no.of pages17 article in press · possibility to reserve products online for purchase in the...

R

A

crcspiai©

K

Bad

ja

h0

ARTICLE IN PRESS+ModelETAIL-551; No. of Pages 17

Journal of Retailing xxx (xxx, 2015) xxx–xxx

Integrating Bricks with Clicks: Retailer-Level and Channel-Level Outcomesof Online–Offline Channel Integration

Dennis Herhausen a,∗, Jochen Binder b,1, Marcus Schoegel a,2, Andreas Herrmann c,3

a Institute of Marketing, University of St. Gallen, Dufourstr. 40a, 9000 St. Gallen, Switzerlandb GIM Gesellschaft für Innovative Marktforschung, Schumannstrasse 18, 10117 Berlin, Germany

c Center for Customer Insight, University of St. Gallen, Bahnhofstrasse 8, 9000 St. Gallen, Switzerland

bstract

This research examines the impact of online–offline channel integration (OI), defined as integrating access to and knowledge about the offlinehannel into an online channel. Although channel integration has been acknowledged as a promising strategy for retailers, its effects on customereactions toward retailers and across different channels remain unclear. Drawing on technology adoption research and diffusion theory, the authorsonceptualize a theoretical model where perceived service quality and perceived risk of the Internet store mediate the impact of OI while the Internethopping experience of customers moderates the impact of OI. The authors then test for the indirect, conditional effects of OI on search intentions,urchase intentions and willingness to pay. Importantly, they differentiate between retailer-level and channel-level effects, thereby controlling fornterdependencies between different channels. The results of three studies provide converging evidence and show that OI leads to a competitivedvantage and channel synergies rather than channel cannibalization. These findings have direct implications for marketers and retailers interested
n understanding whether and how integrating different channels affects customer outcomes.
2015 New York University. Published by Elsevier Inc. All rights reserved.

eywords: Multi-channel management; Omnichannel retailing; Channel integration; Channel synergies; Channel cannibalization; Willingness to pay

Pitricatsn

Introduction

“There was a time when the online and offline businesseswere viewed as being different. Now we are realizing that weactually have a physical advantage thanks to our thousandsof stores, and we can use it to become Number 1 online.”

Raul Vasquez, Walmart.com Chief Executive (cited inBustillo and Fowler 2009)

Retail firms with physical stores, such as Walmart, Macy’s,

Please cite this article in press as: Herhausen, Dennis, et al, Integrating

Online–Offline Channel Integration, Journal of Retailing (xxx, 2015), http

est Buy and Staples have begun to realize that they may have physical advantage over purely online players. However, evi-ence for the success of multi-channel retailers remains scarce.

∗ Corresponding author. Tel.: +41 71 224 2859.E-mail addresses: [email protected] (D. Herhausen),

[email protected] (J. Binder), [email protected] (M. Schoegel),[email protected] (A. Herrmann).1 Tel.: +49 30 240 009 11.2 Tel.: +41 71 224 2833.3 Tel.: +41 71 224 2131.

as

tdMbIti

ttp://dx.doi.org/10.1016/j.jretai.2014.12.009022-4359/© 2015 New York University. Published by Elsevier Inc. All rights reserv

urely online players, such as Zappos and Amazon, tend to dom-nate the market for certain product categories and outperformheir brick-and-click competitors, whereas some multi-channeletail firms with physical stores still struggle to answer themportant yet unresolved question of whether they can createompetitive advantage from a multi-channel strategy (Neslinnd Shankar 2009). We believe that the reason is often due tohe lack of integration between a firm’s Internet and physicaltores. Firms could provide, for instance, in-store online termi-als in a physical store or physical store locators and assortmentvailability information on the Internet. However, many firmseem either unable or unwilling to provide such services.

Traditionally, most multi-channel retailers have siloed struc-ures, where the physical store division and the Internet storeivision operate independently of each other (Gallino andoreno 2014; Rigby 2011). Consequently, real collaboration

etween retailers’ physical stores and digital stores remains rare.

Bricks with Clicks: Retailer-Level and Channel-Level Outcomes of://dx.doi.org/10.1016/j.jretai.2014.12.009

ndeed, a recent survey among 80 retail organizations found thathe majority of multi-channel retailers have siloed their capabil-ties and systems (Aberdeen Group 2012). Other studies have

ed.

dx.doi.org/10.1016/j.jretai.2014.12.009


mailto:[email protected]





ARTICLE IN PRESS+ModelRETAIL-551; No. of Pages 17

2 f Reta

fyemBg

idn(fodsibiiapsciddctcmo2

teactciGa2ttetaba

sr2aoa

piaaatr

•

•

•

roicearntR

D

fTvpvItiJootnegative effect of non-availability for physical store customers(Bendoly et al. 2005), assisted online terminals can complementpersonal service (Glushko and Tabas 2009), and Internet banking

D. Herhausen et al. / Journal o

ound similar results, indicating that channel integration has notet been achieved (Accenture 2010). However, some businessxperts foresee that channel integration will soon become theain focus of retailers and channel managers (Google 2011).ooz & Company (2012, p. 1) even stated, “cross-channel inte-ration is fast becoming a competitive necessity”.

Despite this optimistic appraisal of channel integration,mportant arguments support and oppose the integration ofifferent channels. Promoters of integration state that chan-el integration could enrich the customer value propositionGallino and Moreno 2014) or prevent customer confusion andrustration (Gulati and Garino 2000). However, the advantagesffered by channel integration are not without risks and potentialownsides. Integrating different channels may increase researchhopping – defined as the propensity of consumers to searchn one channel and then purchase through another channel –y more than 25 percent as it reduces channel-specific lock-n effects (Verhoef, Neslin, and Vroomen 2007). This findings critical given that showrooming – whereby customers evalu-te products at one store but buy them elsewhere – is a majorroblem for physical stores (Zimmerman 2012) and Internettores (Ryan 2013). In addition, a firm’s integration activitiesould be counteracted by the inherently missing complementar-ty between the firm’s distribution channels, which due to theirifferent characteristics, such as price and assortment strategies,o not enable achieving synergies (Zhang et al. 2010). Finally,hannel integration may be either a zero-sum game where advan-ages in one channel are offset by disadvantages in anotherhannel (“dissynergies” or “cannibalization”; Falk et al. 2007) oray even harm firms in cases of negative spill-over effects from

ne channel to another (van Birgelen, de Jong, and de Ruyter006).

Thus, channel integration can present both opportunities andhreats to firms, namely, channel integration can be performancenhancing and performance destroying. In light of the costsssociated with channel integration, further insights into theonsequences of channel integration are highly relevant in prac-ical and theoretical terms. In this paper, we explore whetherustomers value integrated channels and whether firms integrat-ng Internet and physical stores benefit from their investments.iven that Internet search and store purchase is acknowledged

s the most common form of research shopping (DoubleClick004; Verhoef, Neslin, and Vroomen 2007), the importance ofhe Internet as an information source (Pauwels et al. 2011), andhe relative lack of research (Bendoly et al. 2005 is a notablexception), our study focuses on online–offline channel integra-ion (OI). In this context, OI is defined as integrating access tond knowledge about the offline channel into an online channely, for instance, providing a store locator or store assortmentvailability information via the Internet store.

To gain insights into the current state of OI, we manuallyearched the websites of 62 leading European multi-channeletailers of fashion, electronics and household goods from March



013 to April 2013. We found that the most common integrationctivities are efficient dealer search (introduced by 19 percentf leading multi-channel retailers), the ability to check productvailability in the physical store via the Internet (34 percent), the

paV

iling xxx (xxx, 2015) xxx–xxx

ossibility to reserve products online for purchase in the phys-cal store (25 percent), and to return products purchased onlinet the offline store (15 percent). The relatively low percent-ge of multi-channel retailers that use online–offline integrationctivities confirms the notion that OI is a recent phenomenonhat deserves more detailed examination. Focusing on customereactions to OI, we address the following questions:

Do customers value an integrated Internet store? Doesonline–offline channel integration lead to a competitiveadvantage for a focal retailer?

How does online–offline channel integration affect customersearch intention, purchase intention and willingness to payin both Internet and physical stores?4 How strong are thecannibalization effects between the two channels?

Does online–offline channel integration have the same effecton all customers? How do individual differences affect theoutcome of online–offline channel integration?

To answer these questions, we used technology adoptionesearch and diffusion theory to explain the potential effectsf channel integration. We conducted three empirical studies tonvestigate how OI affects customer behavior. Importantly, weontrolled for potential synergies and dissynergies across differ-nt channels (Avery et al. 2012; Falk et al. 2007). Although wecknowledge the importance of other channels for omnichanneletailing (e.g., catalogs, mobile devices), we focus on the Inter-et and physical store channels in view of the dominant role ofhese channels in multi-channel firms (Neslin and Shankar 2009,igby 2011).

Conceptual Development

efinition of Online–Offline Channel Integration

Channel integration is defined as the degree to which dif-erent channels interact with each other (Bendoly et al. 2005).here are two basic approaches to channel integration: (1) pro-iding access to and knowledge about the Internet store athysical stores (offline–online channel integration), and (2) pro-iding access to and knowledge about physical stores at thenternet store (OI). Integration can therefore occur either fromhe store to the Internet or from the Internet to the store. Tontegrate online features into offline channels, firms such as.C. Penney and Louis Vuitton provide self-service or assistednline terminals in their physical stores. Researchers studyingffline–online channel integration find that self-service onlineerminals and information on online channels can reduce the


4 Given the tendency of customers to use one channel for search and another forurchase, we believe it is important to differentiate between search intentionsnd purchase intentions in our investigation (e.g., Neslin and Shankar 2009;erhoef, Neslin, and Vroomen 2007; Zhang et al. 2010).



f Reta

ao2filien(rtapbcc

R

oIwBeird

rt(orocmofiics

ucaFoctota2ano

fl

naarTsaosastasrfrwarcaTa

OE

m(atSqorcqstppmcc(we


ssistance to customers in the branch is a valuable part of theffline service experience (Patrício, Fisk, and Falcão e Cunha008). To integrate offline features into their online channels,rms such as IKEA and John Lewis provide a physical store

ocator or physical store assortment availability informationn their Internet stores. The only study that to our knowl-dge investigates OI finds that perceptions of OI moderate theegative effect of non-availability for Internet store customersBendoly et al. 2005). However, given their research design, theesults of Bendoly et al. (2005) are unable to answer the impor-ant yet unresolved question of whether perceived integrationffects channel evaluation or whether channel evaluation affectserceived integration. Thus, we extend the authors’ insightsy investigating how and when OI affects retailer-level andhannel-level outcomes with an experimental design that allowsontrolling for the causality of the observed effects.

esearch Framework

From the customer perspective, OI provides the key potentialf enhancing the Internet store experience (Bendoly et al. 2005).n this respect, OI enables firms to offer customers what theyant at each stage of the buying process (Gensler, Verhoef, andöhm 2012). However, when considering the outcomes in differ-nt channels, OI may also create unwanted consequences. Thisssue can be understood best through using technology adoptionesearch and diffusion theory to explain the desired and potentialissenting effects of OI.

Technology adoption research has identified the mediatingole of user evaluations of a new technology and the impor-ant role of system attributes as antecedents of these evaluationsDavis 1989; Davis, Bagozzi, and Warshaw 1989). The technol-gy adoption framework has been widely used in multi-channelesearch. For example, Falk et al. (2007) find that the effectf a status quo channel attribute (satisfaction with the currenthannel) on new channel adoption decisions of customers isediated by the perceived usefulness and the perceived risk

f the new channel. Montoya-Weiss, Voss, and Grewal (2003)nd that Website attributes, such as the navigation structure, the

nformation content and graphic style, affect customer onlinehannel usage decisions through evaluations of online channelervice quality and risk perceptions.

In our research context, OI is a system attribute that affectsser evaluations of the online channel and subsequent retailer orhannel adoption decisions. Following prior research on retailernd channel adoption decisions (Cronin, Brady, and Hult 2000;alk et al. 2007; Forsythe and Shi 2003), we use evaluationsf perceived service quality and perceived risk of the onlinehannel as parallel multiple mediators between OI and adop-ion decisions. However, these evaluations do not only affectnline channel adoption decisions but also overall retailer adop-ion decisions or adoption decisions of alternative channels, suchs the physical store channel (Montoya-Weiss, Voss, and Grewal



003). In this context, user evaluations of the online channel acts a reference point for adoption decisions of the primary alter-ative channel. Increasing service quality or decreasing the riskf the online channel with OI may increase the reference point

Hss

iling xxx (xxx, 2015) xxx–xxx 3

or the physical store channel, which could therefore lead to aower adoption of this alternative channel (Falk et al. 2007).

Diffusion theory posits that adoption decisions of new tech-ologies are influenced by the characteristics of the technologynd by the personal characteristics of the adopter (Gatignonnd Robertson 1985). In particular, personal patterns of expe-ience moderate the influence of technological characteristics.his is in line with technology adoption research, whichtates that customer experience may interact with attributionalntecedents of perceived usefulness and risk (Davis 1989). Inur research context, channel evaluation and adoption deci-ions are based on subjective evaluations of a channel’s relativedvantage and the compatibility of that channel with per-onal experience. Customers’ online shopping experience mayherefore affect the impact of online channel attributes, suchs OI, on user evaluations of the service quality and risk

of the online channel (Nysveen and Pedersen 2004). The cor-esponding research framework is presented in Fig. 2. In theollowing section, we summarize specific arguments for the indi-ect effects of OI on search intention, purchase intention, andillingness to pay (WTP) through the perceived service quality

nd the perceived risk of the Internet store, and the moderatingole of the Internet shopping experience. Importantly, in a multi-hannel context, customers not only choose a distinct retailer butlso one or more channels from a retailer’s alternative channels.hus, we differentiate between the effect of OI on retailer-levelnd channel-level outcomes.

nline–Offline Channel Integration and Customervaluations

Service quality perceptions are defined as overall assess-ents of the perceived performance of the Internet store

Zeithaml, Parasuraman, and Malhotra 2002). Risk perceptionsre defined as overall assessments of the uncertainty and poten-ially adverse consequences of the Internet store (Dowling andtaelin 1994). An integrated Internet store increases serviceuality and decreases risk because it helps combine online andffline channels according to their specific ‘conspicuous or expe-iential capabilities’ (Avery et al. 2012). In particular, an offlinehannel may complement an online channel in terms of serviceuality (i.e., by highlighting the possibility of a rich, multisen-ory brand experience and the ability of salespeople to increasehe service experience) and risk (i.e., referring to the physicalresence of the retailer and the possibility of touching and feelingroducts before purchasing them to reduce uncertainty). Argu-ents for such complementarities are provided by studies of

ustomer satisfaction deriving from the availability of multipleustomer touch points (Hitt and Frei 2002). Likewise, Stewart2003) shows that an Internet store decreases risk perceptionshen it signals its association with a physical store. Thus, we

xpect that:


1. Online–offline channel integration (a) increases perceivedervice quality and (b) decreases perceived risk of the Internettore.



4 D. Herhausen et al. / Journal of Retailing xxx (xxx, 2015) xxx–xxx

Online-Offline Channel

Integration

Perceived Servi ce Qual ity of the Internet Store

Perceived Risk of the

Internet Store

Internet Shopp ing Exper ience

Overall (Retailer-Level)

Search Intention Purchase Intention Willingness to Pay

Physical Sto re (Channel-Level)

Sea rch Inte ntion Purchase Inte ntion Willingness to Pay

Internet Store (Channel-Level)

Sea rch Inte ntion Purchase Inte ntion Willingness to Pay

Multi-Channel Att ributes

Multi-Channel Out comes

Customer Eva lua tions

arch

C

lOullfpr

cisVaBF2barattcit

ir(2aoeI

baf2(Rin(adaaGsc

Fig. 1. Rese

ustomer Evaluations and Multi-channel Outcomes

We consider three different customer reactions on the retailer-evel and channel-level as indirect multi-channel outcomes ofI: search intention, defined as the likelihood that customersse a focal retailer (retailer-level) or a certain channel (channel-evel) for searching a product, purchase intention defined as theikelihood that customers use a focal retailer or a certain channelor purchasing a product, and WTP, defined as the maximumrice at which a customer will purchase a product from a focaletailer or a certain channel.

In line with the technology adoption framework, research onhannel design perceptions emphasizes that online service qual-ty and online risk are important drivers of overall consumeratisfaction, regardless of channel choice (Montoya-Weiss,oss, and Grewal 2003). In turn, customer satisfaction has beenssociated with higher search and purchase intentions (Cronin,rady, and Hult 2000) as well as a higher WTP (Ratchford 2009).urthermore, both overall customer usage intentions (Baker et al.002) and WTP (Grewal et al. 2010) are positively affectedy service quality and negatively affected by risk. Likewise,lthough each channel may offer a unique value proposition,esearch findings suggest that online evaluations may drive over-ll retailer choice (Pauwels et al. 2011). Thus, we hypothesizehat the perceived service quality of the Internet store consti-utes a positive effect and the perceived risk of the Internet store



onstitutes a negative effect on overall customer reactions, thats, customers’ search intention, purchase intention, and WTP onhe retailer-level.

toc

framework.

Prior research further indicates that online service qualityncreases the intention to use the online channel, while onlineisk has the opposite effect and decreases online usage intentionBadrinarayanan et al. 2012; Falk et al. 2007; Forsythe and Shi003; Park and Jun 2003). Following these results and the abovergumentation, we expect that service quality (risk) perceptionsf the Internet store resulting from OI have a positive (negative)ffect on customer Internet store reactions, that is, customers’nternet store search intention, purchase intention, and WTP.

In the literature, both complementarity and substitutionetween alternative channels of the same retailer have beendvocated. While some scholars find complementarity of dif-erent channels (Avery et al. 2012; Wallace, Giese, and Johnson004), others find that different channels substitute each otherFalk et al. 2007; Montoya-Weiss, Voss, and Grewal 2003).esearch on cross-channel dissynergies states that satisfaction

n one channel reduces the positive assessment of other chan-els due to increasing expectations of the alternative channelsFalk et al. 2007). Following this view, the online channel mayct as a reference point for store channel assessment and usageecisions. In this context, higher online channel service qualitynd lower risk may raise expectations and lower the positivessessment of the offline channel (Montoya-Weiss, Voss, andrewal 2003). As a result, the increase (decrease) in online

ervice quality (risk) resulting from OI may negatively affectustomers’ physical store reactions. Thus, channel cannibaliza-


ion may occur, as in, for instance, the partial or full reductionf an offline channel’s sales due to the integration of an onlinehannel.



D. Herhausen et al. / Journal of Retailing xxx (xxx, 2015) xxx–xxx 5

Fig. 2. Discrete retailer choice in Study 1. Note: N = 107. Participants were asked to choose one of the following options: [1] purchase (participation) at Retailer 1,[ rete rei s).

tti2oaLeesta

arr

isIprp

2] purchase (participation) at Retailer 2 or [3] no purchase (participation). Discntegrated (INT) and non-integrated (NON) retailers (all p < .01, two-tailed test

Conversely, research on cross-channel synergies finds thathe negative direct effect of higher expectations of the alterna-ive channel is offset by a stronger positive indirect effect throughmproved service experience (Wallace, Giese, and Johnson004). In addition, valuable brand associations attributed tone channel (Ailawadi and Keller 2004) and positive associ-tions formed by the knowledge of one channel (Kwon andennon 2009) may transfer to other channels by way of a haloffect. Following this view, the alternative channel may ben-fit from investments in a different channel. Thus, the higher



ervice quality and lower risk of the online channel may increasehe positive assessment of the offline channel (Wallace, Giese,nd Johnson 2004). In this case, channel synergies may occur

Hn

tailer choice for purchase and gift voucher lottery participation differs between

nd the increase (decrease) in online service quality (risk)esulting from OI may favorably affect customers’ physical storeeactions.

Taken together, we expect that potential dissynergies result-ng from increased expectations of the physical store andynergies resulting from an increased service experience in thenternet store offset each other, and that service quality and riskerceptions of the Internet store have no effect on physical storeeactions, that is, customers’ physical store search intention,urchase intention, and WTP:


2. Increasing the perceived service quality of the Inter-et store (a) positively affects overall retailer-related customer



6 f Reta

rtc

Hp(Ir

T

qIbaesKtfSIe2O(Iootce

Hpset

emtsaYuasmlec

D

l22AHupwpAt

owidaotaptuao

lsipueitrotAtamatr

hst


eactions, (b) positively affects channel-related customer reac-ions to the Internet store, and (c) does not affect channel-relatedustomer reactions to the physical store.

3. Decreasing the perceived risk of the Internet store (a)ositively affects overall retailer-related customer reactions,b) positively affects channel-related customer reactions to thenternet store, and (c) does not affect channel-related customereactions to the physical store.

he Moderating Effect of Internet Store Experience

Given the potential advantages of OI outlined above, a keyuestion is whether all customers respond in a similar way to OI.n multi-channel management, customer segments are largelyased on specific usage patterns of different channels (Neslinnd Shankar 2009). Thus, we use customers’ Internet shoppingxperience – defined as the greater knowledge and experience ofearching and purchasing products online in general (Menon andahn 2002) – to investigate differences in customer responses

o OI. Customers with more Internet experience are likely toeel comfortable with a retailer’s online channel (Forsythe andhi 2003; Montoya-Weiss, Voss, and Grewal 2003). Moreover,nternet experience has been identified as an important mod-rator in determining online channel assessments (Falk et al.007). Thus, customers with high Internet experience will judgeI as “just another addition to their existing Internet services”

Falk et al. 2007, p. 150). On the contrary, customers with lessnternet experience will not be as comfortable with a retailer’snline channel and OI will thus provide more value in termsf service quality and risk reduction. Therefore, we expect OIo be more important in predicting service quality and risk per-eptions of the Internet store for customers with low Internetxperience:

4. Customer’s Internet shopping experience (a) decreases theositive effect of online–offline channel integration on perceivedervice quality of the Internet store and (b) decreases the negativeffect of online–offline channel integration on perceived risk ofhe Internet store.

A further hypothesis is implicit in the reasoning of the directffects, namely that the multiple mediating effect of OI onulti-channel outcomes through service quality and risk percep-

ions of the Internet store is moderated by customers’ Internethopping experience (Hayes 2013). We thus predict moder-ted mediation (Edwards and Lambert 2007; Muller, Judd, andzerbyt 2005). The corresponding model in Fig. 1 enablesnderstanding the effects of OI, given that Judd, Yzerbyt,nd Muller (2014, p. 654) note, “a full theoretical under-tanding of an effect of interest involves both understandingechanisms (the question of mediation) and understanding



imiting conditions (the question of moderation), and the knowl-dge gained from both of these assessments ultimately mustonverge”.

I[pt


Study 1

esign, Participants, and Procedure

The purpose of Study 1 was to test whether and how OIeads to a competitive advantage for retailers. The study used a

× 2 ([Retailer 1: integrated versus non-integrated] × [Retailer: integrated versus non-integrated]) between-subject design.s the sensory product of interest, we chose Team Germanyavaianas Flip-Flops, an important product category given thepcoming 2014 World Cup in Brazil. A market research com-any assembled a sample of 107 German participants (50 percentomen, Mage = 43.92 years). Participants were financially com-ensated and entered into a lottery for five 20 Euro gift vouchers.s an incentive-aligned treatment, participants chose one of the

wo retailers for the gift voucher.Participants were consecutively presented the Internet stores

f two anonymized multi-channel retailers, each either with orithout OI. The non-integrated version was based on a typ-

cal retailer’s homepage and included features such as productescriptions, product ratings and reviews from customers, avail-ble sizes and colors, and a buy button. Based on the results ofur exploratory study of multi-channel retailers, OI was opera-ionalized by offering a physical store search function (includingvailability check and reserving products in the store) and theossibility of returning products bought in the Internet storeo any physical store. Screenshots and text descriptions weresed to visualize the different versions of the Internet store. Tovoid confounding effects, the Internet stores of the two retailersffered exactly the same features.

Initially, participants were asked to imagine that they wereooking for Team Germany Havaianas Flip-Flops and werehown a picture and the description of the flip-flops. The partic-pants then indicated whether they knew the Havaianas brandrior to the study, their involvement with the world cup, the prod-ct category, and the brand Havaianas, and whether they hadver bought flip-flops from Havaianas or another brand. Partic-pants were subsequently confronted with the Internet stores ofhe two retailers. Both retailers were described as multi-channeletailers with an Internet store and several physical stores, somef which were in the hometown of the participants. They werehen randomly assigned to one of the four possible conditions.fter participants had finished reading each scenario, they rated

he perceived service quality and risk of the Internet store, over-ll search intention, purchase intention and WTP, together with aanipulation check. The evaluation order (i.e., dependent vari-

bles first vs. manipulation check first) was randomized to ensurehat results would not be affected by a particular order. Allespondents provided answers for both retailers.

Participants indicated perceived service quality (“Overall,ow satisfied are you with the service offered in the Internettore of retailer [x]?”), perceived risk (“The risk concerninghe product characteristics (e.g., size and color) is high when

would purchase the flip-flops in the Internet store of retailer


x].”), purchase intention (“How likely is it that you wouldurchase the flip-flops from retailer [x]?”), and search inten-ion (“How likely is it that you would search in the Internet or



f Reta

pIaaddmmrt(a(

tcnftvavatao

A

iarotcd(aTpos

tp

wwlccaruc

Rβ

fqRpppt

2m(bcotipCtc

pfltcFtpdiatdwplfltpc


hysical store of retailer [x] next time you look for flip-flops?”).n line with previous multi-channel research (Verhoef, Neslin,nd Vroomen 2007), we used single items to measure mediatorsnd dependent variables to reduce the timing of the researchesign, which was already relatively lengthy due to the intro-uction of both retailers and the repeated measurement.5 Toeasure WTP, we employed an open-ended question (“Howuch would you pay, in Swiss Francs, for the flip-flops at

etailer [x]?”) adapted from van Doorn and Verhoef (2011). Fur-hermore, we measured discrete retailer choice for purchasingVerhoef, Neslin, and Vroomen 2007). Here, respondents weresked to choose only one retailer for their purchase decisionincluding a no-choice option).

In addition to the above-mentioned controls, we asked par-icipants to indicate their monthly disposable income and priceonsciousness (Wakefield and Inman 2003), and their inter-al reference price for similar flip-flops of the same qualityrom a comparable brand (Lichtenstein and Bearden 1989). Fur-hermore, participants rated the credibility of the scenario (notery credible/credible), the difficulty of the task (not difficultt all/very difficult), and chose one of the retailers for the giftoucher lottery. Finally, the participants provided their gender,ge and what they thought was the aim of the study. None ofhe participants correctly guessed the purpose of the study (mostssumed we were investigating WTP for world cup merchandizer flip-flops).

nalysis and Results

Manipulation and confound checks. Participants in the exper-mental groups rated the perceived integration of the Internetnd physical store (“The services and functions offered atetailer [x]’s Internet and physical store complement eachther”) higher than those in the control groups (Retailer 1:(105) = 6.99, p < .01; Retailer 2: t(105) = 8.80, p < .01). Theonfound checks indicated no significant differences across con-itions for credibility (p = .24; Mcredibility = 5.69) and difficultyp = .72; Mdifficulty = 1.22). The high (low) means suggest thatll scenarios were considered credible and easy to understand.hese analyses showed no effects for the evaluation order (all

> .15), and the data were collapsed across the two differentrder conditions. The means, standard deviations, and effectizes for all studies are reported in Appendix.



Direct effects. We used regression analysis including all con-rol variables to test our hypotheses. We found that OI has aositive effect on perceived service quality (R1: β = .43, p < .01;

5 We conducted reliability analyses on data collected during a pre-test studyith 53 students on a multi-item measure of search intention (e.g., how likelyill you use the Internet store next time you look for [product category]?/how

ikely will you consult the Internet store for future information about [productategory]?/how likely is it that you would search in the Internet store for [productategory] in the future?). Given the high value of Cronbach’s alpha (α = .93)nd considering that the reflective indicators are homogeneous and semanticallyedundant (Diamantopoulos et al. 2012), we concluded that although not optimal,sing a single item to measure the dependent variable appears to be a reasonablehoice.

pDO(capnc

D

t


2: β = .53, p < .01) and a negative effect on perceived risk (R1: = −.29, p < .01; R2: β = −.40, p < .01), thus providing supportor H1a and H1b. We found that an increase in perceived serviceuality increases overall search intention (R1: β = .37, p < .01;2: β = .61, p < .01), overall purchase intention (R1: β = .42,

< .01; R2: β = .56, p < .01), and overall WTP (R1: β = .26, < .05; R2: β = .28, p < .05), supporting H2a. A decrease inerceived risk did not significantly increase overall search inten-ion, purchase intention, and WTP; thus H3a is not supported.

Mediation. We used Model 4 of the PROCESS macro (Hayes013) including all control variables to test for the multipleediating effects of perceived service quality and perceived risk

Zhao, Lynch, and Chen 2010). The analyses conducted throughootstrapping (5,000 bootstrap samples) with 95 percent bias-orrected confidence intervals indicated indirect positive effectsf OI on overall search intention (R1: β = .16, 95 percent CI = .07o .30; R2: β = .33, 95 percent CI = .20 to .52), overall purchasentention (R1: β = .20, 95 percent CI = .10 to .33; R2: β = .31, 95ercent CI = .18 to .47), and overall WTP (R1: β = .11, 95 percentI = .02 to .23; R2: β = .17, 95 percent CI = .05 to .32). Together,

hese results indicate that the indirect effect of OI on overallustomer reactions is mediated by perceived service quality.

Discrete retailer choice. While customers may use multi-le retailers for shopping, they can choose only one retaileror a specific purchase situation. We used the separately col-ected variables for discrete purchase and the gift voucher lotteryo account for this unique choice. The distribution of retailerhoices in the different experimental conditions is shown inig. 2, and differences were assessed using a multinomial logis-

ic regression analyses. The overall fit indices suggest that theurchase model has a good fit (χ2(18) = 68.67, p < .01; McFad-en’s R2 = .32). Sixty-nine percent of the predicted choices weren line with the actual customer responses (Kendall’s Tau = .49nd Somer’s D = .50). The marginal effects analysis suggestshat the OI of R1 increased (M.E. = .49, p < .01) and of R2ecreased (M.E. = −.30, p < .01) the likelihood that participantsould choose R1, and that the OI of R1 decreased (M.E. = −.30,

< .01) and that of R2 increased (M.E. = .40, p < .01) the like-ihood that participants would choose R2 for purchasing theip-flops. The OI of R1 decreased (M.E. = −.26, p < .01) and

hat of R2 did not significantly affect the likelihood that partici-ants would choose the no choice option. The gift voucher lotteryhoice model (χ2(18) = 50.78, p < .01; McFadden’s R2 = .30; 71ercent correct predicted choices; Kendall’s Tau = .48; Somer’s

= .49) replicates the retailer choice results and indicates thatI of R1 increased (M.E. = .41, p < .01) and that of R2 decreased

M.E. = −.33, p < .01) the likelihood that participants wouldhoose R1, and that OI of R1 decreased (M.E. = −.35, p < .01)nd that of R2 increased (M.E. = .38, p < .01) the likelihood thatarticipants would choose R2 for the gift voucher lottery. OI didot significantly affect the likelihood that participants wouldhoose the no participation option.


iscussion of Study 1

The results of Study 1 support the notion that OI has posi-ive effects on overall purchase intention, search intention, and



8 f Reta

WrtitcipdtmrcTfc

D

wpdso1cbwoc

PionaapIcamwws1i

tots

WWtoi

mIbmdcc[[wMl

pitcapiepn

A

egWcMp

tpaA(nii


TP, and thus that OI provides a competitive advantage foretailers. More specifically, and in line with technology adop-ion research, we found support for the indirect effects of OI: OIncreases perceived service quality and reduces perceived risk ofhe Internet store, perceived service quality in turn increases pur-hase intention, search intention, and WTP for a retailer with anntegrated Internet store. However, contrary to our expectations,erceived risk does not directly affect relevant outcomes. Theiscrete retailer choice analysis further revealed that OI increaseshe likelihood that customers choose a distinct retailer from

ultiple alternatives. However, given that this study focused onetailer-level effects and overall customer reactions, examininghannel synergies or cannibalization was not part of the design.o further extend insights into customer reactions across dif-erent channels, Study 2 explores whether OI has the expectedhannel-level effects in the Internet and physical store.

Study 2


The purpose of Study 2 was to test whether, how, and underhat conditions OI affects customer reactions in Internet andhysical stores. The study used a generic between-subjectsesign, with the level of integration in the Internet store as aingle two-dimensional treatment variable. The sensory productf interest is Wooden sunglasses from Woodfellas. A total of29 Swiss undergraduates participated in the study (50.8 per-ent women, Mage = 24.85 years), all of whom neither knew therand nor bought its products.6 As an incentive, participantsere entered into a lottery for five movie gift cards and had thepportunity to directly buy the sunglasses using an incentive-ompatible mechanism (Becker, DeGroot, and Marschak 1964).

The experimental procedure is comparable to the first study.articipants were presented one of two versions of a hypothet-

cal Swiss Internet store for Woodfellas – either an integratedr a non-integrated version based on the brand’s German Inter-et store. As in Study 1, OI was operationalized by offering

physical store search function (including availability checknd reserving products in the store) and the possibility to returnroducts bought in the Internet store to any physical store.nitially, participants indicated their percentage of Internet pur-hases from overall purchases of products such as sunglassesnd apparel (% Physical stores, % Internet stores [includingobile], % other channels [e.g., catalog or telephone]), whiche used to operationalize the Internet shopping experience. Theyere then asked to imagine that they were looking for wooden



unglasses, the brand Woodfellas and specific sunglasses worth59 Swiss Francs were introduced (without showing any pricenformation). The participants then rated whether they knew the

6 Woodfellas does not sell its products in Switzerland. However, three par-icipants indicated that they knew the Woodfellas brand prior to the study, andne had already bought Woodfellas sunglasses. We excluded these participantso avoid the confounding effects of prior knowledge of the Woodfellas Internethop and pricing.

pdssqo(ii


oodfellas brand and whether they had bought products fromoodfellas prior to the study, how important sunglasses were to

hem, and whether they had bought wooden sunglasses of anyther brand. Participants were then confronted with either thentegrated or the non-integrated Woodfellas Internet store.

After participants finished reading the scenario, they rated aanipulation check, perceived service quality and risk of the

nternet store, search intention, purchase intention, and WTPoth for the Internet and the physical store. As consumers can useultiple channels for purchase and search decisions, all respon-

ents provided answers for both channels. We measured discretehannel choice for purchasing where respondents chose only onehannel for their purchase decision (with the following options:1] Physical store, [2] Internet store, [3] Other online store, or4] None of the options). Furthermore, we captured overall WTPith an incentive-compatible mechanism (Becker, DeGroot, andarschak 1964). The order of evaluation between the manipu-

ation check and dependent variables was randomized.In addition to the aforementioned controls, we controlled for

articipants’ monthly disposable income, price consciousness,nternal reference price, multichannel self-efficacy, defined ashe confidence and ability to use different types of distributionhannels (Compeau and Higgins 1995), need for touch, defineds the preference for physically experiencing and evaluating aroduct before purchase (Peck and Childers 2003), need fornteraction, defined as the importance of interacting with a realmployee (Dabholkar 1996), age and gender. We again askedarticipants what they thought was the aim of the study, butone of them guessed correctly.

nalysis and Results

Manipulation and confound checks. Participants in thexperimental groups perceived the Internet store as more inte-rated than those in the control groups (t(127) = 19.08, p < .01).e found no significant differences across conditions for

redibility (p > .97; Mcredibility = 6.23) and difficulty (p > .90;difficulty = 1.52), and no effects for the order of evaluation (all

> .12).Direct effects. We used regression analysis including all con-

rol variables to test our hypotheses. We found that OI has aositive effect on perceived service quality (H1a: β = .58, p < .01)nd a negative effect on perceived risk (H1b: β = −.32, p < .01).n increase in perceived service quality increases overall WTP

H2a: β = .31, p < .01), while a decrease in perceived risk haso significant effect on overall WTP (H3a). We found that anncrease in perceived service quality increases Internet searchntention (β = .41, p < .01), Internet purchase intention (β = .38,

< .01), and Internet WTP (β = .38, p < .01), supporting H2b. Aecrease in perceived risk has no significant effects on Internetearch intention, purchase intention, and WTP; thus H3b is notupported. We further found that an increase in perceived serviceuality has no effect on store search intention but positive effects


n store purchase intention (β = .31, p < .01) and store WTPβ = .29, p < .01), providing mixed support for H2bc. A decreasen perceived risk has no significant effects on store searchntention, purchase intention, and WTP; supporting H3c. As



f Reta

pOi(yipoe

mmJbiq(ti(t(wm9cC.β

anCcC.β

rcae

dccc(rTlIt(twnss

D

eanfiidnWsWoaoaitoI

D

pABuwp1pt1rcibbcafam

bo(ri


redicted, we found a negative effect of the interaction betweenI and Internet shopping experience on perceived service qual-

ty (β = −.16, p < .05) and a positive effect on perceived riskβ = .17, p < .05), supporting H4a and H4b. The simple slope anal-ses showed that the effect of OI on perceived service qualitys stronger when Internet shopping experience is low (β = .74,

< .01) than when it is high (β = .42, p < .01), and that the effectf OI on perceived risk is only significant when Internet shoppingxperience is low (βlow = −.49, p < .01; βhigh = −.15, ns).

Moderated mediation. We use Model 7 of the PROCESSacro (Hayes 2013) including all control variables to test foroderated mediation (Edwards and Lambert 2007; Muller,

udd, and Yzerbyt 2005). Bootstrapping analyses with 5,000ootstrap samples and 95 percent bias-corrected confidencentervals indicated overall mediation effects of perceived serviceuality in the relationships between OI and overall WTPβ = .11, 95 percent CI = .01 to .24), Internet search inten-ion (β = .23, 95 percent CI = .13 to .40), Internet purchasentention (β = .22, 95 percent CI = .10 to .38), Internet WTPβ = .21, 95 percent CI = .11 to .37), store purchase inten-ion (β = .18, 95 percent CI = .09 to .31), and store WTPβ = .16, 95 percent CI = .07 to .29). These mediation effectsere stronger at minus one standard deviation below theean Internet shopping experience score (βoverall WTP = .15,

5 percent CI = .01 to .32; βInternet search intention = .32, 95 per-ent CI = .15 to .54; βInternet purchase intention = .29, 95 percentI = .13 to .52; βInternet WTP = .28, 95 percent CI = .14 to

48; βstore purchase intention = .24, 95 percent CI = .11 to .42;store WTP = .21, 95 percent CI = .08 to .39) and weakert plus one standard deviation above the mean Inter-et shopping experience score (βoverall WTP = .07, 95 percentI = .01 to .18; βInternet SEARCH INTENTION = .15, 95 per-ent CI = .08 to .30; βInternet purchase intention = .14, 95 percentI = .06 to .29; βInternet WTP = .13, 95 percent CI = .06 to

28; βstore purchase intention = .11, 95 percent CI = .06 to .24;store WTP = .10, 95 percent CI = .04 to .22). Together, theseesults show that the indirect effects of OI on customer out-omes mediated by perceived service quality are significant,nd that Internet shopping experience moderates these indirectffects.

Discrete channel choice. We used the separately collectediscrete channel choice variable to account for the unique pur-hase choice. The distribution of channel choice in differentonditions is shown in Fig. 3, Panel A. Differences in channelhoice were assessed using a multinomial logistic regressionχ2(27) = 52.85, p < .01; McFadden’s R2 = .20; 68 percent cor-ect predicted choices; Kendall’s Tau = .28; Somer’s D = .38).he marginal effects analysis suggested that OI increased the

ikelihood that participants would make purchases in the brand’snternet store (M.E. = .18, p < .01) and decreased the likelihoodhat participants would make purchases in other online storesM.E. = −.09, p < .05). The effect on the likelihood that par-icipants would make purchases in the brand’s physical store



as insignificant. Thus, additional sales in an integrated Inter-et store are mainly drawn from other, third-party online stores,uggesting that OI does not significantly cannibalize the physicaltore.

tito



Study 2 finds evidence that OI has positive (negative) directffects on perceived service quality (risk) of the Internet store,nd positive indirect effects on overall WTP, as well as on Inter-et search intention, purchase intention, and WTP. We furthernd that Internet shopping experience moderates the direct and

ndirect effects of OI. As in Study 1, perceived risk does notirectly affect the relevant outcomes. Moreover, OI does notegatively affect store search intention, purchase intention, andTP. Contrary to our expectations, we also found positive and

ynergistic effects of OI on store purchase intention and storeTP. Thus, we found no evidence of unintended effects of OI

n the physical store. In addition, a discrete channel choicenalysis revealed that OI diverts mainly customers that wouldtherwise purchase from other, non-integrated online stores ton integrated Internet store. Thus, OI does not result in signif-cant cannibalization across channels. In Study 3, we furtherest for the conditional indirect effects of OI with a samplef actual customers and a different operationalization of thenternet shopping experience based on archival firm data.

Study 3


We conducted Study 3 together with two retail firms: a Euro-ean manufacturer and retailer of sporting equipment (Retailer) and a German fashion manufacturer and retailer (Retailer B).oth retailers operate their own physical stores, sell their prod-cts in multiple channels (both directly and indirectly), competeith third-party online stores that sell the same products at fixedrices, and do not constitute OI to date. At Retailer A, a total of38 Swiss newsletter recipients participated in the study (20.9ercent women, Mage = 38.29 years). As an incentive, the par-icipants were entered into a lottery for an outdoor outfit worth,000 Swiss Francs. At Retailer B, a total of 577 Germans whoeceive the firm’s newsletter participated in the study (79.4 per-ent women, Mage = 51.87 years). Gift vouchers were used asncentives for participation. All participants of Retailer A hadought products from the firm’s physical store prior to the studyut not from the Internet store (and belong to a “physical store”ustomer segment according to Retailer A’s management) whilell participants of Retailer B had bought the majority of productsrom the firm’s Internet store prior to the study (and belong to

“multi-channel” customer segment according to Retailer B’sanagement).We used OI as a two-dimensional treatment variable in

oth data collections. In the integrated version, OI wasperationalized by offering a physical store search functionincluding availability check and reserving products). Theetailers’ objections that such a feature could lead to increas-ng expectations of customers that could not be fulfilled in


he near future prevented us from including the possibil-ty of returning products purchased in the Internet store tohe physical store. Thus, together with the different levelsf Internet shopping experience (Retailer A = Store channel




Fig. 3. Discrete channel choice in Study 2 and Study 3a. Note: NStudy 2 = 129; NStudy 3a = 138. Participants were asked to choose one of the following options: [1]purchase in the Physical store of the retailer, [2] purchase in the Internet store of the retailer, [3] purchase in other online stores or [4] none of the above options.Discrete channel choice for the Internet store (p < .01) and other online stores (p < .05) differs between integrated and non-integrated Internet stores in both studies(

cBevhtj

sifsp

fmcp

A

i

two-tailed tests).

ustomers with low Internet shopping experience; Retailer = Internet channel customers with high Internet shoppingxperience), Study 3 uses a 2 × 2 ([Internet store: non-integratedersus integrated] × [Internet shopping experience: low vs.igh] between-subject design. The procedure was comparableo the previous studies and the sensory product of interest was aacket from the respective brand.

We first measured perceived integration, then perceivedervice quality and risk of the Internet store, purchase and searchntention for both the Internet and physical store, and controlled



or participants’ monthly disposable income, experience withearching and purchasing online and offline, proximity to ahysical store, multichannel self-efficacy, need for touch, need

tpec

or interaction, product category involvement, brand involve-ent, age, and gender. In addition, we measured discrete channel

hoice for purchasing at Retailer A. Again, none of the partici-ants correctly guessed the purpose of the study.

nalysis and Results

Manipulation and confound checks. Participants in the exper-mental groups perceived the Internet store as more integrated


han those in the control groups (Retailer A: t(136) = 10.07, < .01; Retailer B: t(575) = 8.28, p < .01), and participants’ gen-ral Internet shopping experience differed between store channelustomers and Internet channel customers (t(713) = 5.86,



f Reta

pfd

tapts(inap(itfnpstIir(

mmsii(tmmlpCrβ

rmI

coapDtrl(lpS

D

2plergopcinacwospct

S

m2obegltOoiTpsoottntw

T


< .01). We found no significant differences across conditionsor credibility (pA > .19; MA = 5.96; pB > .43; MB = 5.74) andifficulty (pA > .93; MA = 2.02; pB > .32; MB = 1.82).

Direct effects. We used regression analysis including all con-rol variables to test our hypotheses. We found that OI has

positive effect on perceived service quality (H1a: β = .58, < .01) but not on PR; thus H1b is not supported. We foundhat an increase in perceived service quality increases Internetearch intention (β = .41, p < .01) and Internet purchase intentionβ = .20, p < .01), supporting H2b. A decrease in perceived riskncreases Internet search intention (β = −.10, p < .01) and Inter-et purchase intention (β = −.17, p < .01). We further found thatn increase in perceived service quality has no effect on storeurchase intention but a positive effect on store search intentionβ = .12, p < .01), providing mixed support for H2bc. A decreasen perceived risk has no significant effects on store search inten-ion and purchase intention, supporting H3c. As predicted, weound a negative effect of the interaction between OI and Inter-et shopping experience on perceived service quality (β = −.13,

< .05) and a positive effect on perceived risk (β = .13, p < .05),upporting H4a and H4b. The simple slope analyses showed thathe effect of OI on perceived service quality is stronger whennternet shopping experience is low (β = .45, p < .01) than whent is high (β = .12, p < .01), and that the effect of OI on perceivedisk is only significant when Internet shopping experience is lowβlow = −.32, p < .01; βhigh = .01, ns).

Moderated mediation. We use Model 7 of the PROCESSacro (Hayes 2013) including all control variables to test foroderated mediation. Bootstrapping analyses with 5,000 boot-

trap samples and 95 percent bias-corrected confidence intervalsndicated overall mediation effects of perceived service qualityn the relationships between OI and Internet search intentionβ = .08, 95 percent CI = .05 to .11) and Internet purchase inten-ion (β = .04, 95 percent CI = .02 to .06), but no significantediation effects of perceived risk nor on store outcomes. Theediation effects of perceived service quality were stronger for

ow Internet shopping experience (βInternet search intention = .19, 95ercent CI = .12 to .27; βInternet purchase intention = .10, 95 percentI = .06 to .16) and weaker for high Internet shopping expe-

ience (βInternet search intention = .05, 95 percent CI = .02 to .09;Internet purchase intention = .03, 95 percent CI = .01 to .05). Theseesults show that the indirect effects of OI on Internet outcomesediated by perceived service quality are significant, and that

nternet shopping experience moderates these indirect effects.Discrete channel choice. The distribution of discrete channel

hoice for Retailer A is shown in Fig. 3, Panel B. The fit indicesf a multinomial logistic regression suggest that the model has

good fit (χ2(26) = 76.52, p < .01; McFadden’s R2 = .34; 78ercent correct predicted choices; Kendall’s Tau = .45; Somer’s

= .56). The marginal effects analysis reveals that OI increasedhe likelihood that participants would make purchases in theetailer’s Internet store (M.E. = .16, p < .02) and decreased theikelihood that they would make purchases in other online stores



M.E. = −.09, p < .10), but did not significantly affect the like-ihood that participants would make purchases in the retailer’shysical store. Thus, we find the same pattern of results as intudy 2.

oari



The results of Study 3 replicate the patterns observed in Study regarding OI’s indirect positive effects on Internet search andurchase intention, and support the notion that OI does notead to channel cannibalization. We further provide additionalvidence for the moderating role of the Internet shopping expe-ience. In this respect, Study 3 increases the plausibility of theeneral mediated moderation model of OI. Although the patternf results for the hypothesized paths is fairly consistent with therevious study, there are some discrepancies to note. Specifi-ally, 6 of the 8 hypothesized results from Study 2 are replicatedn Study 3. The two exceptions in Study 3 are as follows: OI doesot affect perceived risk of the Internet store (H1b not supported),nd perceived risk of the Internet store decreases search and pur-hase intention in the Internet store (H3b supported). In addition,e found mixed support for the expected non-significant effectf perceived service quality on store outcomes (H2c) and noignificant indirect effects of OI on store outcomes. We proposeotential theoretical reasons for these findings in the general dis-ussion section and provide ideas of how future research couldest these propositions.

General discussion

ummary of Findings

Researchers have explored the causes and consequences ofulti-channel shopping (Venkatesan, Kumar, and Ravishanker

007) and provided first insights regarding the moderating rolef channel integration for non-availability and firm switchingehavior (Bendoly et al. 2005). However, there is no thoroughxamination of the consequences of OI to date. Addressing thisap, this research is the first to determine whether and how OIeads to a competitive advantage for retailers, whether OI leadso synergies or cannibalization among channels, and whetherI has the same effects on all customers. Table 1 gives anverview of the results obtained. In a series of independent stud-es, evidence is provided for most of the hypothesized effects.he findings support the premises (1) that OI directly increaseserceived service quality of the Internet store; (2) that perceivedervice quality of the Internet store increases overall and Internetutcomes; (3) that OI indirectly increases overall and Internetutcomes via perceived service quality of the Internet store; (4)hat the direct and indirect effects of OI are moderated by cus-omers’ Internet shopping experience; and (5) that OI does notegatively affect the physical store. These findings have impor-ant implications for multi-channel theory and practice, whiche discuss next.

heoretical Implications and Extensions

Creating a competitive advantage with OI. To the best of


ur knowledge, our research is the first to examine whethernd how OI leads to a competitive advantage for retailers. Theesults of our studies provide evidence that customers value anntegrated Internet store and that OI leads to more favorable




Table 1Overview of results.

Study 1 Study 2 Study 3

H1a: OI → Perceived service quality (+)√ √ √

H1b: OI → Perceived risk (−)√ √ √a

H2a: Perceived service quality → Overall outcomes (+)√ √

H2b: Perceived service quality → Internet store outcomes (+)√ √

H2c: Perceived service quality → Physical store outcomes (no effect) Mixed MixedH3a: Perceived risk → Overall outcomes (−) ns nsH3b: Perceived risk → Internet store outcomes (−) ns

√H3c: Perceived risk → Physical store outcomes (no effect)

√ √H4a: OI × Internet shopping experience → Perceived service quality (−)

√ √H4b: OI × Internet shopping experience → Perceived risk (−)

√ √

N d.rnet st

bmp2tsOWacawioaeaiOwt(ewn

sfbwaifscnIppOwf

ar

nWe(atipatr

oltctaOsacoimmvh

M

rdfi

ote: OI = online–offline channel integration (dummy coded), ns = not supportea Online–offline channel integration does not affect perceived risk of the Inte

ehavior toward a retailer and its Internet store. Our experi-ental design enabled us to inform the customer satisfaction

erspective of multi-channel management (Neslin and Shankar009) by providing first empirical evidence of the causal rela-ionship between OI, service quality perceptions of the Internettore, and retailer-level and channel-level customer outcomes.I not only enhances search intention, purchase intention, andTP in the Internet store but is also a source of competitive

dvantage for the whole firm. To date, research on channel per-eptions focused solely on the channel level (Verhoef, Neslin,nd Vroomen 2007). In taking the analysis to the retailer level,e provide evidence that customers choose retailers offering

ntegrated online channels over retailers with non-integratednline channels. Given that only brick-and-click retailers areble to integrate their channels (but not purely online retail-rs), OI is a promising route for these retailers to establishing

competitive advantage. Moreover, we find the service qual-ty of the Internet store fully mediates the relationship betweenI and multi-channel outcomes. This finding is in accordanceith technology adoption research (Davis 1989), which posits

hat attributes (OI) only indirectly influence customer behaviorsearch intention, purchase intention, and WTP) through influ-ncing customer evaluations (perceived service quality). Thus,e also inform multi-channel theory by providing the mecha-isms of how OI leads to a competitive advantage.

No significant cannibalization of the physical store. Somecholars cautioned that OI may lead to negative spillover effectsrom one channel to another or be a zero-sum game withoutenefits (e.g., Verhoef, Neslin, and Vroomen 2007). However,e found no negative effects of OI on physical store outcomes

cross the studies, and only weak and insignificant cannibal-zation of the physical store resulting from OI. Thus, we findurther support that Internet channels complement rather thanubstitute physical channels (Avery et al. 2012). We not onlyonfirm prior findings but also extend them to integrated Inter-et stores. Although OI increases the service quality of thenternet store, customers do not show lower search intention,urchase intention, and WTP in the physical store. Further sup-ort is provided by the analysis of discrete channel choice.



I mainly attracts additional customers to the Internet storeho would otherwise purchase in non-integrated online stores

rom competing retailers. These findings underline that OI is

mist

ore in the high Internet shopping experience condition.

potential source of competitive advantage for multi-channeletailers.

No undesired effects of OI on customers’ WTP across chan-els. We found positive effects of OI on overall WTP, InternetTP, and store WTP. These effects extend theoretical knowl-

dge on the antecedents of online and offline retail pricingGrewal et al. 2010). OI is a channel factor that increases WTPcross channels and thus also customer perception and accep-ance of higher prices across channels. In particular, our resultsn Study 2 suggest that OI can be used as an instrument to sellroducts with non-digital attributes (such as sunglasses) offlinend online at competitive prices. This finding further supportshe notion that higher service quality of the Internet channelesulting from OI does not negatively affect physical channels.

The effects of OI vary across customers with different levelsf Internet shopping experience. By considering the individualevel of Internet shopping experience as a moderator, we informheory on whether the effect of OI is stronger or weaker forertain types of customer segments. These findings are impor-ant given that retailers need to allocate marketing resourcescross customer channel segments (Neslin and Shankar 2009).ur results show that customers with higher levels of Internet

hopping experience are less influenced by OI. These findingsre in line with technology adoption research, which states thatustomer experience may interact with attributional antecedentsf perceived usefulness (Davis 1989). Thus, we provide newnsights on a contingency condition of OI, which suggests that

ulti-channel retailers should focus OI efforts on customer seg-ents with lower Internet experience because these customers

alue integrated channels more that customer segments withigh Internet experience.

anagerial Implications

Our work has direct practical relevance for multi-channeletailers since physical retail and Internet retail are often con-ucted through separate entities or divisions within the samerm. Thus, integrating these separate channels is first and fore-


ost a cost-intensive investment and a key question for managerss whether OI pays off. We respond to this question and offereveral implications for retail managers. First, although conven-ional wisdom holds that channel integration influences overall



f Reta

bnsimutscgonhimsflsbiectaalca

DD

(frIpoa

nnt(nntdIn(tfiitr

ttabesmrti

po(apsrntentIpliobhoacp2rc

tpptpaapSmstoutcomes for first-time customers but not for repeat customers.In addition, van Birgelen, Jong, and de Ruyter (2006) comparedcross-channel effects for non-routine and routine services and


ottom-line outcomes, limited published evidence supports thisotion – particularly regarding different channels. The presenttudies therefore provide important evidence of the effect ofntegration on consumer attitudes and intentions, suggesting that

anagers from brick-and-click retailers should feel confident insing OI as a means of achieving a competitive advantage inhe marketplace. Second, our findings suggest that OI createsynergies rather than cannibalization. The results of the discretehannel choice analyses suggest that additional sales in inte-rated Internet stores are generated from customers that wouldtherwise buy from non-integrated, third-party online stores andot from physical stores. This is an encouraging finding that mayelp retail managers “sell” integration strategies to their phys-cal stores. Third, our results are particularly encouraging for

ulti-channel retailers that operate Internet and physical storesince OI leads to a 28 percent increase in overall WTP for flip-ops in Study 1 and a 35 percent increase in overall WTP forunglasses in Study 2. Moreover, OI also increases Internet WTPy 22 percent and store WTP by 7 percent in Study 2. Thus, byntegrating online and offline channels, brick-and-click retail-rs may be able to reduce price competition with purely onlineompetitors and charge higher prices online. Finally, based onhe results from the moderation analysis, retail managers shouldnalyze the channel usage behavior of target customer groupsnd accordingly focus OI efforts on customer segments withower Internet experience. Such a strategy should help multi-hannel retailers increase the benefits of OI and offset the costsssociated with channel integration.

ifferences Across the Studies and Future Researchirections

While the results across the studies are fairly consistentTable 1), there are some interesting discrepancies to note. Dif-erences across studies include (1) the effect of OI on perceivedisk of the Internet store, (2) the effects of perceived risk of thenternet store on Internet store outcomes, and (3) the effects oferceived service quality of the Internet store on physical storeutcomes. We conducted some additional post hoc analyses toddress these inconsistencies.

First, we used a multigroup analysis to investigate theon-significant effect of OI on the perceived risk of the Inter-et store in Study 3. While OI decreases perceived risk ofhe Internet store for physical store customers of Retailer Aβ = −.33, p < .01), it does not affect perceived risk of the Inter-et store for Internet store customers of Retailer B (β = .01,s). Thus, the insignificant effect of OI on perceived risk ofhe Internet store may be explained by the fact that respon-ents from Retailer B already bought products from the retailer’snternet store. The low mean for the perceived risk of the Inter-et store across the integrated and non-integrated conditionsMNo Integration = 3.21; MIntegration = 3.19 for Retailer B comparedo MNo Integration = 4.47; MIntegration = 3.55 for Retailer A) con-



rms that respondents from Retailer B have a high level of trustn the Internet store regardless of OI. Indeed, perceived risk ofhe Internet store for the non-integrated condition is lower forespondents from Retailer B than for all other respondents across fi


he different studies (all p < .01). These findings suggest thathe channel-patronage behavior of customers (Inman, Shankar,nd Ferraro 2004) may be an additional moderating conditioneyond customers’ Internet shopping experience that strength-ns and weakens the effects of OI. Customers that already haveufficient positive purchase experience with the Internet storeay show higher initial confidence in this channel. Thus, the

isk reduction aspect of OI could be significantly less relevant forhese customers. Future research should examine this interplayn more detail.

Second, we addressed the non-significant effects of theerceived risk of the Internet store on overall and Internet storeutcomes in Study 1 and Study 2. We followed prior researchFalk et al. 2007) on channel adoption decisions and includedn additional path in our conceptual model, namely, from theerceived risk to the perceived service quality of the Internettore to account for serial multiple mediation (Hayes 2013). Theationale behind this path is that the service quality of the Inter-et store will be rated lower by customers if its usage entailsaking a risk. We then recalculated all analyses using structuralquation modeling with maximum-likelihood estimation and aon-parametric bootstrapping routine, thereby accounting forhe effect of perceived risk on perceived service quality of thenternet store.7 The results suggest that all reported regressionarameters remained stable (i.e., no changes in the significanceevels) and that perceived risk decreases customer outcomesndirectly via perceived service quality (for all indirect effectsf perceived risk β ≤ −.05, p < .05). In other words, customerehavior is not directly affected by risk perceptions but thatigher risk leads to lower service quality, which in turn decreasesverall and Internet store search intention, purchase intention,nd WTP. This indirect path of perceived risk may explainonflicting empirical results concerning the direct effect of riskerceptions on channel usage decisions (e.g., Kwon and Lennon009; Montoya-Weiss, Voss, and Grewal 2003). However, futureesearch is needed to clarify the ambiguous relationship betweenhannel service quality perceptions and risk perceptions.

Third, we investigated the potential reasons of why OI leadso synergies in Study 2 (i.e., significant indirect effects of OI viaerceived service quality of the Internet store on physical storeurchase intention and WTP) but not in Study 3. Notably, par-icipants in Study 2 had never bought a product from the retailerrior to the experiment (and did not know the retailer’s brand indvance), while participants in Study 3 were frequent customerst the respective retailer prior to the experiment. Thus, the sam-le of Study 2 consists of first-time customers, and the sample oftudy 3 consists of repeat customers. First-time customers areore likely than repeat customers to require a more sophisticated

hopping experience (Avery et al. 2012). Given that OI enricheshe shopping experience, one could expect that it affects store


7 Structural equation modeling results for all studies are available from therst author.



1 f Reta

fcfibkAfNt

R

somoisrassOvoo

iaaVmtaw

stclegaecat

easawari

ctcFaec

rsacsncsppws

Mdetdif

S

R

PqPOiOintentionOverallwillingness to pay(in D )

12.45a 10.65 16.04b 6.12 d = .40

4 D. Herhausen et al. / Journal o

ound that complementary effects between the Internet and storehannel can only be observed for non-routine services. For therst-time customer, buying a product from a new retailer maye a “non-routine task”, whereas a (re-)purchase from a well-nown retailer may be a “routine task” for the repeat customer.gain, this could be a potential explanation for the observed dif-

erences regarding the synergetic effects of OI across the studies.evertheless, investigating the conditions under which OI lead

o synergies holds promise for further research.

esearch Limitations and Future Research Directions

Our findings should be viewed as a first step toward under-tanding the effects of OI on retailer-level and channel-levelutcomes. Further research is needed to extend the conceptualodel and to overcome the following limitations. Due to the

bjections of the cooperating retailers, we excluded the possibil-ty of returning products bought in the Internet store to the offlinetore from our operationalization of OI in Study 3. Although weeplicated most of the findings from Study 2 with different oper-tionalizations of OI, we cannot rule out that customers valueome elements of OI more highly than others. Future researchhould therefore investigate how different design elements ofI such as the ability to check availability in the physical storeia the Internet and the ability to return products purchasednline to the offline store, and their interactions, affect customerutcomes.

Using a prototype design for OI compelled us to use behav-oral intention as a proxy for actual search and purchase behaviorcross channels. Although the use of intentions is generallyccepted as an indicator of behavior (Badrinarayanan et al. 2012;erhoef, Neslin, and Vroomen 2007), similar studies could makeore insightful claims by adopting a longitudinal research set-

ing and including actual behavior. A potential approach for such study would be to accompany the introduction of OI at a retailerith a quasi-experimental, longitudinal design.We also relied on a self-reported general measure of Internet

hopping experience in Study 2 and a retailer-specific opera-ionalization of online shopping experience provided by theooperating retailers in Study 3 since we were unable to col-ect objective data on participants’ general Internet shoppingxperience. Future research should try to measure participants’eneral Internet shopping experience with objective data thatccounts for all online purchases regardless of specific retail-rs, for example, using actual transactions data from customerredit cards. Moreover, we employed single-items for purchasend search intentions; future research should strive to improvehe measurement of these outcomes.

Although we studied the effects of OI in several product cat-gories, we only considered sensory products with non-digitalttributes (flip-flops, sunglasses, jacket). Further research shouldtudy additional, non-sensory product categories with digitalttributes such as books or CDs (Lal and Sarvary 1999). One



ould expect that OI is less important for such products as theirttributes can readily be communicated on the Web and do notequire physical inspection. Testing this assumption provides annteresting avenue for future research. A potential moderating


ondition beyond customers’ Internet shopping experience andhe product attributes that may affect the effects of OI is howustomers’ pre-perceptions of the retailer affect the findings.or example, would the results equally hold if the retailer were

well-established brick-and-mortar retailer that launches an-commerce site and if the retailer were a well-established e-ommerce site that launches brick and mortar stores?8

Furthermore, in light of the trend toward omnichanneletailing, channel integration may appear in combination byimultaneously providing online terminals in physical stores and

physical store locator in mobile channels. Reactions to jointhannel integration may differ from our results. Future researchhould investigate whether the addition of offline–online chan-el integration affects our findings. Given the challenge to designhannels to grow or at least maintain a retailer’s loyal researchhopper base and to ensure a retailer obtains its share of com-etitive research shoppers (Neslin and Shankar 2009), anotherromising avenue for future research constitutes the question ofhether and how OI affects channel lock-in and cross-channel

ynergies among different channels.

Acknowledgments

The authors thank Eric T. Anderson, Paul Liu, Antoniooreno, and Rick Wilson for helpful comments on an earlier

raft of this manuscript. In addition, they gratefully acknowl-dge the constructive guidance of J. Jeffrey Inman and thankhe three anonymous reviewers for their constructive commentsuring the review process. Special thanks go to the cooperat-ng retailers for creating the stimulus material and their valuableeedback.

Appendix. Mean (SD) Results Across ExperimentalConditions

tudy 1

etailer 1

No integration (n = 56) Integration (n = 51)

Mean SD Mean SD Effect size

erceived serviceuality

3.70a 1.39 5.22b 1.47 d = .94

erceived risk 4.54a 1.68 3.49b 1.67 d = .60verall search

ntention2.98a 1.66 4.08b 1.86 d = .60

verall purchase 3.16a 1.58 4.43b 1.81 d = .71


8 We thank an anonymous reviewer for this suggestion.



f Reta

S

R

P

PO

O

O

S

I

P

PI

I

I

S

S

S

O

S

R

PPIISS

S

R

P

PI

I

SS

Nafcwta

c

A

A

A

A

B

B

B

B

B


tudy 1

etailer 2



erceivedservicequality

3.55a 1.53 5.41b 1.34 d = 1.09

erceived risk 4.71a 1.74 3.26b 1.63 d = .79verall searchintention

3.21a 1.61 3.98b 1.93 d = .43

verallpurchaseintention

3.25a 1.47 4.17b 1.92 d = .52

verallwillingnessto pay (inD )

12.53a 8.50 16.06b 7.36 d = .44

tudy 2



nternet shoppingexperience (inpercent ofoverallpurchases)

29.12a 28.51 29.84a 29.51 d = .02

erceived servicequality

3.46a 1.38 5.28b 1.36 d = 1.10

erceived risk 5.23a 1.33 4.32b 1.44 d = .63nternet search

intention3.06a 1.50 3.80b 1.80 d = .44

nternet purchaseintention

2.40a 1.26 2.94b 1.64 d = .37

nternetwillingness topay* (in SwissFrancs)

82.54a 64.62 100.69a 71.86 d = .26

tore searchintention

3.66a 1.86 3.95a 1.86 d = .16

tore purchaseintention

3.78a 1.66 3.59a 1.68 d = .11

tore willingnessto pay (in SwissFrancs)

98.35a 66.15 105.61a 71.29 d = .11

verallwillingness topay (in SwissFrancs)

68.00a 59.07 92.12b 61.09 d = .40

tudy 3

etailer A: Store channel customers

No integration(n = 67)

Integration(n = 71)


erceived service quality 4.79a 1.37 5.92b .84 d = .90erceived risk 4.47a 1.44 3.55b 1.37 d = .64nternet search intention 5.54a 1.34 6.01b 1.02 d = .39



nternet purchase intention 2.99a 1.46 4.44b 1.57 d = .86tore search intention 5.31a 1.34 5.66a 1.23 d = .27tore purchase intention 5.69a 1.40 5.72a 1.36 d = .02

B

C


tudy 3

etailer B: Internet channel customers

No integration(n = 293)

Integration(n = 284)


erceived servicequality

5.68a 1.33 6.01b 1.07 d = .27

erceived risk 3.21a 1.36 3.19a 1.42 d = .01nternet search

intention5.71a 1.12 6.01b .99 d = .44

nternet purchaseintention

4.96a 1.54 5.36b 1.37 d = .27

tore search intention 4.94a 1.61 4.90a 1.62 d = .02tore purchaseintention

3.98a 1.96 4.39b 1.87 d = .21

ote: For each dependent variable, means not sharing a common subscript differt p < .05 (two-tailed tests). If not specified otherwise, the possible range of scoresor listed values is 1–7, with higher values indicating more positive responses. Wealculated effect sizes with the following formula: Cohen’s d = (MA − MB)/σ,here MA and MB are the two means, and σ refers to the standard deviation for

he population. For Cohen’s d, values of .8, .5, and .2 represent large, medium,nd small effect sizes.* The effect of OI on Internet willingness to pay becomes significant whenontrolling for participants income and reference price.

References

berdeen Group (2012), “The 2012 Omni-Channel Retail Experience,”(accessed December 31, 2013), [available at http://www.aberdeen.com/aberdeen-library/7101/RA-omni-channel-retail.aspx]

ccenture (2010), “Cross Channel Integration: The Next Step forHigh Performing Retailers,” (accessed December 31, 2013), [avail-able at http://www.accenture.com/nl-en/Documents/PDF/Accenture CrossChannel Integration Retail.pdf].

ilawadi, Kusum L. and Kevin L. Keller (2004), “Understanding Retail Brand-ing: Conceptual Insights and Research Priorities,” Journal of Retailing, 80(4), 331–42.

very, Jill, Thomas J. Steenburgh, John Deighton and Mary Caravella (2012),“Adding Bricks to Clicks: Predicting the Patterns of Cross-Channel Elastic-ities Over Time,” Journal of Marketing, 76 (5), 96–111.

adrinarayanan, Vishag, Enrique P. Becerra, Chung-Hyun Kim and SreedharMadhavaram (2012), “Transference and Congruence Effects on PurchaseIntentions in Online Stores of Multi-Channel Retailers: Initial Evidence fromthe US and South Korea,” Journal of the Academy of Marketing Science, 40(4), 539–57.

aker, Julie, A. Parasuraman, Dhruv Grewal and Glenn B. Voss (2002), “TheInfluence of Multiple Store Environment Cues on Perceived MerchandiseValue and Patronage Intentions,” Journal of Marketing, 66 (April), 120–41.

ecker, Gordon M., Morris H. DeGroot and Jacob Marschak (1964), “MeasuringUtility by a Single-Response Sequential Method,” Behavioral Science, 9 (3),226–32.

endoly, Elliot, James D. Blocher, Kurt M. Bretthauer, Shanker Krishnan andM.A. Venkataramanan (2005), “Online/In-Store Integration and CustomerRetention,” Journal of Service Research, 7 (4), 313–27.

ooz & Company (2012), “Cross-Channel Integration in Retail Creating aSeamless Customer Experience,” (accessed December 31, 2013), [availableat http://www.booz.com/global/home/what-we-think/reports-white-papers/article-display/cross-channel-integration-retail-creating].

ustillo, Miguel and Geoffrey A. Fowler (2009), “Wal-Mart Sees Stores as


Online Edge,” The Wall Street Journal, December 15,ompeau, Deborah R. and Christopher A. Higgins (1995), “Computer Self-

Efficacy: Development of a Measure and Initial Test,” MIS Quarterly, 19(2), 189–211.


http://www.aberdeen.com/aberdeen-library/7101/RA-omni-channel-retail.aspx

http://www.aberdeen.com/aberdeen-library/7101/RA-omni-channel-retail.aspx

http://www.accenture.com/nl-en/Documents/PDF/Accenture_Cross_Channel_ Integration_Retail.pdf

http://www.accenture.com/nl-en/Documents/PDF/Accenture_Cross_Channel_ Integration_Retail.pdf

http://refhub.elsevier.com/S0022-4359(14)00097-9/sbref0015




























































































































































http://www.booz.com/global/home/what-we-think/reports-white-papers/article-display/cross-channel-integration-retail-creating

http://www.booz.com/global/home/what-we-think/reports-white-papers/article-display/cross-channel-integration-retail-creating











































1 f Reta

C

D

D

D

D

D

D

E

F

F

G

G

G

G

G

G

G

H

H

I

J

K

L

L

M

M

M

N

N

P

P

P

P

R

R

R

S

v

v

V

V

6 D. Herhausen et al. / Journal o

ronin, J. Joseph, Michael K. Brady and G. Tomas M. Hult (2000), “Assessingthe Effects of Quality, Value, and Customer Satisfaction on ConsumerBehavioral Intentions in Service Environments,” Journal of Retailing, 76(2), 193–218.

abholkar, Pratibha A. (1996), “Consumer Evaluations of New Technology-Based Self-Service Options: An Investigation of Alternative Models ofService Quality,” International Journal of Research in Marketing, 13 (1),29–51.

avis, Fred D. (1989), “Perceived Usefulness, Perceived Ease of Use, and UserAcceptance of Information Technology,” MIS Quarterly, 13 (3), 318–40.

avis, Fred D., Richard P. Bagozzi and Paul R. Warshaw (1989), “User Accep-tance of Computer Technology: A Comparison of Two Theoretical Models,”Management Science, 35 (8), 982–1003.

iamantopoulos, Adamantios, Marko Sarstedt, Christoph Fuchs, Petra Wilczyn-ski and Sebastian Kaiser (2012), “Guidelines for Choosing BetweenMulti-Item and Single-Item Scales for Construct Measurement: A PredictiveValidity Perspective,” Journal of the Academy of Marketing Science, 40 (3),434–49.

oubleClick (2004), “Multi-Channel Shopping Study,” (accessedDecember 31, 2013), [available at http://www.doubleclick.com/us/knowledge central/research/email solutions/].

owling, Grahame R. and Richard Staelin (1994), “A Model of Perceived Riskand Intended Risk-Handling Activity,” Journal of Consumer Research, 21(1), 119–34.

dwards, Jeffrey R. and Lisa Schurer Lambert (2007), “Methods for Integrat-ing Moderation and Mediation: A General Analytical Framework UsingModerated Path Analysis,” Psychological Methods, 12 (1), 1–22.

alk, Thomas, Jeroen Schepers, Maik Hammerschmidt and Hans H. Bauer(2007), “Identifying Cross-Channel Dissynergies for Multichannel ServiceProviders,” Journal of Service Research, 10 (2), 143–60.

orsythe, Sandra M. and Bo Shi (2003), “Consumer Patronage and Risk Per-ceptions in Internet Shopping,” Journal of Business Research, 56 (11),867–75.

allino, Santiago and Antonio Moreno (2014), “Integration of Online andOffline Channels in Retail: The Impact of Sharing Reliable Inventory Avail-ability Information,” Management Science, 60 (6), 1434–51.

atignon, Hubert and Thomas S. Robertson (1985), “A Propositional Inven-tory for New Diffusion Research,” Journal of Consumer Research, 11 (4),849–67.

ensler, Sonja, Peter C. Verhoef and Martin Böhm (2012), “UnderstandingConsumers’ Multichannel Choices across the Different Stages of the BuyingProcess,” Marketing Letters, 23 (4), 987–1003.

lushko, Robert J. and Lindsay Tabas (2009), “Designing Service Systems byBridging the ‘Front Stage’ and ‘Back Stage’,” Information Systems and E-Business Management, 7 (4), 407–27.

oogle (2011), “Influencing Offline – The New Digital Frontier,”(accessed December 31, 2013), available at http://www.google.com/think/research-studies/influencing-offline-the-new-digital-frontier.html].

rewal, Dhruv, Ramkumar Janakiraman, Kirthi Kalyanam, P.K. Kannan, BrianRatchford, Reo Song and Stephen Tolerico (2010), “Strategic Online andOffline Retail Pricing: A Review and Research Agenda,” Journal of Inter-active Marketing, 24 (2), 138–54.

ulati, Ranjay and Jason Garino (2000), “Get the Right Mix of Bricks & Clicks,”Harvard Business Review, 78 (3), 107–14.

ayes, Andrew F. (2013), Introduction to Mediation, Moderation, and Condi-tional Process Analysis: A Regression-Based Approach, New York: GuilfordPress.

itt, Lorin M. and Frances X. Frei (2002), “Do better Customers utilize Elec-tronic Distribution Channels? The Case of PC Banking,” ManagementScience, 48 (6), 732–48.

nman, J. Jeffrey, Venkatesh Shankar and Rosellina Ferraro (2004), “The Roles ofChannel-Category Associations and Geodemographics in Channel Patron-age,” Journal of Marketing, 68 (2), 51–71.

udd, Charles M., Vincent Y. Yzerbyt and Dominique Muller (2014), “Mediation



and Moderation,” in Handbook of Research Methods in Social and Person-ality Psychology, Reis Harry T. and Judd Charles M., eds. Cambridge, UK:Cambridge University Press, 653–76.

W


won, W. and S. Lennon (2009), “What Induces Online Loyalty? OnlineVersus Offline Brand Images,” Journal of Business Research, 62 (5),557–64.

al, Rajiv and Miklos Sarvary (1999), “When and How is the InternetLikely to Decrease Price Competition?,” Marketing Science, 18 (4),485–503.

ichtenstein, Donald R. and William O. Bearden (1989), “Contextual Influ-ences on Perceptions of Merchant-Supplied Reference Prices,” Journal ofConsumer Research, 16 (1), 55–66.

enon, Satya and Barbara Kahn (2002), “Cross-Category Effects of InducedArousal and Pleasure on the Internet Shopping Experience,” Journal ofRetailing, 78 (1), 31–40.

ontoya-Weiss, Mitzi M., Glenn B. Voss and Dhruv Grewal (2003), “Deter-minants of Online Channel Use and Overall Satisfaction with a Relational,Multichannel Service Provider,” Journal of the Academy of Marketing Sci-ence, 31 (4), 448–58.

uller, Dominique, Charles M. Judd and Vincent Y. Yzerbyt (2005), “WhenModeration is Mediated and Mediation is Moderated,” Journal of Personalityand Social Psychology, 89 (6), 852–63.

eslin, Scott A. and Venkatesh Shankar (2009), “Key Issues in MultichannelCustomer Management: Current Knowledge and Future Directions,” Journalof Interactive Marketing, 23 (1), 70–81.

ysveen, Herbjørn and Per E. Pedersen (2004), “An Exploratory Study ofCustomers’ Perception of Company Web Sites Offering Various Interac-tive Applications: Moderating Effects of Customers’ Internet Experience,”Decision Support Systems, 37 (1), 137–50.

ark, Cheol and Jong-Kun Jun (2003), “A Cross-Cultural Comparisonof Internet Buying Behavior: Effects of Internet Usage, PerceivedRisks, and Innovativeness,” International Marketing Review, 20 (5),534–53.

atrício, Lia, Raymond P. Fisk and João Falcão e Cunha (2008), “DesigningMulti-Interface Service Experiences,” Journal of Service Research, 10 (4),318–34.

auwels, Koen, Peter S.H. Leeflang, Marije L. Teerling and K.R. Eelko Huiz-ingh (2011), “Does Online Information Drive Offline Revenues? Only forSpecific Products and Consumer Segments!,” Journal of Retailing, 87 (1),1–17.

eck, Joann and Terry L. Childers (2003), “Individual Differences in HapticInformation Processing: The Need for Touch Scale,” Journal of ConsumerResearch, 30 (3), 430–42.

atchford, Brian T. (2009), “Online Pricing: Review and Directions forResearch,” Journal of Interactive Marketing, 23 (1), 82–90.

igby, Darrell (2011), “The Future of Shopping,” Harvard Business Review, 89(12), 65–76.

yan, Tom (2013). “Is Showrooming a Problem For E-Tailers?,” (accessedDecember 31, 2013), [available at http://www.retailwire.com/news-article/17132/is-showrooming-a-problem-for-e-tailers].

tewart, Katherine J. (2003), “Trust Transfer on the World Wide Web,” Organi-zation Science, 14 (1), 5–17.

an Birgelen, Marcel, Ad de Jong and Ko de Ruyter (2006), “Multi-Channel Service Retailing: The Effects of Channel PerformanceSatisfaction on Behavioral Intentions,” Journal of Retailing, 82 (4),367–77.

an Doorn, Jenny and Peter C. Verhoef (2011), “Willingness toPay for Organic Products: Differences Between Virtue and ViceFoods,” International Journal of Research in Marketing, 28 (3),167–80.

enkatesan, Rajkumar, V. Kumar and Nalini Ravishanker (2007), “MultichannelShopping: Causes and Consequences,” Journal of Marketing, 71 (April),114–32.

erhoef, Peter C., Scott A. Neslin and Bjorn Vroomen (2007), “Multi-channel Customer Management: Understanding the Research ShopperPhenomenon,” International Journal of Research in Marketing, 24 (2),129–44.


akefield, Kirk L. and Jeffrey J. Inman (2003), “Situational Price Sensitivity:The Role of Consumption Occasion, Social Context and Income,” Journalof Retailing, 79 (4), 199–212.







































































































































































http://www.doubleclick.com/us/knowledge_central/research/email_solutions/

http://www.doubleclick.com/us/knowledge_central/research/email_solutions/



















































































































































































































http://www.google.com/think/research-studies/influencing-offline-the-new-digital-frontier.html

http://www.google.com/think/research-studies/influencing-offline-the-new-digital-frontier.html

































































































































































































































































































































































































































































































































































http://www.retailwire.com/news-article/17132/is-showrooming-a-problem-for-e-tailers

http://www.retailwire.com/news-article/17132/is-showrooming-a-problem-for-e-tailers





























































































































































f Reta

W

Z

Z

Zand Kenny: Myths and Truths about Mediation Analysis,” Journal of Con-


allace, David W., Joan L. Giese and Jean L. Johnson (2004), “CustomerRetailer Loyalty in the Context of Multiple Channel Strategies,” Journalof Retailing, 80 (4), 249–63.

eithaml, Valarie A., Arun Parasuraman and Arvind Malhotra (2002), “Service



Quality Delivery through Web Sites: A Critical Review of Extant Knowl-edge,” Journal of the Academy of Marketing Science, 30 (4), 362–75.

hang, Jie, Paul W. Farris, John W. Irvin, Tarun Kushwaha, Thomas J. Steen-burghe and Barton A. Weitz (2010), “Crafting Integrated Multichannel

Z


Retailing Strategies,” Journal of Interactive Marketing, 24 (2), 168–80.

hao, Xinshu, John G. Lynch Jr. and Qimei Chen (2010), “Reconsidering Baron


sumer Research, 37 (2), 197–206.immerman, Ann (2012), “Showdown Over ‘Showrooming’,” (accessed

December 31, 2013), [available at http://online.wsj.com/news/articles/].













































































































http://online.wsj.com/news/articles/