Available online at www.sciencedirect.com

ScienceDirect

www.elsevier.com/locate/ijpromanInternational Journal of Project Management 32 (2014) 529–539

Project planning practices based on enterprise resource planningsystems in small and medium enterprises — A case study from the

Republic of Macedonia

Frosina Tasevska a, Talib Damij a, Nadja Damij b,⁎

a Faculty of Economics, University of Ljubljana, Kardeljeva ploscad 17, SI-1000 Ljubljana, Sloveniab Faculty of Information Studies, Ulica Talcev 3, SI-8000 Novo mesto, Slovenia

Received 8 May 2013; received in revised form 25 July 2013; accepted 1 August 2013

Abstract

This paper examines whether Macedonian SMEs plan for the implementation of ERP projects and studies the effect of project planningpractices on project success. Four project planning measures were taken into consideration: business case development, scope planning, baselineplan development and risk planning along with three measures of project success; customer satisfaction, perceived quality of the project andsuccess of the implementation process. The study was based on a survey that was conducted on 30 SMEs in the Republic of Macedonia. Datadimensionality was reduced through factor analysis and relationships between the two sets of variables were analyzed by correlation and regressionanalyses. The findings demonstrated that Macedonian SMEs implemented general project planning practices, even though they did not consider theplanning process as a separate phase of the ERP implementation. However, they did not use any particular project planning tools, such as the Ganttchart or WBS. Of the project planning practices that were surveyed, the most practiced were the development of a business case, project scope andbaseline plan. The least practiced were risk planning practices. Considering the success of the ERP implementations, this study demonstrated thatmost of the companies' representatives perceive this undertaking as successful in terms of client satisfaction and perceived quality measures. Ahigher percentage of respondents found their ERP implementations unsuccessful in terms of implementation process measures, when compared tothe previous two success parameters.© 2013 Elsevier Ltd. APM and IPMA. All rights reserved.

Keywords: Project management; SMEs implementation; ERP project; Project success

1. Introduction

Enterprise Resource Planning (hereinafter: ERP) systems havebeen developed to enable overall integration of business processeswith the end result being the efficient deployment of resources andthe effective management of the whole enterprise (Leon, 2008).The implementation of such systems is certainly not an easy task.It requires companies to commit a significant amount of resourcesand to implement a large scale changes that will affect every aspectof the functioning of the company (Kumar et al., 2002). This

⁎ Corresponding author.E-mail addresses: frositasevska@yahoo.com (F. Tasevska),

tdamij@gmail.com (T. Damij), nadja.damij@gmail.com (N. Damij).

0263-7863/$36.00 © 2013 Elsevier Ltd. APM and IPMA. All rights reserved.http://dx.doi.org/10.1016/j.ijproman.2013.08.001

causes various organizations (Roncevic and Makarovic, 2010,2011; Roncevic and Modic, 2011) to experience difficultiesresulting in their ERP systems being implemented late or overbudget. Little (2011) and Pinto and Prescott (1990) argued thatproject planning is a significant facilitating factor in implementingprojects successfully. It directly affects the cost and scheduleperformance of projects (Hamilton and Gibson, 1996; Wang andGibson, 2010). Dvir (2005) argued further that the amount of effortinvested in defining the project goals, functional requirements andspecifications has a positive effect on the project success.Aladwani (2002) also contended that IT project planning plays amajor role in achieving success in IT projects. Al-Mashari et al.(2003), Glenn (2008) and Tchokogué et al. (2005) acknowledgedthis for ERP projects in particular. Many IT projects fail at the

530 F. Tasevska et al. / International Journal of Project Management 32 (2014) 529–539

beginning rather than at the end, because of insufficient planning(Phillips, 2011). Good planning is in fact halfway to success.Therefore the implementation of ERP systems should also firststart with the planning of the system, before addressing the higherstages (Shanks et al., 2003).

Furthermore, Glenn (2008) argued that project planning is oneof the five common factors that can determine the success of anERP implementation. Taking into consideration the high costsassociatedwith ERP implementation and the cumbersome processof realization, the importance of the planning issues cannot beoveremphasized (Chen, 2001). Mabert (2003) in their study of theUS manufacturing sector, also discovered that companiesemphasize the importance of planning an ERP implementation.

Ngai et al. (2008) conducted a literature review on CriticalSuccess Factors (hereinafter: CSFs) in the implementation of ERPsacross 10 different regions. When considering project manage-ment, they state that a clear and defined project plan includinggoals, objectives, strategy, scope, schedule, and so forth wasfrequently cited in CSFs for ERP implementation in almost all ofthe regions and countries examined in their study. Furthermore,ERP projects involve significant levels of different types of risk(Iskanius, 2009). These should be taken into account andappropriate mitigation strategies and contingency plans should bedeveloped. Therefore, based on the evidence provided in theliterature, the main hypothesis is that project planning practiceshave a positive effect on project success.

Even though ERP systems were first developed andimplemented in the developed countries, companies fromdeveloping countries are also embracing these systems. Theyalso account for CSFs when implementing their ERP projects.As the study of Mooheba et al. (2010) indicated, projectmanagement is of similar importance for companies from bothdeveloping and developed countries. Large companies werealso the first to implement ERP systems, but as O'Leary (2002)assumed, ERP systems can benefit both large scale companiesas well as small and medium-sized enterprises (hereinafter:SMEs). Many ERP software packages have been developedrecently to suit the needs of SMEs, specifically in terms of costsand functional scope.

Macedonian companies are certainly following the trend ofimplementing ERP systems mainly for lower scale ones, primarilybecause of resource constraints. Many of them implement ERP inorder to improve their business processes, but some of them tocomply with legislative requirements. Usually, only a fewmodulesare implemented, mainly for finance, inventory and accountingpurposes. Santa (2010), in his case study conducted in a smallMacedonian company, found that ERP implementation was runintuitively based on the business experience of the owner, withoutemployment of any particular project management practices.However, Ordanoski (2010), a Macedonian programmer andentrepreneur, realized the importance of having developed aproject and assigned a team to work on an ERP implementation,and recommended these practices to other Macedonian managers.

This study investigated whether Macedonian SMEs in generalimplement project planning practices, andwhether they influencedthe success of ERP implementation. The initial assumption wasthat successful ERP implementation depends on implementing

sound project planning practices, as suggested by the literature.Therefore the primary research question was whether IT projectplanning has an impact on project success. The findings wereexpected to bring some positive value to a topic on which verylittle research has been done and to incite more research in thefuture. This contribution refers especially to the literature andindustry in developing countries where scholars have significantinterest in the CSFs of the ERP implementations (Amid et al.,2012). The survey was employed as a technique for primary datacollection. Both the planning and the success of the project weremeasured via several dimensions. A combination of exploratoryand confirmatory factor analysis was used to reduce the datacollected through Likert-type items and to produce thesedimensions. After performing this analysis, correlation andregression analyses were used in order to describe the relationshipsbetween the project planning practices and project successmeasures (Dvir, 2005).

2. The study

2.1. Methodology

The project planning practices were taken as independentvariables to be tested. Twenty one variables, as shown in Table 1,were used to measure the planning effort that companies put inwhen implementing ERP. They were all organized along fourdimensions: business case development practices, scope planningpractices, baseline plan development practices and risk planningpractices.

The first dimension, business case development, measurespractices involved in the initial planning that result in the creationof a business case. Salomo et al. (2007) used a scale of nine itemsto measure this dimension. The same items were also used in thisstudy, except that two of them (Alternative market scenarios andFit with core competences) were omitted as they were notconsidered applicable to this case. Since no comprehensive scaleswere identified for the scope planning and baseline plandevelopment practices, original scales were developed in accordwith the recommendations of Saunders et al. (2003) and Hair etal. (2010). The risk planning dimension was measured on threeitems used in the study of Salomo et al. (2007).

Respondents were asked to say to what extent they agreed ordisagreed with each statement that indicated usage of a certainplanning practice (1 — strongly disagree; 7 — strongly agree).Hakkinen and Hilmoli (2008), as well as Aladwani (2002),utilized the seven-point Likert scale when they measuredproject planning items. Therefore, the same seven-point Likertscale was used in this case, too.

In order to judge whether companies implemented certainplanning practices or not, variables were recoded and valuesclassified into two groups: not implemented (encompassinganswers from 1 — completely disagree to 4 — neutral) andimplemented (encompassing answers from 5— partially agree to7 — completely agree). The “Recode into different variable”option in SPSS was used as suggested by Brace et al. (2003). Thesame option was used for variables that were expressed innegative terms to reverse code them so that high or low values

Table 1Project planning measurement dimensions and their items.Source: S. Salomo et al., NPD Planning Activities and Innovation Performance: The Mediating Role of Process Management and the Moderating Effect of ProductInnovativeness, 2007, p. 302.

Business case (BC) Scope planning (SP) Baseline plan (BP) Risk planning (RP)

Overall, the analysis we conducted beforedeciding to implement the ERP wasthorough and methodical (BC1)

We defined the goals that we wanted toachieve with the ERP implementation (SP1)

We defined all the activities needed toexecute the ERP implementation (BP1)

We conducted analysis of risksand their consequences (RP1)

We identified the main value drivers of theERP implementation (BC2)

We defined all the outcomes that shouldhave been delivered during implementation(SP2)

We did not define the sequence ofactivities (BP2)

We created detailed plans foruncertainty reduction (RP2)

We conducted systematic identification ofalternative ERP solutions (BC3)

We did not define the most significantevents that should have occurred duringimplementation (SP3)

We defined the duration of the activities(BP3)

We created detailed risk responseplans (RP3)

We conducted systematic selection of thepreferred ERP solution (BC4)

We defined the requirements that the softwareshould have fulfilled (SP4)

We did not define the resources neededfor the execution of activities (BP4)

We evaluated the fit between the ERPimplementation and the corporate strategy(BC5)

We did not consider all the constraints wehad to cope with during the implementation(SP5)

We established a detailed schedule forERP implementation (BP5)

Relevant departments participated in theplanning process (BC6)

We established a detailed budget forERP implementation (BP6)

Team/responsible person was committed toproject goals (BC7)

531F. Tasevska et al. / International Journal of Project Management 32 (2014) 529–539

indicated the same type of response on every item (Grace-Martin,2012).

Project success as a dependent variable was operationalized byusing a scale consisting of 12 itemsmeasuring three dimensions ofproject success: client satisfaction, perceived quality and successof the implementation process (see Table 2). This scale was usedand empirically tested by Mahaney and Lederer (2006) in theirstudy that specifically analyzed information system projects, sinceit had been previously tested and its reliability reported by severalauthors in the Project Management Journal. The items themselveswere slightly rephrased to suit the needs of this study. Accordingto Mahaney and Lederer (2006), the first dimension attempts tomeasure the level of acceptance of the project with its intendedbenefits by the users. The secondmeasures the effect of the projectin terms of improved performance. The third tries to find outwhether the project was completed on time, within schedule andwhether it met its technical goals.

Table 2Project success measurement dimensions and their items.Source: R. C. Mahaney and A. L. Lederer, The Effect of Intrinsic and Extrinsic Rew

Client satisfaction (CS) Perceived quality

The ERP software that was implemented works (CS1) The implementedamong the set of a

The ERP software is used by its intended users (CS2) Use of this ERP sor more effectivefor the users (PQ2

This ERP software directly benefited the intended userseither through increasing efficiency or employeeeffectiveness (CS3)

This software hasmake use of it (PQ

Important users, directly affected by the ERP software,make use of it (CS4)

The results of the imrepresent a definiteperform these activ

We are confident that non-technical start-up problems wereminimal, because the ERP software was readily acceptedby its intended users (CS5)

Respondents were asked to say to what extent they agreed ordisagreed with each statement on a seven-point Likert scale (1—strongly disagree; 7 — strongly agree). In order to discoverwhether the projects were successful or not, the variables wererecoded and answers classified into two groups: not successful(encompassing answers from 1 — completely disagree to 4 —neutral) and successful (encompassing answers from5— partiallyagree to 7 — completely agree).

Consequently, the hypotheses of this study were:

– Hypothesis 1: The level of effort put on business casedevelopment is positively related to project success in termsof client satisfaction, perceived quality and implementationprocess.

– Hypothesis 2: The level of effort put on scope planning ispositively related to project success in terms of clientsatisfaction, perceived quality and implementation process.

ards for Developers on Information Systems Project Success, 2006, p. 52.

(PQ) Implementation process (IP)

ERP software was the best choicelternatives (PQ1)

The ERP implementation came withinits original schedule (IP1)

oftware directly led to improveddecision making or performance)

The ERP implementation came withinits original budget (IP2)

a positive impact on those who3)

I was satisfied with the process by whichthe ERP software was completed (IP3)

plementation of this ERP softwareimprovement in the way the usersities (PQ4)

532 F. Tasevska et al. / International Journal of Project Management 32 (2014) 529–539

– Hypothesis 3: The level of effort put on baseline plandevelopment is positively related to project success in termsof client satisfaction, perceived quality and implementationprocess.

– Hypothesis 4: The level of effort put on risk plan developmentis positively related to project success in terms of clientsatisfaction, perceived quality and implementation process.

Two instruments were used for collecting primary data: astandardized questionnaire and in-depth interviews. Before con-structing the questionnaire, in-depth interviews were conductedwith the owners of three small companies. The purpose of theseinterviews was to gain an overview of the way the companiesapproached ERP implementation and the goals they pursued whendoing this. Furthermore, information on whether project planningwas practiced and how they assessed the success of their ERPimplementation was collected. During the interviews, the initialquestionnaire questions were also discussed and the terminologythat managers used when talking about ERP and planning ingeneral was examined. Three open-ended questions were askedinitially, but during the discussion several others were added tohelp in reaching the research objectives.

The questions were stated as follows:

– How did you decide to implement an ERP system in general?– What planning practices did you undertake before the actualimplementation?

– How were you satisfied with the ERP system after it wasimplemented?

A questionnaire was used to gather standardized data fromrepresentatives of Macedonian SMEs by asking questionsabout their opinion on the project planning and projects successmeasures in the manner explained above. The questionnairewas structured into six sections. The first section was dedicatedto general type questions. The next four sections collectedinformation about the level of implementation of each of the ITproject planning dimensions. The sixth section contained itemsmeasuring project success. The questionnaire was distributedelectronically and in person during July 2012.

Factor analysis was applied in order to confirm the dimensionsdefining both project planning and project success, and to designthe scales to be used in further analysis. The independent effect ofthe project planning dimensions on the project success dimen-sions was then measured by simple correlation analysis using thePearson coefficient. The effect of all project planning dimensionson each of the project success dimensions was analyzed byregression analysis. This analysis was used in order to betterunderstand the relationship between project planning factors andproject success factors as suggested by Dvir (2005).

A combination of convenience and snowball samplingapproaches was used to select 30 SMEs from different industries.This particular non-probability sampling approach was chosenbecause of the lack of any comprehensive list that encompassesall Macedonian SMEs that have introduced ERP solutions.SMEs, including micro enterprises, according to MacedonianCompany Law (Macedonian Stock Exchange, MSE, 2012), are

defined as enterprises having 10 (micro), 50 (small) and 250(medium) employees.

2.2. Results

Firstly, the interview results demonstrated that the represen-tatives of the sampled Macedonian SMEs did not use the termERP to refer to the ERP software solution they possess, butinstead they used the term “Software” or “Computer program”.When asked to clarify the functionalities of their software, itwas clear that they used an ERP solution. Only one of therepresentatives used the term ERP since he had been presentedwith it by his vendor. Hence, advice was provided to addclarification to the term ERP in the questionnaire by indicatingthat it referred to the software used for internal materialsmanagement or software for managing internal operations.

The responses to the first interview question showed thatcompanies had clear goals when they implemented ERP, assuggested by the literature. Usually they wanted to integrate alltheir data so that reporting would be facilitated. Anotherimportant reason was that they had to introduce some softwarefor accounting and finance, since it was required by legislation.

Based on the responses to the second question it was concludedthat the companies used planning practices, but did not recognizethem as such; i.e. they did not consider planning as a separatephase of the ERP project undertaking. Furthermore, they did notuse any particular tools, such as the Gantt chart or WBS, and theirrepresentatives were not even familiar with these terms. Forexample, one of the interviewees said that their finance managerwas responsible for analyzing the costs of several offers, and theirpart-time IT technician for analyzing the features of the solutions,but no responsibility matrix was developed, nor was a Gantt chartused that would demonstrate these activities, their interdepen-dences or durations. The fact that the companies' representativesdid not have a project management background was probably themain reason for such an approach. Munns and Bjeirmi (cited inYanwen, 2012) recognized the lack of topmanagement awarenessabout project management as one of the problems in developingcountries. Furthermore, the formal methodology of implementa-tion was not mentioned by the companies' representatives, exceptthat the vendors helped them install the system at the verybeginning and provided them with brief instructions for its use.Besner and Hobbs (2008) identified the difference betweengeneral project management processes, and specific tools andtechniques. Tools and techniques such as WBS or project charterwere used by practitioners for executing particular projectmanagement processes. Therefore, the questionnaire askedquestions about general project planning practices and avoidedgoing into details about specific tools used as was done by Besnerand Hobbs (2008).

Regarding the third question, two of the company represen-tatives considered their ERP implementation as successful. Oneof them reported being satisfied with the opportunities the ERPbrought to them, but was not satisfied with the implementationprocess itself, since it took too long for the software to beimplemented and accepted by all the users.

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

100.0%

SP1 SP2 SP3* SP4 SP5*

10.0 6.713.3 13.3 13.3

90.0 93.386.7 86.7 86.7

Not implemented Implemented

Fig. 2. Frequency of implementation of scope planning (SP) practices. *Valuesof these variables were reverse coded.

533F. Tasevska et al. / International Journal of Project Management 32 (2014) 529–539

In order to draw conclusions on the frequency of implemen-tation of the analyzed planning practices, the questionnaireresponses were first examined through frequency analysis and theresults were graphically represented by bar charts. The graphsdemonstrate that the Macedonian companies that were includedin the sample did implement project planning practices, assuggested by the literature. At least 86.7% of all respondents saidthat their companies implemented each of the planning practicesinvolved in the development of a business case (see Fig. 1),project scope (see Fig. 2) and baseline plan (see Fig. 3). However,the study showed that risk planning practices were the leastimplemented compared to the other project planning practices(see Fig. 4). Detailed plans for uncertainty reduction (RP2) anddetailed risk response plans (RP3) were implemented by only46.7% of the respondents. On the other hand, this finding is inline with the argument of Kwak and Stoddard (2004), whopointed out that most project managers consider riskmanagementactivities as extra work and expense, and therefore avoidimplementing them. Analysis of risks and their consequences(RP1) was made by 73.3% of the respondents. It seems that thecompanies involved in the case study did conduct analysis, butconsidered the actual development of plans as extra work andexpense, as argued by Kwak and Stoddard (2004).

These findings are contradictory to a Macedonian studyconducted in the ERP field which argued that Macedoniancompanies do not follow project planning (Santa, 2010). However,the questionnaire was developed in such a way that it did not askfor any particular project planning tools and did not include anyparticular project management terminology. Instead, the findingsfrom the interview phase were followed, and for example it wasasked whether the companies had identified alternative ERPsystems or whether they had identified the value of ERPimplementation, rather than whether they had developed abusiness case. Therefore, it was assumed that the questions inthis case were clearer for the respondents and they were able toprovide more realistic answers.

With regards to the project success evaluations, most of therespondents (at least 85%) assessed the projects as successful,based on client satisfaction and perceived quality dimensions (seeFigs. 5 and 6), whereas the last dimension, measuring whether the

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

100.0%

BC1 BC2 BC3 BC4 BC5 BC6 BC7

3.3 3.310.0 6.7 6.7 6.7 6.7

96.7 96.790.0 93.3 93.3 93.3 93.3

Not implemented Implemented

Fig. 1. Frequency of implementation of business case (BC) development practices.

project was implemented on time and within budget, gave morenegative responses (see Fig. 7). Most negative responses weregiven to the IP2 variable (33.3%), which measures whether theproject was completed within budget. Based on these findings,the conclusion could be drown that the major issues facing thecompanies during ERP implementation were related to budgetoverruns, and less frequently to schedule overruns. These findingsare consistent with other literature research, one of them beingZhang et al. (2005), who argued that companies very oftenexperience budget and schedule overruns when implementingERP.

2.2.1. Factor analysisConfirmatory factor analysis (hereinafter: CFA) was per-

formed on the project success dimensions i.e. constructs sincean a priori pattern of factor loadings on each of the projectsuccess constructs was known from theory (Hair et al., 2010).Project success was defined by the three latent constructs aspreviously mentioned. The overall model was over-identifiedby having 78 unique variance/covariance terms, 27 parameters

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

100.0%

BP1 BP2* BP3 BP4* BP5 BP6

6.7 10.0 6.7 6.7 10.0 13.3

93.3 90.0 93.3 93.3 90.0 86.7

Not implemented Implemented

Fig. 3. Frequency of implementation of baseline plan (BP) practices. *Values ofthese variables were reverse coded.

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

RP1 RP2 RP3

26.7

53.3 53.3

73.3

46.7 46.7

Not implemented Implemented

Fig. 4. Frequency of implementation of risk planning (RP) practices.

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

PQ1 PQ2 PQ3 PQ4

10.0 10.0 10.0 10.0

90.0 90.0 90.0 90.0

Unsuccessful Successful

Fig. 6. Project success evaluation based on perceived quality (PQ) variables.

534 F. Tasevska et al. / International Journal of Project Management 32 (2014) 529–539

to be estimated (12 factor loadings, 12 error variances and 3covariances), and consequently 51 degrees of freedom. Allconstructs in CFA were exogenous and therefore only thecovariance relationships were hypothesized among the threelatent constructs (Brown, 2006; Hair et al., 2010; Schreiber etal., 2006). The recommendation of having 5 to 10 cases perestimated parameter (Brown, 2006; Schreiber et al., 2006) or atleast 300 cases (Tabachnick and Fidell, 2007) was not met.Therefore there might be over fitting of data because of thelarger number of variables relative to the sample size(Tabachnick and Fidell, 2007). One of suggestions of Kline(cited in Harrington, 2008) in this case is to use factor loadingsgreater than 0.6. Furthermore, Marsh et al. (1988, p. 396) arguethat χ2, as one of the most cited indicators of goodness-of-fit,does not vary with the sample size if the model is true.

The χ2 value of 54.378 and p-value of 0.347 demonstratedgood correspondence between the observed and expectedcovariance matrices. CFI with a value of 0.989 and RMSEAwith a value of 0.048 and a 90% confidence interval between0.000 and 0.131 also showed a good model fit. However, NFIand AGFI were below the recommended level of 0.95. RMR,with a value of 0.086, did not indicate a very good model fiteither. However, the standardized residuals provided in Table 3did not exceed the recommended level of |4.0|, suggesting nomodel changes were needed. The modification indices provided

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

100.0%

CS1 CS2 CS3 CS4 CS5

6.7 10.0 13.3 10.0 10.0

93.3 90.0 86.7 90.0 90.0

Unsuccessful Successful

Fig. 5. Project success evaluation based on client satisfaction (CS) variables.

in Table 4 do suggest some paths be freed, but since they arenot much higher than 4.0, changes need not be based solely onthem, as suggested by Hair et al. (2010). Therefore, furtheranalysis was made on the basis of this initial model.

The model demonstrated good convergent validity (seeTable 5). All factor loadings appeared statistically significantand all standardized factor loadings were above 0.5, as suggestedby Kline (cited in Harrington, 2008) and Schreiber et al. (2006).All communalities were above the recommended level of 0.5. TheAVE extracted by all three factors was above 0.5 and constructreliability measures were above 0.7, confirming the constructvalidity of the model. All of the squared correlations were lowerthan the AVEs, confirming the discriminant validity of the model.

Exploratory factor analysis (hereinafter: EFA), specificallyPrincipal Component Analysis (hereinafter: PCA), was performedon all the variables measuring project planning constructs, sinceno a priori knowledge existed about the factors as they were usedin this study. To justify the utilization of EFA, a strong correlationamong the variables was needed. Visual inspection of thecorrelation matrix showed high Pearson coefficients (higher than0.8) between many of the variables. The suggestion of Hair et al.(2010) was above 0.3 for these coefficients. The second indicatorevaluated was the measure of sampling adequacy (hereinafter:MSA), which is an index for which values of 0.5 are considered asacceptable, both for individual variables as well as for the overallmodel (Hair et al., 2010). The individual MSAs in this case

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

IP1 IP2 IP3

23.3

33.330.0

76.7

66.770.0

Unsuccessful Successful

Fig. 7. Project success evaluation based on implementation process (IP) variables.

Table 3Standardized residuals (Amos output).

IP3 IP2 IP1 PQ4 PQ3 PQ2 PQ1 CS5 CS4 CS3 CS2 CS1

IP3 .000IP2 − .011 .000IP1 .060 − .012 .000PQ4 − .189 .127 .864 .000PQ3 .162 − .427 .646 .030 .000PQ2 .367 .147 .997 − .036 − .050 .000PQ1 − .611 −1.135 .121 − .023 .006 .063 .000CS5 − .720 .125 − .398 − .108 − .236 − .058 − .493 .000CS4 − .358 − .107 − .415 .017 − .312 .062 − .075 .059 .000CS3 − .206 .199 − .198 − .199 − .590 − .301 − .814 .257 − .165 .000CS2 − .361 .299 .216 .219 − .063 .574 .008 − .032 − .016 − .033 .000CS1 .225 .627 .257 1.310 1.037 .731 .681 − .188 .118 − .228 .089 .000

535F. Tasevska et al. / International Journal of Project Management 32 (2014) 529–539

satisfied the criteria of being higher than 0.5, and the overall MSAwith a value of 0.678 also satisfied the criteria. The Bartlett test ofsphericity with a Chi-square value of 515.927 was statisticallysignificant with a p-value of 0.000 at the 0.05 level ofsignificance, providing evidence that the correlation matrix wasstatistically different from the identity matrix.

The analysis resulted in four factors extracted witheigenvalues higher than one. To better distribute the varianceamong them, Varimax rotation was used. Since two variables(SP2 and BP5) persistently cross-loaded even after rotation,they were deleted one by the other. The structure of thevariables, after the deletion of the two variables, resembled thegrouping of items made in the questionnaire. The factorsaccounted for 78.371% of the total variation in the variables,which was above the minimum of 60% (Hair et al., 2010). Allvariables had a communality higher than |0.5|, meaning that thesolution explained more than 50% of their variation. Theirfactor loadings on the corresponding factors were also higherthan |0.5|, as was sought in CFA, but very low on the otherfactors. This finding demonstrated the independence of the fourfactors (Dvir, 2005). The rotated component matrix andcommunalities are presented in Table 6.

Table 4Modification indexes (Amos output).

Covariances: (Croup number 1 — Default model)

M.I. Par changee11 b–N Perceived quality 4.020 − .225e10 b–N Perceived quality 4.425 .217e6 b–N Implementation process 4.106 − .168e6 b–N e11 4.262 − .122e1 b–N Perceived quality 4.428 .168

Variances: (Croup number 1 — Default model)

M.I. Par change

Regression Weights: (Croup number 1 — Default model)

M.I. Par changePQ1 b–- Implementation process 4.331 − .144PQ1 b–- IP2 5.845 − .125CS1 b–- PQ4 4.686 .231CS1 b–- PQ3 4.294 .252

As CFA and EFA confirmed the structure of the factorsemployed, summated scales were created on their basis beforecontinuing with further analysis. They enabled creation ofsingle composite measures and thereby captured the multipleaspects of each factor represented by their indicator variables(Hair et al., 2010). Cronbach's alpha for each scale, as ameasure of internal consistency and reliability, is presented inTable 7. Values above 0.7 were sought as suggested by Hair etal. (2010). The high construct reliability measures estimated inCFA were consistent with the high Cronbach's alpha.

2.2.2. Correlation and regression analysisThe results of this study provide only partial support to the

main claim that planning has a positive effect on projectsuccess. It was found that not all the planning practices have anequal effect on project success. This finding is in line with theconclusions drawn in the study conducted by Zwikael andGloberson (2006). Based on the correlation analysis (seeTable 8), the development of a business case is positivelyrelated to all of the project success measures. Therefore, it canbe expected that if more effort were made in developing abusiness case, ERP implementation would be more successfulin terms of client satisfaction, perceived quality and implemen-tation process.

Multiple regression was used to measure the relative impor-tance of each planning measure in predicting the project successmeasures. The criteria of having five cases per independentvariable included in the model (Hair et al., 2010) was satisfied asone dependent and four independent measures were used eachtime. The model summary and ANOVA results from running astep-wise regression analysis for the three dependent variables areprovided in Table 9. All three models appear valid, according toANOVA analysis (p b 0.05) (Hair et al., 2010). Based on thevalue of R2, the first model explains 50.1% of the total variation inthe client satisfaction. The second model explains 38.5%, whereasthe third model has the least explanatory power (22.7%). Thedifferences between R2 and adjusted R2 are small, indicating thatno redundant variables are present in the model and that there is noover fitting of the data.

When business case is strictly defined as a predictor variableof the project success measures, as in the regression analysis

Table 5Validity measures of project success constructs.

Unstandardized factor loadings p-Value Standardized factor loadings Communality Delta

Client satisfaction CS1 1.000 .000 .801 .575 .425CS2 1.403 .000 .920 .849 .151CS3 1.308 .000 .806 .787 .213CS4 1.300 .000 .877 .851 .149CS5 1.373 .000 .920 .847 .153Σ 4.324 3.909 1.092AVE .782Construct reliability .945

Perceived quality PQ1 1.000 .000 .924 .754 0.246PQ2 1.034 .000 .869 .853 0.147PQ3 .929 .000 .920 .847 0.153PQ4 1.061 .000 .922 .769 0.231Σ 3.635 3.223 .777AVE .806Construct reliability .944

Implementation process IP1 1.000 .000 .887 .649 .351IP2 1.177 .000 .921 .847 .153IP3 .891 .000 .759 .641 .359Σ 2.567 2.137 .863AVE .712Construct reliability .884

536 F. Tasevska et al. / International Journal of Project Management 32 (2014) 529–539

(see Table 10), it turns out that it only has a significantcontribution to implementation process success.

Based on the regression coefficient, the conclusion can bemade that for every increase of one point in the attitude ofrespondents regarding business case development, the percep-tions of the success of the implementation process will increaseon average by 0.856 points on the 7-point Likert scale. That iswhy the first hypothesis is only partially accepted.

The baseline plan is a significant predictor of clientsatisfaction and perceived quality, as hypothesized. As thesign of the coefficient of the baseline plan demonstrates in theclient satisfaction variate, for every increase of one point in the

Table 6Rotated component matrix and communalities for project planning measures.

Businesscase

Scopeplanning

Baselineplan

Riskplan

Communalities

BC1 .783 .185 .185 .220 .730BC2 .735 − .187 .432 .104 .772BC3 .804 − .033 .251 − .096 .720BC4 .865 .139 .153 − .063 .795BC5 .825 .131 .279 .137 .794BC6 .792 .312 .092 .283 .814BC7 .533 .344 .160 .400 .587SP1 .156 .937 .148 − .038 .926SP3 .215 .806 .242 − .238 .810SP4 .051 .876 − .070 .099 .784SP5 .043 .809 .244 .043 .719BP1 .318 .221 .854 .174 .909BP2 .210 .027 .890 .239 .894BP3 .196 .166 .886 .174 .882BP4 .284 .172 .744 .432 .850BP6 .239 .152 .782 − .074 .697RP1 − .033 − .134 .294 .880 .880RP2 .038 − .068 .223 .879 .828RP3 .366 .093 − .012 .738 .687

attitude of respondents regarding baseline plan development,client satisfaction will increase on average by 0.583 points onthe 7-point Likert scale. Perceived quality, on the other hand,will increase on average by 0.475 points. The correlationanalysis also confirms these positive relationships since thebaseline plan is most strongly correlated with client satisfactionand perceived quality when compared to the other projectplanning measures. Furthermore, it is also positively related tothe implementation process, with r = 0.410 being significant atthe 0.05 significance level. However, the regression model ofthe implementation process demonstrates that planning theschedule and budget does not lead to a project being completedon time and within budget as hypothesized, but that anothervariable, in this case business case development, is a betterpredictor. Therefore, the hypothesis that states that baselineplan development has a positive effect on project success is alsopartially accepted.

The contradictory findings based on the regression analysisindicating that baseline plan development does not affect theimplementation process could be explained in several ways.Firstly, the companies under study may indeed consider achiev-ing client satisfaction and reaping benefits from the ERP to bemore important than completing the implementation on time and

Table 7Scale reliability.

Scale Cronbach's alpha

Business case (BC) .915Scope planning (SP) .906Baseline plan (BP) .934Risk planning (RP) .836Client satisfaction (CS) .934Perceived quality (PQ) .948Implementation process (IP) .890

Table 8Correlations between project planning composite measures and project successcomposite measures.

BC SP BP RP CS PQ

SP Pearson correlation .315 1Sig. (2 tailed) .090

BP Pearson correlation .561 ⁎⁎ .326 1Sig. (2 tailed) .001 .079

RP Pearson correlation .342 − .008 .384 ⁎ 1Sig. (2 tailed) .065 .966 .036

CS Pearson correlation .420 ⁎ .370 ⁎ .708 ⁎⁎ .168 1Sig. (2 tailed) .021 .044 .000 .375

PQ Pearson correlation .507 ⁎⁎ .342 .621 ⁎⁎ .095 .440 ⁎ 1Sig. (2 tailed) .004 .064 .000 .616 .015

IP Pearson correlation .476 ⁎⁎ .268 .410 ⁎ .202 .380 ⁎ .223Sig. (2 tailed) .008 .152 .025 .284 .038 .236

⁎⁎ Correlation is significant at the .01 level (2-tailed).⁎ Correlation is significant at the .05 level (2-tailed).

537F. Tasevska et al. / International Journal of Project Management 32 (2014) 529–539

within schedule. As a result, they might plan all their activitiesand resources in order to achieve that. Unfortunately, this issuewas not studied and could be an interesting topic for furtherresearch. Secondly, due to the small sample size, the partialcorrelation of baseline plan might not have been statisticallysignificant when compared to business case development, andthus the baseline plan was removed from the model.

The scope planning measure was not included in any of theregression equations because it was considered to have a weakeffect on the project success measures. The correlation analysis,on the other hand, showed that scope planning is weakly butstatistically significantly correlated only to the client satisfac-tion project measure. As the analysis in this case did notprovide support to any of the aforementioned statements, thesecond hypothesis claiming that the definition of project scopehas a positive effect on the project success is rejected.

The regression analyses in this study omit risk planning in allthree equations, demonstrating its insignificance in predictingproject success. Correlation analysis also confirms this statementsince risk planning does not exhibit a positive correlation to anyof the project success measures. Therefore the fourth hypothesisis also rejected.

The contradiction in the results in the case of scope planningand risk planning might be explained by the limitations broughtby sample size. As Hair et al. (2010) argue, a smaller samplesize can make even strong correlations appear statisticallyinsignificant. Since the sample used in this study included only30 cases, it is reasonable to conclude that some results may beinsignificant on that account. The signs of the correlation

Table 9Models' summary and ANOVA.

Model summary ANOVA

R2 Adjusted R2 F Sig.

CS .501 .483 28.088 .000PQ .385 .363 17.532 .000IP .227 .199 8.214 .008

coefficient, on the other hand, reveal that there is a positiveassociation between scope planning and project success, as wellas between risk planning and project success, as suggested bythe literature. However, because of their statistical insignifi-cance, it is not possible to claim that these linear relationshipsdefinitely exist.

3. Conclusion

The results of this study demonstrate that Macedonian SMEsimplement general project planning practices, even though theydo not consider the planning process as a separate phase of ERPimplementation. However, they do not use any particular projectplanning tools, such as the Gantt chart or WBS. From the projectplanning practices that were surveyed, the most practiced are theones involved in the development of a business case, project scopeand baseline plan. The least practiced are risk planning practices, afinding which is consistent with the literature. Regarding thesuccess of ERP implementations, this study demonstrates thatmost of the companies' representatives perceive this undertakingas a success in terms of client satisfaction and perceived qualitymeasures. A higher percentage of respondents find their ERPimplementations unsuccessful in terms of implementation processmeasures when compared to the previous two success dimensions.

However, the study reveals that not all IT project planningpractices have the same effect on project success. Developmentof a baseline plan appears positively related to all of the projectsuccess measures used in this study. Based on regressionanalysis, it appears to be a significant predictor of two of thesemeasures, client satisfaction and perceived quality. Thus, theconclusion can be drown that the development of a baselineplan may improve the likelihood of client satisfaction andperceived quality. Business case development also appears tobe positively related to all of the success measures. However,regression analysis demonstrates that it can only be consideredas a significant predictor of the implementation process successmeasure. Thus, developing a business case may improve thelikelihood of the success of the implementation process.

Scope planning and risk planning practices were alsoconsidered in this study as they were indicated by the literatureto have a positive effect on project success as well, but thisstudy fails to support this claim. Both correlation and regressionanalysis did not indicate significant relationships between thesetwo measures and the project success measures. Therefore, itcan be concluded that if Macedonian companies want toachieve successful ERP implementations, they should put moreeffort into developing a business case and developing a baselineplan.

Based on the results of this study, the main recommendationwould be that even though every project planning practice isimportant for the success of the project, as indicated by theliterature, more effort should be put into the development of abusiness case and a baseline plan. These two measures appearto have a major effect on ERP project success in theMacedonian companies that were included in the sample. Thisis most probably so because both encompass more tangible andmeasurable activities that are easier to understand when

Table 10Regression coefficients.

BC SP BP RP

Intercept B Beta Sig B Beta Sig B Beta Sig B Beta Sig

CS 2.940 .583 .708 .000PQ 3.572 .475 .621 .000IP .033 .856 .476 .008

538 F. Tasevska et al. / International Journal of Project Management 32 (2014) 529–539

compared with scope planning and risk planning activities. Forexample, the baseline plan expresses all the decisions that havebeen made during scope planning into clearly defined activitiesthat have to be executed. This method might be betterunderstood and followed by the implementation team or theperson responsible for implementation, than when they areexpressed in terms of goals or deliverables. This is especiallyvalid for companies that do not possess human resources withappropriate project management skills, as in this case study.

Because of the small sample size, the results obtained in thisstudy are sample-specific and cannot be generalized. However,they bring empirical evidence (that can be validated throughbigger sample size) to enhance the theories related to ERPplanning and ERP project success. A bigger sample could alsoovercome potential over-fitting that might be present in thisstudy. The outcomes of the study might also be of interest toconsultants in suggesting areas on which to focus whenworking on a plan for ERP implementations in developingcountries. Furthermore, as the study validated an existinginstrument for measuring project success and developed a newone for measuring project planning, they both might be used infurther studies with more confidence. It seems necessary thatfuture studies investigate further the effect of risk and scopeplanning on the ERP project success as they were found to haveno effect in this case-study.

References

Aladwani, A.M., 2002. IT project uncertainty, planning and success — anempirical investigation from Kuwait. Inform. Technol. People 15 (3),210–226.

Al-Mashari, M., Al-Mudimigh, A., Zairi, M., 2003. Enterprise resource planning:a taxonomy of critical factors. Eur. J. Oper. Res. 146 (2), 352–364.

Amid, A., Moalagh, M., Ravasan, Z.A., 2012. Identification and classification ofERP failure factors in Iranian industries. Information Systems 37 (3), 227–237.

Besner, C., Hobbs, B., 2008. Project management practice, generic or contextual: areality check. Proj. Manag. J. 39 (1), 16–33.

Brace, N., Kemp, R., Snelgar, R., 2003. SPSS for Psychologists: A Guide toData Analysis Using SPSS for Windows, Versions 9, 10 and 11, 2nd ed.Lawrence Erlbaum Associates, Inc., Mahvah, NJ.

Brown, T.A., 2006. Confirmatory Factor Analysis for Applied Research. TheGuilford Press, New York.

Chen, I.J., 2001. Planning for ERP systems: analysis and future trend. Bus.Process. Manag. J. 7 (5), 374–386.

Dvir, D., 2005. Transferring projects to their final users: the effect of planningand preparations for commissioning on project success. Int. J. Proj. Manag.23 (4), 257–265.

Glenn, G., 2008. Enterprise Resource Planning 100 Success Secrets — 100Most Asked Questions: The Missing ERP Software, Systems, Solutions,Applications and Implementations Guide. Emereo Pty Ltd., Queensland.

Grace-Martin, K., 2012. An easyway to reverse code scale items. Retrieved August12, 2012, from http://www.theanalysisfactor.com/easy-reverse-code/ (nd).

Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E., 2010. Multivariate DataAnalysis. Pearson Prentice Hall, New Jersey.

Hakkinen, L., Hilmoli, O.-P., 2008. Life after ERP implementation: long-termdevelopment of users perceptions of system success in an after-salesenvironment. J. Enterp. Inf. Manag. 21 (3), 285–310.

Hamilton, M.R., Gibson, J.G.E., 1996. Benchmarking project planning effort.Journal of Management in Engineering 12 (2), 25–33.

Harrington, D., 2008. Confirmatory Factor Analysis. Oxford University PressInc., Oxford.

Iskanius, P., 2009. Risk management in ERP project in the context of SMEs.Eng. Lett. 17 (4), 266–273.

Kumar, V., Maheshwari, B., Kumar, U., 2002. Enterprise resource planningsystems adoption process: a survey of Canadian companies. Int. J. Prod.Res. 40 (3), 509–523.

Kwak, Y.H., Stoddard, J., 2004. Project risk management: lessons learned fromsoftware development environment. Technovation 24 (11), 915–920.

Leon, A., 2008. Enterprise Resource Planning, 2nd ed. Tata McGrew-HillPublishing Company Limited, New Delhi.

Little, B., 2011. The principles of successful project management. Hum.Resour. Manag. Int. Dig. 19 (7), 36–39.

Mabert, V.A., Soni, A., Venkataramanan, M.A., 2003. The impact oforganisation size on enterprise resource planning (ERP) implementationsin the US manufacturing sector. OMEGA Int. J. Manag. Sci. 31, 235–246.

Macedonian Stock Exchange (MSE), 2012. Retrived July 3, 2012 from http://www.mse.mk/Page.aspx?ContentID=31&lang=mk.

Mahaney, R.C., Lederer, A.L., 2006. The effect of intrinsic and extrinsicrewards for developers on information systems project success. Proj.Manag. J. 37 (4), 42–54.

Marsh, H.W., Balla, J.R., McDonald, R.R., 1988. Goodness-of-Fit Indexes inConfirmatory Factor Analysis: The Effect of Sample Size. PsichologicalBulletin 103 (3), 391–410.

Mooheba, M.R., Asem, A., Jazi, M.D., 2010. A comparative study of criticalsuccess factors (CSFs) in implementation of ERP in developed anddeveloping countries. Int. J. Advancements Comput. Technol. 2 (5),99–109.

Ngai, E.W.T., Law, C.C.H., Wat, F.K.T., 2008. Examining the critical successfactors in the adoption of enterprise resource planning. Comput. Ind. 59 (6),548–564.

O'Leary, D.E., 2002. Enterprise Resource Planning Systems: Systems, LifeCycle, Electronic Commerce, and Risk. Press Syndicate of the University ofCambridge, United Kingdom.

Ordanoski, V., 2010. Moderen informativen system— ERP (Modern informationsystem — ERP). Retrieved July 2, 2012, from http://instantekspert.blogspot.com/2010/03/erp.html.

Phillips, J. (n.d.). Project Management Definition and Solutions. CIO.Retrieved November 27, 2011, from http://www.cio.com/article/40342/Project_Management_Definition_and_Solutions.

Pinto, J.K., Prescott, J.E., 1990. Planning and Tactical Factors in the ProjectImplementation Process. Journal Of Management Studies 27 (3), 305–327.

Roncevic, B., Makarovic, M., 2010. Towards the strategies of modern socialprocesses. Innov. Eur. J. Soc. Sci. Res. 23 (3), 223–239.

Roncevic, B., Makarovic, M., 2011. Societal steering in theoretical perspective:social becoming as an analytical solution. Pol. Sociol. Rev. 176 (4),461–472.

539F. Tasevska et al. / International Journal of Project Management 32 (2014) 529–539

Roncevic, B., Modic, D., 2011. Regional systems of innovations as socialfields. Sociologija i prostor 49 (191), 313–333.

Salomo, S., Weise, J., Gemunden, H.G., 2007. NPD planning activities and innovationperformance: the mediating role of process management and the moderatingeffect of product innovativeness. J. Prod. Innov. Manag. 24 (4), 285–302.

Santa, M., 2010. Implementation of enterprise resource planning software insmall company in Macedonia — case study. Yearbook of the Faculty ofEconomics-Skopje.

Saunders, M., Lewis, P., Thornhill, A., 2003. Research Methods for BusinessStudents, 3rd ed. Financial Times Prentice Hall, Harlow.

Schreiber, J.B., Stage, F.K., King, J., Nora, A., Barlow, E.A., 2006. Reportingstructural equation modeling and confirmatory factor analysis results: areview. J. Educ. Res. 99 (6), 323–337.

Shanks, G., Seddon, P.B., Willcocks, L.P., 2003. Second-Wave EnterpriseResource planning Systems: Implementing for Effectiveness. The PressSyndicate of the University of Cambridge, Cambridge.

Tabachnick, B., Fidell, L.S., 2007. (5th ed.). Using Multivariate Statistics.Pearson Education, Inc., Boston, MA.

Tchokogué, A., Bareil, C., Duguay, C.R., 2005. Key lessons from theimplementation of an ERP at Pratt & Whitney Canada. Int. J. Prod. Econ.95 (2), 151–163.

Wang, Y.-R., Gibson, J.G.E., 2010. A Study of preproject planning and projectsuccess using ANNs and regression models. Automation in Construction 19(3), 341–346.

Yanwen, W., 2012. The study on complex project management in developingcountries. Phys. Procedia 25, 1547–1552.

Zhang, Z., Lee, M.K.O., Huang, P., Zhang, L., Huang, X., 2005. A frameworkof ERP systems implementation success in China: an empirical study. Int.J. Prod. Econ. 98 (1), 56–80.

Zwikael, O., Globerson, S., 2006. From Critical Success Factors to CritisalSuccess Processes. International Journal of Production Research 44 (17),3433–3449.