Determinants of User Participation in Pre-Implementation ERP Development

(1)

in Pre-Implementation ERP

Development

Master thesis, MscBA, specialization Change Management University of Groningen, Faculty of Management and Organization

10 June 2008 R.J. Moek Student Number: 1552902 Wulp 21 9502 WJ Stadskanaal (0031)615445995 rmoek@hotmail.com Supervisor University B. Emans

(2)

SUMMARY

The IT sector tends to solely focus on performing a project within time, within budget, and according to the quality requirements that were preset. Nevertheless, many authors emphasize user acceptance as a pivotal factor in IT project success (Westhuizen and Fitzgerald, 2005). User participation is often mentioned as the most important tool to increase user acceptance. This thesis aims to provide knowledge about why organizations do, or do not, invest in user participation. This research is demarcated to ERP projects because these projects require a change in the working routines and attitudes of users. This increases the importance of user acceptance and user participation. The research question is as following:

“What are the determinants of the user participation set up that aims to increase user acceptance of ERP projects?”

Based on a literature review, several hypotheses were set up. To answer the research question and test the hypotheses, nine projects were analyzed by interviewing the project leaders. For each project, two interviews were set up: one to determine the context of the project and one that aimed to map the determinants of the user participation set up. The interviewees were chosen based on their high position and decision power in the projects. Based on the interviews six user participation interventions were set up, these will be shortly elucidated: 1. Requirements definition: this refers to setting the conditions for an IT system that has to be developed. 2. Mailings: this is, for example, a weekly newsletter. 3. Information sessions: these are meetings of (project) management with users of the system. Mailings and information systems are both a form of communication. 4. Training: this intervention aims to teach users how to work with the new IT system. 5. Sand box environment: this enables users to practice with the new IT system in an informal setting. 6. Acceptance testing: this aims to check if the IT system that was built fulfills the needs of the organization.

(3)

Requirements definition:

• Type of organization; very formal, top down organizations give users

less freedom in their participative role in the project.

• Complexity of the organizational change; highly complex projects

could benefit from a set up of user participation that focuses more on creating commitment among users.

• Commitment of users to the project; when commitment towards the

project is low, more user participation increases commitment.

• Knowledge of users about the work processes; this knowledge is

needed to map the current processes and set quality measures for the system development.

• Expectations of users on the project; in order to get to know what

users expect from a new system, more users participate and they participate by doing.

Mailing and Information sessions:

• Chance of resistance; when chance of resistance among users is

high, top management should be involved in the communication efforts.

• Knowledge of users about the project; when knowledge on the

project and it effects are low, the communication efforts should become more intensive.

Formal training:

• Time pressure; the training moment near the live moment of a

system is vital for success. Only in extreme situations this moment will be changed.

• Knowledge of users about the system; the degree of knowledge that

the users have about the new system influences the set up of the training. A whole new ERP system requires more training than an upgrade of an ERP system that is already in use.

• Need for diversity; every individual benefits from a particular training

(4)

Sand box environment

• Belief in the benefits of a sand box environment; the benefits of

practicing in a sand box environment is doubted among the participants. Someone that does not believe in the benefits will less likely invest in a sand box environment.

• Organizational difficulties in setting up a sand box environment; it is

very hard to set up a sand box environment and enabling users to work with it. This requires investments in time, costs and labor. The interdependencies in an ERP system also require that many users simultaneously practice.

Acceptance testing

• Risk management; risk management is the main influence on the set

up of acceptance testing. The risk that a particular area bears is scored and put against the costs of testing that area. Some organizations test almost every aspect of a new system, whereas others only test the high risk areas. This makes the set up of acceptance testing a very personal one.

• Earlier experiences; the set up of the acceptance testing is partly

determined by earlier experiences. Positive experiences with a particular approach will quickly lead to adoption of that same approach in future projects.

• Project importance; this refers to the fact that for many

(5)

TABLE OF CONTENT

LIST OF TABLES... 7

INTRODUCTION... 8

CHAPTER 1: PROBLEM STATEMENT AND RESEARCH QUESTION... 9

1.1PROBLEM STATEMENT... 9

1.2BACKGROUND INFORMATION... 11

1.2.1 User Acceptance ... 11

1.2.2 Users ... 12

1.2.3 ERP Systems ... 12

1.3OBJECTIVES AND RELEVANCE OF THIS THESIS... 13

1.4STRUCTURE OF THE THESIS... 13

CHAPTER 2: USER PARTICIPATION IN ERP DEVELOPMENT ... 14

2.1USER PARTICIPATION... 14

2.2INTERVENTIONS IN USER PARTICIPATION... 15

2.2.1 Participative Decision Making ... 16

2.2.2 Communication ... 18 2.2.3 Training ... 19 2.2.4 Acceptance Testing... 20 CHAPTER 3: METHODOLOGY ... 23 3.1INTERVIEW I ... 23 3.2INTERVIEW II... 24

CHAPTER 4: RESEARCH RESULTS... 26

4.1PROJECTS... 26 4.1.1 Achmea ... 26 4.1.2 CBS ... 26 4.1.3 Gasunie ... 26 4.1.4 Grontmij ... 27 4.1.5 NS... 27 4.1.6 Praxis ... 27 4.1.7 Randstad... 28 4.1.8 TNT ... 28

(6)

4.2.1 Interview Results Regarding Determinants of Requirements Definition ... 29

4.2.2 Conclusions on User Participation in Requirements Definition ... 37

4.2.3 Interview Results Regarding Determinants of Communication ... 41

4.2.4 Conclusions on User Participation in Communication ... 45

4.2.5 Interview Results Regarding Determinants of Training ... 47

4.2.6 Conclusions on User Participation in Training ... 54

5.2.7 Interview Results Regarding Determinants of Acceptance Tests... 55

4.2.8 Conclusions on User Participation in Acceptance Tests... 59

CHAPTER 5: CONCLUSIONS ... 61

5.1VALIDITY OF THE RESEARCH... 61

5.2CONCLUSIONS HYPOTHESIZED DETERMINANTS OF USER PARTICIPATION SET UP. 61 5.3CONCLUSIONS ON DETERMINANTS OF USER PARTICIPATION SET UP... 63

5.4DISCUSSION... 65

5.5GENERAL CONCLUSIONS... 65

(7)

LIST OF TABLES

TABLE I Beer and Nohria’s Framework 18

TABLE II User Participation in Requirements Definition at Achmea 30

TABLE III User Participation in Requirements Definition at CBS 31

TABLE IV User Participation in Requirements Definition at Gasunie 32

TABLE V User Participation in Requirements Definition at Grontmij 33

TABLE VI User Participation in Requirements Definition at NS 34

TABLE VII User Participation in Requirements Definition at Praxis 35

TABLE VIII User Participation in Requirements Definition at Randstad 36

TABLE IX User Participation in Requirements Definition at TNT 37

TABLE X User Participation in Communication at Achmea 41

TABLE XI User Participation in Communication at CBS 42

TABLE XII User Participation in Communication at Gasunie 42

TABLE XIII User Participation in Communication at Grontmij 43

TABLE XIV User Participation in Communication at NS 43

TABLE XV User Participation in Communication at Praxis 44

TABLE XVI User Participation in Communication at Randstad 44

TABLE XVII User Participation in Communication at TNT 45

TABLE XVIII User Participation in Training at Achmea 48

TABLE XIX User Participation in Training at CBS 48

TABLE XX User Participation in Training at Gasunie 49

TABLE XXI User Participation in Training at Grontmij 50

TABLE XXII User Participation in Training at NS 51

TABLE XXIII User Participation in Training at Praxis 52

TABLE XXIV User Participation in Training at Randstad 53

TABLE XXV User Participation in Training at TNT 54

TABLE XXVI User participation in Acceptance Testing at Achmea 56

TABLE XXVII User Participation in Acceptance Testing at CBS 56

TABLE XXVIII User Participation in Acceptance Testing at Gasunie 57

TABLE XXIX User Participation in Acceptance Testing at Grontmij 57

TABLE XXX User Participation in Acceptance Testing at NS 58

TABLE XXXI User Participation in Acceptance Testing at Praxis 58

TABLE XXXII User Participation in Acceptance Testing at Randstad 58

TABLE XXXIII User Participation in Acceptance Testing at TNT 59

(8)

INTRODUCTION

(9)

CHAPTER 1: PROBLEM STATEMENT AND RESEARCH QUESTION

This chapter discusses the subject that was stated in the introductory chapter. First the problem statement is being discussed, which leads to a preliminary research question. Section 1.2 provides background information on the most important elements of the research question, user acceptance, users, and ERP. The objectives and relevance of this research are also discussed in this chapter and a description of the structure of this thesis is given.

1.1 Problem Statement

Lack of user acceptance has long been an impediment to the success of new information systems (Davis, 1993). The IT sector tends to solely focus on performing a project within time, within budget, and according to the quality requirements that were preset. Nevertheless, many authors emphasize user acceptance as a pivotal, fourth factor in IT project success (Westhuizen and Fitzgerald, 2005).

User participation is often mentioned as the most important tool to increase user acceptance. Much research has been done in the area of the effectiveness of user participation (Legris et al., 2002). This thesis aims to provide knowledge on why organizations do, or do not, invest in user participation interventions that aim to increase user acceptance. In contrast to earlier research, this research focuses on the input of the user participation interventions rather than on the output of it. Based on the above, the preliminary research question is:

“What are the determinants of the user participation set up that aims to increase user acceptance of ERP projects?”

(10)

figure I) is taken in account. The software development life cycle (SDLC) is a structure imposed on the development of an information system. It describes a variety of tasks that take place during the development. The software development life cycle phases will now be briefly elucidated.

Project planning refers to the planning of the project. Resources are assigned to the various phases of the SDLC.

Requirements definition is the phase were the requirements for the new system are set. Requirements are set for what the system should be able to perform and for how the system should operate.

Design refers to putting the requirements into a design of the new system. A plan is developed for the new system.

Development is the phase where the design is actually built.

Integration and test refers to the assembly of the system and the tests to see if it works according to the requirements that were set.

Installation and acceptance is the phase where the system is put in use and accepted by the user organization.

FIGURE I

(11)

1.2 Background Information

To facilitate reading the theory section, the use of the concepts of user acceptance, users, and ERP in this thesis has to be unambiguous. Therefore, this section provides background information about the subject of this research and the context in which it is relevant. Information and definitions will follow on user acceptance, users, and on ERP projects. User participation will be discussed more thoroughly in the theory section of this thesis.

1.2.1 User Acceptance

User acceptance must be seen in the context of project success. The main determinants of project success are the criteria of time, budget, and quality (Wateridge, 1998). A project that is successful on these areas is called project management success (Baccarini, 1999). Wateridge (1998) questions if this should be the sole criteria for IT developments to be scored on. According to Baccarini (1999) project success is the sum of project management success and project product success. Project product success focuses on the project’s end product and the effect it has on the client organization. This refers to the added value of the product to the organization and the user acceptance of the project product. The subject of this research, user participation, aims to increase this user acceptance of the project product.

Now that the context of user acceptance in this thesis has been discussed, the concept of user acceptance will be discussed next. This thesis applies the definition of user acceptance as it was developed by Dillon and Morris (1996). They define user acceptance as:

“The demonstrable willingness to use the project product for the purpose for which it was developed”.

The reason why this definition is applied is because it regards user acceptance in the project success context that was just mentioned.

(12)

project, which refers to the users’ attitude or motivation towards the new system. Cognitive outcomes are largely knowledge based and refer to the fact if users are capable to work with the information system. In contrast to the affective outcomes, cognitive outcomes are impressionable by technology related aspects like system design (Davis, 1993). Similar to the cognitive outcomes, behavioral outcomes are also impressionable by technology related aspects. The behavioral outcomes refer to the degree to which users actually make use of the system’s functionality.

1.2.2 Users

Ives and Olsson (1984) distinguish primary and secondary users in ERP systems. Primary users are those employees who work with the output of the systems. Secondary users deliver the input into the systems. Secondary users are often the clerical workers, primary users are often the line managers. Both these classes are included in this research because they directly work with an ERP system.

1.2.3 ERP Systems

This research is demarcated to user participation in ERP development. Therefore, it is important to briefly discuss ERP and its specific characteristics. According to Calisir (2004) , a usable ERP system can be defined as:

“a comprehensive management software package that supports the effective and efficient completion of tasks in a given work context”.

(13)

1.3 Objectives and Relevance of this Thesis

Davis’ (1993) and Keens’ (1991) comments about the lack of user acceptance in IT development initiated this research. This thesis aims to provide knowledge about the determinants of the set up of user participation in ERP development.

The target groups of this thesis are IT project managers and consultants that are already involved in ERP implementations or think about engaging in an ERP project. This thesis should provide managers and consultants with more knowledge about the determinants of the set up of user participation in the development phase of ERP systems. Knowing why particular choices are made can help IT project managers and consultants in setting up user participation in future projects.

1.4 Structure of the Thesis

(14)

CHAPTER 2: USER PARTICIPATION IN ERP DEVELOPMENT

Now that the problem has been stated, this chapter discusses the core of this thesis: user participation during the development of ERP systems. This chapter elucidates user participation by addressing several sources of literature about the subject. The first paragraph discusses the concept of user participation. There are several interventions to let users participate during the development of an ERP system. These will be discussed in the second section. This second section will also lead to phrasing the hypothesized relationships between particular determinants of the decision of organizations whether or not – and how – to set up user participation.

2.1 User Participation

User participation in this thesis is regarded as being pivotal to increase user acceptance. Particularly in ERP projects, user participation is a way to shape a positive attitude of users towards the change that is caused by the new ERP system (Muntslag, 2001). User participation is a hard to define concept. In this research Hartwick and Barki’s (1994) view will be used. User participation is defined as:

“The behaviors, assignments, and activities which users or their representatives perform during the information systems development process”.

Behaviors, assignments, and activities refer to situations in the project SDLC (see section 1.1). In many cases not all users participate in the project. Like mentioned in the definition it is often a representation of the total user group that participates.

(15)

This latter example means that users participate in the decision whether or not to proceed with the project and how to proceed. 3. Doing refers to a situation where users are actively participating in a design group of an ERP project or are the official liaison of the group. This threefold categorization is a solid tool in this research for assessing how user participation is set up. When none of these three classes is applied, it can be automatically concluded that there is no user participation.

2.2 Interventions in User Participation

Now that the concept of user participation has been discussed, it is time to look at how users actually participate during the SDLC of an ERP system. Based on several literature sources and contact with IT project managers, four interventions are chosen to be subject of this research. First a brief description of the interventions is given, after which the motives for choosing these particular interventions are given.

The first intervention is participative decision making (PDM). This refers to the level of influence employees have in the decision making process during the SDLC (Alegre and Chiva, 2007). A second possible intervention is communication towards users. In this research communication is seen as sharing and spreading information towards users by project management and top management. The third influence is the training for users, which is required to let them get acquainted with the new system and teach them how to work with it. Acceptance testing is the last intervention that will be discussed. Acceptance testing aims to map and, if necessary, improve the usability of the system. These tests solely aim to determine whether the system is user friendly.

(16)

of the training can be quite diverse. The reasons for a particular approach will be studied in this thesis.

In discussing the different interventions (2.2.1 PDM; 2.2.2 Communication; 2.2.3 Training; 2.2.4 Acceptance testing), hypotheses will be set up. The hypotheses will presuppose a relationship between a particular determinant and the set up of user participation. It has to be regarded that the nature of the hypotheses slightly differs from each other. Some hypotheses presuppose a relationship between a determinant and the choice whether or not to set up user participation, whereas other hypotheses presuppose a relationship between a possible determinant and how user participation is set up. In contrast to the first, this latter already presupposes that there is user participation.

2.2.1 Participative Decision Making

PDM in the SDLC can cover several phases. The goal of PDM is to increase input of users into management decisions that are related to their jobs (Ives and Olson, 1984). Locke and Schweiger (1979) found that there is a strong, positive relationship between PDM and user acceptance.

Users are not likely to have the knowledge to contribute to, or make decisions through every phase of the SDLC. Requirements definition and testing (see section 1.1) of the new system are phases where PDM can be useful (McKeen and Guimareas, 1997). In these phases knowledge of current work processes and the desired situation are gathered. The users are the ones who posses this knowledge. The project needs this knowledge to optimize the project success. Therefore, user participation in these phases is pivotal to project success. Acceptance testing will be discussed separately in section 2.2.4, this section discusses requirements definition. The advantages of user participation in the requirements definition are (McKeen and Guimareas, 1997):

• More complete and accurate requirements,

• Using knowledge about organization (unit) were the new system is

implemented,

• Better user understanding of the system,

• More realistic user expectations,

• Creates feeling of ownership,

(17)

• Creating commitment among users.

These advantages show that user participation in the requirements definition is necessary for achieving a solid project product, as well as it being a possibility to create commitment among users towards the project product. Two determinants are hypothesized to have a relationship with the setup of the requirements definition.

From Alegre and Chiva (2007) it can be derived that type of organization is an important determinant of the set up of user participation in the requirements definition phase. Organizations can be divided into many different types. In this research Beer and Nohria’s (2000) distinction in the change approach of organizations is applied. They divide change approaches into a theory E and a theory O approach (see table II for their characteristics). The distinction of Beer and Nohria is chosen because of the contrasting role of the users in the change process. In a theory E organization users are guided much more by tight procedures than in a theory O organization. To achieve the purpose of theory O, developing organizational capabilities, empowerment of users is required. Alegre and Chiva (2007) argue that a theory O organization will invest more in user participation by PDM, in order to achieve user empowerment. The role of consultants also shows the distinction in user participation: in a theory E approach the consultant has a knowledge driven, expert role. Theory O organizations however, rely on their employees to develop organizational capabilities and make use of consultants who mostly lead employees through the accompanying change processes. Based on these arguments the following hypothesis is stated:

Hypothesis I: A theory E type organization will have less user participation in the requirements definition phase than a theory O type organization.

(18)

argue that ERP projects that adopt a “vanilla” approach are most likely to be successful. Like Schaffer (2000), they mention that large projects should be divided into smaller projects. According to Schaffer, these smaller projects enable more effective user participation. The participation of users can become more effective because they can focus on a particular area. In complex ERP projects there are many interdependencies between different organization processes. For employees, who often work in just one of those processes, it is hard to apply a helicopter view and regard the bigger picture of the project. This leads to the following hypothesis:

Hypothesis II: Greater complexity of the ERP project leads to a less participative approach in the requirements definition phase.

TABLE I

Beer and Nohria’s Framework of Theories of Change

Theory E Theory O

Purpose Maximize economic value Develop organizational capabilities

Leadership Top down Bottom up

Focus Structure and systems Culture

Planning Programmatic Emergent

Motivation Incentives lead Incentives lag

Consultants Knowledge driven Process driven

Source: Beer and Nohria: Breaking the code of change (2000)

2.2.2 Communication

(19)

This paragraph discusses the determinants that have an impact on the set up of communication. Based on academic writings, the risk of resistance among users against change is hypothesized to have a relationship with the set up of communication. When resistance increases among users, top management commitment can take away this resistance. In successful ERP implementations Shanks and Parr (2000) identified top management commitment during the design and reengineering phase as being pivotal to a successful ERP implementation. In these successful implementations top management was actively involved and promoted the benefits of the change to the users that were directly affected by the change. Top management in successful efforts “walk the talk”. This is the basic assumption for the next hypothesis:

Hypothesis III: Greater risk of resistance among users requires more top management commitment in the communication towards users.

2.2.3 Training

The importance of training is emphasized in many academic writings (Bransford et al., 1989; Baltzer, 1991; Marler et al., 2006). According to Klein et al. (1990), training serves a twofold goal. The first is that it should facilitate learning how to use the new ERP system. Secondly, it is a possibility to influence the perceived benefits of the ERP system. Therefore, training is not just a necessary evil to let users get acquainted with the IT development. It is also an effective and efficient way to influence the attitude of users towards that development. Marler et al. (2006) make this clear by stating that training is not just about acquiring skills. Besides teaching users how to work with the new system, training is an avenue for managing the users perception about available resources to practice and master use of the new system.

(20)

strict guidance. This informal training is often referred to as working in a ‘sand box environment’.

The first determinant of the communication set up that is being discussed is time pressure. The ideal training moment is just before the live moment of a new ERP system; nevertheless, many projects fail to work according to schedule. In this research it is presupposed that time pressure could be a determinant of the moment of the formal training. Project managers could prefer an on time delivery above the learning process of the users of the new ERP system, or decide not to shift the training moment when a project is delayed. It should be taken into account that the training moment is seen as a relative point in the SDLC, not as an absolute time of date. This reasoning leads to the following hypothesis:

Hypothesis IV: High time pressure results in the fact that the formal training moment is not close before the live moment of the ERP system.

A second variable has already been mentioned in section 2.2.1, the type of organization. This section will also adopt the distinction by Beer and Nohria (2000) in theory E type organizations and theory O type organizations. A formal training will be present in (almost) all ERP implementations because the users have to get familiar with the new system and the altered way of working. However, informal training is not likely to be present in every ERP implementation. The extent to which organizations are prepared to invest in informal training opportunities could depend on whether an organization adopts a theory E or a theory O approach. A presumption is that a theory O organization is more likely to set up an informal training moment. An informal training benefits the purpose of a theory O organization, developing organizational learning capabilities, because users get acquainted with the new system and are able to discuss it. This leads to the following hypothesis:

Hypothesis V: Theory O organizations are more likely to invest in informal training opportunities than theory E organizations.

2.2.4 Acceptance Testing

(21)

correctly into the system and whether the system is user friendly. The goal is to test if the system lives up to the expectations of the users. Li (1990) argues that two types of testing provide the opportunity to let users participate: software acceptance testing and final acceptance testing. The distinction is that software acceptance testing focuses on functionality (does the system meet the requirements that were set up), final acceptance testing is more about usability (user acceptance). In this thesis both types are regarded.

Acceptance testing is part of the SDLC. However, there are several moments in time when acceptance tests can be performed. Acceptance testing was mentioned by Davis (1993) as a pivotal tool for increasing user acceptance. A premise is that the test function should be involved early on in the SDLC. Three reasons that show the importance of acceptance testing early in the SDLC are: key decisions about the design of the system are made, only a small fraction of development costs have been incurred, and there is greater flexibility when modification of the design is necessary” (Davis, 1993).

Acceptance testing is done by many organizations, the set up can be quite different however. It can be considered if one lets many users participate, or if one works with external consultants. Next to the question who participates, there is also the question what to test. Test procedures can be very extensive and cover almost every area of the ERP system, it can also be less extensive and aim to test just the areas that bear most risk. Next, the hypothesized determinants of a particular set up will be discussed.

Conversations with many specialists in the test sector led to the idea that costs are the main determinant of why organizations do not invest in early acceptance testing. Huq (2000) performed a research into acceptance testing at the end of the SDLC and acceptance testing at the end of every phase of the SDLC. A conclusion was that costs do not differ between the two methods, and the latter method actually prevented errors in an earlier stage. This makes costs an interesting area to investigate and leads to the next hypothesis:

Hypothesis VI: Increasing costs lead to less investment in acceptance testing.

(22)

(1991) argue that the benefits of acceptance testing are not always clear to organizations. Negative earlier experiences with acceptance testing are likely to have influence on the set up of acceptance tests. The hypothesis is:

(23)

CHAPTER 3: METHODOLOGY

This chapter discusses the research methodology applied for testing the hypothesized relationships in this thesis. For this research expert interviewing was applied for nine ERP projects. The goal of the interviews was to gather information from people that are well informed about the project and the organization where the project took place. The information was gathered from people who performed at a high level in the specific ERP projects. The interviewees of this research are the managers that were the decision makers about the set up of user participation interventions in the ERP projects.

The selection of projects was done by contacting IT departments of big organizations. After asking if there were any recently implemented ERP projects, the subject of research was elucidated. The interviews took place at organizations from different branches, so that the chance on different types of organizations in the research increased. The types of projects were also diverse in order to increase the chance on projects with differing complexity. Every project had one main interviewee; however, at some of the projects (additional) information was obtained through other interviewees.

To test the hypotheses that were set up in the previous chapter, two interviews for every project were set up. The first was conducted through the telephone; the second was a face to face interview. In total, eighteen interviews took place for this research. The interviews will now be discussed.

3.1 Interview I

The aim of this interview was to gather information about the context of the projects. The first questions concerned the project goal and mission, as well as the role of the interviewee in the particular project. This information was needed to see if the project was suitable for this research and if the interviewee was in a role where the decisions about the set up of user participation were made.

(24)

reengineering of the work processes as a result of the project. To assess top management commitment the interviewees were asked to give their thought about the top management commitment in their particular project. This information helped in setting up the second interview in such a way that the hypotheses could be tested.

Another goal of this interview was to map the concrete actions that were taken to let users participate in the projects. This led to a specification of the interventions mentioned in the previous chapter. The interventions that were set up by the organizations were categorized. For example, in communication, newsletters, websites, and mailings were shared under the same heading. This led to the following interventions that were discussed:

• Requirements definition

• Mailings

• Information sessions

• Training

• Practicing in a sand box environment

• Acceptance tests

From this first interview some opinions could be derived about the set up of user participation. These opinions were used as input for the second interview. The other information of this first interview was also used as input for the second interview.

3.2 Interview II

This second, ‘live’ interview reached the core of this thesis. This interview aimed to explore the determinants of the set up of user participation. The interview was set up based on the information gathered in the first interview. The interview was divided into six parts. These six parts all discussed one of the six interventions of user participation mentioned in the preceding section. For every intervention the determinants were discussed.

(25)

questions were set up by using information from the first interview. It was questioned if the interviewee thought the hypothesized variable could be a determinant of the set up of user participation, but also if the variable was indeed a determinant in this project.

(26)

CHAPTER 4: RESEARCH RESULTS

This chapter discusses the results of the research and its conclusions. The first section summarizes the projects that took part in this research. Section two discusses the results of the interviews. The results will be divided into four sections: participative decision making, communication, training, and acceptance testing. The results are summarized in a table for every analyzed project. After the tables, the conclusions for each specific intervention are given.

4.1 Projects

The interviews took place at eight organizations and discussed nine projects. This section will briefly discuss the projects and the interviewees’ role.

4.1.1 Achmea

At Achmea the interviewee was Jacques de Rooij, project manager. The project resulted from the merger of Achmea and Interpolis. It concerned the standardization of the current ERP systems. Interpolis’ system was adopted by Achmea. The users in this project are the employees of the ‘old’ Achmea and Interpolis. The complexity of the ERP project was not that high, the merger made it more complex. Initially top management kept aloof; however, after some problems occurred they showed more commitment to the project.

4.1.2 CBS

The project at CBS concerned the implementation of a SAP I-procurement module. Rogier Groenendijk works at the functional control department at CBS. The complexity of the project was not very high. Top management was not very involved in the project because it did not affect the core business of CBS.

4.1.3 Gasunie

The interview at Gasunie was with Egbert Renkema. Two projects were discussed. Renkema was involved at both projects.

(27)

among employees however. Therefore, the organizational change due to this project was not that high. Top management was sponsor and client of the project.

Project II: This second project concerned Human Resources and aimed to improve the system for several HR arrangements for employees. The ERP part of this project is a consequence of a procedural change. The IT part of this project was less complex than that of the first project; however, the organizational change part of this second project was more complex. Top management was, like the first project, sponsor and client. However, they were more actively involved in the communication in this project.

4.1.4 Grontmij

This project concerned the implementation of Oracle, the financial and the HR modules. The interview was with Jeroen Musters, project leader. The implementation influenced 2.500 employees of Grontmij. The change concerned the financial and HR departments the most. The IT complexity did not seem to be very high, the organizational change was more complex. Because the customization of the ERP package was low, many procedures had to be changed to fit the standard ERP package. Top management commitment was initially very high; there was actually a member of the board of advisory included in the steering committee of the project. However, after a while other tasks were given a higher priority.

4.1.5 NS

The NS project was initiated by the CFO. The absence of employees was very high which resulted in high costs for NS. Jan Opfergelt, the interviewee, was appointed project leader. The project goal was standardization of the management reports on absence. The users were mostly employees of the Human Resource department. The complexity of the IT project was not very high. The organizational change was more complex. As a result of the initiation by the CFO, top management was very committed to the project.

4.1.6 Praxis

(28)

staff departments. The IT complexity was not that high. In contrast, the complexity of the organizational change was high. Top management showed to be very committed and grasped the importance of the project.

4.1.7 Randstad

Jolanda Vellekoop was implementation manager at Randstad Netherlands during the implementation of the Mondriaan package. The project affected the 2.500 employees of Randstad. The users were line management and interceders. The business goal was to create a more centralized system, so different locations could use each other’s information. Both the IT project and the organizational change did not appear to be very complex. Top management commitment initially was high but faded later on.

4.1.8 TNT

The interview at TNT was with Peter Vos, IT project manager. This particular project concerned the implementation of a SAP HR module as replacement of an outdated system. The users in this case were: line management; employees of the HR Shared Service Center; and HR advisors. The complexity of the IT project was small, the organizational change was more complex. The project was not of high priority to top management, as a result they were not actively involved.

4.2 Research Results per Intervention

This section discusses the results of the interviews. Per intervention the determinants of the set up are summarized in tables for each project. First the hypothesized determinants to have a relationship with the set up of user participation are discussed. The second half of the table discusses the determinants of the set up of user participation that were not discussed in the hypotheses. These determinants were mentioned by the interviewees. The determinants of the user participation set up were categorized where this was possible. This was done in order to draw conclusions more easily. Based on the information in the tables conclusions are given on the set up of each user participation intervention.

(29)

elucidation is based on the opinion of the interviewee. When reading the tables it should be taken into account that it is not fact but opinion.

4.2.1 Interview Results Regarding Determinants of Requirements Definition

This section contains eight tables that summarize the results of the interviews that were conducted. This section discusses the results on the determinants of the set up of the requirements definition. The eight tables reflect nine projects, because the two Gasunie projects are summarized in one table.

In chapter three two hypothesized determinants on the set up of requirements definition are mentioned: type of organization and complexity of the project. During the interviews it became obvious that this latter should be divided into complexity of the IT project and complexity of the organizational change. This distinction will be elucidated in the conclusions on the requirements definition.

Caption for the tables

(30)

TABLE II

User Participation in Requirements Definition at Achmea

REQUIREMENTS DEFINITION Participants View

Relationship Elucidation

Type of Organization +++

The type of organization was not a determinant of the set up of the requirements definition. The benefits of user participation could have been greater when the type of organization would have been regarded in the set up of the requirements definition. Looking back the users would have gotten less freedom in their tasks because of the theory E approach at Achmea.

Complexity IT 0

Complexity of IT is no determinant of the set up of requirements definition. There is not more or less user participation due to the IT complexity. Users only participate in areas they have knowledge of, the IT complexity only affects the IT department.

Complexity Organizational

Change +++

Due to the high organizational complexity of the change, the requirements definition was set up in such a way that it was a mean to create

commitment. More concrete this led to more user participation with a focus on participation by "having a say" and by "doing". Therefore, it can be said that the complexity of the organizational change does not lead to less user participation, but to a different set up of user participation.

Knowledge of Users +++

Local knowledge of employees is a determinant of the requirements definition set up because the knowledge is necessary for a solid requirements definition. This belief results in a set up of the requirements definition with a focus on "participation by doing".

Expectations of Users ++

The expectations of users are regarded as pivotal input for designing a good system. It is a determinant of the set up of requirements

definition. The focus is on user participation on the levels of "having a say" and "doing".

Commitment of Users +++

The commitment of users is a determinant of the set up of the requirements definition. If

commitment is low the set up of the requirements definition has an emphasis on creating commitment. This determinant is connected to the complexity of the organizational change.

(31)

TABLE III

User Participation in Requirements Definition at CBS*

REQUIREMENTS DEFINITION Participants View Relationship Elucidation

The theory E approach at CBS resulted in very little (almoste none) user participation in requirements definition. A more participative approach would have been adopted in an type O organization.

Complexity IT 0

Change 0

Although complexity of the organizational change is a determinant of user participation, this does not show in the phase of requirements definition.

Project Planning +++

Involving users slows down the project process. Partly because of its emphasis on reaching consensus, partly because of the lower education level of users. To keep up with the project planning no users participated in the requirements definition.

Knowledge Users +++

The knowledge of users is regarded as not being sufficient to participate in the requirements definition. Therefore, instead of letting users participate, external consultants were hired.

Project Importance +++

(32)

TABLE IV

User Participation in Requirements Definition at Gasunie

The type of organization does not lead to less investments in user participation in the

requirements definition, it leads to a different set up. Gasunie is more a theory O organization than a theory E organization. This influenced the set up of the requirements definition, the users got more freedom in performing their tasks.

Complexity IT 0

Change +++

The complexity of the organizational change of project II was higher than in project I. This influenced the set up of user participation in the requirements definition. The emphasis in the second project was more on "participation by doing".

O th e r D e te rm in a n t Expectations of Users +++

The expectations of users are regarded as pivotal input for designing a good system. It is a determinant of the set up of requirements

definition. The focus is on user participation on the levels of "having a say" and "doing".

(33)

TABLE V

User Participation in Requirements Definition at Grontmij

Grontmij seems to be between a type E and a type O organization, this influenced the set up of the requirements definition. Although reaching consensus was a main goal, there were strong final decision makers. The influence of type of

organization is thus mainly on the amount of freedom users get in requirements definition, not on the degree of investements in user participation in the requirements definition.

Complexity IT ++

The high IT complexity asked for more specific work process knowledge, which is possessed by users. Instead of a less particpative approach, a more participative approach was set up in the requirements definition.

Change +++

The complexity of the organizational change did not directly imply less user participation, it resulted in another set up. Requirements definition was set up in such a way that it was a mean to create

commitment. More concrete this led to a user participation by "having a say".

Multisite Organization +++

The fact that Grontmij is a multiple site organization influenced the set up of the requirements definition. Users from every site had to be involved in order to not neglect one site.

Strategic Project Approach +++

(34)

TABLE VI

User Participation in Requirements Definition at NS

In an E type organization a more top down approach to the set up of requirements definition would have been adopted. At NS emplyees got a fair share of freedom in their jobs, partly to reach consensus. Therefore, a type E organization does not lead to less user participation but it is a determinant of the set up of it.

Complexity IT +++

Complexity of the IT project is a determinant of the set up of user participation, not the level of participation. The high complexity caused project management to set up tight boxes in which users could work in the requirements definition phase.

Change +++

The complexity of the organizational change did not directly imply less user participation, it resulted in another set up. Requirements definition was set up in such a way that it is a mean to create

commitment. More concrete this led to user participation by "having a say".

Multiple Business Units +++

(35)

TABLE VII

User Participation in Requirements Definition at Praxis

Type of Organization 0

In any other organization the same approach to user participation in the requirements definition would have been adopted. Therefore, the type of

organization is not a determinant of the level of user participation.

Complexity IT +++

Complexity of the IT project is a determinant of the set up of user participation, not of the level of participation. The complexity caused project management to set up tight boxes in which users could work in the requirements definition phase.

Change +++

The complexity of the change on organizational side did not lead to less participation, it stressed the need for user participation to create commitment. Therefore, the focus of the set up was different and led to user participation by having a say and by doing.

Expectations of Users +++

The expectations of users are regarded as pivotal input for designing a good system. It is a determinant of the set up of requirements definition. The focus is on the levels of "having a say" and "doing".

(36)

TABLE VIII

User Participation in Requirements Definition at Randstad

Type of organization is a determinant of the set up of user participation in the requirements definition, not of the degree of user participation. The difference is that in a theory E the amount of freedom users get to perform their tasks is smaller.

Complexity IT +++

Complexity of the IT project is a determinant of the set up of user participation, not of the level of participation. The complexity caused project management to set up tight boxes in which users could work in the requirements definition phase.

Change +++

Expectations of Users +++

The expectations of users are regarded as pivotal input for designing a good system. It is a determinant of the set up of requirements definition. The focus is on the levels of "having a say" and "doing".

(37)

TABLE IX

User Participation in Requirements Definition at TNT

Type of organization is a determinant of the set up of user participation in the requirements definition, not of the investments in user participation. The difference is that in a theory E the amount of freedom users get to perform their tasks is smaller.

Complexity IT 0

Change +++

Project Importance +++

The project did not concern the core business. This made it more difficult to free people from their regular tasks.

Local knowledge of employees influences the requirements definition set up because the knowledge is necessary for a solid requirements definition. This belief results in a set up of the requirements definition with a focus on "participation by doing".

H y p o th e s iz e d D e te rm in a n ts O th e r D e te rm in a n ts

4.2.2 Conclusions on User Participation in Requirements Definition

The tables in the previous section show the results of the interviews. This section discusses the conclusions that can be drawn from the information in the tables. First the hypothesized determinants of user participation in the requirements definition will be discussed. Secondly, the other determinants of the set up of the requirements definition that were mentioned during the interviews will be discussed.

The first hypothesis on user participation in the requirements definition was:

(38)

From the first round of interviews it could be derived that many organizations let users participate in the requirements definition phase, only CBS was an exception. Because the projects were conducted at both theory E organizations as well as at theory O organizations, this already indicates that the hypothesis needs to be adjusted. It can be said that the investments in user participation in the requirements definition do not decrease in a theory E organization. However, in a theory E organization the users receive a more limited freedom for performing their tasks than the user of a theory O organization. The project at Achmea is particularly suited to illustrate this conclusion. Interpolis is a theory O type organization, Achmea a theory E type. After the merger, the way of working at Interpolis was adopted in this project. This resulted in problems for the employees of Achmea who could not cope with the given freedom. It appears that in a theory O organization, user participation is a combination of participation by doing and by having a say. In contrast, a theory E organization lets users participate more by a combination of having a say and by doing.

One reservation has to be made to this conclusion. Although almost all organizations regard type of organization as a determinant of how to let users participate in the requirements definition, when asked almost none of them actually took in account the type of organization for the set up of user participation in the requirements definition. This indicates a difference in the actual situation versus the preferable situation. This seems to refer to espoused theory, a notion developed by Argyris and Schön (1974). This can be referred to as the words we use to convey what we do, or what we would like others to think we do. Therefore, it can be concluded that it is preferable that type of organization is regarded when setting up user participation, in reality however this often is not the case.

The second hypothesis regarded project complexity:

Hypothesis II: Greater complexity of the ERP project leads to a less participative approach in the requirements definition phase.

(39)

Hypothesis IIa: Greater IT complexity of the ERP project leads to a less participative approach in the software development life cycle.

Hypothesis IIb: Greater organizational complexity of the ERP project leads to a less participative approach in the software development life cycle.

On hypothesis IIa the opinions of the interviewees were quite opposite to each other. Half of the interviewees think that IT complexity is not a determinant of the degree of user participation. To support this, the most mentioned argument is that the IT complexity is a concern of the IT department. It was also mentioned that user participation is needed, regardless the complexity of the IT project. The other half of the interviewees take another point of view: IT complexity is a determinant of the set up of user participation in the requirements definition. The influence is that greater complexity requires tighter procedures for the users to work with. Especially in complex ERP projects it is hard to take a helicopter view, users mostly look at their own part of the project and do not see the bigger picture. Regarding the hypothesis it can not be confirmed that greater IT complexity leads to a less participative approach. Nevertheless, the set up is influenced by the degree of IT complexity. In highly complex IT projects, more tight procedures are set up for users. For the users this decreases the freedom of performing their tasks. This looks like adopting a ‘vanilla’ approach towards the ERP project. According to Shanks and Parr (2000), vanilla projects are more suited for user participation.

Looking at the difference of opinion on the hypothesis, it can be said that individual preferences and experiences play part in this. The interviewees that implied that there is no relation between the set up of the requirements definition and the complexity of the IT project, based their opinion on earlier experiences and their own beliefs about the ability of users to participate in the requirements definition. This shows that the user participation setup has a very personal touch to itself.

(40)

projects there were additional actions taken to increase commitment to the project. For example, an extra group of users was initiated to continuously review the project and its progress.

Regarding the hypothesis it can be said that the complexity of the organizational change does not lead to a less participative approach in the requirements definition but that it is a determinant of the set up of user participation. Much more focus is on creating commitment. Looking at Rengersens (2005) classification of degrees of user participation, it can be said that a complex project on the organizational change side tends to focus more on user participation by having a say and by doing. These two classes give users the idea that they really influence the project and its outcome and therefore are the most suited for increasing commitment to the project (Ives and Olsson, 1984). In contrast, less complex projects on the organizational change side focus more on user participation by having a say and by being informed. In this case the focus is more on informing the users what they can expect. In contrast to participation by doing they can not shape the expectations about the new ERP system.

Before the hypothesized determinants were discussed in the interview, the interviewees named determinants of the set up of user participation in the requirements definition in general. From the interviews it can be derived that three determinants were recognized by most organizations: commitment of users; knowledge of users; and expectations of users.

When commitment is low, or not high enough, this influences the set up of user participation in the requirements definition. The set up of the requirements definition becomes more focused on creating commitment. This results in user participation by doing and user participation by having a say.

Knowledge of users is the second determinant that is mentioned. When the knowledge of users is needed to optimize the project outcome, more users are participating in the requirements definition. Furthermore, when organizations believe that users have knowledge that is crucial to project success, users are participating by doing.

(41)

4.2.3 Interview Results Regarding Determinants of Communication

This section contains several tables that summarize the results of the interviews that were conducted. This section discusses the results on the determinants of the set up of communication. Within communication, two interventions were subject of research: mailings and information sessions. The first is communication through technical means, information sessions are more personal. The eight tables reflect nine projects, because the two Gasunie projects are summarized in one table.

Caption for the tables

0 No relation + Weak relation ++ Moderate relation +++ Strong relation - Not applicable TABLE X

User Participation in Communication at Achmea

COMMUNICATION Participants View Mailing Information

sessions

Relationship Relationship Elucidation

H y p o th e s iz e d D e te rm in a n t Chance of Resistance +++ +++

A high chance on resistance among users towards the change requires more top management commitment in communication. Users are more receptive for communication from top

management.

Knowledge of Users about the Change

+++ ++

The knowledge of users about the change is a determinant of the set up of communication. In both mailings as well as in the information sessions it is aimed to inform users about the change.

Commitment of

Users + +++

(42)

TABLE XI

User Participation in Communication at CBS

sessions

A high chance on resistance among users towards the change does not require more top management commitment in communication. Communication is less in order to not distress the users.

Organizations

standard +++ +++

At CBS a standard approach for communication is applied. This approach is adopted throughout the organization and is based on former experiences.

Knowledge of Users about the Change

+++ ++

The knowledge of users about the change is a determinant of the set up of communication. In both mailings as well as in the information sessions it is aimed to inform users on the change.

O th e r D e te rm in a n ts TABLE XII

User Participation in Communication at Gasunie

sessions

management. Knowledge of

Users about the Change

+++ +++

The knowledge of users about the change is a determinant of the set up of communication. In both mailings as well as in the information sessions it is aimed to inform users on the change. Number of

Employees +++ +++

The amount of employees is a determinant of communication. When a large group is affected by the change, mailing become the prevalent way to communicate. In smaller groups information sessions might be preferred.

(43)

TABLE XIII

User Participation in Communication at Grontmij

COMMUNICATION Participants View Mailing Information sessions Relationship Relationshi p Elucidation H y p o th e s iz e d D e te rm in a n t Chance of Resistance +++ +++

+++ +++

The knowledge of users about the change is a determinant of the set up of communication. In both mailings as well as in the information sessions it is aimed to inform users on the change.

Geographical

Dispersion +++ +++

The geographical dispersion of users at Grontmij was a determinant of the set up of communication. Mailings are a good way to overcome the problem of distance. The set up of information sessions is also influenced. In a less dispersed organization the emphasis in communication would be more on information sessions, a more personal approach.

Commitment of

Users + +++

When commitment is low, more intensive communication is needed. Communication is seen as a good way to create commitment. O th e r D e te rm in a n ts TABLE XIV

User Participation in Communication at NS

+++ +

The knowledge of users about the change is a determinant of the set up of communication. In both mailings as well as in the information sessions it is aimed to inform users on the change. Commitment of

Users + +++

(44)

TABLE XV

User Participation in Communication at Praxis

+++ +

The knowledge of users about the change is a determinant of the set up of communication. In both mailings as well as in the information sessions it is aimed to inform users on the change. Commitment of

Users + +++

When commitment is low, more intensive communication is needed. Communication is seen as a good way to create commitment. O th e r D e te rm in a n ts TABLE XVI

User Participation in Communication at Randstad

management.

Size of Change +++ 0

Randstad Nederland consists of several business units. This caused the need for a website (categorized as mailing here). On this website every affected organization could read about the change. They could also communicate themselves as well. Knowledge of

+++ +

The knowledge of users about the change is a determinant of the set up of communication. In both mailings as well as in the information sessions it is aimed to inform users on the change. Experiences of

Users 0 +++

Earlier experiences showed that just communication through mailing (paper, website) was not sufficient. Therefore, a more personal way of communication was adopted in this project as well.

Commitment of

Users + +++