SUCCESSFUL INNOVATION: What is required?

(1)

SUCCESSFUL INNOVATION:

What is required?

A case study research in the small software developing industry.

Graduation Thesis

Janette Hagelaars

(2)

SUCCESSFUL INNOVATION:

What is required?

A case study research in the small software developing industry.

J.M.L. Hagelaars

S1710044

January 2012

University of Groningen

Faculty of Economics and Business

MSc Business Administration

Specialization Business Development

First supervisor:

Drs. H.P. van Peet

Second supervisor:

Dr. J.D. van der Bij

(3)

Preface

This thesis has been written on behalf of graduating for the master Business Administration, specialisation Business Development, at the University of Groningen. I had the opportunity to do my research and write this thesis at Innocision, a consultancy company located in Capelle aan den IJssel. During this research at Innocision, I did not only learn about innovation in software developing companies or how to conduct proper research, I also gained a lot of insights on business issues and life in general. I want to thank everyone at Innocision for contributing to the pleasant time I have had at Innocision, it has been a time that I will never forget. I am especially thankful to Bart Weekhout and Rik van Nieuwkerk for giving me the opportunity to graduate at their company. Another special thanks goes out to Maarten van der Hooft, who was my direct supervisor at Innocision, for his help and support.

I also want to take this opportunity to thank all the companies who participated in this research and in particular the companies who contributed to the case studies. Without those companies, doing this research would have been impossible.

Of course I also want to thank my first supervisor from the University of Groningen, Drs. H.P. van Peet. She guided me throughout this research, and with her guidance, critique and support this thesis was raised to a higher level.

Lastly, I want to thank all my close friends and family for their support throughout the whole trajectory. Without them, writing this thesis would have been a whole lot harder.

Janette Hagelaars January 2012.

(4)

Executive Summary

There is a large body of literature dealing with the successfulness of innovation projects. Still, many innovation projects in small software developing companies are not completed successfully. Therefore, this study tries to uncover the still unknown factors influencing innovation project success at software developing companies. The following research question was derived:

“What are the main reasons behind the phenomenon that some innovation projects fail, while other innovation projects are successful, with respect to innovation projects in small companies in the software developing industry?”

In order to answer this question, a case study research was executed at 30 small software developing companies, being customers of Innocision, a consultancy company. The performance of the innovation projects at those small software developing companies was measured along two innovation project performance dimensions: operational performance and product performance. In those two dimensions, four project performance criteria were being distinguished, being: time, costs, quality and revenue.

All small software developing companies of Innocision’s customer base were evaluated alongside those four performance criteria. The best performing companies together with the worst performing companies were selected in order to seek potential differences. As such, six cases were selected for a case study visit. At those visits, in-depth interviews in combination with observation served as data for seeking the differences between the best and the worst performing innovation projects.

A ‘within-case’ analysis in combination with a ‘cross-case’ analysis uncovered the factors which have a direct influence on project success. Those factors are shown in Table 1.

(5)

Table 1 Project success factors

Factors that are found to be positively influencing project success in:

- Three cases Innovation strategy Skilled project manager Project outline Overview of project status Weekly task division

- Two cases Input employees Skilled employees Innovation process Sales perspective Courses Social Media

Factors that are found to be negatively influencing project success in:

- Three cases Tasks loosely divided No courses

- Two cases Little end-user involvement Innovation no priority No project overview

A comparison to literature showed that most of the factors are supported by literature, as literature stresses the importance of proper project management. However, the importance of having an innovation strategy is less validly showing from literature, meaning this factor could be underestimated.

Most factors shown in the table are supported in general by literature on innovation projects. Therefore, it seems that the failure of innovation projects does not entirely lie at the lack of knowledge in literature of the success factors, but more likely, the project managers do not act upon those factors, ignore them or simply do not know which factors influence their innovation project success.

Overall, the study results in the following propositions:

1) Having a project planning, skilled staff and an innovation strategy are the most important factors positively influencing innovation project success in the software developing industry, in particular for small companies.

2) The factors influencing success or failure of innovation projects in small companies are largely similar as those for big companies, such as having a skilled project manager, and proper project planning including project status tracking.

(6)

3) Existing literature describes already most of the factors influencing innovation project success; however, small companies are not acting upon those factors. 4) When measuring innovation project success, the time and cost criteria are

valued as being less important than the product performance measurement criteria by project managers.

(7)

1 Introduction

1.1 Innocision

Innocision is a Dutch consultancy company in the area of technical innovation. The company is located in Capelle aan den IJssel and eight people are currently employed at Innocision. It is a relatively young company, operating for almost two years and actively seeking for growth. This growth will be accomplished by expanding their current core business by attracting additional customers and by seeking new business opportunities in relevant areas.

As for now, Innocision offers various services for companies who seek to innovate. Their most important consultancy activity concerns giving advice on the financing of innovation. The customer base of Innocision consists of a diverse set of companies but most of their customers are small software developing companies. Innocision is able to support them by assisting them in receiving a subsidy from governmental institutions for their technological and process innovations. In order to do so Innocision acts as an intermediary between the government institutions and Innocisions’ customers. Currently, they are mainly focused on one fiscal incentive measure for innovation, being the WBSO (Wet Bevordering Speur- en Ontwikkelingswerk1_{). By means of the WBSO,} the Dutch government compensates (partly) for Research & Development (R&D) labour costs incurred by companies engaging in innovation projects that are technically new to the company and involve a certain risk.

However, Innocision is eager to extent their services by supporting small sized companies throughout their whole innovation process, instead of concentrating on supplying their customers with financial resources.

1_{This is a Dutch regulation.}

(10)

1.2 Research scope

The software developing industry is considered a typical high technology industry. This industry is characterized by innovation-driven market growth, rapidly shrinking product and technology life cycles and high knowledge intensity (Nambisan, 2002). Because of these characteristics, the software industry sector is confronted with high innovation rates, which triggers high spending on research and development (R&D) in this area.

In the European Union, gross expenditure on R&D in the information, communication and technology (ICT) sector reached up to €35.9 billion in 2005 (European Commission, 2009). Nevertheless, not all this money is spent in an effective way, as there are many information technology (IT) projects that are cancelled halfway during the project due to e.g. cost and/or time overruns. Of all IT projects that are being initiated, 5 to 15 percent will be abandoned before delivery, because they are perceived to be inadequate and / or crossing time or cost limits (Charette, 2005). Apart from that, of the IT projects that are completed and deliver actual software, in many cases the software performs below standard and / or below expectations when in use, which results in abandoning the software after delivery. Another reason for software projects to fail is the fact that requirements change during the project or after delivery, which results in either cancellation of the project, before the project has ended, or just after. Taken together, 26 to 34 percent of all IT projects are either delivered to perform unsuccessfully or cancelled before delivery (El Emam and Koru, 2008). In other words, by far not all R&D expenditures that are aimed to improve the IT business through innovation result in actual successful innovation projects. This leads to the question why some innovation projects fail to meet the desired outcomes, while other projects do meet these outcomes and are considered to be successful.

Many researchers (e.g. Cerpa and Verner, 2009; Charette, 2005; Chow and Cao, 2008; Cozijnsen et al., 2000; Cooke-Davies, 2002; De Jong and Marsili, 2006; El Emam and Koru, 2008; Johnson et al, 2001; Jones, 2004; Procaccino et al., 2002; Reel, 1999;

(11)

of failure. However, despite all the research that has been done, apparently still many innovation projects eventually lead to failure. Reasons for this could be the fact that the factors that are found in literature are hard to actually implement, as they are abstract and thus not practible. Another reason could be that the set of factors, which are to be found in literature, that lead to success or failure, is still incomplete and that there are other factors that should be taken into account. However, even if the list of factors influencing the innovation project with respect to successfulness were to be complete, it could be that companies which are undertaking those projects do not have the knowledge of those factors, or skills to implement them, and thus there will not be acted upon.

In this research paper, the focus will be on innovation projects in small companies. Many factors found in literature are actually hard to apply to small companies as many researchers look at fairly large companies when defining those factors. This is a strange phenomenon, as small companies are considered very well suited for innovation, due to their flexible structures and minimal bureaucracy. Furthermore, small companies fulfil an important role in today’s economy, which is emphasized by the fact that many governmental policies designed to stimulate innovation are mainly focused on small companies. What is more, in Europe, the IT sector tends to be dominated by small companies (Laporte et al., 2008). Therefore, it is highly relevant to investigate which factors have an influence on the innovation projects of small companies. This study aims to investigate why some innovation projects succeed, while other projects fail, looking at small firms in the software developing industry. This sector is considered highly innovative with a failure rate of on average 30 percent. For that reason, it is especially interesting for the IT sector to find out what causes such a high failure rate. The central research question of this study will be:

“What are the main reasons behind the phenomenon that some innovation projects fail, while other innovation projects are successful, with respect to innovation projects in small companies in the software developing industry?”

(12)

Answering this question will be of value twofold. First, it will provide Innocision with more insights on the innovation projects of their customers. As factors which distinguish between success and failure will be defined, Innocision will be able to act on that, and support their customers where needed. Furthermore, from an academic point of view, this study will contribute to the existing literature on innovation projects performed in small software developing companies. By means of a case study, factors that have an influence on the success of an innovation project will be uncovered. Those factors will be compared to existing literature in order to see if there is a match or if there are new factors found.

In order to answer the central research question of this thesis, first it should be defined when an innovation project is considered to be successful. Therefore, the following sub- question is formulated:

- When is an innovation project considered to be successful, with respect to innovation projects in small companies in the software developing industry?

1.3 Paper outline

The remainder of this paper is structured as following. First, relevant theoretical concepts will be outlined to give a more clear insight on the research topic. In the end of that section, following from literature, an answer will be proposed to the sub-question. After that in the methodology section the outline of the case study research will be given. Subsequent the collected data will be described and analysed. The outcome of the analysis results into a conclusion, in which an answer will be proposed to the research question. The final part of the paper will deal with the discussion, limitations of the study and insights for future research.

(13)

2. Theoretical concepts

In order to clarify the research question and to give an answer to the sub-question, relevant theoretical concepts will be covered in this chapter. A ‘small software company’ will be defined, as well as ‘innovation’ and ‘innovation project’, resulting in a definition of a successful innovation project, which will be defined in the last paragraph of this chapter.

2.1. Small software developing company

In this study, the focus will be on small software developing companies. In the European Union recommendation, a small company is defined as a company with 10 to 50 employees (The European Commission, 2003). Companies smaller than this, with up to 10 employees, are defined to be micro companies. The IT sector in Europe mainly consists of micro companies, as 85 percent of all companies operational in the IT sector in Europe have 1 to 10 employees (Laporte et al., 2008). This is also true for the customer base of Innocision that mainly consists of micro companies. Since micro companies are such an important part of the software industry, those companies should not be left out when referring to small companies in this study. Therefore, in this study a small company will be a company consisting of 1 to 50 employees. It should be noted however, that the population of this study is largely consisting of micro companies. A software developing company carries out software processes that can be described as: ‘a set of activities, methods, practices and transformations that are used to develop and maintain software and the associated products’ (Raman, 2000: 33). When referring to a software developing company in this study it could be either:

1. a company developing and selling products to the market with no customer-specific modifications or

2. a company developing software based on a specific order, selling customer-specific solutions to their customer.

However, many companies combine those traits, exploiting one of the options as being the dominant one (Cusumano, 2004). In this study the term software developing company will refer to those companies which involve both in software development for the mass market and in developing customer-specific solutions.

(14)

2.2 Innovation

In order for firms to survive and grow in the marketplace, it is of crucial importance for them to engage in innovation. Without innovating, the chances are high that eventually their business will be outdated and out performed as others do innovate. However, what exactly is innovation? Innovation is a widely used term, and different researchers define innovation in a different way. Baregheh et al. (2009) have combined those many different definitions of innovation found in literature and came up with a general and integrative definition of innovation, which encompasses the different perspectives and aspects of innovation and captures its essence.

Their research has lead to the following definition:

‘Innovation is the multi-stage process whereby companies transform ideas into new/improved products, service or processes, in order to advance, compete and differentiate themselves successfully in their marketplace’.

When referring to innovation in this study, it will refer to the above definition of innovation. Innovation has a noteworthy impact on many industries, as market demands change with time and location and if firms do not respond to those changes, they will not be able to sell their products or services in the long term. This is even more so the case in the software industry, as technology-lifecycles are shrinking and getting shorter and shorter. However, responding to changes in the market is not an easy task. Though engaging in innovation is inevitable, many companies do not succeed in doing so. Apart from that, small companies could be hesitant to engage in innovation as innovating brings a lot of uncertainties and risks, and chances of failure are high (Cozijnsen et al., 2000). However, innovative activities do create value for new and established small companies as the benefits from engaging in innovation generally outweigh the costs of doing so (Rosenbusch, 2011).

(15)

2.3 Innovation project

Schwalbe (2007) defines a project as a temporary effort undertaken to create a unique product, service, or result. Innovation projects are a specific type of projects, and combining the definitions of an innovation and a project, a definition for an innovation project arises. An innovation project is: ‘a temporary effort in which through a multi-stage process ideas are transformed into new/improved products, service or processes, in order to create a unique product whereby companies advance, compete and differentiate themselves successfully in their marketplace’. Put differently and following Blindenbach-Driessen et al. (2010), an innovation project is a project ‘in which a new product, service or combination thereof is developed and commercialized for a range of customers’. As for Innocision, all customers of Innocision are executing innovation projects dealing with software development. However, as Innocision’s customer base consists out of small companies often the business of those small companies evolves around one innovation project. Those projects are dealing with developing, improving and upgrading that innovation. Thereby most customers of Innocision are innovating on a continuous base, rather than undertaking a temporary effort.

2.4 Successful innovation projects

Regarding the definition of an innovation, a successful innovation should lead to new, improved and higher quality products, services or processes. When looking at an innovation project however, not only the outcome of the project has an influence on the successfulness of the project, being the new, improved product/service or process, but also the progression of the innovation process itself. A successful innovation project should have both successful operational performance and successful product performance (Tatikonda and Montoya-Weiss, 2001).

2.4.1 Operational performance

An innovation could be a success in terms of having developed a new/improved product that finds market acceptance, but the innovation project could still be a failure if that same innovation was delivered way over the scheduled time and/or the innovation costs were a lot higher than budgeted. Johnson et al. (2001), consider a software application development-project successful when it is completed on time, on

(16)

budget and with all the features and functions originally specified. Thereby they are explicitly looking at the operational performance of a project. Operational performance reflects how the innovation project is executed (Blindenbach-Driessen et al., 2010). In other words, the operational performance measures the success in terms of the internal project. In order to have a successful innovation project, the operational performance should be at least satisfactory or sufficient.

2.4.2 Product performance

Since innovation projects are a specific subset of projects, it is not enough to look only at the operational performance. An innovation project differentiates itself from normal projects by the implementation of the outcome. To define the success of an innovation project, it is important to look at both the operational performance, and the product performance. Product performance evaluates the commercial outcome of an innovation project (Blindenbach-Driessen et al., 2010). An innovation project can only be successful, if the product performance is sufficient.

If both operational performance and product performance are considered sufficient, the innovation project is considered to be successful. However, different people in different roles are likely to have a diverse point of view with respect to the successfulness of a project. A project leader might consider his project successful, based on certain criteria, while the customer might think otherwise. Amongst different stakeholders of the innovation projects, there is no consensus as to what represents ‘project success’ or ‘project failure’ (Ika, 2009). However, in the software development industry a strong agreement has been found between the perspective of the software developer and the end-user with respect to project success. Both parties value the same criteria when it comes to successful project development and a successful final software system (Procaccino and Verner, 2009). Even though, many other stakeholders are likely to value other criteria. Therefore, it is important to identify specific measures to define when an innovation project is considered to be successful, and when it is considered to be a failure. Those measurement criteria should be as objective as possible, so that little

(17)

2.4.3 Measurement criteria

Innovation projects at diverse companies serve different purposes and have different objectives depending on the strategy of that company. Therefore, according to Griffin and Page (1996), the measurement criteria on which an innovation project is evaluated should depend on the firm strategy. However, as this study aims to compare the successfulness of different innovation projects at different companies, one uniform set of success criteria should be distinguished, which can not be dependent on the individual firm strategies. The success criteria will be formed along the two dimensions described in the former paragraph: operational- and product performance. To measure the success of the operational- and product performance of an innovation project, the two concepts as described above should be defined in more detail.

Operational performance

Blindenbach-Driessen et al. (2010) define operational performance success of an innovation project in terms of being; adherence to schedule, adherence to budget, perceived quality and captured knowledge. However, the captured knowledge item does not appear to have a significant influence on operational performance (Blindenbach-Driessen, 2010). Furthermore, three of items, being adherence to schedule, adherence to budget and quality are widely known in literature as being ‘The Iron Triangle’ and those are believed to be the most important indicators of software developing project success (Agarwal and Rathod, 2006; Bernroider and Ivanov, 2011; Huckman et al., 2009; Johnson et al., 2001). Therefore, in this study only those three items will be taken into account when looking at operational performance.

Figure 2.1: The Iron Triangle With respect to the Iron Triangle (Figure 2.1), it is important to find the right balance between those three factors in order achieve success. Some projects require more quality of the deliverable (when demanded by the customer/ end-user), which is likely to result in a longer time span and higher costs. The other way around, a tight budget often means that in terms of quality, a less qualitative product will be arrived at.

(18)

When the right balance is found, which is agreed upon by all stakeholders of the project, chances of successfully completing the project will be higher. Even though this triangle is widely used, there have been critics on this triangle, mainly dealing with the fact that there are far more factors to take into account when defining project success. Ideally, more factors would be taken into account (e.g. satisfies the project team, meets strategic objectives), however for this study that is not feasible. This is because of pragmatic reasons; there is not enough time or information available to study more factors. Furthermore, Innocisions’ customers should not get too disturbed by this study as that might discourage them to participate and as the good relationship between Innocision and their customers should be sustained.

Even so, not only the Iron triangle is taken into account, as for the innovation projects not only operational performance is measured, but as well product performance. Furthermore, as mentioned before, the ‘Iron Triangle’ is perceived to consist out of the most important indicators of software developing project success. Therefore, the ‘Iron Triangle’ is perceived to be adequate for measuring operational project success in this particular study.

The usability of the ‘Iron Triangle’ for measuring project success in the software developing industry is nevertheless not optimal. The reason for this is that in the software developing industry, there is a lack of reliable models to estimate the projects costs and time with precision (Agarwal and Rathod, 2006; Sudhakar et al., 2011). In its essence, many activities performed by small software developing companies are experimental by nature and therefore potential hiccups are often unforeseen. This leads to the fact that it is hard to estimate the time and costs accurately. Relying too much on those criteria would therefore not adequately measure project success. However, they should be taken into account, as greatly exceeding planned project duration or actual costs that are far above approved budgets will not lead to a successful project.

The most important remaining factor for measuring project success on an operational level is quality. When looking at software professionals (programmers/ developers/

(19)

(Agarwal and Rathod, 2006). Furthermore, Blindenbach-Driessen et al. (2010) found that when it comes to an innovation projects, quality is relatively the most important factor for determining operational performance.

So in order to measure if the operational performance of an innovation project was successful, three criteria will be taken into account: costs, time schedule and quality.

Product performance

Apart from operational performance, to complete an innovation project successfully also product performance should be sufficient. However, product performance is influenced by operational performance, as efficient operational performance leads to better market results (Swink et al., 2006). This implies that a part of product performance is thus already measured in operational performance. To avoid measuring this part twice, in this study when measuring innovation project success, product performance is weighted less than operational performance.

Product performance can be measured in terms of the adherence to profit goals, adherence to revenue goals, market share, gained reputation and competitive advantage (Blindenbach-Driessen et al., 2010). In the case of operational performance, three items are measured, as those three together form operational performance. In the case of product performance however, the different items (adherence to profit goals, adherence to revenue goals, market share, gained reputation and competitive advantage) cover the same conceptual domain, which makes the items to a certain extent interchangeable (Blindenbach-Driessen et al., 2010). This fact, together with the pragmatic reasons of time and information shortages leads to the decision that for product performance only one measurement indicator will be used.

This indicator will be the revenue. If sufficient revenue is raised from the innovation, this means that the product satisfies the customers. So measuring product performance through revenues indirectly indicates how well the software solution is accepted by the end-user. This is an important aspect of innovation project success, as an integral element of innovation is the fact that the completed product, service or process should be able to advance, compete and differentiate successfully in the marketplace, thereby

(20)

gaining acceptance from customers. Furthermore, according to Blindenbach-Driessen et al. (2010), revenue is one of the best determinants for determining product performance. If the revenue has reached its targets after the introduction of a software solution resulting from the innovation project, the project will be considered successful with respect to the product performance.

To summarize and give a direct answer to the sub-question; an innovation project in the software developing industry is considered to be successful if both the operational performance and the product performance are considered to be successful. This implies that an innovation project should be considered satisfactory with respect to time, costs, quality and revenue.

(21)

3. Factors influencing innovation project success

The degree to which an innovation project will be successful depends on different factors that have an influence on the success criteria. In project management literature, many factors are found which have an impact on the success of a project. However, much research on the factors influencing new product success is incoherent and is lacking to the point conclusions on which factors should gain the most attention (Montoya-Weiss and Calantone, 1994). Furthermore, not all factors that are found in literature have an impact on every type of company. Some factors only make sense to project management in large companies. For example, a factor positively influencing project performance is the use of cross-functional teams (Love and Roper, 2009). However as small companies often do not have functional departments, the concept of using cross-functional teams with respect to innovation projects in small companies is irrelevant.

What is more, according to Montoya-Weiss and Calantone (1994), preliminary evidence suggests that the set of key determinants of performance may depend on the type of innovation. Therefore, as this study specifically aims at software developing companies, factors which are found in similar studies are the most relevant to discuss. Even so, other factors that are found in studies, which are not exclusively aimed at software developing companies, could be relevant for them as well. Therefore, initially, through an extensive literature study, a broad set of success factors is derived. A set of factors resulting from the literature study will then be further discussed primarily for small software developing companies.

3.1 Success factors for innovation projects

Many articles have been written on factors that have an influence on innovation projects. Table 3.1 provides an overview of the most cited factors found in literature. All factors could potentially have an impact on the success of an innovation project of a small software developing company, so all those factors will be taken into account when conducting the case studies. However, looking at the table, a few factors seem to be more essential than others are, as those factors are cited far more often than others are. Those factors are: customer/end-user involvement, project planning, project team and

(22)

senior management. In the next paragraph, those factors will be elaborated on explicitly for small software developing companies.

(23)

3.2 Success factors for small software developing companies

End-user involvement is one the main factors influencing the success of software development projects (Chow and Cao, 2008; Johnson et al., 2001; Jones, 2004; Nolan, 1999; Procaccino and Verner, 2009; Zwikael and Globerson, 2006). Another cited success factor is the use of project planning (Johnson et al., 2001; Jones, 2004; Procaccino et al., 2002; Zwikael and Globerson, 2006). Other important factors include good personnel recruitment and senior management (Chow and Cao, 2008; Johnson et al., 2001; Procaccino et al., 2002; Procaccino and Verner, 2009; Reel, 1999; Zwikael and Globerson, 2006). Software developing companies with projects that are not successful or fail typically do not perform well at those practices. The impact of the four factors on small companies will be explained in more detail in the upcoming paragraphs.

3.2.1 End-user Involvement

When innovating, interacting with, and orientating on customers has a positive impact on new product success (Gruner and Homburg, 2000; Kujala, 2003; Salomo et al. 2003). Developers and managers share this view as they feel a reasonable level of user involvement during the project, leads to the project being more likely to be successful (Procaccino et al., 2002). When developing software it is therefore important to involve the end-users/customers during the project. Without involving the end-users in any stage of the project, chances are high that the eventually developed product will not become a market success, as the product will likely not fit with the market needs. Involving the end-users and potential customers could prevent this and lead to useful information with respect to the user/customer needs regarding the product. Increasing direct contact with the users ensures that the product requirements are based on relevant information regarding the user’s needs. Furthermore, if adequate knowledge about the user’s needs is missing, misunderstandings about the systems purpose are likely to occur (Kujala et al., 2005), which leads to less qualitative products. Therefore, involving users early on in the project seems to be a good way of improving requirements quality and project success (Kujala et al., 2005). For small companies it is ever so important to involve customers in the innovation project, as a developed product that fails because of market unacceptance is likely to be devastating for the continuance of the company.

(24)

3.2.2 Project planning

In order to execute successful projects, it is important to plan the project well. Proper planning is an essential part of good performance. Project planning is a process of balancing (Muther and Nadler, 2011):

- the product – the purpose or results wanted - the process – the way or procedures to follow - the project – the people, resources and timing.

Among the most common factors of failure of software projects are unrealistic or unarticulated project goals, inaccurate estimates of needed resources and badly defined system requirements (Cerpa and Verner, 2009; Charette, 2005; Chow and Cao, 2008: Johnson et al. 2001). These could be prevented by planning the project accurately. For small companies this is of importance since they often do not have enough resources to cover for unplanned activities. However, proper project planning for small companies engaging in software developing is a tough task, as those projects are often pioneer projects. Often when it comes to developing, problems arise along the way, which have an influence on the planning, and result in a longer time span than planned. Furthermore, software developers are often trying to ‘hit a moving target’, meaning that specifications change during the project which again influences project time and costs. However still it is important to have a project planning, in order to be able to review the project status and to be able to intervene when schedules are likely to escalate.

3.2.3 Personnel Recruitment

The most important asset a software developing company has, are their human resources. Without qualified developers, the development project will never lead to the desired results and quality requirements. Therefore, it is important to recruit capable personnel, with the right knowledge for the project execution. This is obviously the fact for all companies, but even so more for small software companies as their whole business is reliant on the capabilities of their developing staff. When looking at the staff, various literature sources stress the importance of having: sufficient, skilled, experienced and motivated staff members to increase the chances of successfully

(25)

factor influencing the success in a software-developing project is attrition (Reel, 1999). When a qualified software developer leaves a small company, this could have tremendous results. Replacement personnel, who are hard to find in the first place, must quickly get a grip on software that is not complete, not tested and probably not well documented yet. Furthermore, employees left behind will use the person leaving as scapegoat for every problem from then on (Reel 1999).

3.2.4 Senior Management

In small companies, most often only one or two people occupy the management function. That is why in small companies the role of the manager is especially important when engaging in innovation (Marcati et al., 2008).

Leaders in small firms can successfully encourage enthusiasm for innovation in the behaviour of others in their company (Jong and Marsili, 2006). However, in how far the management is encouraging innovative behaviour partly depends on degree of innovative behaviour of the management itself. This degree of innovative behaviour is, amongst other factors, dependent on personality factors. A person’s personality can be classified along five broad dimensions, being: openness, neuroticism, agreeableness, conscientious and extraversion. Those dimensions are generally known as being the ‘Big-Five’ and have become the most widely used and extensively researched factors of personality (Gosling et al., 2003). The five factors all have an indirect influence on the success of a software development project, as they all influence the performance of a project manager, and the project manager has an influence on the success of the project (Wang, 2009). Openness, agreeableness, conscientious and extraversion are all indirectly positively influencing project success, whereas neuroticism has an indirect negative influence on project success. Extraversion is even found to be directly influencing project success (Wang, 2009). This means that indeed the personality characteristics of a project manager will be of influence for the chances of successfully completing an innovation project. This influence is considered higher when it comes to small firms, as direct contact between the company employees is high.

Another often cited factor in literature is management support. For every project is important to have management support as for the project team to gain sufficient

(26)

resources and appreciation. A lack of support is likely to lead to a failing project. In small companies however, management is very likely to support the project, as the project is most often crucial to the core business.

The factors mentioned above all are considered to have an effect on the success of innovation projects in software developing companies. However, as still many projects fail, there is a possibility that there are other, unknown, factors that have an influence on those specific projects. By means of a case study, factors will be uncovered which have had an influence on the success of a certain innovation project. Those factors will be compared to the above factors, in order to find out if indeed those factors are the main/only factors influencing innovation project success in software developing companies, or if there are other factors that are (more) relevant.

(27)

4. Methodology

The aim of this chapter is to describe the procedures and research methods used in order to give an answer to the research question. This research question is: “What are the main reasons behind the phenomenon that some innovation projects fail, while other innovation projects are successful, with respect to innovation projects in small companies in the software developing industry?”

There are several research cycles that can be used to execute a study. This study is outlined following the exploratory research cycle based on Eisenhardt (1989) (Figure 4.1). In this type of research, the research is based on a business phenomenon that is not fully addressed yet in academic research. In other words, the research is based on a gap in the current literature. Based on this literature gap, a research question is formulated, which in this study deals with the factors behind the successfulness of innovation projects in small companies (see above). In order to answer this question case studies are conducted, which serve to study the phenomenon. Using case studies is useful in this type of study as through case studies deeper, hidden, insights on phenomena can be revealed. After the case study observations, the results that are gained will be analysed and compared to existing literature dealing with the same phenomenon. Ideally, some new insights with respect to the topic are revealed. The research concludes by presenting propositions on those new insights, which could serve for future research. These propositions are changes or additions to the existing literature.

Figure 4.1: The exploratory research cycle

In the upcoming paragraphs, a more detailed insight will be given on the execution of the case study projects.

(28)

4.1 Case study design

For the purpose of this study, case studies were conducted. The unit of analysis regarding those case studies is innovation projects. Innovation projects from different companies serve as individual cases. The population of companies where the cases were drawn from is a segment of the customer base of Innocision. Of this customer base, only the software developing companies were taken into account, which fulfil the definition of a small company; a company consisting of 1 to 50 employees. Many of those operate on a business-to-business base. The management of Innocision contacted all their small software developing customers to see if they would be willing to participate in this research. However, the decision to participate could be influenced by the research topic. Since the research is concerning project success, and thereby measuring project performance, good performing customers could be more willing to participate than bad performing customers, and thereby the customers would be self-selecting. To avoid this, in the initial contact between Innocision and their customers, no insights on the research topic were given. Furthermore, all customers were guaranteed anonymity when participating. Resulting from the initial contact, twenty-five customers appeared to be willing to participate in this research. Therefore, the total population of companies from which the case studies were selected consists of twenty-five companies. From this base, the cases were selected, in which at the most one case per company can be used for a case study.

4.1.1 Selection of the cases

Van Aken et al. (2007) define two grounds on which case selection can take place; pragmatic grounds and theoretical grounds. In this study the selection of the cases from the case population is based on theoretical grounds, as cases were selected based on the project success criteria derived from literature. All cases of the population were evaluated and a unique value was assigned to all cases, reflecting the successfulness of the projects. This value is based on scores that were assigned to the measurement criteria: cost, time, quality and revenue.

(29)

The data that was used for assigning the scores to the cases were gathered through telephone interviews. Ideally, the data which is collected is as objective as possible, so that the scores and values can be assigned to the different cases as objective as possible. By means of a pre-test, a concept version of the interview was developed and tested on a random sample of the population (Appendix 1). This in order to find out if the questions were adequate, answerable and if they measured what was intended to be measured. In this concept-interview the questions were mainly focused on attaining quantitative data, in order to gain the answers (data) as objective as possible. In ‘Appendix 2’ can be found how the scores and corresponding values can be calculated resulting from the data gained through the concept interview shown in ‘Appendix 1’. For the pre-test, two companies from the population were contacted and the project managers2_{were interviewed by telephone, in order to test the usefulness of the concept} interview. Only two companies were included in the pre-test, as this is already nearly 10% of the total population and optimally as many companies as possible remain available for the actual interview in case the concept interview turns out not to be adequate. During those couple of interviews, it became evident that none of the interviewees could accurately answer the questions on time, costs and revenue. The reason for this was that they do not have strict time schedules nor do they have strict budgets. Therefore, they do not possess exact figures on those criteria, which left the majority of the questions unanswered. Furthermore, they appeared to have little knowledge on actual figures with respect to revenue. Thereby the concept version of the interview proved be impractical. Consequently, the corresponding computation of the scores and values were also not valid.

As the concept interview was not applicable, the questions were revised in such a way that the respondents are able to answer them satisfactorily. As planned, the data that was used for assigning the scores to the cases were still gathered through telephone interviews. Also still, from each company from the population the project manager was interviewed by telephone, regarding their project performance with respect to time,

2_{As the cases are conducted at small companies, which often evolve around one innovation project, employees}

are likely to fulfil multiple roles. The interviewee, the project manager, is therefore in some cases also the company owner as well as a technical expert/software developer.

(30)

costs, quality and revenue. As the respondents were not able to give an accurate answer to the initial set of questions, the questions were modified in order make sure the respondents were able give satisfactory answers. It was decided to ask the respondents questions that are more subjective. In the new format, the respondents should rate their successfulness based on their perceived performance. In ‘Appendix 3’ the outline of the revised interview is given, along with a short explanation. The scoring on the criteria and the computation of corresponding value was also changed in order to fit with the new interview questions (Appendix 4). As this new set of questions is more subjective, respondent bias is likely to increase. Therefore, in order to minimize respondent bias, apart from the telephone interviews, discussions were held with the consultants of Innocision. In those discussions, all innovation projects were discussed with respect to their performance, i.e. the performance criteria. Based on their expert knowledge on the customers and their innovation projects, the consultants of Innocision reviewed the scores given by the respondents to see if they are likely to represent reality. In case the consultants regarded the scores given by the respondents as untrue or unrealistic, those cases were excluded for the case selection.

With the remaining scores, the values assigned to the projects were computed. The projects with the lowest values were considered to be the least successful projects of the population, the projects with the highest values were considered to be the most successful projects of the population. In Table 4.2 an overview of the scores and the corresponding values that were assigned to the different innovation projects is given.

(31)

(32)

In this study multiple cases were selected, which enables to compare cases and examine cross-case patterns. Both successful and unsuccessful cases were selected to investigate the differences. Often when conducting case studies, the selection of the cases relies on theoretical sampling. The goal of theoretical sampling is to choose cases that are likely to extend the emergent theory, and therefore cases are chosen for theoretical reasons (Van Aken et al., 2007). This is also true for this study, as the cases that were selected represent the extremes that were present in the population; it was not a random sample. The extremes consist of the most successful and least successful innovation projects. By choosing the extremes, the chances are highest that actual differences between the projects will be exposed, and so success factors will be uncovered. Being the extremes, the 10 percent most successful projects and the 10 percent least successful projects were selected in order to make a decent comparison between the two groups. This results in six cases, three best performing innovation projects and three worst performing innovation projects. If multiple innovation projects are assigned a same value, and this value proves to be one of the highest/ one of the lowest values, one of those innovation projects is selected at random.

The selected cases were innovation projects from: Company F, Company K, Company M, Company R, Company V and company W.

4.1.2 Data collection

Once the cases were selected, the data was collected. The data that was collected is data regarding the success factors of an innovation project. For the data collection, field visits to the selected companies, and their innovation projects, were made. Prior to those field trips data was collected through documentation. Documentation is the use of existing documents as a source of information (Van Aken et al., 2007). Documents that were used to attain information were the company websites and the WBSO-application forms3_{. From those data sources, general information about the companies was gained} alongside with more detailed information on the innovation projects. After the documentation phase, the companies of the six cases were visited. All cases were visited

(33)

once, and during those visits, the main data were collected. The on-site data collection was done by means of in-depth interviews and observations.

4.1.2.1 In-depth Interview

For the in-depth interviews, the project leader was interviewed and questions were asked about the innovation project. How did the project evolve? What factors have lead to the success/failure? The interviewer beard the potential success factors for innovation projects mentioned in chapter three in mind during the interview. Might those factors have been an influence? A list of possible interview questions was made regarding the potential success factors. This list is given in ‘Appendix 5’. It should be noted that the questions on the list only served as a guidance for the interviewer and it was not distributed to the interviewees. It does not represent the outline of the actual interview. At the end of the in-depth interview, the interviewee was asked to fill out a small survey with respect to their personalities. This to find out about which different personality traits the interviewees possessed, to see if this might be a factor influencing the innovation project. In ‘Appendix 6’, the results of this survey are presented, alongside with a short explanation. If possible, other members of the staff were also interviewed (shortly) in order to get information as unbiased as possible and to attain information on the innovation project from different viewpoints. For this study, only one researcher was available, in order not to miss out on any information given by the project manager, the interview was recorded with a voice recorder in compliance with the project leader. In this way, the researcher could fully focus on the interview, without being distracted by taking too much notes.

4.1.2.2 Observation

Next to documentation and in-depth interviews, the researcher collected data through observation. Informal, spontaneous communication has been shown to be crucial to the performance in innovative projects (Brenner 2007). This informal communication can be stimulated by the way the workplace is designed. A workspace that is supportive for internal communication is thereby indirectly supportive for innovation. Therefore, the workspace of the companies visited is observed. A workspace supportive for informal communication is considered to be a workspace in which all members of the company are co-located (Collaborative Work Environment (CWE)). Furthermore, working

(34)

together in the same room also triggers communication between team members, as well as ‘meeting points’ like a coffee machine. That is why the workspace at the cases is observed to see how supportive it is for informal communication. The researcher took direct field notes on the observations made.

4.1.3 Data analysis

The data were analysed through two methods. First, each case was analysed individually through a ‘within case’ analysis. A ‘within case’ analysis involves detailed case study write-ups, which are often pure descriptions. These descriptions are essential to the generation of insight as they help to cope with the large amount of data collected (Eisenhardt, 1989). The data that was collected on site at a case were integrated into one file concerning that particular case. This file was translated into a case description, which was based on the interview. The case description was then examined thoroughly, focussing on the individual success/fail factors of that case. This was done by means of inductive reasoning; following from the data, factors that are probable to have an influence on the successfulness of the case are summarized. This process was done for each case individually. Focussing on each case individually allows for unique patterns of each case to emerge before heading for a generalized pattern across the different cases (Eisenhardt, 1989).

After the ‘within case’ analysis a ‘cross case’ analysis was conducted. A ‘cross case’ analysis allows data to be viewed in divergent ways, from different viewpoints. In order to find ‘cross case’ patterns, firstly the most successful cases were compared to each other, by pairing the cases. As such, three comparisons were made.4_{In those} comparisons, the potential success factors found in the within-case analyses are compared, to see what the similarities or differences are between the cases (Eisenhardt, 1989 5_{). Afterwards the least successful cases were compared to each other in the same} way. Finally, the factors of the most successful cases were compared to the factors derived from the least successful cases in order to seek potential differences. By executing a ‘cross case’ search for patterns, the probability of capturing novel findings,

(35)

analysis were compared to the literature described in chapter three. This gives insights to what degree the success factors found in this research are supported by literature sources and to what degree the success factors resulting from this study are unique.

4.2 Quality criteria

In order to evaluate the quality of a research, quality criteria have been defined. The most important research-oriented quality criteria are controllability, reliability and validity (Swanborn, 1996; Yin, 1994). These criteria are important because they provide the basis for inter-subjective agreement on research results (Habermas, 1981; Swanborn, 1996). In the upcoming paragraphs, each quality criterion will be discussed in more detail in order to elaborate on the quality of this research.

4.2.1 Controllability

A research should be controllable, so that other researchers could replicate the study to see if the same outcomes are obtained. To make a research controllable, it has to be revealed how it was executed (Van Aken et al., 2007). In this study, this methodology section makes the controllability high as it gives detailed insights on how the study was executed, and therefore other researchers could replicate the study. Furthermore, in the appendices, the concepts and data collection methods are even further explained, increasing the controllability.

4.2.2 Reliability

The second quality criterion that is discussed is reliability. The results of a study are considered reliable when they are interdependent of the particular characteristics of that study (Yin, 1994; Swanborn, 1996). Four potential sources; the researcher, the respondent, the instrument and the situation (Van Aken et al., 2007), can bias a study. Research results are more reliable when they are independent of the person who has conducted the study (Van Aken et al., 2007:159). Researchers can be biased due to the influence of interests, motivations and emotions and results can be influenced by a confirmation bias. In order to avoid researcher bias, ideally multiple researchers should execute the study. However as this research is an academic research, the research should and has been executed by one person, decreasing the researcher reliability. Still

(36)

researcher reliability can be considered relatively high, as the researcher is completely objective towards the research, and does not in any way has any interests or motivations to deliberately manipulate the outcome.

Respondent reliability is not optimal in this study. Research results are more reliable when, within a case, respondents from different roles are taken into account. This is to decrease the subjectivity of the data. In this study, only the project leaders of the innovation projects were interviewed for the in-depth interviews. No other respondents have been interviewed due to time restrictions. However, as noted before, as the study concerns small software developing companies, one person is likely to fulfil different roles in their project. Moreover the outcome of this study has no direct influence on the respondents, which makes them more likely to give truthful answers and thereby be reliable.

Instrument reliability is enlarged in this study by using the concept of triangulation. Triangulation is the use of multiple instruments to conduct the research, and combine the results that are gained (van Aken et al., 2007). In this research, the multiple research instruments that are used are; a structured interview for selection of the cases, documentation and a combination of an in-depth interview with observation during the company visits.

Reliability with respect to the situation is enlarged by conducting the in-depth interviews at the cases at different moments in time on different days, in order to gain results independent of the circumstances.

In general, the reliability of this study is further enlarged by adding a case study protocol, which is outlined at the beginning of this chapter.

4.2.3 Validity

The last quality criterion that is discussed is validity. A research result is valid if they way that it is generated provides good reasons to believe that the research result is true or adequate (Van Aken et al., 2007). Van Aken et al. (2007) categorize validity into three

(37)

Construct validity refers to the extent to which a measuring instrument measures what it is intended to measure. Thereby, a concept should be covered completely and the measurement should not have components that do not fit the meaning of the concept (Van Aken et al., 2007). Construct validity is in this study assured by basing the meaning of the used concepts on a thorough literature study. However, construct validity would be higher if more success criteria would have been measured which in this study was impossible due to limitations described in paragraph 2.4.3.

The results of a study are internally valid when conclusions about relationships are justified and complete (Van Aken et al., 2007: 164). In this study internal validity is strengthened in two ways. Firstly by doing both ‘within case’ and ‘cross case’ analyses, which enforces the researcher to look at the data from different viewpoints before drawing conclusions. Furthermore, the results from the case analyses have been compared to existing matching and conflicting literature, which is generally known as pattern matching. The pattern matching carried out in this study further strengthens the internal validity of this study.

If the research results and conclusions are generalizable to other people, organizations, countries and situations the research scores high on external validity (Van Aken et al., 2007). A factor positively influencing the external validity of this study is the fact that multiple case studies are conducted. Additionally, the cases that are conducted are generalizable to a wider population as they are highly similar: they are all small software developing companies, which makes the results applicable to a wider population of small software developing companies. Furthermore, the results are compared with available literature, which sharpens generalizability (Eisenhardt, 1989).

(38)

5. Data analysis

In this chapter, the data that were gathered during the field trips are analysed. First, every case will be analysed individually by means of a ‘within case’ analysis. Secondly the cases will be analysed through a ‘cross case’ comparison. All information that is presented in the case descriptions is gained through the interview with the project manager of each respective company.

5.1 Within case analysis

5.1.1 Case 1: Company F

5.1.1.1 Case description Company F

Looking at the scores assigned, company F turns out to be the most successful company of the companies investigated with respect to innovation project performance. Unlike most other companies,

company F manages to perform very well with respect to delivering the project within the expected time schedules. The innovation project was executed much faster than expected. Also, compamy F obtained a high score on both product quality and revenue, which makes their innovation project the most successful project of the projects investigated in this study (Table 5.1).

Table 5.1: Scores on success criteria company F

Company F sees innovation at company F is seen as a pure necessity. For them innovation is extremely important, as without innovating they see no way to survive in the competitive software developing business. Company F has a clear goal; sell as much

(39)

room for developing for customer demand too, but they only carry out this work when it does not interfere with their innovation projects and their innovation road map. Innovation ideas in company F come solely from the project manager’s ideas. Those emerge out of prior customer demand, trends in the market, problems ventilated in the market, social media and forums. Furthermore, the project manager is highly engaged in connecting to other (business) people, visiting other companies and being amongst other people talking around, which gives him new insights and input in the generating of innovation ideas. When starting an innovation project, the sales opportunities are always taken into account first; to whom, how much and how can the product be sold? Company F has no strict project schedule for innovation. How fast the developers need to work, depends on the market demand; the urgency to finish the innovation quickly. To gauge that, the owner of company F always puts a concept version of his innovative idea online, which makes it seem that the product is already available. Based on demand resulting from this, the urgency to finish the project is altered. Every week the project manager contacts each of his employees, discussing and reviewing the weekly project planning. Those discussions are always one on one, since the developers all work individually at home, having individual tasks. Some developers are only working on innovation projects, others only on customer-specific projects. As all developers of company F work at home, they can determine their own time schedules based on how much work needs to be done. Furthermore, they are all able to work when they prefer to and they do not have to wait for inspiration at ordinary working hours. The developers are so called ‘super-nerds’ and enjoy working late in the evenings, or even at night. They are highly skilled which results in quality software. However, they are not socially skilled and rarely leave their home, which is why the project manager does not consult them (anymore) for the generation of innovative ideas; they do not spend enough time in the outside world. The project manager on the other hand, has in depth knowledge on the market as well as a excellent technical background, which makes him able to understand what the developers are doing, what is needed, and where he has to intervene if necessary.

SUCCESSFUL INNOVATION: What is required?

SUCCESSFUL INNOVATION:

What is required?

A case study research in the small software developing industry.

Graduation Thesis

Janette Hagelaars

SUCCESSFUL INNOVATION:

What is required?

A case study research in the small software developing industry.

J.M.L. Hagelaars

S1710044

January 2012

University of Groningen

Faculty of Economics and Business

MSc Business Administration

Specialization Business Development

First supervisor:

Drs. H.P. van Peet

Second supervisor:

Dr. J.D. van der Bij

Preface

Executive Summary

Table of Contents

1 Introduction

2. Theoretical concepts

3. Factors influencing innovation project success

4. Methodology

5. Data analysis