Information Technology Governance Mechanisms and Processes: The Missing Link Between IT Governance and Firm Performance

(1)

Information Technology Governance Mechanisms and Processes:

The Missing Link Between IT Governance and Firm Performance

University of Groningen Faculty of Economics and Business

MSc BA: Change Management

BERT-JAN BUTIJN

S2799243

Supervisor RuG: Prof. Dr. E.W. Berghout Co-Assessor RuG: Dr. B. Mueller

(2)

1

Abstract

Prior studies have proposed that Information Technology Governance facilitates firms in enhancing their performance. Notwithstanding the valuable insights these studies have yielded, it is still unclear how ITG affects firm performance because a detailed analysis of the relationship between IT Governance processes, and mechanisms, and firm performance remains unexplored. Furthermore, although previous studies have suggested that the quality or maturity of a firm’s ITG is an important factor that influences it’s capability to enhance a firms performance, few studies have investigated how firms increase and assure the quality of their ITG. By reviewing the ITG literature, this study has identified five processes to with regard to governing IT investments: an identification, justification, realization, exploitation and evaluation process, and three types of ITG mechanisms to support these processes. Using data collected from 97 companies in the Netherlands, this study found that a higher maturity of each of these ITG processes is positively related to firm performance. By performing a mediation analysis this study found that a justification process to select IT investments does not mediate the identification process to identify potential IT investments. The results of the other two mediation analysis conducted for this study showed that firms can improve and assure the maturity of their realization and exploitation process only through a mature evaluation process. As such, this research contributes to our understanding of how ITG can facilitate firms in enhancing their performance. Besides extending our understanding of how ITG processes and mechanisms are related to one another and to firm performance, this study contributes to the current body of knowledge regarding ITG by examining the relation between ITG processes.

Keywords: IT governance, structural mechanisms, relational mechanisms, performance

(3)

2

1. Introduction

Recently, researchers have started to investigate how IT Governance (ITG) can enhance a firms’ performance. Luftman and Kempaiah (2007) suggest that ITG facilitates the alignment a firms’ business and IT strategy. According to these authors, ITG will ensure alignment of business and IT strategy by enforcing mutual consultation between the business and IT domain with regards to the IT investments proposals of the firm. In turn, alignment between business and IT strategy will ensure that the IT investments made by the firm will support its business strategy to gain a competitive advantage. Other authors such as Ali, Green and Robb (2015) suggest that by governing their IT, companies will gain more insight in the cost and benefits of their IT investments. This will consequently enable firms exert more control over their IT investments, resulting in a better return on their investments. Previous studies (Luftman, 2003; De Haes, Van Grembergen, & Debreceny, 2013) found that the quality of a firms ITG influences its capability to enhance firm performance. A study by Luftman (2003) suggests that the quality of a firm’s ITG increases as processes become less ad-hoc organized and more structured, which in turn increases the firms capability to align their business and IT strategy. These findings are supported by other previous studies that demonstrated that firms that have a superior quality of IT governance are more likely to gain a high return on their investment (Weill and Ross, 2014) and to have superior organizational performance (Wu, Straub and Liang, 2015).

(6)

5

proposed literature gap mentioned previously two research questions have been formulated. To determine the relation between IT Governance and firm's performance the following research question has been formulated:

RQ1: What is the relationship between the maturity of IT Governance processes and firm

performance?

To address the research gap concerning the development of a firm’s ITG maturity:

RQ2: How do firms increase and assure the maturity of their IT governance processes?

This thesis is organized as follows: In the first and second paragraph of the literature review the relationship between full life cycle governance, ITG processes, and mechanisms will be discussed. In the paragraphs thereafter the relation between ITG processes, mechanisms and firm performance will be discussed. Next, the methodology and data, which has been obtained from senior managers working in 97 different companies in the Netherlands will be introduced. Thereafter the results of this study will be discussed. Finally, after indicating the limitations of this study and suggesting future research opportunities, the conclusions of this study and its implications for practice and research will be presented.

2. Literature Review

(7)

6

and IT function1_{. This chapter begins by reviewing the ITG literature with regard to ITG}

processes. It will then go on to review the literature concerning IT Mechanisms in ITG. Next, the role of ITG mechanisms during ITG processes will be reviewed.

2.1. IT Governance Processes

De Haes and Van Grembergen (2009) argue that formal processes can be referred to as: “the processes that govern strategic decision making, planning and monitoring to ensure that business needs are aligned with IT policies” (De Haes & Van Grembergen, 2009, p. 124). Researchers have begun to investigate which processes can be discerned. For instance, research by Weil and Ross (2004) and Xue, et al. (2008) has investigated how firms govern the decision making processes with regard to IT investments. Other studies have examined how ITG facilitates the realization process of IT investments (Jiang et al., 2004; Ali et al., 2015). Berghout and Nijland (2002) suggest that five ITG processes can be discerned which are intricately related to each other: First, during the identification process an IT asset is identified and proposed as an investment opportunity. Next, during the justification process the costs and benefits of the IT investment proposal are assessed and weighed against that of other investment proposals. Third, during the realization process the IT asset is developed. During the fourth process, the exploitation process, the developed IT asset becomes operational and a company can reap the benefits of their investment. Finally, once an IT asset has been discarded its performance should be evaluated to assess how the previously mentioned ITG processes can be improved, by carrying out an evaluation process. Although prior research (Weil & Ross, 2004; Xue, et al., 2008; Jiang et al., 2004; Ali et al., 2015) has made valuable contributions to investigate the relation between ITG processes and firm performance, these studies have only investigated how firms govern a specific ITG process (e.g. an decision making process). Furthermore, although it can be argued that Berghout and Nijland (2002) provide a more holistic view of which ITG processes can be discerned, the relation between these processes and firm performance remains empirically untested.

1_{Because ITG does not solely govern Information Systems or just technical infrastructure and both can have an}

(8)

7

2.2. IT Governance Mechanisms

De Haes and Van Grembergen (2009) suggest that structural and relational mechanisms should support ITG processes to ensure that the relevant stakeholders actively participate in the processes that govern the decision making process and monitoring of IT. They argue that structural mechanisms can be referred to as the division of roles and responsibilities in IT decision making (De Haes & Van Grembergen, 2004). According to De Haes and Van Grembergen (2009), relational mechanisms can be referred to as the communication and shared learning between the IT function and the business. Ali and Green (2012) argue that performance measurement mechanisms such as investment appraisal methods, IT chargeback systems and service level agreements should also be an integral part of a firms ITG in order to effectively govern IT. These performance measurement mechanisms encompass a set of metrics that provide a firms management with an accurate analysis of how IT is performing, of the potential of IT investments (Ali and Green, 2012). Taken together, the studies reviewed in this section indicate that ITG governance mechanisms play an important role in supporting ITG processes.

2.3. IT Governance Process Maturity

(9)

8

The evidence presented in this section suggests that a firms can deploy ITG mechanisms to formalize communication means, measurement metrics, roles and responsibilities, while carrying out an ITG process. Which in turn leads to a higher quality of these processes. Therefore, throughout this research, the maturity of an ITG process will be referred to as: the extent to which a company has deployed structural, relational and performance measurement mechanisms to support the governance of an ITG process. Furthermore, in the literature presented in this section, the maturity of ITG processes has been associated with a firm’s capability to coordinate business and IT efforts. As such, the maturity of ITG processes might play a pivotal role in enhancing a firm’s performance. The current study examines the relation between the maturity of ITG processes and firm performance by drawing upon the ITG processes identified by Berghout and Nijland (2002). Firstly, in accordance with a study by Hoffman and Broekhuizen (2010), the research model for this study is presented in Figure 1.

Figure 1: Conceptual model

Maturity Identification process Maturity Justification process H2A H1 H2 Maturity Realization process H5A Maturity

Evaluation process Firm Performance

(10)

9

2.4. Conceptual Model

In this model it is proposed that the maturity of the identification, realization and exploitation processes as identified by Berghout and Nijland (2002), are related to firm performance. While prior research has acknowledged the importance of ITG process maturity (e.g. Luftman, 2003), and its relation to firm performance (e.g. Peterson, 2004), these studies focused on investigating the effect of the maturity of a combination of processes on firm performance without discerning these processes. Contrary, other studies (Xue, et al., 2008; Jiang et al., 2004) have examined how the governance specific ITG processes is related to firm performance. These studies, however, have not accounted for the effects that ITG processes might have on each other. There has been little research conducted to view ITG processes more holistically with regard to the relationships amongst them and to extend their relation to firm performance. Yet, Berghout and Nijland (2002) suggest that these processes are related to another: therefore, the nature of the relation between these processes can be important in determining how the maturity of each of these process enhances firm performance. This study attempts to contribute to the ITG literature by filling these gaps. Hence, it is postulated in the conceptual model that the maturity of the justification process mediates the effect of the identification process. Furthermore, in the conceptual model it is proposed that the maturity of the evaluation process mediates the effect of the maturity of the realization and exploitation process. In the following paragraphs, the structural and relational mechanisms to support these processes will be reviewed. Furthermore, the relationship between these processes will be further elaborated upon.

2.5. Identification Process for IT investments

(11)

10

2.5.1. Performance measurement mechanisms to support the identification process. The literature suggest that companies should be aware of their environment to identify

threats and new IT investment opportunities (Wade, 2015). Therefore, a prerequisite to identify IT investments is that firms examine the developments in their environment (Berghout and Nijland, 2002). In addition, leveraging new IT investment opportunities might entail that firms need to revise their business strategy (Downes & Nunes, 2013). Consequently, it might also be necessary for firms to revise their IS strategy since IS can play a pivotal role in supporting a business strategy (Chan, Huff, Barclay, & Copeland, 1997; Sabherwal & Chan, 2001). Chen, Mocker and Preston (2010) argue that from this point of view an IS strategy can be seen as determining the position of IS to support the business. In addition, Chen et al. (2010) argue that an IS strategy can also be viewed as a master plan for the IS function that focusses on the identification of the required IT assets, human capital and structure to enable business strategy. In this perception IS strategy should provide the IT function with a plan that can serve as a basis for the management of IT assets. Part of this plan is the describing which new IT investments are required to support the business (Chen et al., 2010). Furthermore, the plan will indicate whether an IS system is no longer able to support the business. Subsequently, a decision should be made to either invest in improving the IS system or to discard them. Berghout and Nijland (2002) suggest that as a part of this plan firms should periodically check whether or not it is necessary to discard or improve operational information systems.

2.5.2. Relational and structural mechanisms to support the identification process.

(12)

11

role of IT within organization which decision making approach (top-down or bottom up) is more appropriate; A bottom up approach for decision making is appropriate for firms that focus on an IS strategy that helps to foster innovation. Contrary, a top down decision making approach is more suitable when the firms pursue a more conservative IS strategy. Firms that pursue a conservative IS strategy do not seek innovate new IT but rather focus on the effective implementation and management of IT that is already available on the marketplace. Furthermore, A study by Weil and Ross (2005) substantiates the point made by Chen et al. (2010); their study found that firms take on different approaches to organize their IT Governance. They found that some organizations organize their decision making process concerning IT bottom up, while others take on a more top down approach depending on their corporate strategy. From this reasoning it follows that both top management and employees can be regarded as a source for new IT investment proposals.

However, Earl (1989) and Smith & Tushman (2005) argue that identifying business opportunities and concomitant IT proposals can be done top-down, bottom-up but also inside-out. To illustrate this point: Earl (1989) suggests that suppliers and business partners can also help to identify business opportunities and concomitant IT proposals. For this study both internal and external actors will be regarded as potential sources for the identification of IT proposals. Although companies might differ in their approach to identify IT investment proposals, it is important that one person (or department) is responsible for the identification of IT investment proposals (Rau, 2004). Irani and Love (2002) suggest that firms should provide guidelines how investment proposals should be drawn up. They argue that these guidelines are important to assure that the IT investment proposals attain the quality that the firm desires. These guidelines increase the quality of IT investment proposals by specifying which financial metrics should be stated in the proposal, or what content has to be addressed. Furthermore, by specifying the metrics and content which have to be addressed in IT investment proposals, these guidelines assure uniformity of IT investment proposals. Without uniformity of the IT investment proposals cannot be compared. Therefore they argue that these guidelines are needed to make comparison between IT investments proposals possible.

2.5.3. The relation between the identification process and firm performance. Prior

(13)

12

of Entrepreneurial Alertness. Which they define as “the capability of a firm to explore its marketplace, detect areas of market-place ignorance, and determine opportunities for action” (p. 250). The concept of entrepreneurial alertness formulated by Sambamurthy et al. (2003) closely resembles the definition of hyperawareness as proposed by Wade (2015) who states that: “Hyperawareness is an organizations capability to recognize future trends that will impact an organization”(p. 11). Berghout and Nijland (2002) suggest that during an identification process firms should examine their environment to detect IT investment opportunities. Therefore, one can argue that by examining their environment during an identification process, firms will be better able to recognize threats and opportunities that might be present in their environment. Furthermore, one can argue that an identification process which addresses capturing IT investment proposals from suppliers and customers, employees and top management (bottom-up, top-down and inside-out) will help firms to detect IT investment opportunities. Therefore, it can be argued that the maturity of a firms identification process is associated with its performance. This leads me to suggest the following hypothesis:

H1: A higher maturity of the IT investment identification process is positively associated with

firm performance.

2.6. Justification Process of IT Investments

(14)

13

2.6.1. Performance measurement mechanisms to support the justification process.

The ITG literature suggests several performance measurement mechanisms to determine which of the identified IT investment opportunities is the most profitable to their firm. For instance, Chen and Small (1994) suggest when companies want to be able to appraise their IT investment choices, that these firms should first determine which benefits they obtain. Next these firms should determine the criteria to measure these benefits. The benefits of an investment in IT can be divided into three categories: strategic, tactical or operational (Irani & Love, 2002). Operational benefits, and to a lesser extent tactical benefits, can be directly measured because they are more directly related to cost reduction and improved business performance (e.g. customer satisfaction). However, capturing strategic benefits has been proven to be difficult (Irani & Love, 2002) because the effects of these investments cannot always be distinguished. However, in order to make a comparison between IT investments possible, it is important to develop distinctive criteria by which operational, tactical, and strategic benefits can be measured (Irani & Love, 2002). Another reason why accounting for investment costs in IT has proven to be difficult is because it is arduous to directly relate the benefits that are obtained by an IT asset to the IT asset. In general the benefits gained by an IT asset are gained by the business they support (Parker et al., 1988). A third reason why accounting for IT assets has been proven to be difficult is that investments might relate to one another. To illustrate this point; When investing in IT infrastructure such as servers or an network, the net worth of the asset might not be evident because it does not directly supports the business. However, these investments might be necessary to enable future investments or to support currently deployed information systems. Besides considering how the proposed IT investments add value, firms should also take into account how these proposals relate to one another (Nijland, 2004).

(15)

14

IT provides a more complete overview of the anticipated costs and benefits. However, to make comparison possible these intangible costs need to be quantified (Money, Tromp & Wegner, 1988). In addition, to accurately reflect the costs incurred during the IT assets lifespan an estimation should be made of not only the initial investment to realize the asset but also its exploitation. Finally, clear project goals and objectives contribute significantly to a common understanding about what the system should be able to do. Berghout and Nijland (2002) argue that formulating project goals has three main goals: First, the formulated goals serve as a benchmark during the ex-post assessment of an IT asset. Second, establishing project goals enables the organization to assess ex-post whether or not an IT asset reached its goal. Third, if the investment in an asset did not reach its goals, the formulated goals will help determine why this was the case.

2.6.2. Relational and structural mechanisms to support the justification process.

Whenever a proposals meets the criteria it is ready to be appraised. This raises the question who should appraise the proposal. A study by Vessey and Scott (2002) found that involvement of strategic management, in the form of a steering committee or a board dedicated to decision making concerning IT, has been found to contribute to a higher chance of successful implementation. According to Ehie and Madsen (2005): “Top management support is very invaluable in ensuring that ERP projects come to fruition”(p. 554). This support might include providing strategic direction by being actively involved in various high-level cross-functional implementation teams. By involving the strategic management of the firm support is gained for the investment proposal. Moreover, the IT domain gets an insight in how the proposal might relate to other parts of the business domain. It must be noted however, that both the study from Ehie and Madsen (2005) and Vessey and Scott (2002) focus on the implementation process of an ERP system.

(16)

15

regards to the goals the IT asset should achieve. Furthermore, it is paramount that the client provides insight about the composition of tasks and processes that might be associated with the realization of the asset (Kearns & Sabherwal, 2006). Bassellier, Reich and Benbasat (2001) found that managers tend to fall short of skills and knowledge to make the right decisions with regards to IT. The IT domain have an important role in advising the business domain in the (im)possibilities of information technologies. After the information technology has been realized the department that requested the investment will be working with the newly developed IT. As such, besides being the owner of an IT asset, line managers can also be viewed as the users. Bala and Venkatesh (2016) found that involving users during the decision-making process about the features, functionalities and form of information technology increases the chance that the users will see the IT as an opportunity rather than as a threat.

2.6.3. The relation between the identification and justification process of IT investments. Organizational politics can play an important role in the decision making

processes (Pettigrew, 2014). As such, the process of decision making does not always yield results that are in the best interest of the organization but rather that of the most powerful stakeholders (Pettigrew, 2014). A study by Farbey, Land et al. (1999) found that in decision making concerning IT, politics can also play a pivotal role in the selection of IT investments. Furthermore, they found that to diminish the negative effects of organizational politics on the decision-making process of IT investments, managers will often force their subordinates to conduct an ex ante evaluation to justify why they propose to invest in IT. By doing so, these managers attempt to exert more influence over the decision making process concerning the selection of IT investments. Likewise, a study by Langley (1989) found that senior executives often impose formal ways of information analysis as a means to exert control over organizational decision making and organizational politics. She found that by doing so, executives attempt to force their subordinates to justify the decisions they make. Furthermore, she found that in some cases formal information analysis is used as a means by line management to justify only their projects. She argues that in these cases: “There is often no pretense here that information was a motive for the analysis, and the communicator's own criteria for the decision may be openly quite different from that of the principal target” (Langley, 1989, p. 606).

(17)

16

making processes in their organization. One can argue that when firms become better at identifying IT investment proposals and the number of identified proposals increases, management will feel a greater need to exert control over the selection of IT investments. Which will in turn result in an increase of the mechanisms deployed to support the justification process. Furthermore, the literature reviewed in the previous paragraph (see 2.3.2.) highlights the importance of guidelines for IT investment proposals. These guidelines assure the quality and uniformity of IT investment proposals. Therefore, it can be expected that not providing guidelines for the members of their organization about how they should draw up IT investment proposals will have two consequences: Firstly, important information required to compare and appraise IT investments will be missing. Secondly, comparison between IT investments will prove to be difficult because the IT investment proposals do not contain uniform content. Combined, these arguments suggest that the maturity of the identification process is related to the maturity of the justification process.

H2: A higher maturity of the identification process is positively associated with the maturity of

the justification process.

2.6.4. The maturity of the justification process as mediator. This study proposes an

(18)

17

2015, p. 12). Wade suggests that governance processes to provide information and to organize cross functional deliberation can help organizations to make more informed decisions (Wade, 2015). In similar vein as Wade, Berghout and Nijland (2012) suggest that line, IT and strategic management should be involved in the justification process. Furthermore, they argue that by carrying out an justification process, companies obtain the information they need to make informed decisions with regards to the IT investment proposals they have identified.

Farbey et al. (1999) suggest that decision makers should justify why the IT investments they are proposing add value to their firm. They argue that neglecting this process might result in not selecting the IT investments that support their business strategy. Instead the most powerful stakeholder will decide which investments will be made. Furthermore, Farbey et al. (1999) suggest that decision makers should justify their IT investment proposals by carrying out an ex ante evaluation. They state that an ex-ante evaluation of proposed IT projects provides an organization “with a good estimate as to what the outcomes of maintaining and installing the system in question might be, and second to evaluate the outcomes in terms of the organizations interests”. Therefore it provides a firm with an objective basis to determine whether the investment is worthwhile (Farbey et al., 1999). In addition, an ex-ante evaluation will provide decision makers with the required information to weigh up their IT investment proposals and select the most profitable ones (Farbey et al., 1999). Overall, the studies presented in this section highlight the importance of an justification process to select the most profitable IT investments after identification. Therefore, it is expected that the effect of the maturity of the identification process on firm performance is mediated by the maturity of a firms justification process.

H2A: The maturity of the justification process mediates the effect of the maturity the

identification process on firm performance.

2.7. The Realization Process of IT investments

(19)

18

delivery (Lee & Xia, 2010). From these examples it becomes clear that particular attention should be paid to the realization process of an IT asset. Indeed, various studies (Vessey & Scott, 2002; Ehie & Madsen, 2005; Markus, 2004) have shown that the process of realizing IT assets can be a perilous one. However, Jiang et al. (2004) found that a more mature realization process of IT assets is positively associated with project performance. Determining which ITG mechanisms support the realization process can prove to be important in achieving higher success rates. In the following sections the literature with regards to the realization of IT assets will be reviewed.

2.7.1. Performance measurement mechanisms to support the realization process.

Vessey and Scott (2002) suggest that before commencing the development of information technology an ex ante evaluation should be carried out. Indeed, several authors (Barki, Rivard & Talbot, 2015; Jiang et al., 2004; Stefanou, 2001) suggest that in order to successfully anticipate unwanted risks that could jeopardize the realization of the IT asset an assessment of these risks is needed. A risk assessment should not solely concern the technical aspects of the realization process, but also the human factors that can influence the realization of the asset (Jiang et al. 2004). This is needed because as Vessey and Scott (2002) state; “The majority of project failures do not stem from technology issues per se but from management issues surrounding the implementation” (p. 79).

(20)

19

969). Furthermore, it provides organizations with an overview to coordinate and execute their activities.

By creating milestones firms can gain more insight in which goals should be reached and at what costs. They can also serve as a roadmap by which managers and participants can make sense of complex projects (Yakura, 2002). Furthermore, these milestones can serve as specific check points during the project to decide whether or not to continue with the realization of the project, or to make adjustments if necessary (Ward, Taylor & Bond, 1996). If for instance, the project has ran over budget for a particular milestone, initially planned functionalities could be discarded to reduce costs. Another example would be that during a project it has taken more time to reach a milestone than planned. To deliver the project on the final deadline it might be necessary to compromise in time with regards to other activities. Moreover, Kotter (1995) suggests that short-term wins are vital to keep a change initiatives on track. Milestones can indicate which short-term wins there are to be made during the project and thus give the participants a sense of achievement. However, to in order to decide whether a milestone is met continuous monitoring of cost, time and functionality throughout the project is needed. This helps making sure that the project stays within budget, is delivered in time and encompasses the desired functionalities.

2.7.2. Relational and structural mechanisms to support the realization process.

During the project both technical and human related problems might emerge. Both can result in the expenditure of extra costs or the project running over time (Vessey & Scott, 2002). According to Berghout and Nijland (2002) problems that arise during the realization of the IT asset should be documented. This serves two purposes: First, if companies encountered these problems before and formulated a solution, the company does not need to reinvent the wheel when encountering similar problems during the realization of future projects. Second, the report about the problems that have emerged can serve as a basis to inform the board. However, a mechanism should be in place that actually enforces notification of the board in case of problems. White and Fortune (2002) actually found that a lot of problems during projects tend to escalate because no one in management is warned about the problems encountered. As a result, management cannot act upon the problems that arose.

(21)

20

approaches is the Waterfall method. These method have been criticized on the basis of their inability to respond to changing user requirements and time/cost overruns (Lee & Xia, 2010). As a response to the inability of traditional methods to adequately manage software development, Agile methods such as Scrum, Extreme Programming and DSDM have recently become popular amongst software developers (Lee & Xia, 2010). Lee & Xia (2010) found that agile methods indeed provided software developers with more flexibility to cope with emerging changes. They argue that this can partially be explained by the fact that agile method provide a clear project management structure yet leaves room for emergent changes.

2.7.3. The realization process and firm performance. Previous studies (McHugh &

Hogan, 2011; Lee & Xia, 2010) have shown that project and software methodologies can help firms to diminish the risks resulting in project failure. Other studies (Barki, Rivard & Talbot, 2015; Jiang et al., 2004; Stefanou, 2001) have demonstrated that carrying out an ex ante evaluation helps to anticipate unwanted risks. Some of these risks are for instance, failing to gain the support from top management, ineffective resource management and ineffective communication (Cule, Schmidt, Lyytinen & Keil, 2000). A study by Barki, Rivard and Talbot (1993) found that when the approach taken to carry out an IT project becomes more formalized and mature, the chance of meeting the objectives of the project increases. These findings are in line with a study by Jiang et al. (2004) which showed that firms that have a more mature software development process will be better able to diminish the risks associated with the realization of IT assets. Furthermore, their study found that a more mature software development process is positively associated with project success. Therefore I expect that firms that have more mature ITG processes to guide the realization process of their IT assets have a higher chance that the realization of their IT projects will succeed. Which in turn leads to better firm performance because failed IT asset development projects are associated with high costs, missed strategic opportunities (Laudon & Laudon, 2004) and in some cases bankruptcy of the firm that commenced the software development project (Vessey & Scott, 2002). Therefore, it is proposed:

H3: A higher maturity of the realization process of IT assets is positively associated with firm

(22)

21

2.8. The Exploitation Process of Information Technology

Once a project has successfully been finished and the IT asset has been realized, a firm can start to reap the benefits of the asset. In order to do this, companies need to keep the IT asset operational during its economic lifespan (Banker and Slaughter 1997). In the last two decades a large body of literature on the topic of software maintenance has been established that suggests that software needs to evolve during its lifespan in order to keep it operational, and to ensure that it meets user requirements (Goltz et al., 2015). Evolving information systems entails the need of regular software maintenance. This maintenance mainly takes place by adding new functionalities to the system to keep up meeting requirements. Fixing bugs and improving the systems performance are an important incentive to perform maintenance to the software in order to keep it operational (Goltz et al., 2015). In the next sections, the literature with regards to the exploitation of IT investments will be reviewed.

2.8.1. Performance measurement mechanisms to support the exploitation process.

To determine when maintenance is needed, the performance requirement of each IS system should be registered. By registering the functionalities of the system firms can determine when there is a discrepancy between what the system in its current state is capable of and what it should be able to achieve (Banker & Slaughter, 1997; Goltz et al., 2015). Furthermore, because a firm has to take into account the costs that are being incurred during the exploitation of an IT asset it should also embed ITG mechanisms that govern the registration of these costs. Therefore the allocation of a budget to keep the asset operational and functioning is recommended (Chapin et al., 2001). Benefits that are derived from exploiting an asset will become salient within the business domain. By contrast, costs are likely to be directly incurred when the IT domain undertakes actions to support the exploiting of the IT asset. Because IT assets are developed and exploited to facilitate the business domain it is proper to allocate IT costs to that part of the business domain that profits from the asset (Earl, 1989).

(23)

22

that are a more attractive substitute because they can (out) perform the operational asset at a lower cost (Berghout, Nijland & Powell, 2011). When the costs of an asset start to outweigh its perceived benefits a firm might decide to discard the altogether, whether or not a substitute is available. Decision -making to discard, substitute or improve an asset has been proved to be problematic if the exploitation costs of each system are not registered (Berghout, Nijland & Powell, 2011).

2.8.2. Relational and structural mechanisms to support the exploitation process.

Although this method of cost allocation can be regarded as common sense it creates an agency problem; As pointed out in the previous paragraph, one of the tasks of the IT domain is to advice the business domain of the possibilities that IT can bring and manage its exploitation. One of the main reasons that the IT domain has this role is because in general managers lack the capabilities and insight to make decisions concerning IT themselves. In literature such a situation would be referred to as an agency problem (Eisenhardt, 1989). Jensen and Meckling (1976) found that in an attempt to control such situations firms tend to engage in a (social) contract with their agent. By doing so the firm has more control over the outcome of the agents actions. Contracts with the specific purpose of defining the relation between the principle and the agent for IT assets are commonly known as Service Level Agreements (SLA). Goo et al. (2009) found that these contracts generally consist of three components: First, a SLA defines who bears the process ownership of the information system, the service level objectives and the service level contents. Second, it establishes how and when software maintenance and innovation takes place. Third, the SLA provides encompasses the legal clauses concerning conflict arbitration and enforcement. The findings of the study suggest that firms that have a well-defined SLA which encompasses all three components will be managing the exploitation more harmonious and more adequate.

2.8.3. Firm performance and the exploitation process. Banker and Slaughter suggest

(24)

23

spending during the exploitation process. Therefore, mechanisms that help firms to provide insight in the incurred costs during the exploitation phase can be of value. Furthermore, without registering the most important functionalities of the system firms will be unable to determine when maintenance is needed. Neglecting needed maintenance increases the chance that an information system will no longer be able to deliver benefits as it becomes less efficient, or incapable to carry out its tasks altogether (Goltz et al., 2015). Therefore, an effect of neglecting to perform maintenance is that firms will be less able to reap the benefits of their assets (Goltz et al., 2015). A study by De Haes et al. (2011) conducted at an Dutch airline company suggests that combined these mechanisms help firms to better control the exploitation process of their IT assets. Their study found that when the airline company re organized their ITG processes, and deployed structural and relational mechanisms they gained a better grasp on the costs of their continuously deployed IT systems. These results suggests that firms that have a high ITG maturity with regards the exploitation phase of an IT assets lifecycle are therefore more likely to be able to actually reap the benefits of their investments and control their maintenance costs. The accompanying research hypothesis is:

H4: A higher maturity of the exploitation process of IT assets is positively associated with firm

performance.

2.9. The Evaluation Process of IT Assets

(25)

24

2.9.1. Performance management mechanisms to support the evaluation process. To learn about the processes that govern their IT investments, firms ought to periodically analyze their complete IT cost/benefit management. Reviewing the reports that have been drawn up during the realization and exploitation phase of their IT assets will allow firms to assess the differences between the expected and actual costs (Davern & Kauffman, 2000). In addition, Berghout and Nijland (2002) suggest that firms should conduct both a quantitative and a qualitative analysis of the investments’ benefits when evaluating their realization and exploitation process. Furthermore, they argue that the exploitation results of IT assets should be compared to their goals. By evaluating these differences companies can lay bare which improvements can be made to their ITG (Farbey et al., 1999).

2.9.2. Relational and structural mechanisms to support the realization process.

Strategic management should be involved in the evaluation process because the results of the evaluation provides them with the opportunity to learn how they can improve their firm’s ITG processes. Another reason to involve strategic management in the evaluation process is that they are responsible for the IS strategy (Chen et al., 2010), and they have the formal authority to change the governance of the realization and exploitation process when desired. Peterson (2004) argues that to successfully change their IT governance, companies should compare the current state of their official ITG model to its desired state.

2.9.3. The relation between the realization and exploiting process of IT investments and the evaluation process. The literature reviewed in the previous sections suggest that

(26)

25

process they will perhaps not drawn up reports with regards to the performance of their assets during exploitation.

H5A: A higher maturity of the realization process of IT assets is positively associated with the

maturity of the evaluation process.

and,

H5B: A higher maturity of an exploitation process of IT assets is positively associated with the

maturity of an evaluation process.

2.9.4. The maturity of the evaluation process as mediator. Although it can be argued that ex post evaluation of IT investments is a practice which enables firms to learn about their ITG, it does not explain how firms can profit from learning about their ITG. A study by Zhang, Zhao and Kumar (2016) found that high ITG maturity has a positive effect on developing superior IT capabilities. IT capabilities can be defined as “a firm’s ability to mobilize and deploy IT-based resources in combination or co-present with other resources and capabilities” (Bharadwaj, 2000, p. 171). In turn, according to Zhang, et al. (2016), superior IT capabilities will result in higher firm performance (Aral and Weill, 2007; Bharadwaj, 2000). Adding to the argument by Zhang et al. (2016), one can argue that superior IT capabilities if captured by an ex post evaluation process can also increase the maturity of the ITG processes. In turn this will lead to increased firm performance. Aral and Weill (2007) refer to IT Capabilities as a combination of practices and IT management competencies. IT management competency can be regarded as the competence of individuals or groups to manage or accomplish organizational tasks (Aral & Weill, 2007). They suggest that: “Competencies are developed through learning and the repeated performance of contextual activities” (Aral & Weill, 2007p. 767). As such, individuals or groups within an organization built IT competence by learning from carrying out activities related to their tasks (Aral and Weill 2007). According to Aral and Weill (2007) a set of repeating activities can be referred to as a practice. By learning from the process of conducting these activities individuals or groups increase their IT competence and know-how. In turn this enables them to adjust their practices to become more effective and efficient in accomplishing their tasks. As such, the combination of practices and IT competency mutually reinforce and support each other (Aral and Weill, 2007).

(27)

26

assets gain IT competency because the individuals’ IT competence or know-how is to a large extent tacit (Aral and Weill, 2007). Edmondson, Winslow, Bohmer, and Pisano, (2003) found that tacit knowledge is not easily transferred within an organization. This would imply that an individual’s IT competence that could be used to improve the realization and exploitation process of IT assets might not be widely available to every member of the organization. This stresses the importance of ex post evaluation because it will facilitate a larger number of organizational members to learn about the realization and exploitation process of their organization and how to make improvements (Davern & Kauffman, 2000). Peterson (2004) highlights another important aspect of an evaluation process; He argues that over time gaps may emerge between actual and desirable behaviors with regards to a firms ITG decision making. For instance, some executives may delegate their decision rights to subordinates which is not in line with the companies IT policies. Therefore, a firms formalized IT governance model might not reflect the actual decision making in practice. Peterson (2004) argues that therefore: “a critical activity in designing effective IT Governance is architectures is devising a diagnostic system to assess the actual and the intended IT Governance model, and its effectiveness” (p.18). The arguments presented in this section suggest that an evaluation process is important to facilitate members of an organization to learn about the realization and exploitation process of their organization, and how to make improvements. Furthermore, the arguments presented by Peterson (2004) suggest that it is important for firms to regularly evaluate their ITG. Neglecting this process might entail that a company has deployed several ITG mechanisms on paper, yet in practice these mechanisms will not be used properly. Thus, the effect of the maturity of a firm’s realization and exploitation process might be mediated by the maturity of their evaluation process.

H6A: The maturity of the evaluation process mediates the effect of the maturity the realization

process on firm performance.

and,

H6B: The maturity of the evaluation process mediates the effect of the maturity the exploitation

(28)

27

3. Methodology

3.1. Data Collection Procedure

For this thesis, data from surveys filled in by senior managers from 2010 to 2017 has been used. The data has been collected during an executive course in IT control which these senior managers attended. In total, 132 senior managers working in various industries filled in the survey. The survey used for this study is based on a quickscan instrument which has been developed by Nijland and Berghout (2002) to indicate the maturity of a firms’ ITG processes. Respondents were asked to fill in the survey to indicate how they assessed the maturity of their companies’ IT governance processes. In addition, the respondents were also asked to provide information about the number of full time equivalent employees that work at their company. The results of the survey were immediately available and accessible to each respondent after having been filled in. By providing these results respondents obtained an indication of the maturity of their firms ITG processes to govern: (1) the identification of IT investments (2) the justification of IT investments (3) realization of IT investments (4) the exploitation of IT investments and (5) the evaluation of IT investments. After the collection of the surveys, the results have been discussed and assessed during a classroom meeting. There were some respondents however, that filled in the surveys that did not want to participate in the discussion and assessment of their firms ITG. In these cases, the data concerning the firm on which the respondent reflected have not been stored, and as such were not part of the initial data set. The researcher that conducted this study has not been part of the initial data collection process.

(29)

28

department they were working for. Consequently these respondents did not indicate the maturity of these processes from a firm level perspective. Since this study focusses on the maturity of a firms’ ITG processes on a firm level, only surveys for which respondents reflected on the companies IT Governance from a firm-level perspective have been selected.

The initial dataset did not contain any financial information, or information about the industry type of the firms that were selected for this study. However, this financial information was required to determine the performance of each firm. Furthermore, information about the industry group of the companies was needed. Therefore, after selecting the surveys that were eligible for this study, these surveys have been matched with the financial data for each company of the respective year in which the survey has been filled in by the professional (2010-2017). In addition, the surveys that were eligible for this research have also been matched with data about the industry group for each company. As shown in Table 1, the financial data and information about the industry group of each firm has been retrieved from a secondary data source: CompanyInfo, a website that provides access to the financial statements of the Dutch chamber of commerce. After matching the selected surveys with the required data, a total of 97 firms were selected to study for this research. The information obtained from the financial statements showed that the firms selected for this study were categorized into eight industry groups by the Dutch Chamber of Commerce. A complete overview of the descriptive statistics of the sample characteristics can be found in Appendix B.

Table 1: Summary of Key Informants/Data sources

Information on: Key informant/data source:

Maturity identification process Senior Managers

Maturity justification process Senior Managers

Maturity realization process Senior Managers

Maturity exploitation process Senior Managers

Maturity evaluation process

Amount of FTE employees working at firm

Senior Managers Senior Managers

Financial information: ROA, ROE, Return on Revenues Industry group

(30)

29

3.2. Measurements

3.2.1. Maturity of ITG processes. As mentioned before, for this research the relation

between the maturity of ITG processes and firm performance is investigated by drawing upon the ITG processes identified by Berghout and Nijland (2002). For this research, the maturity of the identification, justification, realization, exploitation and evaluation process is defined as the extent to which a company has deployed structural, relational and performance measurement mechanisms to support the governance of each of these processes. To measure whether a company has deployed a particular structural, relational or performance measurement mechanism, 42 items have been used from the quickscan instrument developed by Berghout and Nijland (2002). In the survey used to collect the data, each ITG mechanism is represented by an item which can be answered with a yes, to indicate that a mechanism has been deployed by a firm to support an ITG process, or a no which indicates that the firm has not. A total of seven items has been used to measure the maturity of the identification process. Nine items have been used to gauge the maturity of the justification process. To measure the maturity of the realization process, ten items have been included in the survey. Eight items have been used to measure the maturity of the exploitation process. The maturity of the evaluation process is measured by eight items. However, the initial dataset which has been offered to the researcher conducting this research did not contain the answers given by respondents for each item. In the data set offered to the researcher, the total number of mechanisms deployed by a firm per process were given, which was used to measure the maturity of an ITG process. Table A1 in Appendix A provides an overview of all items included in the survey.

3.2.2. Firm performance. For this study, firm performance is defined as the financial

(31)

30

by the total assets minus liabilities. A composite index has been created by calculating the mean of each firms ROA, ROE and profit margin.

3.2.3. Control variables. Research with regards to ITG has not always concluded that

the maturity of a firms ITG has a direct or positive relation with firm performance and these inconsistent results are perhaps due to lack of control variables in the analysis (Chan & Reich, 2007). Hale and Cragg (1996) found, for example, that in larger organizations, managers more commonly introduce formal processes and structures to ensure alignment between the IT and business function. Larger firms tend to have more slack, which can be devoted for alignment purposes. While firm size is usually treated as one of the antecedents to maturity of a firms ITG, for this study it is modelled as a control variable directly affecting the maturity of the justification and evaluation process. By doing so, this study tries to rule out rival explanations for the increase in the maturity of the justification process other than the maturity of the identification process. In addition, firm size is also modeled as a control variable for the maturity of the evaluation process to rule alternative explanations for the relation between the maturity of the realization and exploitation process, and the maturity of the evaluation process. Furthermore, prior research (Tan, 1995; Khanna & Rivkin, 2001) has found that a firm’s industry group can have an effect on its performance. Therefore, industry group has been added as a control variable for firm performance. Because the sample used for this study has been collected from small- to large-sized firms across different industry groups, firm size and industry group will be specified as control variables for ruling out alternative explanations in the research model. The total number of full-time equivalent employed by a firm was used as a proxy of firm size. As mentioned before, the firms selected to study for this research can be categorized into eight different industry groups according to the Dutch Chamber of Commerce. However, because the selected firms were divided into eight industry types, each industry group encompassed only a small number of firms. This can potentially decrease the statistical power of tests to compare the differences between groups (Cohen, 1992; Mahoney & Magel, 1996)2. Therefore, in line with other previous studies (Tan, 1995; Khanna & Rivkin, 2001), the selected surveys have been re-categorized into five industry groups. A complete overview of the categorization of industry types used for this study can be found in Appendix B, Table B4.

2_{To be more complete: when comparing more than two groups for differences, an one way Analysis Of Variance}

(32)

31

3.4. Data Analysis and Results

3.4.1. Reliability and validity. As mentioned before, the maturity of the identification,

justification, realization, exploitation and evaluation process has been measured using the items derived from the quickscan instrument proposed Berghout and Nijland (2002). The quickscan has been developed, using a Delphi method to assure sufficient content validity, by interviewing 27 senior business and IT managers from 9 firms3. Subsequently, based on the findings of these interviews, a set of 42 ITG mechanisms have been identified that can support the governance of the identification, justification, realization, exploitation and evaluation process. Berghout and Nijland (2002) suggest that each of these mechanisms serve specific or multiple goals in the challenges offered while carrying out one of these ITG processes. The ITG literature has been reviewed to validate this suggestion.

In this thesis three mediating effects are hypothesized. For a mediation effect to exist it is a necessity that there is collinearity between the independent, mediating and dependent variable (Baron & Kenny, 1989). In such cases one can argue that a multi-collinearity problem might make the coefficient estimates of a model much more sensitive to minor changes in the model or increase the variance of the coefficient estimates (Grewal, Cote & Baumgartner, 2004). Therefore it is recommended to assess the influence of multicollinearity between the constructs proposed for the mediation analysis (Baron & Kenny, 1989). Multicollinearity between the constructs has been assessed by examining the variance inflation factors (VIFs) in each test. The highest VIFs which have been found and assessed, while testing the hypotheses can be found in Table 2. The VIFs range from 1.000 to 2.039, which is well below the 3.3 guidelines recommended by Fox and Monette (1992), and Chatterjee and Hadi (2015). The data used for this study has been collected from 2010 to 2017 over a span of seven years. As such, autocorrelation of the data collected for this study might bias the results of the tests performed for this study (Savin & White, 1977). To assure that autocorrelation does not bias the results of the tests performed, the Watson statistics has been assessed (see Table 2). The Durbin-Watson statistics of the tests which have been conducted range from 1.760 to 2.517, indicating that the error terms are not correlated according to the guidelines provided by Field (2009). In

3_{The study has been conducted by Nijland for his master thesis: Grip op kosten en baten van Informatie}

(33)

32

addition, the inter correlations between the constructs has also been tested the results of these tests can be found in Appendix B, table B3.

3.4.2. Hypothesis testing. For this study the proposed research model has been tested

by utilizing several ordinary least squares regression analyses. Firstly, the hypothesized association between the maturity of the identification process and the justification process has been tested. Next, the association between the maturity of the realization and exploitation process, and the maturity of the evaluation process will be examined. Based on previous literature, organizational size (measured in FTE), has been added to the models as a control variable for the maturity of the justification process, and for the maturity of the evaluation process. Thereafter, the direct effect of the maturity of each ITG process on firm performance will be examined. Lastly, the effect of industry group membership on firm performance will be investigated. The results of these tests are presented in Table 2.

Table 2: Relationship Maturity ITG Processes and Firm Performance

Maturity Justification process Maturity Evaluation process Firm Performance Independent variables: Model 1 Model 2 Model 3 Model 4 Model 5

Mat. Identification process 0.617** 0.600** 0.124**

Mat. Realization process 0.220* 0.215* 0.009*

Mat. Exploitation process 0.568** 0.583** 0.009**

Mat. Justification process 0.046*

Mat. Evaluation process 0.308**

Firm size 0.130 -0.061

R2 0.380 0.364 0.398 0.382 0.164

F-value 58.25 30.91 31.01 20.76 3.573

Highest VIF 1.000 1.016 1.245 1.306 2.039

Durbin-Watson value 2.517 2.493 1.760 1.772 2.055

Notes: In the first five rows the values presented are the standardized beta’s the coefficients.

** Significant at 0.01; * Significant at 0.05

(34)

33

process decreases to 0.364 percent (see Table 2, model 2). In addition, the results of the regression analysis indicate that firm size is not significantly associated with the maturity of the justification process (p>0.05). As can be seen in model 3 of Table 2, combined the maturity of the realization and exploitation process explain 0.398 percent of the variance in the maturity of the exploitation process. The maturity of the realization process is associated with the maturity of the evaluation process (β = 0.220, p<0.05). Furthermore, the maturity of the exploitation process has also been found to be significantly associated with the maturity of the evaluation process (β = 0.568, p<0.01). Combined these results provide support for hypothesis 5A and 5B that a higher maturity of both the realization and exploitation process is associated with a higher maturity of the evaluation process. In addition, the association between the maturity of the realization and exploitation process, and the maturity process has been tested while adding firm size to the model (see Model 5, Table 2). However, when adding firm size as a control variable the explained variance of the maturity of the evaluation process decreased from 0.398 to 0.382. Furthermore, firm size has not been found to be significantly associated with the maturity of the evaluation process (p>0.05).

(35)

34

determine whether firms in different industry groups perform different, a Kruskal-Wallis test has been carried out4_{. The results of the Kruskal-Wallis test showed that there was no statistical}

difference between firm performance of firms from different industry groups (χ2_{= 2,479, p>}

0,05).

3.4.3. Testing the mediating effect of the justification process. To assess the

hypothesis 2A, which posits that the justification process mediates the effect of the identification process and firm performance is mediated by the governance of the justification process, a Sobel test5 has been performed (Baron & Kenny, 1989; Sobel, 1982). This test examines both the path coefficients the standard errors of the direct paths between the identification process (independent variable), the justification process (mediating variable) and firm performance (dependent variable). Surprisingly, the results of the Sobel test indicate that the association between the identification process and firm performance is not significantly mediated by the justification process (Z = 1.079 p = 0.28) (see Appendix C for calculation of Sobel test). Hypothesis 2A has also been tested by conducting another Sobel test, using the PROCESS macro developed by Preacher and Hayes (2008). This macro conducts a Sobel test after performing a bootstrap approximation. For this study the macro was set to perform a bootstrap approximation of 5000 samples. The results of the Sobel test after bootstrapping confirm that the association between the identification process and firm performance is not significantly mediated by the justification process (Z = 1,043 p = 0.29) (see Appendix C for results Sobel test after bootstrap approximation). Hence, both the results of the Sobel test and the bootstrap approximation suggest that there is no support for hypothesis 2A.

3.4.3. Testing the mediating effects of the evaluation process. The results of a second

Sobel test suggest that the effect between the realization process and firm performance is significantly mediated by the evaluation process (Z = 2,611 p<0.01) (see Appendix D for calculation Sobel test). Thus, supporting hypothesis 6A, that the evaluation process mediates the effect of the realization process on firm performance. A third Sobel test has been performed

4_{Because the dependent variable (firm performance) was not normally distributed in each of the five industry}

groups, the assumptions required to carry out an ANOVA were violated. Consequently, a nonparametric test is more appropriate since a normal distribution is not required to carry out these tests, in this case a Kruskal-Wallis test (Keller & Gaciu, 2012).

5_{There are three principal versions of the Sobel test (MacKinnon et al., 2002). For this research the Aroian version}

(36)

35

to test the hypothesis that the maturity of the evaluation process also mediates the maturity of the exploitation process (see Appendix E for calculation Sobel test). This test also yields significant results (Z = 2.630 p<0.01), supporting hypothesis 6B. Hypothesis 6A and 6B have also been tested by carrying out two Sobel tests using the Preacher and Hayes PROCESS macro, with a bootstrap approximation of 5000 samples. The results of the first Sobel test conducted after a bootstrap approximation suggest that the effect of the realization process on firm performance is mediated by the evaluation process (Z = 2,554 p<0.01). Furthermore, the results of the second Sobel test conducted after a bootstrap approximation suggests that the evaluation process mediates the effect of the exploitation process (Z = 2,519 p<0.01) on firm performance. Taken together, the results of the Sobel tests suggest that the evaluation process mediates both the effects of the realization and exploitation process on firm performance. The results of the Sobel test after bootstrapping for the mediation effect of the evaluation process can be found in Appendix D and E.

To determine whether the evaluation process partially or fully mediates the relation between the realization process and firm performance the steps as proposed by Baron and Kenny (1986) have been carried out. These steps have been repeated to determine whether the evaluation process fully or partially mediates the link between the exploitation process and firm performance. Table 3 shows these four steps for both mediating effects.

The first of these four steps is to carry out step a simple linear regression to determine whether a relation exists between the independent variable and the dependent variable. The results of

Table 3: Regression Models for Mediating Effect Evaluation Process

Step 1: Dependent = f(independent) Step 2: Mediator = f(independent) Step 3 and 4:

Dependent = f(independent and mediator)

Independent variable: Coefficients for the independent Variable Coefficients for the independent Variable

Coefficients for the independent

Variable

Coefficients for the Mediator

Variable

Mat. Realization process 0.201* 0.372* 0.011 0.069*

(0.04) (0.129) (0,149)

Mat. Exploitation process 0.280** 0.621** 0.070 0.335**

(0.078) (0.386) (0.148)

Notes: Figures in brackets are the coefficients of determination (R2_{) for each regression model.}

** Significant at 0.01; * Significant at 0.05

The independent variables are stated in the left-hand column

Information Technology Governance Mechanisms and Processes: The Missing Link Between IT Governance and Firm Performance