagement team), net differentiation emphasis (i.e.,relative emphasis on differentiation versus innovation),

and corporate citizenship (i.e., level of corporate socialresponsibility).

Family-named family firms, given the greater visibil-ity of their family-firm link, are expected to place a greateremphasis on ensuring that their customers have a favor-able image of the firm and a positive perception of thecorporate brand. This makes it especially important forFN family firms to have a customer advocate in the formof a chief marketing officer (CMO) present in their topmanagement team. FN family firms are also likely to havea greater net differentiation emphasis than NFN familyfirms because innovation is inherently risky and FNfamily firms given the greater visibility of the family-firmlink may be more averse to taking the risk of innovation.A greater visibility of the family-firm linkage for FNfamily firms also suggests that these firms may be morelikely to safeguard their firms’ reputation. Hence FNfamily firms are likely to have more social strengths (i.e.,launch more positive social initiatives) and to have fewersocial weaknesses (i.e., get involved in fewer negativesocial activities). Finally, to the extent that prior researchhas shown firms’ greater corporate citizenship levels toimprove their performance (e.g., Luo and Bhattacharya2006) and that the stock market reacts favorably when afirm increases its net differentiation emphasis (Mizik andJacobson 2003), the authors hypothesize that FN familyfirms perform better than NFN family firms and that FNfamily firms’ better performance is mediated by (a) theirrelatively more social strengths (b) their relatively fewersocial weaknesses, and (c) their relatively higher netdifferentiation emphasis.

Findings based on a multi-industry sample of 130publicly listed U.S. family firms over a five-year period(2002–2006) broadly support the authors’ hypotheses.These results show that compared to NFN family firms,FN family firms have significantly higher levels of corpo-rate citizenship and representation of customers’ voice(i.e., presence of a chief marketing officer) in the topmanagement team. FN family firms also have a higher netdifferentiation emphasis (i.e., relative emphasis on valueappropriation versus value creation) compared to NFNfamily firms. Furthermore, FN family firms perform bet-ter than NFN family firms and their superior performanceis partially mediated by their higher corporate citizenshiplevels and higher net differentiation emphasis.


This research has a number of theoretical and practi-cal implications. With regards theoretical contributions,this research challenges the extent assumption that familyfirms are a homogenous group displaying similar kinds ofbehavior. This research also adds to the stream of reputa-tion and identity by providing an understanding of howdifferent degrees of concern for owners’ reputation and

identity can motivate firms to behave differently in strate-gic terms. With regards practical contributions, a keymessage the authors give to family firm practitioners isthat regardless of the way the firm is named, it pays toavoid social transgressions and to have an increasedemphasis on differentiation versus innovation. Refer-ences are available upon request.

For further information contact:Saim Kashmiri

McCombs School of BusinessUniversity of Texas at Austin1 University Station, B6200

Austin, TX 78712Phone: 512.287.1830



THE EFFECTS OF NEGATIVE EMOTIONS ON AVOIDANCE BEHAVIORIN SERVICESCAPE

Jung Ok Jeon, Pukyong National University, Korea.Hee Young Han, Pukyong National University, Korea

SUMMARY

Recent research has shown that emotions experi-enced in the store environment significantly affect cus-tomers’ shopping behavior from the environmental psy-chology perspective (Baker et al. 1992; Kaltcheva andWeitz 2006; Sherman et al. 1997; Sweeney and Wyber2002). Although many studies have shown that physicalenvironments’ influence on shopping behavior mediatescustomers’ emotions in-store, there is very little empiricalstudy to examine the relationship between negative emo-tions and avoidance behavior. In addition, although themanagement of negative emotions, which cause avoid-ance behavior, is a very important part, most of previousresearch has evaluated negative emotions generally insteadof presenting concrete factors of negative emotions inservice environments.

Therefore, the purposes of this study are as follows:First, the authors classify the dimensions of departmentstores’ servicescape, which is extended on the basis ofBaker’s (1986) model. The authors adopt environmentalpsychology literature to introduce a three-dimensionalframework that effectively discusses the “physical envi-ronment” on the basis of ambient, design, and socialfactors. In addition, the category of “social factor” isdivided by two separate factors: employee factor andother customer factors. Second, the authors investigatewhich dimensions of negative emotions customers feel ina servicescape and evaluate them through an exploratorystudy. Third, based on an exploratory study, the authorsestablish a research model regarding the influential rela-tionship among a department store’s servicescape, nega-tive emotions, and customer’s avoidance behavior.

The exploratory study also derives dimensions ofnegative emotions in servicescapes and reveals threedimensions of negative emotions, including tightness,discomfort, and aversion. The data, collected from a

sample 344 shoppers in Korea, indicate that negativefactors of a servicescape have a pronounced effect onnegative emotions and that negative emotions serve ascritical mediators in the servicescape – avoidance behav-ior relationship.

SPSS 12.0 and AMOS 4.0 statistic packages wereadopted for hypotheses testing. The results of this studyare as follows. First, two dimensions of servicescapes hadrelatively different influences on the three dimensions ofnegative emotions. More specifically, of the three dimen-sions in the physical environment, design had no signifi-cant impact on any other negative emotions, while ambi-ent and facility factors positively influenced the tightnessand discomfort emotions. Furthermore, of the two dimen-sions in the social environment, employee and othercustomer factors had an influence on tightness, discom-fort, and aversion emotions. Thus, the findings show thatnegative emotions were induced by negative environmen-tal factors in the servicescape. Second, there were differ-ences in negative emotional responses, which influencedthe type of environmental stimulus. There were alsodifferences in behavioral responses, which were influ-enced by the type of negative emotions. It is supposed thatwhen customers felt tight, they did exploration avoidancebehavior. If customers felt discomfort in a servicescape,they did both exploration and communication avoidancebehaviors. Finally, when customers felt averse in aservicescape, they did communication avoidance behav-ior. In addition, there was empirical support for positiverelationships between tightness and exploration avoid-ance, discomfort and search avoidance, and aversion andcommunication avoidance.

In conclusion, this study provides the crucial insightinto the understanding of customers’ avoidance behaviorwith empirical results and emphasizes that firms shouldcare about environmental elements, which stimulate nega-tive emotions in servicescapes.

REFERENCES

Baker, Julie, Michael Levy, and Dhruv Grewal (1992),“An Experimental Approach to Marketing RetailStore Environmental Decisions,” Journal of Retail-ing, 68 (Winter), 445–60.

Kaltcheva, V.D. and B.A. Weitz (2006), “When Should aRetailer Create an Exciting Store Environment?”Journal of Marketing, 70 (1), 107–18.

Sherman, Elaine, Anil Mathur, and Ruth B. Smith (1997),“Store Environment an Consumer Purchase Behav-ior: Mediating Role of Consumer Emotions,” Psy-


chology and Marketing, 14 (July), 361–78.Sweeney, Jillian C. and Fiona Wyber (2002), “The Role

of Cognitions and Emotions in the Music-Approach-

Avoidance Behavior Relationship,” Journal of Ser-vices Marketing, 16 (March), 51–69.

For further information contact:Hee young Han

Pukyong National University8421, Daeyeon 3-dong,

Nam-gu, BusanKorea 608–737

Phone: 82.51.629.5730Fax: 82.51.629.5720



FROM HOLLYWOOD TO BROADWAY: A FIELD STUDY OF THEEFFECTIVENESS OF PRODUCT PLACEMENTS

Rick T. Wilson, Hofstra University, HempsteadBrian D. Till, Saint Louis University

SUMMARY

Much of the research pertaining to product place-ments has been laboratory based and has relied on the useof partial movie or television clips (Balasubramanian,Karrh, and Patwardhan 2006). Furthermore, most studieson product placements have focused on the film andtelevision media at the expense of alternative entertain-ment media. A notable exception is video games (e.g.,Nelson 2002). We conduct two field studies to assess thecognitive responses (in the present research, recall) toproduct placements. In the first study, we assess thememory of product placements within four movies cover-ing a variety of genres (comedy, drama, historical action,and biography) using moviegoers in a movie theater. Thesecond study assesses the memory of product placementsin a Broadway musical using theatergoers in a New YorkBroadway theater.

Hypotheses

Based on an extensive literature review, we presentand test the following hypotheses.

H1: Brand recall will be highest for combined audio andvisual placements, next highest for audio only place-ments, and lowest for visual only placements.

H2: Brand recall will be higher for placements that appearin the first-half of entertainment media and lower forbrands that appear in the second-half of entertain-ment media.

H3: Brand recall will be equal for positively-, negatively-,and neutrally-portrayed placements.

H4: Brand recall will be higher for extended on screenplacements as compared to fleeting placements.

H5: Brand recall will be higher for prominent placementsas compared to subtle placements.

H6: Brand recall will be higher for placements in whichan actor interacts with the brand as compared to thosein which an actor does not interact with the brand.

H7: Brand recall will be higher for placements with twoor more verbal mentions, next highest for placements

with one verbal mention, and the lowest for place-ments with no verbal mentions.

Methods

For Study One, data were collected in a Midwesttheater over the course of 17 days. A total of 1,176 surveyswere handed out equally divided between aided andunaided recall. An overall 52 percent response rate wasrealized. For Study Two, data were collected after oneshow of a popular comedic Broadway show in New YorkCity. A total of 114 surveys were completed.

Results

Data were tested using t-tests and one-way analysesof variance. The results for Study One for product place-ments within four movies are as follows: Combined audioplacements generated higher rates of aided and unaidedrecall than either visual only or audio only placements. Nodifferences were found between audio and visual onlyplacements. Extended placements, prominent placements,and placements in which the actor interacted with thebrand each generated higher rates of aided and unaidedrecall than did fleeting placements, subtle placements,and placements in which the actor did not interact with thebrand. In addition, product placements with two or moreverbal mentions in a movie had significantly greater aidedand unaided recall than those with one or no verbalmentions. The nature of the portrayal or whether theplacement appeared in the first or second half of the moviehad no effect on either form of recall.

For Study Two, product placements within a Broad-way show, the results are as follows: Combined audioplacements generated higher rates of aided and unaidedrecall than either visual only or audio only placements. Nodifferences were found between audio and visual onlyplacements. Prominent displays and placements in whichthe actor interacted with the brand each generated higherrates of aided and unaided recall than did subtle place-ments and placements in which the actor did not interactwith the brand. In addition, product placements with twoor more verbal mentions during the show had higher aidedrecall than those with one or no verbal mentions (unaidedrecall was not significant). The nature of the portrayal,length of the portrayal, and whether the placement appearedin the first or second half of the Broadway show had noeffect on either form of recall.


Conclusion

Through our two studies we have attempted to answerother researchers’ calls for more non-laboratory-orientedresearch in an effort to better understand the real-worldeffects of product placements (Balasubramanian, Karrh,and Patwardhan 2006). In doing so, we have also addressedanother gap in the literature by assessing the effectivenessof product placements in another entertainment mediaformat – live musical theater or Broadway. Our field

research has confirmed most of the propositions putforward and tested by other researchers in a laboratorysetting. However, several discrepancies remainunanswered between laboratory and field research, whileother discrepancies have been discovered between differententertainment media. It is with these results in mind thatwe encourage other researchers to undertake additionalfield studies and to conduct them across other types ofentertainment media. References are available uponrequest.

For further information contact:Rick T. Wilson

Hofstra University124 Weller Hall

Hempstead, NY 10036Phone: 516.463.5175



DOES CONSUMPTION KNOW NO LIMITS? THE MODERATING ROLEOF SATIATION IN DETERMINING CUSTOMER REPURCHASE

Glenn B. Voss, Southern Methodist University, DallasAndrea L. Godfrey, University of California, Riverside

Kathleen Seiders, Boston College

SUMMARY

Despite widespread recognition that customerbehavior is subject to myriad contingencies, there islimited conceptual or empirical research that addressesmoderating influences on repurchase behavior. In thisstudy, the authors theoretically explicate two types ofmoderating effects, and they develop theoretical argumentsfor predicting systematic differences in moderating effectsacross different purchase categories. When satiation effectsare weak, moderating variables can complementsatisfaction effects to increase repurchase. When satiationeffects are strong, habituation can lead to substitutionbetween moderating effects and satisfaction. An empirical

test combining survey and longitudinal purchase datafrom two distinct purchase categories –fashion appareland automobile service – provides a remarkable degree ofsupport for the hypotheses. In the fashion apparel purchasecategory, six customer, relational, and marketplacecharacteristics exert complementary effects on repurchase.In the automobile service purchase category, involvement,income, relationship-building programs, and convenienceexert substitution effects on the satisfaction-repurchaserelationship. The results offer new theoretical insights andprovide substantive guidance for managers in effectivelyallocating resources to initiatives that complement orsubstitute for customer satisfaction to increase repurchase.

For further information contact:Andrea Godfrey

School of Business AdministrationUniversity of California, Riverside

233 Anderson HallRiverside, CA 92521Phone: 951.827.6192



Abdul-Muhmin, Alhassan G. 147Alabdi, Yaser F. 99Andrews, Melinda 123Andrus, Deborah 389Angulo, Fernando 266Areni, Charles S. 344Arkes, Hal R. 53Ashley, Christy 60Athaide, Gerard A. 228Atuahene-Gima, Kwaku 389Austermann, Dominik 373Avlonitis, George J. 1, 91Backhaus, Christof 336Badrinarayanan, Vishag 391Baker, Thomas L. 10Ball, A. Dwayne 64Barbulescu, Adina 103Bartholomew, Darrell 361Bauer, Hans H. 334Baumann, Jasmin 352Becerra, Enrique P. 213Bello, Daniel 266Belz, Frank-Martin 377Bernard, Elena Kiryanova 75Blatter, David 342Blut, Markus 143, 202Bodderas, Mareike 129Bolton, Ruth 97Bornemann, Torsten 188Boulding, William 387Boyd, D. Eric 385Brettel, Malte 95, 207, 373Brocato, E. Deanne 60Brock, Christian 202Brown, Brian P. 385Brunk, Katja H. 32Bucerius, Judith 54Burns, Alvin C. 217Burton, Scot 41Bush, Ronald F. 217Buys, Sebastian 121Caemmerer, Barbara 149Calter, Dayenne 304Carlson, Brad D. 111Cavusgil, S. Tamer 27

AUTHOR INDEX

Chandy, Rajesh 226Chang, En-Chung 145Chang, Wei-Lun 260Chapa, Sindy 213Chaudhuri, Arjun 365Chen, Cuiping 326Chen, Jie 145Chiba, Takahiro 19Cho, Seong-Do 28Choi, Pilsik 234Christensen, Glenn L. 105Cobbs, Joe 338Codita, Roxana 377Comer, Lucette B. 6Coulter, Keith S. 234Coulter, Robin 252Crittenden, Victoria L. 263Crittenden, William F. 263Cronin, Jr., J. Joseph 194Cummins, Shannon 36Daun, Winfried 125de Almeida, Stefânia Ordovás 293de Jong, Ad 109de Ruyter, Ko 109Dellande, Stephanie 371DeRosia, Eric D. 105Dewitt, Tom 111Dholakia, Utpal 293Donavan, D. Todd 111Dong, Maggie Chuoyan 382Drollinger, Tanya 6Eccles, Sue 275Eggert, Andreas 324Eilert, A. Meike 306Eisend, Martin 211Eiting, Alexander 380Elbeck, Matt 222Engelen, Andreas 95, 373Enke, Margit 30Erenburg, Grigori 348Ertekin, Selcuk 161Evanschitzky, Heiner 149, 202Evirgen, Cuneyt 27Fabrize, Robert O. 358Falk, Tomas 334


Fang, Xiang 361Fassnacht, Martin 192Feng, Shan 190Finnegan, Carol 27Fischer, Peter Mathias 186Fitzgerald, Jennifer 54Floh, Arne 73Flynn, Leisa Reinecke 69Folse, Judith Anne Garretson 43Gabisch, Jason 81Gammoh, Bashar S. 363Gao, Shanxing 39Gao, Tao (Tony) 383Garnefeld, Ina 324Geigenmueller, Anja 30Gencturk, Esra F. 268Germann, Frank 334Gilliam, David A. 71Gleim, Mark R. 194Go, Frank 304Godfrey, Andrea L. 399Goldsmith, Ronald E. 69, 123Gounaris, Spiros P. 91Govind, Rahul 107Grabner-Kräuter, Sonja 313Grégoire, Yany 115Groza, Mark 338Guha, Abhijit 387Gustafsson, Anders 97Haas, Alexander 332Haenlein, Michael 79Halbrook, Stephen 213Hammerschmidt, Maik 334Han, Hee Young 395Hanna, Richard 263Hassan, Louise May 151, 198Helm, Sabrina 322Hennigs, Nadine 141, 310Herhausen, Dennis 350Herm, Steffen 67Herrmann, Lars 117Herzog, Walter 186Hess, Silke 54Hild, Sebastian 95Hill, Ronald Paul 45Hoffmann, Stefan 159, 250, 360Hofstetter, Reto 342

Hogreve, Jens 324Hohl, Nikolaus 205Homburg, Christian 157, 188Huber, Adele J. 332Huggins, Kyle A. 330Hughes, Mathew 232Hughes, Paul 232Ingene, Charles A. 107Jenkins, Rebecca 275Jeon, Jung Ok 395Jiang, Jianfeng 107Jiang, Xu 39Jiménez, Fernando R. 361Johansson, Johny K. 23Johnson, Jean L. 115Jung, Gangok 28Kandemir, Destan 268Kang, Jikyeong 99Kaplan, Andreas M. 79Karniouchina, Ekaterina V. 348Kashmiri, Saim 393Kassubek, Martin 141Katsikeas, Constantine 312Kawakami, Tomoko 93Kellarism, James J. 224Kim, Daekwan 28Kim, Kyoungeun 28Kim, Namwoon 389Kirpalani, Nicole 182Kishiya, Kazuhiro 93Koh, Anthony A. 363Kohli, Ajay 77, 103Koritos, Christos D. 91Krohmer, Harley 65, 215Kuester, Sabine 54, 121Kyriakopoulos, Kyriakos 232Labrecque, Lauren I. 258Langerak, Fred 340Lawson, Stephanie J. 194Le Bon, Joel 354Le Meunier-FitzHugh, Kenneth 352Le Meunier-FitzHugh, Leslie 352Lee, Olivia F. 230Lee, Thomas R. 105Lee, Yun 58, 248, 367Lefroy, Kathryn 308Leischnig, Alexander 30


Leschnikowski, Katja 215Li, Cong 265, 274Li, Ning 153Liebermann, Susanne C. 360Lin, Xia-Liang 260Liu, Sandra S. 145Loulakis, Andrea 45Lucas, Christian 380Mac, Lancy 173Madhavaram, Sreedhar 127, 391Madupu, Vivek 155Mahadevan, Jochen 192Mahajan, Vijay 393Mahr, Dominik 37Mai, Robert 159, 250Majid, Kashef A. 23Malaer, Lucia 65Malms, Oliver 356Matsui, Kenji 328Mazzon, José Afonso 293McColl-Kennedy, Janet R. 200McDonald, Robert E. 127, 391Meuter, Matthew L. 230Meyer, Tracy 10Michaelidou, Nina 198Michaelis, Manuel 336Miller, Klaus 342Milne, George R. 258Min, Kyeong Sam 53Minor-Cooley, Delonia 155Mishra, Debi P. 375Moeller, Sabine 369Molesworth, Mike 275Möller, Jana 67Möller, Lise 34Montoya, Mitzi M. 346Mort, Gillian Sullivan 127Muenkhoff, Eva 324Murphy, William H. 153Naskrent, Julia 205Neto, Hugo Müller 293Nijssen, Edwin 340Nyer, Prashanth U. 371Nyffenegger, Bettina 65O’Brien, Matthew 326Ono, Akinori 19Ordanini, Andrea 340

Oyedele, Adesegun 256Page, Albert L. 39Page, Christine 119Panagopoulos, Nikolaos G. 1Pankalla, Lars 141Parry, Mark E. 93Paswan, Audhesh K. 25Patel, Chirag 226Patterson, Paul G. 200Penz, Elfriede 254Peterson, Robert A. 263Pillai, Kishore Gopalakrishna 312Pocsay, Sandra 205Prabhu, Jaideep 226Price, Linda 252Prior, Diego 266Raggio, Randle D. 43Rahman, Kaleel 344Rapp, Adam 354Reeh, Marc-Oliver 310Reimann, Martin 272Reinecke, Sven 186Rekom, Johan van 304Rialp, Josep 266Rindfleisch, Aric 37Rivadeneyra, Raul 226Robinson, Stacey G. 262Rosen, Stefanie 306Ross, Jr., William T. 334Roy, Subhadip 363Saqib, Najam 56Scammon, Debra L. 196Schaefers, Tobias 369Schepers, Jeroen 109Schilke, Oliver 207, 272Schmidt, Svenja 236Schmitz, Christian 3, 356Schögel, Marcus 350Schramm-Klein, Hanna 205Schwarz, Uta 360Schwass, Joachim 34Seiders, Kathleen 399Seo, Joon Yong 196Shen, Anyuan 64Sheth, Jagdish 73Shi, Linda Hui 383Shiu, Edward Man Kee 151, 198


Shu, Chengli 39, 62Simpson, Dayna 308Sinaga, Pan Dapotan 28Singh, Rakesh 184Singh, Ramendra 184Sisodiya, Sanjay R. 115Six, Bjoern 113Slotegraaf, Rebecca J. 37Smirnova, Maria 250Smit, Willem 34Smith, Gerald 366Smith, Krystina 119Sohi, Ravipreet S. 36Sorescu, Alina 226Stacey, E. Craig 69Staelin, Richard 387Stock-Homburg, Ruth Maria 113, 117Stöttinger, Barbara 254Strizhakova, Yuliya 252Suh, Taewon 271Sun, Qin 25Surachartkumtonkun, Jiraporn 200Suri, Rajneesh 190Szymanski, David M. 21Tang, Chuanyi 326Tangari, Andrea Heintz 41Tarasi, Crina 97Thomas, Jacquelyn 272Tieslau, Margie 25Till, Brian D. 397Tomczak, Torsten 125Totzek, Dirk 157Townsend, Janell D. 346Tran, Trang Phuc 209

Troilo, Gabriele 77Troy, Lisa C. 21Tsarenko, Yelena 308Tseng, Chung-Hui 12Uslay, Can 348Varelmann, Dieter 310Varki, Sajeev 56, 366Veeck, Ann 217Vlasic, Goran 77Voorhees, Clay 202Voss, Glenn B. 399Voss, Kevin E. 71Walker, Beth 97Walsh, Gianfranco 151Wang, Chih-Ping 285Wang, Liz C. 8Wang, Wenzhi 230Wang, Xiuli 265Weathers, Peter “Danny” 217Weerawardena, Jay 127Wentzel, Daniel 125White, Darin W. 330Wiedenfels, Gunnar 207Wiedmann, Klaus-Peter 141, 310Willach, Anne 322Wilson, Rick T. 397Woisetschläger, David M. 143, 336, 380Wolter, Jeremy S. 204Xie, Yu Henry 271Zentes, Joachim 205Zhang, Haisu 62Zhang, Jason Q. 228Zhu, Yimin 100

marketing theory and applications · charles a. ingene, jianfeng jiang, rahul govind 107 team...

Documents