the phases of erp software implementation and maintenance

8/3/2019 The Phases of Erp Software Implementation and Maintenance

1/10

THE PHASES OF ERP SOFTWAREIMPLEMENTATION AND MAINTENANCE:A MO DEL FOR PREDICTING PREFERR ED E RP USEALAN R. PESLAKPenn State UniversityDunmore. PA 18512

GIRISH H. SUBRAMANIANPenn State Harrisburg GEORGE E. CLAYTONPenn State Harrisburg

ABSTRACTNearly all information technology educators and practitionersare familiar with the concept of a systems development life cycle(SDLC). The SDLC model is mainly ba.sed on new software de-velopmenl. The stages vary based on methodology, but mostlyinclude planning and requirements definition, analysis, design ofthe new system, implementation, and post-implementation sup-port such as maintenance and security. Less explored and known

are the stages involved in commercial off the shelf installation.The actual implementation of commercial off the shelf softwareERP (enterprise resource planning) systems is an area that is sug-gested to have a life cycle of its own. The importance of specificphases of this ERP implementation life cycle is the subject ofihis study. By surveying SAP enterprise resource planning imple-mentations at two divisions of a major manufacturing company,a coniirmatory factor analysis confirms the existence of fourdistinct phases of ERP implementation preparation and train-ing, transition, performance and usefulness, and maintenance. Inaddition, the overall preferred ERP use is studied and comparedwith these factors. It was found that the two significant phasesv^hich directly influenced preferred ERP use were preparationand training phase, and performance and usefulness phase. Nei-ther transition nor maintenance was found to significantly affectpreferred ERP use. This suggests to practitioners that more focusneeds to be placed on the key determinants of preferred ERP use preparation and training phas e, and performance and useful-ness phase.

Keywords: enterprise resource planning, systems develop-ment life cycle, preparation, training, performance, usefulness.INTRODUCTION

One of the most important and costly processes in organiza-tions today is the implementation of enterprise resource planningsystems. An ERP system is an integrated comm ercial off-the-shelf(COTS) software package that can perform all the major businessfunctions of an organization. These functions generally includeall elements of the value chain from raw material purchases, in-ventory management, production, goods shipments, invoicing,accounting, and human resource management. ERP systems hadtheir roots in manufacturing, including material resource plan-ning, but quickly grew to include all the other related businessfunctions. They now serve as the basic business systems for mostof the large and mid-size organizations in the world today. Thekey elements of an ERP system according to Miller [32] are: onelarge real-time database which reduces data redundancy and im-proves accuracy; integrated business process that cut across busi-ness functions such as supply chain management; and seamlesstransitions between business transactions.

The importance and size of the ERP market have grown sig-nificantly. On the low end, it is estimated that ERP systems willexceed $11.9 billion in sales in 2001 [4]. Others suggest a 20(14market as high as $79 billion in 2004 |6]. According to Bingi,Sharma and Godia |9 ], seventy percent of the Eortune 10C)0 haveor will have ERP systems. Fisher ct al. [201 note that mid-levelorganizations are now a major market for ERP software imple-mentations .With multiyear, multi-firm ERP implementation andfinancial data, firms that invest in ERP tend to show higher per-formance across a wide variety of financial metrics [26]. Clearly,with such a prevalent role in information technology, the impor-tance of successful im plementation of these systems is of great im-portance to practitioners. The massive ERP implementations thatmust take place within organizations are not without challenges.There have been many difficult and costly implementations ofERP systems that have adversely impacted many organizationsincluding EoxMeyer Drug, Dell Computer, Applied Materialsand Dow Chemical [ 16|. Over half of ERP implementations endin failure [8]. Hong and Kim [27] suggest even poorer resultswith 75% of "ERP projects judged to be unsuccessful." Scott andVessey [43] believe that 90% of ERP projects are late. Huang etal . [29] note ERP systems cost "organizations a huge amount ofmoney and manpower. Therefore, even major ERP vendors can-not guarantee the success of implementation." There can be gapsbetween business processes and software [23]. As a result, it isextremely important to understand factors that can influence suc-cess of an implementation.

The acceptance of new sys tems, particularly large syslems thattransform the organ ization, is of vital importance to organizationstoday. As noted by Robey et al.[42| "newer technologies such asenterprise resource planning (ERP) systems continue to be asso-ciated with the agenda of organizational transformation,. .. (yet)despite the transformation agenda accompanying the new system,users initially chose to avoid using it as much as possible (inertia)and later to work around system constraints in unintended ways".Soh, Sia. Boh, and Tang, M. [46] found many misalignment is-sues which can adversely impact ERP acceptance Resistance toacceptance of the new system can adversely affect use and reducethe gains that can be realized with successful implementation. Itis vital that key drivers for ERP system preference be identifiedand measured.

The extent and nature of ER P use does determine ER P success[5] and hence the need to study the determinants of ERP use. Thisstudy then is a review and enhancement of a specific ERP imple-mentation model and an organizational study of how phases inthis model can infiuence and affect preferred use of ERP system s.The study uses systems development theory to build the ERP im-plementation model. Using data on ERP implementation at twodivisions of a manufacturing organization, the important determi-nants of ERP use are extended and empirically confirmed.

Winter 2007-2008 Journal of Computer Information Systems


2/10

LITERATURE REVIEWERP Implementation Models and Phases

Boudreau and Robey [10] as noted, suggest a vital impor-tance to acceptance of ERP systems. Currently they note that ifnot successfully implemented, users may work around the systemand otherwise doom the project to costly duplication of effort,or worse, system failure. A phased implementation approach ishighlighted in Robey et al. [42]. It is important to have a struc-tured approach, similar to systems development, for the imple-mentation and maintenance of ERP systems.Systems development theory uses the concept of a lifecycleand stages in the lifecycle to indicate developmen t of informationsystems. The waterfall model, incremental model. RAD {rapidapplication development) model and spiral model are some ofthe systems development methods prevalent in the literature

[39]. Newer approaches to systems development address com-ponent-based developm ent using off-the-shelf packages, agiledevelopm ent and the unified process for object-oriented softwaredevelopment [39]. The new er approaches have fewer stages in thedevelopment of systems. For example, the unified process whichdraws upon the best practices of conventional software processmodels [39] has inception, elaboration, construction and transi-tion phases. A common aspect of all these models is that theyfocus little attention on implementation and the post implementa-tion of the system.

Empirical research has addressed issues that organizationsface on and after Implementation of systems. Specifically, sev-eral studies have looked at ERP implementation [1], [27], [42].The implementation and performance stage model [31], [15] is auseful tool for understanding the implementation of the ERPtechnology and provides six stages: initiation, adoption, adap-tation, acceptance, routinization. and infusion. This six-stagemodel sets the framework to investigate the implementationand performance issues of utilizing an ERP system withinan organization. The initiation stage analyzes the factors thatinfluence the decision to utilize an ERP system such as incom-patibility, need for connectivity, top management vision, andneed to change. Implementation issues are addressed in theadoption and adaptation stages including: investment decisions.cost/benefit analysis, and choice of appropriate technology.Implementation and performance measures such as systemmodifications, training, integration of functional units, enhancedperformance, user acceptance, flaws corrected, and organization-al integration realized, are identified during the acceptance androutinization stages. Finally, the infusion stage addresses futureinnovations including IT integration at global levels and futureopportunities.

Parr and Shanks [38] review different models of ERP imple-mentation and suggest a PPM/CS F hybrid model that incorporatesa project phase model (PPM) with critical success factors (CSF),The phases included in their model include planning, project, andenhancem ent. O ur w ork i.s both an extension and a testing of theirproject phase model for ERP implementation. Parr and Shanks[38] are the first to suggest that "there is justification for creatinga project phase model (PPM) of ERP implementation which iscentered on the individual, discrete phases of the implementationproject itself rather than one which treats the project as just an-other phase in the whole implementation enterprise." Our modelexplores preference for a new system based on Parr and Shanksmodel but the third phase of project is broken into two separate

phases transition and performance to better understand whattruly influences project acceptance.What happens after ERP implementation, and the benefitsderived from ERP implementation, is specifically addressed byGattiker and Goodhue [22]. The model by these researchers looksat the subunit level of the organization, similar to our study, andlooks at determinants of ERP benefits. Task efficiency, coordina-tion improvements and data quality explain a large amount of theERP benefits of the sub-unit [22].

ERP IMPLEMENTATION ANDMAINTENANCE PHASES OE OUR STUDY

Based on an exploratory study, Rajagopal and Frank [41]revealed that at Owens Coming, marketing and manufacturingeach had its own sales forecasting numbers prior to implement-ing their ERP system. As a result, there were discrepanciesin the information used for strategic decision making in thesame organization. This was a result of the incompatibilityamong the information systems used in various functions atOwens Coming [40]. System incompatibility can also impedeperformance, as observed at Eastman Kodak Company. At onepoint. Eastman Kodak was operating their business on 26(X)different software applications, more than 4000 system inter-faces, and about 100 different programming languages runningon legacy-type mainframe systems [48]. It thus can be conclud-ed that proper preparation and planning is needed for ERP im-plementation. Implementation planning influences ERP systemperformance [56].

User training that included both technical and business pro-cesses, along with a phased implementation approach, helpedfirms to overcome assimilation knowledge barriers [42] notingthe importance of training and a phased approach to ERP imple-mentation. Gupta [24] and Umble and Umbie [53] saw training asone of the most important factors in ERP succ ess. Training is alsonoted as an important factor by Gallivan et al . [21 ] and Barker andFrolick [8 ]. Training is seen as a crucial component in continuousimprovement. Training helps to improve employee participationand involvement in quality programs through propagation of pri-orities and missions of the organization [25|. So, as ERP is tiedclosely with business processes and continuous quality improve-ment, it can be seen that training would influence ERP success."One of the primary challenges is selecting an appropriate planfor end-user education. . . . Managers must avoid this mistake bylooking at various options for end-user ERP training at the begin-ning." [24]. "People are one of the hidden costs of ERP imple-mentation. Without proper training about 30 percent to 40 percentof front-line w orkers will not be able to handle the dem ands of thenew system."[9] So. preparation for ERP implementation is ex-tremely important. In preparation and training phase, we caninclude the initiation, adoption and adaptation phases of CooperandZmud [15].

This model includes the three phases of the Parr and Shanks[38] model but the second phase project is broken into twosubphases transition and performance. Gupta [24] has notedthe importance of performance as a key success factor in ERPimplementation. Al-Mashari. Al-Mudimigh. and Zairi [2] notethe importance of transition suggesting that "it is importantthat an organization approaches the transition of legacy systemcarefully and with a comprehensive plan". Davenport [16] sug-gests legacy systems represent a major issue for new IT projects.Boudreau and Robey [10] identified tninsition as a key factor in


3/10

ERP implementation. Transition is also an important phase in theunified prixress model [39]."A key issue when implementing ERP is performance. Tyingsuch an array of decision support logic to a single database en-gine process can generate huge amounts of load on conventionalsystems. . . . Modem businesses need a much faster and moredynam ic decision structure something that ERP supports but(these) earlier systems cannot. To combat this, ERP vendors are

creating new versions of MRP and MPS under a single umbrelladubbed advanced planning and scheduling (APS). Though someof these engines are still being tested, they promise drasticallyfaster response times and much better business resu lts in the formof accurate inventory planning and precise delivery schedules.With faster background engines like APS combining with ubiq-uitous front-end Web browser access, you have the foundationto build self-service business systems like the insurance examplecited earlier. Com panies that set up limited self-service functionstoday will have a competitive edge." [24]. Nah, Zuckwiler. andLau [36] in their survey of Fortune 1000 CIO s found performanceas a major factor in ERP su ccess. Siau [45] sees quality of serviceas essential in ERP implementations.Usefulness is also related to performance. Perceived useful-ness looks at productivity, job effectiveness, and ease of doingthe job [44], [51] which could be argued as performance relat-ed variables. For example, our performance phase also looks atproductivity and use of information from the ERP system by theuser. Extensive research supports the notion that usefulness andease of use are primary drivers of user intentions to adopt newtechnology [17], [18J, [51], [54]. Howev er, this research has beenconducted primarily in environments in which adoption was vol-untary 112]. Brown et al. [12] found significant impact on tech-nology acceptance in mandated use environments and concludethat "results suggest that TAM (Technology Acceptance Model)and related variations of the model do not generalize directly tomandatory use situations." Hence, we look at the performance

and usefulness phase."The problem of maintaining integrated applications is no meansa simple one and requires an interdisciplinary approach" [33]."Without the understanding of how the system is implemented,and how to maintain the efficiencies and functionality of that(ERP) system, it will be useless to the organization." [7]. Theuse of packaged software is shown to result in decreased soft-ware complexity and software enhancem ent effort [7[ and so it isexpected that ERP packages would have reduced maintenance in

comparison to traditional development. While studies on softwaremaintenance have often considered maintenance as the dependentvariable, the influence of maintenance on preferred use of a sys-tem is not studied. So, the final phase is maintenance.

RELATIONSHIP OF ERP IMPLEMENTATIONPHASES TO PREFERRED ERP USE -RESEARCH MODEL AND HYPOTHESES

An organization has to prepare itself for ERP implementa-tion much before the actual ERP decision is taken [5] and theseresearchers believe that such preparation is key to ERP success.It may take years of build-up and requisite preparation with ITdiffusion and infusion to determine, in large part, whether usersdemonstrate a combination of a positive predisposition to a newsystem when they view it from the standpoint of their work. Ad-ditional support for such requisite build-up comes from Bresna-han & B rynjolfsson [ 11 [, who use the phrase "complementary in-vestment." "We have reason to believe that organizations that arebetter prepared to absorb information technology will be able toleverage the benefits of information technology'"[5]. One of thecritical factors for influencing users" attitudes and involvementcan be understood in terms of the "tuition paid for the learning'[19] that precedes any part of ERP implementations showing theimportance of training for ERP use. Hence, we expect a posi-tive relationship between preparation and training p hase andpreferred ER P use.

Boudreau and Robey |10] identified transition as a key fac-tor in ERP implementation. Transition is also an important phasein the unified process model [39] used in building informationsystems. The unified process model combines all the differentsystems development frameworks into a unified process model.So , we expect a positive relationship between transition ph aseand preferred ERP use.

Zviran et al, [57] note that perceived usefulness infiuencesuser satisfaction in ERP systems. Amaolo-Gympah and Salam [3]show that training and project communication influence sharedbeliefs about benefits of technology and these beliefs influenceperceived usefulness and ease of use of the technology. Job rel-evance, output quality, and result demonstrability significantlyinfluenced user acceptance in the TAM2 model showing theimportance of performance related variables and their influenceon user acceptance [54 ]. Task efficiency, coordination improve -ments and data quality (key performance variables) explain alarge amount of the ERP benefits of the sub-unit [22] pointing

Preparationan dTrainingPhase

TransitionPhase

Performanceand UsefulnessPhase

MaintenancephaseFigure 1 shows the four ERP implementation and maintenance phases.

Winter 2007-2008 Journal of Comp uter Information Systems 27


4/10

to the importance of performance in ERP systems. Zviran et al,[57] report that ERP systems are associated with high levels ofuser satisfaction and perceived usefulness. Perceived usefulness(productivity, job effectiveness, ea se of doing job) is shown to in-fluence predicted future usage (51 ]. Hence, we expect a positiverelationship between performance and usefulness phase andpreferred ER P use.

Staehr. Shanks, and Seddon [49] found strong importance inthe post implementation period of ERP implementation. The useof packaged software is shown to result in decreased softwarecomplexity and software enhancem ent effort |7] and so it is ex-pected that ERP packages would have reduced maintenance incomparison to traditional development. This result could be afactor for users to prefer ERP use as it would involve reducedmaintenance. So, we expect a positive relationship between themaintenance phase and preferred ERP use.

We argue, through our research model, that preparation andtraining, transition, performance and usefulness, and maintenancephases of ERP implementation and maintenance activity in or-ganizations would positively influence the preferred use of ERPsystem.The dependent variable of our study is preferred ERP use.Usage variables are seen as the key variables in systems studies

[50]. End-users" reluctance or unwillingness to adopt or use thenewly implemented ERP system is often cited as one of the mainreasons for ERP failures [30], [36[, [55]. Perceived compatibilityand perceived ease of use have both direct and indirect effects(mediated by attitude) on symbolic adoption, while perceived fitand perceived usefulness influence symbolic adoption by being

fully mediated through attitude [55]. We argue that, in situationsof mandatory use. we need to study the preferred use of the ERPsystem (as a measure of user acceptance of mandated ERP sys-tem) in comparison to the system it replaced.Hypothesis L There are preparation and training, transition,performance and usefulness, and maintenance phases in the ERPimplementation and maintenance activity of an organization.Hypothesis 2. The preparation and training, transition, per-formance and usefulness, and maintenance phases in the ERPimplementation and maintenance activity of an organization havea positive influence on preferred ERP use.Hypothesis 3. There are differences in the influence of thepreparation and training, transition, performance and usefulness,and maintenance phases in the ERP implementation and mainte-nance activity of an organization on preferred ERP use.The first hypothesis explores the idea that ERP project imple-mentation and maintenance consists of phases as proposed byPan and Shanks [38]. Their general model was used as the basisfor developing an approach to answer whether specific phasesof SAP ERP implementation can be identified. Other research-ers such as Cooper and Zmud [15] have also identified an ERP

implementation model. Several other researchers have identifiedspecific phases preparation [56], transition [2]. [10], [39]. per-formance [24], and maintenance [7], [33] in the ERP implementa-tion and maintenance activity of an organization. We would liketo confirm the existence of these phases in ERP implementationand maintenance.The second hypothesis tests the influence of these fourphases on preferred ERP use. The third hypothesis is a logical

Hyp. 1

Preparation& Training

Transition

Performance& Usefulness

Hyp.2 &Hyp. 3

Figure 2. Research Model and Hypotheses


5/10

extension of the second. If a model of ERP implementation canbe developed, it is logical that some steps in this process may bemore important than others. Many studies find varying influencesof independent variables on dependent variables. Chau [13] forexample found viiriability in the influences of computer attitudeand self-efficacy on IT usage b ehavio r.Finally, though variability may exist and be found, some fac-tors may not have statistical significance on the overall depend entvariable of preferred ERP use. Chau [13], as an example, foundno statistical significance at p < .05 of computer attitude on per-ceived ease of use.

M E T H O D O L O G YTo determine the answers to the preceding research questions,a paper based survey was prepared and administered at two divi-sions of a mid-size organization that had implemented an SAPR/3 ERP system. SAP is the German based, largest seller of ERPsystems in the world. The survey population included employ-ees in tbe areas of management, production, human resources,engineering, administration, quality, and maintenance who were

employed at the two companies during the SAP implementa-tion and who used the SAP system. A cover letter and surveyquestionnaire were distributed to employees who were employedduring the ERP implementation and also used the ERP systemin their day-to-day job. Twenty five surveys completed in thefirst division were reported in Subramanian and Hoffer [52] andtwenty eight surveys completed in tbe second division were re-ported in Clayton [14]. Subramanian and Hoffer [52]'s studywas an exploratory study that reported on the twenty five surveysand preliminary results for the first division only. Clayton [I4]'swork reported the data set and some preliminary descriptive dataresults in his Master thesis report. The authors of this study ob-tained these datasets, combined the datasets from Subramanianand Hoffer [52] and Clayton [14] and used the combined datasetof 53 surveys. The literature review, research model, hypotheses,confirmatory factor an alysis, and regression analysis are all origi-nal to this research. The complete set of questions with the abbre-viations used in the data analysis is presented in table I. Fourteenquestions were prepared that suggested issues and steps in theimplementation of an ERP system. An additional question (15)was included to measure preferred ERP use as an indicator of ac-ceptance of the new system. This is the dependent variable in thisstudy.Responses to the questions were measured on a 5 point Lik-ert scale from 1 = strongly disagree to 5 = strongly agree. Allresponses were analyzed using SPSS 10.0 and AMOS 4. Tech-niques included confirmatory factor analysis, scale reliability,

and structural equation modeling.DATA ANALYSIS AND RESULTS

In total, there were 53 valid resp onses to our ques tionnaire. Asnoted, the responses came from two divisions of a manufacturingorganization that had converted to SAP R/3 ERP systems. All ofthe respondents were employed in their respective organizationsduring the transition/implementation of SAP. The length of timethe individuals were employed by their organizations was gener-ally long. Only 27% of the respondents had 10 years or less ofservice. It is suggested that the long time in service would yieldan established pattern of work proce sses that would be difficult toalter. A very high percentage of the respondents (85%) had used

USDTY

RECEIVED

PREPARED

TEAM

SMOOTH

PRODUCT

ACCESS

NEW

DATAENTRDATE

EXPECT

RELEVANC

ACCURACY

UNDERY

PREFER

Table1.

2.

3.

4.

5.

6.

7.

8.

9.10.

11.

12.

13.

14.

15.

1. Survey QuestionsI feel that I understand wh y our com -pany implemented SAP as our Infor-mation System, (understand)I was provided with on-the-job train-ing before we went live with SAP.(training)On tbe day we went live with SAP. 1felt prepared to work on SAP. (pre-pared to work)Everyone worked together to helpthe transition to SAP be successful.(team w ork)From my perspective, the transitionto SAP went smoothly. (Smooth tran-sition)I am able to be more productive withSAP than I was before SAP. (pro-ductivity)I can access information easier withSAP than I could before SAP. (ac-cess information)SAP provides me with infonnationthat was previously unavailable tome. (new information)With SAP. there is less entering ofdata than before SAP. (data entry)With SAP, the data I use is more ac-curate than before SAP. (data accu-racy)I understand what is expected of meto maintain/update/confirm tbe in-formation in SAP. (expectations)I understand how the information Iinput/change/confirm in SAP is rel-evant to our company, (relevancy)1 feel confident that the inform ationI input/change/confirm in SAP is ac-curate and correct, (accuracy)The information ! use from SAPmakes sense to m e. (understandabil-ity)I would rather use SAP than our pre-vious system, (perceived use)

the systems that were previously in place. Finally, the respon-dents had high experience levels with 92% experienced in two ormore systems during their career.!n order to address the tirst hypothesis, confirmatory fac-tor analysis (CFA) was performed on the fourteen SAP ERPimplementation questions. The extraction method was principalcomponent analysis with Varimax rotation. Some questions werediscarded after a few attempts at CFA but 10 questions remainedand were found to measure four unique factors preparationand training, transition, performance and usefulness, and main-tenance. These components are shown in table 2 and 3. Table 2shows that four factors exceeded an eigenvalue of I which is usedfor acceptable cutoff [34]. The four factors represent a very high74 % of the variance in the dependent variable, preference for thenew system.

In addition scale reliability was performed on the factors andWinter 2007-2008 Journal of Comp uter Information Systems 29


6/10

Table 2. Confirmatory Factor A nalysis Total Variance

Component

I23456789

10

InitialEigenvalues

Total

2.6802.2591.3701.097.747.587,456.378J 3 4.191

% o fVariance26.803

22.58713.70410.9757.4695.8724.5583.783X3431.907

Cumulative %

26.80349.38963.09374.06881.53787.40991.96795.75098.093

100.000

ExtractionSums of Squared

Loadings

Total

2.6802.2591.3701.097

% ofVariance

26.80322.58713.70410-975

Cumulative %

26.80349.38963.09374.068

Table 3. PCA (Principal Component Analysis) Rotated Compo nent M atrix

RECEIVEDPREPAREDTEAMSMOOTHPRODUCTACCESSNEWDATEEXCTACCURACYAlpha

ComponentPerformance

and Usefulness-0.039850.0947180.059332-0.006320.7536570.8262320.8842210.6030830.080288-0.00154

.7969

Preparationand training

0.9190880.8283030.1750640.141407-0.17567-0.035140.05052

0.1678180.232754-0.05009

.7674

Maintenance0.1506480.0033790.0764980.139667-0.020850.170766-0.074780.0199280.8631940.910473

.7354

1

Transition0.0776210.2759530.8630910.8308440.270157-0.05761

-0.09560.0313310.1462750.078177

.7314Table 4. Regression Model Sum mary

Model

1

R

.817

R Square

,668

AdjustedR Square.637

Std. Errorof theEstimate.5955all factors were above .70, the minimum for acceptability [37].These are shown along with the factors and Varimax rotation inTable 3. Table 3 also shows the loadings of the questions on thefour factors.

Hypothesis I has been supported. Preparation and training,transition, performance and usefulness, and maintenance phases

are evident in the ERP implementation and maintenance activityof an organization.Since it was found that there were separate phases and factorsin ERP implementation, the second hypothesis requires an analy-sis of the relative influences of each factor. The factors developedin the CFA were saved and entered into SPSS lO.O for multipleregression analysis. The analysis finds that a statistically signifi-cant regression equation is formed. The R squared is relativelyhigh .67 and shown in table shown in Table 4.The detailed results of the model are shown in table 5. Allphases of the ERP implementation have a positive impact on pre-ferred ER P use with the exception of transition. The po sitive coef-ficient of the estimate demonstrates the positive impact. As noted,

transition is marginally negative (-.06) but the table also shows


7/10

Table 5. Regression AnalysisUnstandardized

Model1 (Constant)PreparationTransitionPerformanceMaintenance

CoefficientsBStd.3.544.377

-6.422E-02.691.122

Error.086.089.087.088.086

StandardizedCoefficientsBeta.373

-.065.695.125

t

41.1624.233-.7347.8991.417

Sig.

.000

.000

.467.000,164a Dependent Variable: PREFER

the significance of each phase. Transition is negative but is notsignificant at p < .05. In addition, though positive, maintenanceis not a significant factor in ERP preference. Hypo thesis 2 is par-tially supported. The preparation and training, and performanceand usefulness phases of an ERP implementation have a positiveinfluence on preferred ERP use. The transition and maintenancephases do not.The relative regres.sion weights show large difference be-tween the factors. Performance and usefulness is largest at.691 and significant at p < .0 01 . followed by preparation andtraining at .377 and significant at p < .0 01. The other factorsare much less important and are found not to be significant.Hypothesis 3 has been supp orted. There are differences in theinfluence of the preparation and training, transition, performanceand usefulness, and maintenance phases in the ERP implemen-tation and maintenance activity of an organization on preferredERP use.

LIMITATIONSThe major limitation of the study is the small sample size.Though the results are based on two real world implementationsof SAP software sy.stems. results may not be applicable to allSAP R/3 implementations. A more detailed study involving moreparticip ants in more indus tries is a fruitful avenu e for further re-search. Nevertheless, the results are strong and indicate a direc-tion for both further research and practical use.

DISCUSSIONA confirmatory factor analysis verified the existence of fourdistinct phases of ERP implementation preparation and train-ing, transition, performance and usefulness, and maintenance. It

was also found that the two significant factors which directly in-fluenced preferred ERP use were preparation and training phaseand performance and usefulness phase. Neither transition normaintenance was found to significantly affect preferred ERP use.Bagchi et al |5] believe that ERP preparation is a key to ERP suc-cess. Gupta |24] points to training as an important factor for ERPsuccess. Somers and Nelson [47] found that user training had ahigher than expected importance in ERP implementation. As ERPu.se is a key aspect of ERP success, it is no wonder that prepara-tion and training phase has a significant effect on preferred ERPuse. Usefulness is shown to impact usage [ 17|. |44] and predictedfuture use [51]. Performance variables also influence user atti-tude [54[ and ERP benefits [22]. So, performance and usefulnesshave a significant effect on preferred ERP use. While transition

and maintenance are key phases in the ERP implementation andmaintenance [7]; (33]. they do not have a significant effect on pre-ferred ERP use. It is possible that m aintenance did not play a rolein preference due to the commonality of maintenance issues onboth old and new systems. The need for proper maintenance ex-ists in both situations therefore preference may not be impacted.The lack of effect of the transition phase is a bit more puzzling.This is an area that deserves ftirther exploration.

There are two main implications of this study for researchers.This study has developed specific phases in ERP implementation.This study has found that two phases of ERP implementation aremore important than others for overall preference of the new sys-tem. These areas provide fertile opportunities for further researchin ERP implemen tations but they also provide a framework forfurther studies on other IT implementations. The study can beextended and tested on other COTS packages such as CRM (cus-tomer relationship management), on custom developed applica-tions, and on coordinated outsourced projects. There are largepotential opportunities across a broad spectrum of informationtechnology projects.There are also significant implications for practitioners asa result of this study. Estimates suggest that at least 30% of allIT projects fail and it is suggested that a higher proportion ofERP projects fail, run well over budget, or fail to achieveprojected cost savings and strategic advantages. This study, asnoted, has prepared a model for phases of ERP implementation.It has also found that preparation and performance are the keyinfluences in preference for the new system. If resources andefforts are concentrated in these two phases of the ERP imple-mentation, it is likely that costs will be reduced, projects willexhibit higher levels of success, and greater strategic advan-tages will be garnered. The opportunity to test this approach inother information technology projects may suggest even greater

opportunities.CONCLUSION

The installation of an ERP system can be one of the mostcostly and critical projects that an organization undertakes.Generally, all other information technology projects pale in com-parison to ERP system s. ERP system s iire used to run all standardbusiness processes in an organization. The success or failure ofthe implementation can be vitally important to both current prof-its and the future viability of a business. As a result, understand-ing the processes and phases involved in the implementation ofthis endeavor can pay major dividends. This study has exploreda phased mode! for ERP implementation and found four distinct

Winter 2007-2008 Journal of Computer Information Systems 31


8/10

phases preparation and training, transition, performance andusefulness, and maintenance. The model is an empirical con-firmation of the implementation phased approach of Parr andShanks [38] with the extension of the middle phase of their model("Project") into the two sub-phases of transition [24], [10] andperformance and usefulness [24], [17], [54|. The m odel was em-pirically confirmed at two manufacturing implementations. Thediscovered significance and importance of two particular phasesin the model preparation and training, and performance andusefulness are important results of the study. This study, thoughlimited, has explored an area that can hold significant opp ortunityfor more successful ERP implementations. The authors seek toexpand these preliminary findings in more companies and on awider array of projects.

REFERENCES[1] Akkerm ans, H. and van Helden. K. "Vicious and VirtuousCycles in ERP Implementation: A Case Study of Interrela-tions between Critical Success Factors," European Journal

of Information Systems 11. 2002, 35-46.[2] Al-Mashari. M.. Al-Mudimigh, A. and Zairi. M. "ER P: ATaxonomy of Critical Factors." European Journal of Op-erational Research 146. 2003. 352-364[3] Ama oako-Gyampah, K. and Salam. A. "An extension of thetechnology acceptance model in an ERP implementationenvironment." Information and Management A\.{t). 2004,731-745.[4] ARC Advisory Group. "ER P Market Opportunities ChangeWhile Remaining Strong Overall at $8.9 Billion". 2003.http:/ /www.arcweb.com/Community/arcnews/arcnewsasp?ID=328 (Accessed 5th November 2004).[5] Bagchi. S. Kanungo. S.. and Dasgupta. S. "Modeling useof enterprise resource planning systems: A path analytic

study," European Journal of Information Systems 12:(2),2003. 142+.[6] Bajwa. D., Mooney . T.. and Garcia. J. (2004) "An Integra-[ tive Framework For The Assimilation Of Enterprise Re-r source Planning Systems: Phases. Antecedents. And Out-comes," Journal of Computer information Systems 44:(3).2004.81-90.[7] Banker. R.. Davis. G.. and Slaughter. S. "Software D e-velopment Practices, Software Complexity, and SoftwareMaintenance Performance: A Field Study." Management

Science U.{A), 1998.433-451.[8] Barker. T. and Frolick, M. "ERP Implementation Failure:A Case Study." Information Systems Management 20:(4),2003.43-49.[9] Bingi. P., Sharm a. M.. and Godla. J. "Critical Issues Affect-ing an ERP Implementation." Information Systems Man-agement 16:(3). 1999.7-14.

[10] Boudreau. M. and Robey. D. "Enacting Integrated Informa-" tion Technology: A Human Agency Perspective," Organi-zation Science I6:(l). 2005,3-18.[ Il l Bresnahan T. and Brynjolfsson. E. "Information tech nol-ogy, workplace organization, and the demand for skilled la-bor: Firm-level evidenc e." Quarterly Journal of Economics117:(]), 2000. 339-376.[12] Brown. S.. Massey. M. Mo ntoya-Weiss. M.. and Burk-man, J. "Do I really have to? User acceptance of mandatedtechnology." European Journal of Information Systems 11,

2002. 283-295

[13] Cha u. P. Y. K. "Influence of compute r attitude and self-efficacy on IT usage behavior." Journal of End User Com-puting I3:(l) . 2005. 26-40.

[14] Clayton, G. E, A Ca.se Study and Comparison of EnterprisResource Planning Systems: Implementation and Perfor-mance. Master's Thesis. Penn State Harrisburg, 2005.

[151 Cooper, R. and Zmud. R. "Information technology imple-men tation researc h: a technological diffusion appr oac h,"Management Science 36:{2), 1990, 123-142.

[16] Dav enport, T. "Putting the enterprise into the enterprisesyslem." Harvard Business Review 16:{4). 1998 . 121-131 .

[17] Davis, F. D. "Perceived usefulness, perceived ease of useand user acceptance of information technology." MIS Quar-terly 13:(3), 1989,319-340.

[18] Davis. F., Bagozzi, R.. and Warshaw . P. "User acc eptanceof computer technology: A comparison of two theoreticalmodels." Management Science 55, 1989, 982-100 3.[19] Dewan S. K. and Kraemer. S. "International dimensionsof the productivity paradox." Communications of the ACM41:(9), 1998,56-62.[20] Fisher. D.. Fisher, S., Kiang, M., and Chi. R. "EvaluatingMid-Level ERP Software," Journal of Computer Informa-

tion Systems 45:(1). 2004. 38-46.[21] Gallivan. M.. Spitler, V.. and Koufaris. M. "Does Informa-tion Technology Training Really Matter? A Social Infor-mation Processing Analysis of Coworkers' Influence on ITUsage in the Wo rkplace." Journal of Management Informa-

tion Systems 22:{ 1), 2005. 153-192.[22] Gattiker. T. and Goodhue. D. "What happens after ERPImplementation: Understanding the Impact of Inter-Dependence and Differentiation on Plant-Level O utcom es,"

MIS Quarterly 29:(3). 2005. 559-584.[23] Gulledge. T. and Sommer. R. "Splitting The Sap Instance:Lessons On Scope And usiness Processes." Journal of

Computer Information Systems 44:(3), 2004, 109-115.[24] Gupta. A. "Enterprise Resource Planning: The E mergingOrganizational Value Systems," Industrial Management +Data Systems 100:(3}. 2000, 114+.[25] Hare l. G. and Tzafrir, S. "Th e Effect of Human Reso urceManagem ent Practices on The P erceptions of Organization-al and Market Performance of the Firm," Human ResourceManagement ?>S:(3). 1999, 185-199.[26] Hitt, L., Wu , D. and Zho u. X. "Inve stmen t in EnterpriseResource Planning: Business Impact and Productivity M ea-sures." Journal of Management Information Systems 19:( 1)2002,71-98.[27] Hong. K and Kim, Y. "The Critical Success Factors for

ERP Implementation: An Organizational Fit Perspective,"Information and Management 40:( I) . 2002. 25-40.[28] Hu. L. and Ben tier. P. "Cutoff C riteria for Fit Indexes inCovariance Structure Analysis: Conventional Criteria ver-sus New Alternatives," Structural Equation Modeling 6 i ( l ) ,1999,1-55.[29] Huang. S.. Hung . Y.. Chen. H.. and Ku. C. "TransplantingThe Best Practice For Implementation Of An ERP System:A Structured Inductive Study Of An International Com-pany." Journal of Computer Information Systems 44:(4),2004, 101-110.[30] Krasner, H. "Ensu ring e-business success by learning fromERP failures." IT Pro, 2:(I). 2000. 22-27[31] Kwon, T. and Zmud, R. Unifying the fragmented models

of information systems implementation: Critical issues in


9/10

Information Systertis Research, John Wiley, New York. [45]1987.[32] Miller. B. "What is ERP ?" CIO. 2003. http://www2.cio.con\/aiialyst/report2003.html (accessed 3rd January . 200 5). [46]

[33] Mook erjee, R. "Maintainin g Enterprise Software Applica-tions," Comm unications of the ACM, 48:( 11), 2005, 75-79.[34] Mo ore. J. "One Road to Turnover: An Examination of WorkExhaustion in Technology Professionals," MIS Quarterly [41\

24:(l),2()00. 141-168.[35] Nah . F., Tan. X.. and The, S. "An Em pirical Investigationon End-Users* Acceptance of Enterprise Systems," Infor- [48]mation Resources Management Journal 17:(3), 2004, 3 2-51 . [49]

|36] Nah. E.. Zuckweiler, K. and Lau. J. "ERP Im plementation:Chief Information Officers' Perceptions of Critical SuccessFactors," international Journal of Hu man-Computer Inter-action 16:fl). 2003, 5-22. [50]

[37] Nunn ally. J. C. Psychometric Theory, 2nd ed. New York:McGraw -Hill, 1978.[38i Parr. A. and Shan ks. G. "A model of ER P project imple- [51]

meniaUon," Journal of Information T echnology 15. 2000.289-303.(39] Pressman, R. S. Software Engineering: A Practitioner's [52]Approach, 6th edition. New York: McGraw-Hill, 2004.

[40] Rajagopal, P. and Frank, T. "A compa rative case analysisof enlerpri.se resource planning systems implementationand performance SAP," Annual Conference of Decision [53]Sciences Institute Meeting, 2000.[411 Rajagopal, P. and Frank. T. "Oracle ERP and Network Com- [54]puting Architecture: Implementation and Performance." In -formation Systems Management I9:(2). 2002. 53-68.142] Robey , D.. Ros s. J.. and Bou dreau. M. "Learning lo Imple- [55]ment Enterprise Systems: An Exploratory Study of the Dia-lectics of Change." Journal of Management Information [56]Systems I9:(l). 2002, 17-46

[43] Scott, J. and Vessey . I. "Man aging risks in enterprise imple-mentations," Comm unications of the ACM 45 :{4). 2002,74-81 . [57]

[44| Segars, A. H.. and Grover. V. "Re-examining perceivedease of u.se and usefulness: A confirmatory factor analysis,"MlSQuarterly 17. 1993, 517-525.

Siau, K. "Enterprise Resource Planning (ERP) implemen-tation methodologies." Journal of Database ManagementI5 :( l) . 2004, 1-6.Soh. C . Sia, S., Boh. W.. and Tang. M. "Misalignmentsin ERP Implementation: A Dialectic Perspective," Inter-national Journal of Human-Computer Interaction, I6: (]) ,2003,81-100.Somers, T. and Nelson, K. "A Taxonomy of Players andActivities across the ERP Project Life Cycle," Informationand Management, A\, 2004. 257-278.Stevens, T. "Kodak focuses on ERP," Industry Week246:(15), 1997, 130,Staehr. L., Shanks. G., and Seddon. P. "Analysing ERP Usewith the Structurational Model of Technology," Proceed-ings of Australasian Conference on Information Systems,2002.Straub, D., Limayem. M., and Karahanna-Evaristo. E."Measuring system usage: Implications for IS theory lest-ing," Managetnem Science 4\:iS), 1995, 1328-1342,Subramanian, G. H. "A replication of prerceived useful-ness and perceived ease of use measurement." DecisionSciences, 25:(5/6), 1994. 863-873.Subramanian. G. H. and Hoffer, C. "An Exploratory CaseStudy of Enterprise Resource Planning Implementation."The International Journal of Enterprise Information Sys-tems, ! : ( ! ) . 2005. 23-38.Umble. E. and Umble, M. "Avoiding ERP implementationfailure," Industrial M anagement, 44:(1). 2002. 25-34.Venkatesh. V., and Davis, F. "A theoretical extension ofthe technology acceptance model: Four longitudinal tieldstudies," Management Science 46:(2 ). 2(X)0, 186 -204.Wah. L. "Give ERP a chance," Management Review 89:(3).2000, 20-24.Yang. C , T ing. P. and Wei. C. "A Study of the Factors Im-pacting ERP System Performance from the Users" Perspec-tives." Journal of American Academy of Business, 8:(2),2006,161-167.Zviran. M. Pliskin. N., Levin, R. "Measuring User Sat-isfaction and Perceived Usefulness in the ERP Context,"Journal of Com puter Information Systems 45:(3), 2005,43-52 .

Winter 2007-2008 Journal of Computer Information Systems


10/10

the phases of erp software implementation and maintenance

Documents