managing the fuzzy front end of product and process

DOCTORA L T H E S I S

Department of Business Administration and Social Sciences

Managing the Fuzzy Front End of Product and Process Development

Case Studies of Process Firms

Monika Kurkkio

ISSN: 1402-1544 ISBN 978-91-7439-195-4

Luleå University of Technology 2010

Monika K

urkkio Managing the Fuzzy Front E

nd of Product and Process Developm

ent Case Studies of Process Firm

s

ISSN: 1402-1544 ISBN 978-91-7439-XXX-X Se i listan och fyll i siffror där kryssen är

Managing the Fuzzy Front End of Product and Process Development: Case

Studies of Process Firms

Monika Kurkkio

Printed by Universitetstryckeriet, Luleå 2010

ISSN: 1402-1544 ISBN 978-91-7439-195-4

Luleå 2010

www.ltu.se

To Carl-Johan and Milton

ABSTRACT This thesis addresses different aspects of how to manage product development and process development. Two research studies have been conducted, both within the process industry. In the first study, the focus was on management control in collaborative product development and process development projects, whereas the focus in the second study was on managing the fuzzy front end (FFE) of product development and process development. Content-wise, the thesis comprises an introductory text and five appended papers with the overall purpose to increase our understanding of how process firms can improve the management of product development and process development, with a special emphasis on the fuzzy front end.

The background to the first study was that it is difficult to manage collaborative development projects and that few empirical studies address accounting and control issues in such projects. However, the main focus in this thesis is on the FFE of the innovation process (the second study). The FFE is acknowledged as being one of the most critical stages of the whole innovation process, because the quality of emerging product- or process concepts is determined there. For example, improved product concepts can increase the overall speed of the innovation process as well as cutting costs. The costs of imposing changes in the FFE are still relatively low compared to changes in later stages. However, the FFE is difficult to manage because it implies a significant degree of uncertainty, and it is often characterised by ad-hoc decision-making and conflicting organizational pressures which make it difficult to plan and conduct. The existing research on the FFE has primarily focused on product development, but little is known about how to manage the FFE in other contexts. This thesis aims to shed light on the FFE in non-assembled product development and process development, mainly by means of multiple case studies.

The first study of this thesis (paper I) shows that both informal and formal control mechanisms are important when firms conduct collaborative product development and process development projects. A central contribution in the first study is that it explores how the firms’ motives affect the selection of formal and informal control mechanisms. For example, financial short-term motives (such as sharing development costs) were typically controlled by formal control mechanisms. On the other hand, informal control mechanisms were the main form of control if the motive was to build long-term relations with their customers.

The main findings from the second study (papers II-V) were that the FFE phase of non-assembled product development and process development was conceptualized. These conceptualizations were done at a relatively detailed level and key activities in different sub-phases were thus identified. By conceptualizing the FFE of non-assembled product development and process development, a first and important step towards exploring how the FFE can be better managed was taken. In addition, the second study identified several managerial problems in the FFE, and the negative impact of many of these could arguably be decreased if more formal control mechanisms were introduced and applied. The empirical findings suggest a more formalized process for managing the FFE of non-assembled product development and process development. These findings are important in the light of a growing theoretical and managerial interest in the FFE phase of the innovation process.

ACKNOWLEDGEMENTS Writing this thesis has been an exciting and challenging adventure! I could not have done it without the support and guidance of key persons nearby.

First, I truly want to acknowledge the input from my supervisor Johan Frishammar at Luleå University of Technology (LTU), with whom I have conducted much of my research. You have guided me in the difficult art of designing and conducting scientific research as well as writing articles. I want to express my deepest gratitude for your professional guidance and the possibility to work close together with you. Thank you for inspiring, encouraging and pushing me forward! Henrik Florén at Halmstad University and Jeaneth Johansson at LTU have been my co-supervisors. Thank you, Henrik, for all interesting and reflective discussions! Thank you also for providing me with constructive comments which have improved the quality of my thesis. Jeaneth, thank you for your support, the insightful comments on my manuscript, and your valuable advice!

Several of the papers in this thesis are co-authored and I also want to take the opportunity to thank my co-authors. I believe that good research is often the product of collaboration and I am grateful for the opportunity to work together with Ulrich Lichtenthaler at the University of Mannheim, Germany, and Lena Abrahamson at the division of Industrial Work Environment at LTU. By working closely with you I have learned many things about how research is crafted – things that would have remained tacit to me without your input.

I am also grateful for the financial support from VINNOVA’s LEKA-programme and for the financial support from the participating companies Boliden AB, Höganäs AB, LKAB and SSAB, which made this research possible. Special thanks to all the employees at the companies who participated in the study for their time and input. I also thank Nordea for their financial support that enabled my visit to Copenhagen Business School (CBS). While at the CBS, Jan Mouritsen and John Christiansen provided valuable comments and advice on the early draft of this thesis.

Further, I would like to thank all my colleagues and friends at Accounting and Control, Entrepreneurship, and Industrial Management divisions for making this a fun and interesting workplace! I am also grateful to the research team at Promote for interesting discussions concerning research on process firms. Thanks also to Nicolette Lakemond at Linköping University for acting as “pie-opponent” at my final seminar; I deeply appreciate your feedback. To my mentor, Sara Thorgren, I would like to send special thanks. You have supported me in many ways during my time as a PhD student. Other persons who have helped me in improving my research are the journal editors and the anonymous reviewers. I have learned a lot from addressing your comments!

Finally, the support I get from my loving family and friends have essentially improved my work. Carl-Johan and Milton, sharing my life with you has taught me the true meaning of endless joy, and you are the best thing that have ever happened to me! To all of my family and friends, thanks for believing in me and for being such lovely people as you are! Monika Kurkkio Luleå in December, 2010

LIST OF APPENDED PAPERS IN PART II

Paper I Kurkkio, M. (2009). Control Mechanisms in Collaborative R&D Projects:

The case of supplier and customer relationships in process industry. International Journal of Technology Intelligence and Planning, Vol. 5, No. 1, pp. 73-89.

Paper II Kurkkio, M., Frishammar, J. and Abrahamsson, L. (2010). Antecedents and

Consequences of Firm’s Process Innovation Capability: A literature review and conceptual framework. In review with IEEE transactions on engineering management.

Paper III Frishammar, J., Lichtenthaler, U. and Kurkkio, M. (2010). The Fuzzy

Front End in Non-assembled Product Development: A multiple case study of mineral- and metal firms. In review with Journal of Engineering and Technology Management.

Paper IV Kurkkio, M., Frishammar, J. and Lichtenthaler, U. (2010). Where Process

Development Begins: A multiple case study of fuzzy front-end activities in process firms. In review with Technovation (revise and re-submit)

Paper V Kurkkio, M. (2010). Managing the fuzzy front-end: Insights from process

firms. Accepted for publication in European Journal of Innovation Management.

TABLE OF CONTENT FOR PART I

1. INTRODUCTION.............................................................................................. 1

1.1 MANAGING INNOVATION............................................................................................................. 11.2 THE RESEARCH PROBLEMS AND THE EMPIRICAL CONTEXT OF THE THESIS ......................................... 31.3 THE OVERALL PURPOSE AND RESEARCH QUESTIONS OF THE THESIS .................................................. 61.4 COMMENTS ON THE TERMINOLOGY USED IN THE THESIS ................................................................. 71.5 THE OUTLINE OF THE THESIS....................................................................................................... 8

2. FRAME OF REFERENCE: RESEARCH ON MANAGING PRODUCT AND PROCESS DEVELOPMENT..................................................................................11

2.1 PRODUCT AND PROCESS DEVELOPMENT.......................................................................................112.2 THE FUZZY FRONT END OF THE INNOVATION PROCESS ..................................................................142.3. MANAGEMENT CONTROL IN THE CONTEXT OF PRODUCT AND PROCESS DEVELOPMENT.....................18

3. THE PROCESS INDUSTRY AS A CONTEXT FOR PRODUCT AND PROCESS DEVELOPMENT..................................................................................................23

3.1. THE SPECIFIC CHARACTERISTICS OF PROCESS FIRMS .....................................................................233.2. PRODUCT AND PROCESS DEVELOPMENT IN PROCESS FIRMS............................................................25

4. METHODOLOGY ............................................................................................29

4.1 SCIENTIFIC POSITIONING............................................................................................................294.2 RESEARCH DESIGN ....................................................................................................................29

4.2.1 Research project 1: Management Control in Collaborative Product and Process Development Projects...................... 304.2.2 Research project 2: Managing the Fuzzy Front End of Product and Process Development................................... 30

4.3 THE CASE STUDY APPROACH ......................................................................................................314.4 DATA COLLECTION AND DATA ANALYSIS ......................................................................................324.5 SOME REFLECTIONS ON THE FRAME OF REFERENCE .......................................................................394.6 ASSESSING THE QUALITY OF THE RESEARCH .................................................................................394.7 THE RESEARCH PROCESS AND THE BACKGROUND TO THE APPENDED PAPERS ...................................41

5. SUMMARY OF THE APPENDED PAPERS........................................................43

5.1 PAPER I: CONTROL MECHANISMS IN COLLABORATIVE R&D PROJECTS: THE CASE OF SUPPLIER AND

CUSTOMER RELATIONSHIPS IN PROCESS INDUSTRY ...........................................................435.2 PAPER II: ANTECEDENTS AND CONSEQUENCES OF FIRMS’ PROCESS INNOVATION CAPABILITY: A

LITERATURE REVIEW AND CONCEPTUAL FRAMEWORK .....................................................445.3 PAPER III: THE FUZZY FRONT END IN NON-ASSEMBLED PRODUCT DEVELOPMENT: A MULTIPLE CASE

STUDY OF MINERAL AND METAL FIRMS ..........................................................................455.4 PAPER IV: WHERE PROCESS DEVELOPMENT BEGINS: A MULTIPLE CASE STUDY OF FUZZY FRONT END

ACTIVITIES IN PROCESS FIRMS ........................................................................................465.5 PAPER V: MANAGING THE FUZZY FRONT END: INSIGHTS FROM PROCESS FIRMS ...............................47

6. DISCUSSION....................................................................................................49

6.1. THEORETICAL IMPLICATIONS.....................................................................................................496.2 MANAGERIAL IMPLICATIONS .......................................................................................................526.3. CONCLUSIONS .........................................................................................................................546.4. OVERALL LIMITATIONS .............................................................................................................546.5. SUGGESTIONS FOR FUTURE RESEARCH ........................................................................................55

7. REFERENCES ..................................................................................................57

APPENDIX A: Interview questions on managing collaborative product development ................. 63 APPENDIX B: Interview questions on managing collaborative process development .................. 64 APPENDIX C: Interview questions on managing and organizing the fuzzy front end in product

development ........................................................................................................................ 65 APPENDIX D: Interview questions on managing and organizing the fuzzy front end in process

development ........................................................................................................................ 67

1. INTRODUCTION

This thesis deals with how process firms manage two types of innovations, product and process development. It specifically considers the fuzzy front end (FFE) of the innovation process. In this introductory chapter a background presents the importance of managing innovation and it is followed by a description of the research problem, overall purpose, and the research questions. Finally some comments on the terminology used are provided and the structure of the thesis is outlined.

1.1 Managing innovation

Innovation implies change; either firms take advantage of new opportunities, or they are replaced by other firms who do (e.g. Cooper and Kleinschmidt, 1987; Tushman and Nadler, 1986; Utterback, 1994). The importance of innovation is clear, but how to achieve it is the critical question (Drucker, 1998). Innovation is often portrayed as the process of creating a new product (which can be either a physical product or the service a firm provides) or a change in the way a product is produced (Utterback and Abernathy, 1975; Tushman and Nadler, 1986). The innovation process entails a variety of activities, from idea generation to manufacturing and marketing of a new or improved product or production process, and all these activities need to be managed somehow (Trott, 2005). How do firms manage their innovation processes? In the literature there are many different approaches to how firms can and should manage their innovation processes. In one stream, a structured management approach is argued for, because it is assumed that the innovation processes becomes more efficient. The main argument within this stream is that activities, behaviour and outcomes in the innovation process need to be coordinated and controlled in a structured way (e.g. Cooper; 2008; Cooper and Kleinschmidt, 1987; Nixon, 1998). Maximizing existing resources to achieve cost reductions, lower risks and shortened development cycles by increasing speed and quality in the innovation process are all desirable outcomes of an efficient innovation process (Cooper, 2008; Nixon, 1998). The tactics for increasing speed are planning, coordinating, creating overlap with multifunctional teams, and using reward systems, among other things (Cooper, 2008; Eisenhardt and Tabrizi, 1995). However, other researchers have highlighted the risk that the management becomes too detailed, as it can lead to rigidity and hamper creativity (e.g. Amabile, 1998; Eisenhardt and Tabrizi, 1995). As a consequence, the freedom of engineers is restricted and the overall team performance will be decreased (Bonner et al., 2002; Davila, 2000).

1

A second stream of research criticises the structured approach presented in the first stream (Amabile, 1998; Eisenhardt and Tabrizi, 1995; Jönsson, 2004). Instead, the second stream argues that the innovation process is creative and uncertain, and that imposing structure is therefore not the most appropriate approach. Disciplined problem-solving by the project team, support from top management, and heavyweight leaders are argued for as the approach to increased speed and quality in the innovation process (Brown and Eisenhardt, 1995). The tactics for increasing speed in the innovation process within this stream are improvisation, multiple iterations and a powerful leader (Eisenhardt and Tabrizi, 1995). The attention should be directed towards creating an environment that informally nurtures new ideas and the growth of ideas, as well as quickly building understanding of different options by relying on improvisation and experience (Eisenhardt and Tabrizi, 1995; Griffiths-Hemans and Grover, 2006). These two streams of research represent two different standpoints of how firms should manage their innovation processes, and consequently they provide contradictory findings on what tactics to use to increase speed and quality in the innovation process. However, they indicate that managing innovation involves a trade-off between exercising control and at the same time granting conditions for creativity. A third stream of research has noted the need to balance control and creativity when managing innovations (Bonner et al., 2002; Chiesa et al., 2009; Poskela and Marinsuo, 2009; Simons, 1995). The main argument within this stream is that firms need to simultaneously provide conditions for creativity so that new ideas can flourish, and exercise control to ensure that resources are used in an effective and efficient way. Within this stream it is suggested that individuals should be given high degrees of freedom and interdependence which hopefully will create the required conditions for creativity, and that the management team should create boundaries which will direct the employees to focus on strategically important areas (Simons, 1995). The key question is: what are the tactics that balance control and creativity? Are they a mix of the tactics presented in the first and second stream of research, or does this approach require other tactics? The question of how firms manage their innovation process seems to vary according to the type, degree and stage of the innovation process (Tidd, 2001; Trott, 2005; Chiesa et al., 2009). The focus in the innovation management literature has long been on product development, and this research has made a substantial contribution to understanding the overall innovation process (e.g. Brown and Eisenhardt, 1995; Cooper, 2008). However, other innovation types such as process development and service development have received little attention compared to product development (Goffin and Mitchell, 2005; Pisano, 1997; Reichstein and Salter, 2006). The relative

2

neglect of process development in the literature is surprising as both innovation types are important to achieve long-term competitive advantage in manufacturing firms (Pisano, 1997). Previous research has shown that degree of newness influences how the innovation can and should be managed (e.g. Chiesa et al., 2009; Song and Montoya-Weiss, 1998). For example, incremental and radical innovations differ in how problems are structured and in which information searches are initiated (Reid and de Bretani, 2004). Furthermore, it is found that different stages within the innovation process should be managed differently (e.g. Chiesa et al., 2009; Davila, 2000; Poskela and Martinsuo, 2009). The uncertainty is typically highest in the early stages of the innovation process, often referred to as the fuzzy front end (FFE), and the main task of management is to reduce uncertainty and bring order to the chaos, whereas in the later stages a more structured approach is suitable when the uncertainty is lower (Chiesa et al., 2009).

1.2 The research problems and the empirical context of the thesis

Firms developing and producing non-assembled1 products, i.e. process firms, are the empirical context in this thesis. These firms drive economic development not only through their own operations, but also in industries that are positioned forward in the supply chain. For example, innovations in the products produced by process firms enable innovations in the automotive and construction industries. The process industry is thus an important driver for products positioned in later stages of the supply chain. The process industry stands for a considerable portion of GDP in many countries, which also makes it an important context (van Donk and Fransoo, 2006). Previous studies have shown that process firms differ in many respects from other manufacturing firms (e.g. Barnett and Clark, 1996; Hutcheson et al., 1995; Lager et al., 2010; Lim et al., 2006). For example, the product life cycles are generally very long for non-assembled products, and the demand and pricing for these products are different than for e.g. many assembled products targeted at the consumer market (Sundgren, 2004; Lager et al., 2010). These are some examples of why it is difficult to apply models that are based on assembled product development without adjusting them to the specific context. In Chapter 3 the characteristics of process firms are discussed in more detail.

1 Non-assembled products are composed of one or a few materials (e.g. chemicals, metals, minerals, pulp and paper, and textiles), in contrast to assembled products which are put together from several different components, e.g. computers, cars, furniture etc. The market for assembled products is often consumers, whereas a large part of the non-assembled products are industrial products developed and applied in other firms.

3

This thesis aims to shed light on two types of innovations in the process industry: process development and product development. Two research projects have been conducted to explore product and process development in process firms. In the first research project, the focus is on collaborative development projects, whereas in the second the FFE of the innovation process is the focus of attention. External collaboration in product development has become a way to overcome problems with high investment costs and insufficient resources. These are common reasons to engage in collaborative development projects, but firms can have many other motives as well. Sharing costs, pooling resources, and acquiring resources and skills that otherwise would not be obtainable (Boddy et al., 2000; Perks, 2000) are other benefits of collaborative development projects. However, the flip side is that many of these projects fail. They are generally not legal entities, which makes them difficult to manage, as they involve members from different firms with diverse objectives, values and beliefs –another reason which makes it difficult to manage these projects (e.g. Boddy et al., 2000; Kamminga and Van der Meer-Kooistra, 2007; Trott, 2005). To illustrate, a project leader in the first case study said that one of the major difficulties when managing collaborative development projects was to satisfy all participants. All firms had different objectives with the joint project and it was difficult to ensure that all were content with the objectives. Most difficult was to find the appropriate amount of control that simultaneously allowed financial outcomes as well as building long-term relationships, since these could sometimes be contradictory. The literature on collaborative development is extensive and it outlines both the benefits and the risks, but only a handful of studies have addressed accounting and control issues in these projects (e.g. Dekker, 2004; Kamminga and Van der Meer-Kooistra, 2007). Such studies are important as they are one means to decrease risks and high failure rate of collaborative development projects. More research is needed that explores what control issues are important in managing collaborative development projects. The main focus in this thesis, though, is on how firms can manage the earliest stages of the innovation process, which are explored in the second research project. Previous research has found that although most development projects fail at the end of the development process or during the commercialization, the foundation for successful development is often established in the beginning (e.g. Cooper and Kleinschmidt, 1987; Cooper, 1988). Therefore, firms need to be proficient in managing and organizing the FFE.

4

Nonetheless, previous research shows that despite the importance of the FFE it often receives the least management attention and resources compared to later stages in the development process (Cooper, 1988; Khurana and Rosenthal, 1998). Many firms find it difficult to manage the FFE as it typically involves high levels of uncertainty and equivocality (Chang et al., 2007; Griffith-Hemans and Grover, 2006; Khurana and Rosenthal, 1998). The FFE has been described as a crossroads of complex information processing, tacit knowledge, and conflicting organizational pressures (Khurana and Rosenthal, 1997) which make it difficult to manage the FFE in a structured and systematic manner. From a practical perspective, the case study firms expressed a couple of reasons why they found the FFE difficult to manage. Most of the current ideas for new products and processes were incremental and focused on solving existing problems, rather than exploring radically new ways. Thus, there is a risk that the firms get stuck developing incremental innovations and eventually they will be out-competed. Another difficulty the firms experienced was how to collect and screen ideas. To save money and time it is essential to get rid of bad ideas quickly, but the screening should not be so harsh that it kills good ideas too early. A literature review of the FFE (presented in section 2.2) showed that most research on the FFE focuses on assembled products and research on other innovations is limited. The literature on the front end seems to assume that all innovation types are similar, although in other literature it is shown that different innovation types should be managed differently (e.g. Tidd, 2001; Utterback, 1994). This suggests that the FFE differs depending on the type of innovation that is being created. Important insights have been made on how firms can manage and organize the FFE in assembled product development (e.g. Cooper, 1988; Murphy and Kumar, 1997; Seidel, 2007) and in service development (Alam, 2006) but more research is needed which explores other innovation types. In this thesis the FFE of non-assembled product development and process development is explored. First, the FFE in both contexts is conceptualized to see what constitutes the FFE in process development and non-assembled product development. Several conceptualizations of the FFE exist in the literature, but all of them address assembled product development (e.g. Cooper, 2008; Griffiths-Hemans and Grover, 2006; Khurana and Rosenthal, 1997; 1998; Koen et al., 2001). Conceptualizing the FFE in non-assembled product development and process development is a first important step to find out what constitutes the FFE and how firms can manage this early stage. What are the key activities and the managerial challenges in the FFE of

5

non-assembled product development and process development, and how they can be mitigated, are further explored in this thesis. Managing the FFE is very important for firms that strive to be competitive. Increased knowledge of how firms can manage the FFE is important for several reasons. First, the cost of making changes in the production is still relatively low compared to making changes in the later stages of the innovation process (Smith and Reinertsen, 1998). Second, improved product or process concepts can also increase the speed of the overall development project as well as cutting development costs (Bacon et al., 1994). Thus, the FFE holds great opportunities for firms that strive to make the innovation process more efficient and effective.

1.3 The overall purpose and research questions of the thesis

The thesis consists of two research projects, and its overall purpose is to increase the understanding of how process firms can improve the management of product and process development, with a special emphasis on the fuzzy front end phase. To fulfil and further specify the overall purpose, four research questions have been formulated. The first and second research questions address the challenges of managing the whole innovation process, when creating and developing new products and processes. In the second question the focus is on process innovation, rather than process development. In the third and fourth research questions the focus is on the FFE of product and process development. Research question 1: What control mechanisms are important to consider when managing collaborative product and process development in process firms? (Addressed in paper I) Research question 2: What are the critical factors for managing process innovation? (Addressed in paper II) Research question 3: How can the FFE of product and process development in process firms be conceptualized? (Addressed in papers III, IV and V) Research question 4: How can process firms manage the FFE of product and process development? (Addressed in papers III, IV and V)

6

1.4 Comments on the terminology used in the thesis

A common approach for describing innovation is to show the differences between innovation and invention (e.g. Frishammar and Hörte, 2005; Trott; 2005). A widely used definition in the literature is that innovation is “an iterative process initiated by the perception of a new market and/or new service opportunity for a technology-based invention which leads to development, production, and marketing tasks striving for the commercial success of the invention” (Garcia and Calantone, 2002, p.112). Thus, an invention does not become an innovation until it has passed through production, marketing and commercialization. The concept of innovation has been divided by some researchers into technical and administrative innovations (e.g. Daft, 1978; Van de Ven, 1986). Technical innovations such as products and processes are the centre of attention in the thesis, while administrative innovations, exemplified as new procedures, policies, and organizational forms, are subordinated. It is however recognized that some innovations can include both technical and administrative components. For example, some process development (a technical innovation) includes both new process equipment and new work procedures (administrative innovation). Innovation is further seen as a process encompassing all activities from basic research to market launch, where the outputs can be either products or processes (Brem and Voigt, 2009; Trott, 2005). Thus, product development and process development are sub-processes of the innovation process. The innovation process is sometimes used as a synonym for the product development process in the literature, but not in this thesis, as the focus is on creating both products and processes. In section 2.1 a more detailed description of product and process development is given, as these are key concepts in this thesis. Research and Development (R&D) is another term used in this thesis. In the literature, R&D management is sometimes used as a synonym for innovation management. Thus, R&D is also a broad term that involves several different activities ranging from basic research, applied research, pre-development, product and process development and sometimes technical service (Bröring, 2005; Brem and Voigt, 2009; Trott, 2005). R&D involves the use of both new science and the use of old science to produce a new product (Trott, 2005). Inspired by Bröring (2005) and Brem and Voigt (2009), Figure 1 shows how innovation management and R&D management is classified in this thesis.

7

Innovation Management

R&D management

Basic Research

Applied Research

Pre-development Development Scale-up

productionMarket launch

Figure 1 Classification of innovation management and R&D management

Finally, a comment is needed on why the terms of product development and process development are used instead of product innovation and process innovation. All concepts refer to the output of the innovation process (Tushman and Nadler, 1986; Utterback and Abernathy, 1975). However, when a distinction is made between e.g. product innovation and product development, the former typically signifies a higher degree of novelty (e.g. Garcia and Calantone, 2002). Since the context in this thesis is the process industry, and radical innovations are not common in process firms (Lager et al., 2010), product and process development are used as terms instead of product innovation and process innovation. This will also be elaborated further in section 2.1.

1.5 The outline of the thesis

This thesis consists of two main parts. The first part is an introductory text to the second part. The second part includes five research papers, each with its own specific focus. The main intention of the introduction is to introduce the five individual research papers and to discuss how all these papers together contribute with new knowledge on a more general level. By integrating the individual papers a more general and broader discussion can be made rather than merely presenting the contributions in each individual paper. The contributions in four of the papers (II to V) mostly aim at adding new knowledge to the innovation management literature, whereas one paper (I) as well as this introductory text takes the opportunity to contribute with new knowledge to the management control literature. Overall, this thesis explores management of product and process development and therefore it contributes to both literatures. In the next section the frame of reference is presented. The following section describes the specific characteristics of process firms and how they influence the product and process development process in these firms. In the fourth section the scientific

8

positioning and the methods used in this thesis are discussed. The fifth section summarizes the papers appended in Part II. Finally, a discussion of the findings is made and theoretical and managerial implications are provided as well as conclusions and limitations.

9

2. FRAME OF REFERENCE: RESEARCH ON MANAGING PRODUCT AND PROCESS DEVELOPMENT

This thesis draws on research mainly from the field of innovation management and management control, and consequently both fields are discussed in the theoretical background. This chapter starts by defining product and process development and then the FFE of the innovation process is discussed. Finally, research on management control in the context of innovation is presented.

2.1 Product and process development

As stated in the introduction, product development has long been the focus in the innovation management literature (e.g. Brown and Eisenhardt, 1995; Cooper, 2008; Seidel, 2007) and much less attention has been devoted to other innovations such as process development (Pisano, 1997; Reichstein and Salter, 2006). The discourse around process development is scattered and exists in several fields of research – for example in Technology and innovation management (Reichstein and Salter, 2006; Goffin and Mitchell, 2005; Lager, 2002a), Operations management (e.g. Khazanchi et al., 2007), Total quality management (Prajogo and Sohal, 2006), research on Organizational behaviour (Baer and Frese, 2003; Pearce and Ensley, 2004) and in the general Management literature (Collins et al, 1988; Ettlie and Reza, 1992). However, manufacturing firms need to be proficient in both product and process development in order to achieve long-term competitive advantages (Pisano, 1997). More knowledge is needed on how firms can manage both product and process development; as they are different types of innovations, with different objectives and key activities, they will most likely be managed in different ways (Tidd, 2001; Tushman and Nadler, 1986). Product developments are new outputs introduced into a market ‘external’ to the firm, with the purpose of providing benefits to external customers (Gopalakrishnan and Damanpour, 1997). The goal is to create a new or improved product, and subsequently to diffuse this product into a market (Garcia and Calantone, 2002; Frishammar and Hörte, 2005). Key activities in product development include ideation, idea screening, the creation of a product concept, physical development, testing and validation (e.g. Björk and Magnusson, 2009; Cooper, 2008; Davila, 2000). Compared to product development, process development primarily relates to internal production objectives, such as cost reductions and enhancement of production volumes, and therefore the “market” for process development is internal to the firm (Lager, 2002a; Pisano, 1997). Process development has been defined in several different ways in the existing literature. A broad definition of process development incorporates all the processes within a firm, e.g. the business process, work process, distribution process, and production process. In this thesis, however, process development has been defined

11

in a much narrower way and it only refers to development of the production process. Therefore, key activities in process development are converting raw materials, identifying internal production needs, pre-testing, and transferring development results to full-scale production (Lager, 2000; Reichstein and Salter, 2006). Similarly, Tidd (2001) defines process development as changes in how products are created and delivered. A weakness in the process innovation literature is that a variety of different definitions prevail and many authors address the topic without any explicitly stated definition. Process development focuses on efficiency (doing things right), and the “market” for process development is internal to the firm (Lager, 2002a). Increased efficiency can, for example, involve maximizing existing resources to achieve cost reductions, lower risks and shortened development times (Cooper, 2008; Nixon, 1998). Effectiveness on the other hand is related to doing “the right things”, such as developing innovative products that the customers desire (Khurana and Rosenthal, 1997). Moreover, process development includes deliberate organizational attempts to change or modify the production process. Some process development can occur by serendipity, but this is rarely the case since most process development work is conscious and planned (Baer and Frese, 2003). Another characteristic of process development is that it is systemic (Gopalakrishnan et al., 1999), which implies that changes in the production processes affect other processes in the firm (e.g. product development and manufacturing strategy). In addition, process development involves introduction of new elements into the production process. This can be a variety of things ranging from new task specifications of the work process to new equipment used to produce a product (Gopalakrishnan et al., 1999; Reichstein and Salter, 2006). Finally, process development occurs on a degree of newness continuum (Lager, 2002b; Reichstein and Salter, 2006). This is also the case for product development. Thus, it is important to discuss degree of newness when defining product and process development. Garcia and Calantone (2002) describe innovativeness as the degree of discontinuity that an innovation (product, process or service) can generate in the marketing and/or technological processes. In the literature, innovativeness is considered from a variety of perspectives. The most common perspectives used in the innovation management literature are “new to the firm”, “new to the industry” and “new to the world” (Garcia and Calantone, 2002; Lager, 2002b; Reid and de Brentani, 2004). The highest degree of newness is exemplified by “new to the world”, meaning that the product has never been commercialised on a market previously. Innovations which are considered as “new to the world” often are referred to as radical innovations or discontinuous innovations (Reid and de Brentani, 2004; Verworn et al., 2008). If the outcome of the innovation process is a process instead of a product, the dimension can

12

be termed “newness of process technology to the world”, implying that the process technology is completely new and unknown (Lager, 2002b). “New to the industry” means that the product or process is not new to the world, but new to the industry that the specific firm is competing within (Garcia and Calantone, 2002). However, several researchers suggest that newness should be viewed as a continuum rather than clear-cut categories (Garcia and Calantone, 2002; Pisano, 1997). This latter view is taken regarding degree of newness in this thesis. Thus, product and process development vary within a firm along a continuum from radical innovations to incremental innovations which are smaller changes in the product or in the way the product is produced. As the empirical context in this thesis is the process industry, the term ‘process development’ is used in all empirical papers (I, IV and V) in the thesis, whereas in the conceptual paper (II) innovations with a higher degree of newness, process innovations, are investigated. In mature manufacturing firms (e.g. process firms), process development is often the key to competitiveness, since the focus shifts from product development to process development as firms mature (Hutcheson et al., 1995). In mature industries the production process has been designed to be efficient, as price competition becomes more important in later stages of the product life cycle, which also makes the production process rigid and in-flexible. Further, work tasks become more specialized and formal control mechanisms are used more frequently (Utterback and Abernathy, 1975). However, this does not imply that developing new products is not important in mature manufacturing firms. Rather, both product and process development are essential for achieving competitive advantages. In Table 1, the key characteristics of product and process development are presented.

13

Table 1 Key characteristics of product and process development Product development Process development

Concept definition New outputs introduced into a market ‘external’ to the firm, with the purpose of providing benefits to external customers

Deliberate and systemic development related mainly to internal production objectives, implying the introduction of new elements into the production process with the purpose of creating or improving methods of production

Key activities Ideation, idea screening, the creation of a product concept, physical development, and testing and validation

Converting raw materials, identifying internal production needs, pre-testing, and transferring development results to full-scale production

Driving forces Effectiveness, to create a new or improved product

Efficiency, standardization and rationalization

Overall objective Developing new innovative products that the customers want

Cost reduction, increased speed, and lower risks in the development process

Strategic intent Profits Cost reductions

Key departments Product development, R&D and marketing

Process development, production and R&D

To summarize and further the understanding of the innovation process of different types of innovations, it is important to label the different types and degrees of innovation that are being studied (Garcia and Calantone, 2002; Tidd, 2001). The key concepts of product and process development have been explored and, as they have different objectives, key activities, and main focus, it is suggested that they need to be managed differently (Tushman and Nadler, 1986).

2.2 The fuzzy front end of the innovation process

The main focus in the thesis is on the FFE of non-assembled product development and process development, and therefore the existing literature on the FFE has been reviewed. Previous research describes the FFE as the period between when an opportunity is recognized and when a decision is made regarding whether the idea should be developed further or not (e.g. Khurana and Rosenthal, 1998). This is one of the popular views describing when the FFE begins and when it ends. There seems indeed to be an agreement about when it ends (e.g. Khurana and Rosenthal, 1998; Verworn, 2006) but not about when the FFE starts. Every innovation is based on an

14

idea, which can be originated either inside or outside the firm. External ideas can for example be derived from market and/or technology trends (Brem and Voigt, 2009; Bröring et al, 2006) or an idea can be internally generated, for example at the R&D department or the production department. Thus, the sources of new ideas can be either external (e.g. customers, supplier) or internal. As the sources of new ideas sometimes are external, they are placed outside the FFE, whereas the process of generating new ideas is often a key activity in the FFE and therefore it is part of the FFE. Khurana and Rosenthal (1997) suggest that the FFE starts when the firm first recognizes an opportunity, but it has to be shared with another person so that it not only exists in the mind of a single individual – a position taken in this thesis as well. Hence, the focus is not on the individual level (when an individual has a generated a new idea, before he or she has acted upon it) but on the organizational level (the idea must at least have been shared with another person). Some of the characteristics of the FFE have been outlined in the introduction, but as they are essential they are presented in more detail below. The FFE has been described by previous researchers as ill-defined, uncertain and complex, and it consists of a crossroads of complex information processing (Chang et al., 2007; Kim and Wilemon, 2002; Khurana and Rosenthal, 1998). Moreover, the type of information is mainly qualitative and approximate, the focus of the idea is broad, the budget is typically small or non-existent, and the tasks are performed by a small group or one single individual (Kim and Wilemon, 2002). Tacit knowledge, conflicting organizational pressure and ad-hoc decision-making are other characteristics of the FFE (Khurana and Rosenthal, 1997; Montoya-Weiss and O’Driscoll, 2000). Thus, many firms find themselves to have difficulties in the FFE, since many structured methods and tools for product development are not perceived as adequate for this stage. The fuzzy nature of this stage also makes it difficult for researchers to study the FFE (Murphy and Kumar, 1997). Table 2 below draws on Brem and Voigt, 2009 and Kim and Wilemon, 2002 and reflects my own understanding about the differences between the FFE and the later phases of development.

15

Table 2 Comparison between the FFE and the following development stages of product and process

development The FFE stage The product and process

development process

State of an idea Probable, fuzzy and flexible Clearer, specific and ready for development

Nature of work Experimental, difficult to plan and conduct

Structured and goal-oriented

Commercialization date Unpredictable Definable

Funding Small or none Larger, budgeted

Revenue expectations Often uncertain and often done with a great deal of speculation

Believable with increasingly certainty

Personal involvement Often one individual or a small team

Often a large cross-functional team

Management method Unstructured, experimental and creativity

Structured and systematic, e.g. a Stage-Gate model

Outcome A product/ process concept A product or a process change

The FFE literature has been reviewed and different areas have received varying attention in the existing literature. For example, previous research has conceptualized the FFE (e.g. Cooper, 2008; Griffiths-Hemans and Grover, 2006; Khurana and Rosenthal, 1997; 1998; Koen et al., 2001). General success factors for the front end have been identified (e.g. Koen et al., 2001; Kim and Wilemon, 2002). As the uncertainty in the FFE generally is very high, a handful of scholars have focused on approaches for reducing uncertainties (e.g. Monenart et al., 1995; Verworn et al., 2008). The FFE has been divided into three sub-phases: idea generation, screening of ideas and evaluation of ideas (Björk and Magnusson, 2009; Griffith-Hemans and Grover, 2006). All these sub-phases have been explored, either individually (e.g. Björk and Magnusson, 2009) or by investigating all phases from idea to the final product concept (Khurana and Rosenthal, 1998; Kijkuit and van den Ende, 2007). Idea generation has received much attention in the innovation management literature, and it is indeed a complex and important activity in the FFE. Firms need proficiency in idea refinement and screening of ideas (Griffiths-Hemans and Grover, 2006; Khurana and Rosenthal, 1997), as deficiencies (e.g. poor ideas, too few ideas, or poor

16

screening) often result in costly problems in later stages of the development process (Cooper, 1988). Screening can be done in different areas through business analysis and feasibility analysis, and the tactics can be divided into active and passive screening of ideas (Bröring, 2005; Murphy and Kumar, 1997). Screening of a new product idea is done in terms of its viability as a business proposition, whereas feasibility analysis determines whether a firm can support a subsequent development project with sufficient resources (Murphy and Kumar, 1997). Active screening encompasses active collaboration with external organizations, e.g. customers, competitors, suppliers, research institutes and universities, whereas passive screening involves screening of technical reports, use of patent databases, attendance of conferences, and screening of scientific publications. The final sub-phase is evaluation of a product or process concept. A product definition includes a product concept, which represents the goal for the development project (Montoya-Weiss and O’Driscoll, 2000; Seidel, 2007), and it contains information about target markets, customer needs, and product specifications (Montoya-Weiss and Calantone, 1994) as well as product positioning and product requirements (Cooper and Kleinschmidt, 1987). Proficiency in creating product definitions is crucial as it deeply impacts the go/kill decision prior to formal product development (Cooper, 1988). Evaluation of a product concept is an essential activity in the FFE. Previous literature shows that a cross-functional executive review committee is likely to affect the performance of front end activities positively, as cross-functional competence is needed when evaluating product concepts (Khurana and Rosenthal, 1997) Finally, some attempts have been made to understand what type of management is suitable in the FFE stage. Khurana and Rosenthal (1998) found that firms use either a culture-driven approach or a formal approach when managing the front end. A culture-driven approach involves subtle control, cross-functional integrations, and a deep understanding of product development. This approach is similar in many ways to the second stream of research presented in the introduction, whereas the formal approach is more in line with the structured approach presented in the first chapter. A formal approach according to Khurana and Rosenthal (1998) brings order to the FFE, and a formal process is explicit, widely known, and characterised by clear decision-making responsibilities and specific performance metrics. Further, they found that two types of risks exist when relying on a formalized approach, either an absence of formality or an excessive reliance on it. Recently, researchers have studied the use of management control in the FFE. Poskela and Martinsuo (2009) pose the question whether any type of control has a positive

17

effect on performance in the FFE, and they show that input control and intrinsic task motivation are necessary for strategic renewal in the FFE. Further, they argue that a certain amount of control appears essential in managing the FFE to ensure effective use of resources, and at the same time that the employees require high levels of freedom and independence to execute FFE activities. This short review of the FFE literature shows several things. A first observation is that the literature on the FFE is relatively new compared to research on the whole innovation process, which explains why so many studies are either conceptual or exploratory case studies (e.g. Chang et al., 2007; Kim and Wilemon, 2002; Khurana and Rosenthal, 1998; Verganti, 1997). Another observation is that most studies focus on (assembled) product development (e.g. Cooper, 1988; Koen et al., 2001; Seidel, 2007; Verworn, 2009), while other innovation types such as non-assembled product development and process development have received limited attention. The FFE literature assumes that the FFE phase is similar regardless of whether assembled products, non-assembled products or processes are developed. Finally, only recent researchers have taken into account different contingencies which most likely will affect how the FFE is organized and managed. As an example, Nobelius and Trygg (2002) have criticised the FFE literature for offering a far too simple explanation of how the FFE should be organized and managed, as many of the early studies on the FFE (e.g. Cooper, 1988; Smith and Reinertsen, 1998) do not take into account that different types of innovations most likely require different management. More research is also needed on what type of management is suitable to increase speed and quality of the FFE and at the same time grant conditions for innovation.

2.3. Management control in the context of product and process development

The role of management control in product development has gained in interest during the past decade (Davila, 2000; Poskela and Martinsuo, 2009; Richtnér and Åhlström, 2010). As stated in the introduction, however, there are contradictory findings in the literature about whether management control is suitable to manage innovations or not. The contradictory findings can to some extent be explained by differences in how management control has been defined in the literature. The early classical view of management control suggests that management control makes a foundation for efficient decision-making, mainly based on financial accounting information, with the aim of ensuring that an individual or a group works towards the goals of the

18

organization (Chenhall, 2003; Kirsch, 1996). The classical view of management control implies that it is a systematic approach of controlling outcomes and behaviour (Ouchi, 1979). Thus, the early definitions focused entirely on formal, quantifiable information assisting in managerial decision-making. The definition has evolved over time to take a broader approach to management control. Here information related to markets, customers, non-financial information related to production processes, and informal personal and social controls is included (Chenhall, 2003). Simons defines management control as “the formal, information-based routines and procedures managers use to maintain or alter patterns in organizational activities” (1994, p. 170). This broader definition is typically used when studying management control in the context of innovation (e.g. Abernathy and Brownell, 1997; Chiesa et al., 2009; Davila, 2000). Simons (1995) argues that his view of management control reconciles the conflict between creativity and control. The classical definition of management control that it measures progress against predetermined goals only represents one perspective of a control system, which he calls diagnostic control systems. Business plans and budgets are examples of diagnostic control systems. The other three are: belief systems, boundary systems and interactive control systems (Simons, 1994; 1995). These four main perspectives on management control, all with different characteristics in terms of underlying nature, purpose, and key design variables, are used in practice by managers to assist in formulating and implementing business strategies of the firms. Belief systems are used by managers to define, communicate and reinforce the basic values and mission of the firm. The belief system is a formal system which is communicated through formal documents such as mission statements and statement of purposes. Boundary systems are formal systems to establish limits and rules, which have to be respected. These formal systems are typically stated in negative terms or as minimum standards. The bases for formal boundary systems are codes of business conduct, formal planning systems, and operating directives. The key design variable is to avoid as many risks as possible. Finally, interactive control systems are used by managers to personally and continuously involve themselves in decisions and behaviour of their subordinates (Simon, 1994). The key design variable here which affects the design of interactive control systems is strategic uncertainty. Simons’ studies (1994; 1995) focus on top managers and how they can use management control systems to accomplish strategic renewal. Thus, Simons’ interactive control systems are a way to capture opportunities outside the firm’s boundaries and to guide and give input to innovation (Bonner et al., 2002). However, Simons’ research does not address informal control, which is also important in product development (e.g. Chiesa et al., 2009; Davila, 2000; Poskela and Martinsuo, 2009).

19

Management control is viewed in this thesis according to the stream of literature that argues for a broader definition of management control. Thus the broad definition provided by Simons above is in line with the view on management control in this thesis. But Simons (1994) focuses only on formal control that maintains or changes patterns in organizational activities, while informal control is also an important component of management control in this thesis. In the literature, two broad categories of control are discussed: formal and informal control. Formal control mechanisms are often described as explicitly designed controls, while informal control mechanisms (also referred to as social control, personnel control and clan control) are exemplified by unwritten rules and norms that cannot be designed directly (Abernathy and Brownell, 1997; Dekker, 2004; Ouchi, 1979). Formal control has further been divided into behavioural and outcome control (Kirsch, 1996; Ouchi, 1979), where the latter refers to measuring and monitoring the outputs of operations or behaviour, while the former focuses on specifying and measuring individuals’ behaviour (Abernathy and Brownell; 1997; Kirsch, 1996). Informal or social control mechanisms, on the other hand, are often developed from shared norms, values and beliefs, and are shaped e.g. by self-regulating mechanisms and informal cultures (Dekker, 2004; Chiesa et al., 2009). An example of how firms use informal social control is the selection of appropriate employees and seeing that employees will receive appropriate training which ensures that the “right” employees are obtained (Abernathy and Brownell, 1997). Another explanation for the contradictory findings in the literature is that, within the classical view, the purpose of management control is to reduce goal divergence and to coordinate and control tasks (e.g. Abernathy and Brownell, 1997), while the purpose of management control in an innovation project is to reduce uncertainty (Davila, 2000). Management control is seen as one important source of information but, depending on the type of uncertainty, other alternative sources can be more appropriate, such as experimentation (Pisano, 1997) and informal communication. The concept of uncertainty is divided into three main sources in the product development literature: market-related uncertainty, technological uncertainty and project scope (Kim and Wilemon, 2002; Moenart et al., 1995; Verworn et al., 2008), which shapes the design and use of the type of management control. Davila (2000) found that management control was used to reduce the uncertainty when it was related to the market or project scope, but that other sources of information were used if it was technological uncertainty.

20

Two levels of control exist in innovation projects: project management (the operative level) and the (strategic) top management level (Chiesa et al., 2009; Davila, 2000). Traditionally, management control is viewed as a tool for top management (e.g. Ouchi, 1979); however, by broadening the view of management control to focus on uncertainty reduction in innovation projects, the use of management control is brought down to the project level (Davila, 2000). In this thesis, management control is seen as the formal link between the firm’s (innovation) strategy and the way product and process development projects are conducted (Otley et al., 1995). Finally, the use of management control in the FFE has been explored by a few researchers (Chiesa et al., 2009; Poskela and Martinsuo, 2009). These studies all belong to the stream of research that argues for a balance between creativity and control (introduced in Chapter 1) and they show that flexible control and social control are important in the FFE, as well as defining the front-end task and allocating resources to the FFE. Thus, these studies show that both informal and formal control mechanisms are important in the early stage of the innovation process.

21

3. THE PROCESS INDUSTRY AS A CONTEXT FOR PRODUCT AND PROCESS DEVELOPMENT

Innovation processes, technological opportunities and market conditions differ among industries. These differences are important to clarify for increasing the understanding of how to manage innovation within different industries. As stated in the introduction, the industry focused upon in this thesis is the process industry. The term is commonly used but it is misleading because the process industry involves firms from a variety of industries, rather than firms within one industry. In this section the specific characteristics of process firms are first presented, and then the focus is directed towards how process firms develop new products and processes.

3.1. The specific characteristics of process firms

In the introduction, process firms were briefly introduced and described. In this section a more detailed description is provided. Process firms are a subset of manufacturing firms and the products produced are called non-assembled products, meaning that they are composed of only one or a few materials, compared to assembled products which are put together from several different components (Utterback, 1994). Minerals and metals, chemicals, pulp and paper, textiles, food and beverages are some examples of non-assembled products (Dennis and Meredith, 2000; IVA; 2006; Lager, 2002a; Utterback, 1994). Process industries are some of the oldest and most mature industries in business. They have been historically important for building up the prosperity and civilization in Sweden and elsewhere, and stand for a considerable portion of GDP in many countries (van Donk and Fransoo, 2006). Process firms are suppliers of input materials to many different traditional industries. Steel is one important product that today is a cornerstone of every home and household, and of the entire society (SGU, 2008). The process industry is an industry encompassing both common and industry-specific characteristics (IVA, 2006). The common characteristics are presented first and they are discussed in relation to other manufacturing firms which produce assembled products. The input materials used, the production process, the locus of innovation, and the relationship between product and process development are some of the most important characteristics that will be discussed below, because they influence the management of product and process development. The first major difference concerns the input materials used. In process firms, the input materials are mainly raw materials instead of components from suppliers (Barnett and Clark, 1996; Lager, 2002a). Process firms add value by mixing, separating,

23

forming, and conducting chemical reactions, instead of combining several different components into a product. The products that are produced are homogeneous, rather than heterogeneous as is the case for assembled products (IVA, 2006). Hence, if there are any changes in the input material in process industries, they will affect the entire product group produced and not just a few products. Second, the production process in process firms is often unique for the purpose of producing one particular product. Moreover, it is rigid, capital-intensive, and continuous with on-line control in real time (Barnett and Clark, 1996; Hutcheson et al., 1995; Lager, 2000). Third, the focus of innovation in process firms is primarily on process development rather than product development (Hutcheson et al., 1995; Utterback, 1994). As an industry matures, the focus of innovative activity gradually shifts from product development to process development (Hutcheson et al., 1995) and the nature of non-assembled products makes the production process very critical for achieving competitive advantages. Thus, process development often constitutes a large share of the total R&D budget in these firms, compared to other manufacturing firms (Lager, 2002b). Finally, in process firms, product and process development are deeply intertwined (Barnett and Clark, 1996; Linton and Walsh, 2008; Pisano, 1997), implying that changes in the production process can have a significant impact on the end product, and conversely. For example, by changing one or a few parameters in the production process such as heat and temperature, the product will receive different characteristics. There are other important differences between process firms and other manufacturing firms which affect how product and process development is conducted. For example, the production plants in process firms are generally very large, strongly integrated on one site, often located near the input materials, and they are often situated upward in the supply chain, several steps away from the final customer (Hutcheson et al., 1995; Lager et al., 2010). In many process firms, such as minerals and metals, paper and pulp, chemistry and plastics, the production process is very energy-demanding; consequently, political decisions have a high impact on these firms (SGU, 2008). The demand and pricing, for example of minerals and metals, are very different than for consumer products. The demand of iron ore varies over time and it is linked to all the forthcoming steps in the supply chain2. To summarize, the specific characteristics pointed out above suggest that product development and process development are conducted differently in process firms, rather than when assembled products are developed for consumers.

2 The price of metals (except iron ore) is set at a stock exchange in, for example, London. The price of iron ore is negotiated between mining companies and their largest customers, the steel companies (SGU, 2008).

24

As process firms are a broad group encompassing several types of process firms (Dennis and Meredith, 2000), it is important to clarify that there are also industry-specific differences among the process firms which influence how they manage their innovation processes. Van Donk and Fransoo (2006) describe the variety among process firms: “the process industries cover a wide range of businesses, ranging from continuous facilities in the petrochemical industry, to large batch manufacturing in steel production and glass manufacturing, to small batch manufacturing in the food and pharmaceutical industry” (2006, p. 211). Due to this range of firms, it is likely that different process firms manage their innovation process differently. For example, most of the process firms sell their products to other firms, while a few process firms (e.g. firms in the food and beverage sector) are closer to the end consumers. As the market is different, this will most likely influence how they develop their products. Another example is that the FFE is mostly science-driven in e.g. process firms such as pharmaceuticals and chemicals, while the FFE in consumer-focused food firms is often purely market-driven (Sundgren, 2004; Bröring et al., 2006).

3.2. Product and process development in process firms

As already pointed out, a number of studies have shown that the process industry differs in many respects from other manufacturing industries (e.g. Barnett and Clark, 1996; Lager, 2002a; Pisano, 1997; Utterback and Abernathy, 1975). For example, product development in process firms largely focuses on improvement of product properties and quality (Lager, 2002a) – rather than on developing totally new products, which normally is the case when assembled products are developed for the consumer market. The product life cycles are typically very long (8-10 years) for non-assembled products, e.g. the development of pharmaceuticals or the development of a totally new iron ore compared to assembled products (Sundgren, 2004; Lager et al., 2010). The interdependence between product development and process development has been mentioned as a common characteristic of process firms, and it will be further discussed here as it also affects how product and process development is conducted in process firms. When process firms develop or modify new products, new or modified processes are likely to follow. On the other hand, new or improved processes can sometimes allow new product concepts that were previously impossible to develop (Barnett and Clark, 1996; Lim et al., 2006; Macher and Mowery, 2003). For example, new process technology implemented in the production process can allow the development of new products that previously were impossible to produce in the existing production processes (Lager, 2002a; Lim et al., 2006; Linton and Walsh,

25

2008). Another characteristic of the interdependence of product and process development, which is even more crucial, is that the features of the end product change often when a process development is done (Barnett and Clark, 1996). Therefore, for process-based products, a change in the production process often results in significant changes of end-product features. This implies that the interdependence between product and process development is central for the overall success of an innovation project. Finally, although the FFE of non-assembled product development has been explored previously, only a few papers address this specific context (Bröring et al., 2006; Bröring and Leker, 2007; Elmquist and Segrestin, 2007). Elmquist and Segrestin (2007) have studied the FFE in pharmaceutical R&D. Pharmaceutical companies are under great pressure to innovate and be cost-efficient at the same time, as stricter regulatory demands have led to less product exclusivity. This naturally influences the FFE process. They have looked at actual management practices in the FFE of developing non-assembled products and found that existing research on the FFE is insufficient to understand the FFE of non-assembled products. When new innovative concepts are developed, which are based on new knowledge that the firm does not possess, several product concepts are developed, rather than one optimal candidate which often is the case in assembled product development. This is the strategy for being both innovative and efficient in pharmaceutical development, as the uncertainty is very high and many of the candidates do not make it. Bröring (2005) has also studied the FFE of non-assembled products in her thesis of the FFE in converging industries. The context in Bröring’s study is in the nutraceuticals and functional foods sectors, showing how parts of the food industry and pharmaceutical industry converge. She explores what happens in the FFE when firms from different industries converge. In this context there is another type of uncertainty than the one present in pharmaceutical R&D. Based on these two studies, it can be concluded that the FFE is different if the firm develops pharmaceuticals or if two industries are converging and a totally new product is created. However, more research is needed to explore the FFE in non-assembled product development. The differences among process firms and other manufacturing firms have been outlined above, but it is important to mention that process firms also often face challenges similar to those experienced by other large manufacturing firms. Some of these similarities will be discussed shortly. Process firms stands for a large part of the GDP, which means that these firms export many of their products. Therefore process firms face the increasingly hard pressure on the global market (IVA, 2006) as do other manufacturing firms. The production process is a central source of competitiveness in both process firms and many manufacturing firms, since it is difficult for other firms to

26

copy (Goffin and Mitchell, 2005). Both product development and process development are essential for survival and renewal in both types of firms (Pisano, 1997).

27

4. METHODOLOGY

First the scientific standpoint in this thesis is presented, and it is followed by the overall research design. The main methodological approach used is case studies. The following section describes how the empirical data have been analysed, and then the quality of the research in the thesis is discussed. Finally the research process and the background to the appended papers are presented.

4.1 Scientific positioning

The scientific position taken in this thesis is the foundation for the thesis and it guides the research on the ontological and epistemological level (Guba and Lincoln, 2000). Ontology refers to the set of underlying assumptions about the form and nature of reality. Epistemology concerns the researcher’s assumptions about what one can know, and the underlying ideas of how people understand the world and communicate this knowledge to fellow human beings (Burrell and Morgan, 1979). The underlying ontological and epistemological assumptions will influence how any method or technique is applied and evaluated based on the assumptions upon which it is founded (Guba and Lincoln, 2000). Burrell and Morgan (1979) discuss four competing scientific paradigms: the functionalist, the radical humanist, the radical structuralist and the interpretivist. The standpoint taken in this thesis is similar to what Burrell and Morgan calls the functionalist paradigm. Ontologically, it is assumed that there is a real world existing independently of our attempts to know it (Burrell and Morgan, 1979). Thus, organizations are real and we can study them. Dubin (1982) describes organizations such as entities with structures, boundaries, goals and resources. These organizations also consists of members “whose behaviour results from structured relations among them” (Dubin, 1982, p. 372). Moreover, it is assumed that it is possible to describe and explain both observable and unobservable phenomena. Epistemologically, this thesis builds on the idea that knowledge is a cumulative process in which new insights are added to the existing stock of knowledge (Burrell and Morgan, 1979). Thus, it is possible to gain knowledge of the world as it is open for empirical investigations.

4.2 Research design

Two research projects have been conducted to answer the stated research questions and to fulfil the overall purpose of this thesis. The first research project in the thesis focused on management control in collaborative development projects, while management of the FFE in the innovation process was the centre of attention in the second part of the thesis. See Figure 2 for an overview of the two research projects.

29

Managing innovation

(Innovation Management and Management Control)

Collaborative R&D

In-house R&D

Process development

Process development

Product development

The FFE The FFE

Product development

Paper I Paper II

Paper III Paper IV

Paper V

Figure 2 The research projects within this thesis

4.2.1 Research project 1: Management Control in Collaborative Product and Process Development Projects

Management control in collaborative product and process development projects was first explored (the blue areas in Figure 2). In more detail, control mechanisms and the underlying motives of firms when conducting collaborative product and process development projects were the focus in this initial study. Two case studies were conducted within this research project, one addressing motives and selection of control mechanisms in product development projects, and the other focused on process development projects. The empirical data came from LKAB and external partners (such as two of their main customers, a university and a research centre). This research project resulted in a licentiate thesis (Kurkkio, 2006) and the first appended paper in this thesis.

4.2.2 Research project 2: Managing the Fuzzy Front End of Product and Process Development

The second research project in this thesis focused on the FFE of product and process development (the green areas in Figure 2). Two multiple case studies were conducted, one at the FFE of non-assembled product development (Höganäs AB, LKAB and SSAB were the case study firms) and the other at the FFE of process development (Boliden AB, Höganäs AB, LKAB and SSAB were the case study firms). In this research project the FFE was conceptualized for both product development and process development; key activities, managerial challenges, key problems and success

30

factors were identified and analysed. The second research project resulted in papers II-V appended in Part II of this thesis. In section 4.7 the background to each paper is described, but first the case study approach is discussed.

4.3 The case study approach

In both research projects the case study approach was used as the main research strategy. The perhaps strongest argument for choosing case studies was to ensure methodological fit (Edmondson and McManus, 2007). The idea of methodological fit is to find an appropriate match between research questions, data collection, data analysis, and status of current theory. The status of the existing knowledge of both research projects, “Management control in collaborative product and process development” and “Managing the FFE of product and process development”, is considered as nascent, especially concerning research on the FFE in process development (research project 2). This was the main reason for choosing case studies as the research strategy. Another argument was that case studies are appropriate when the researcher wants to gain a deeper understanding of the specific research area and to investigate social phenomena in real-life contexts (Meredith, 1998; Miles and Huberman, 1994; Yin, 2003). Case studies enable the investigation of formal as well as informal processes and activities, and this was essential in both research projects. For example, in the first research project, management control systems were viewed as consisting of both formal and informal control mechanisms, and the intention was to capture both formal and informal control mechanisms. In the second research project, exploring the FFE of the innovation process, many of the activities were typically done outside formal projects, which shows the importance of identifying informal activities as well as the formal ones. Finally, multiple sources of data can be used within the structure of case study research (Yin, 2003). In both research projects, formal documents such as different types of project reports were used, as well as external information such as annual reports and general information about the firms’ businesses were collected from the Internet. The use of multiple sources of data increases the validity of the conducted research (Yin, 2003). A multiple case study design was also selected instead of a single case study, because a multiple case study design allows comparison of several cases, which is more likely to generate more accurate and generalizable empirical data (Eisenhardt and Graebner, 2007; Yin, 2003).

31

4.4 Data collection and data analysis

When doing case studies, appropriate identification and selection of cases, as well as the definition of units of analysis, is crucial. In both research projects the cases were selected by using theoretical sampling, with the aim of sampling cases that would replicate each other and consequently extend emergent theory (Eisenhardt and Graebner, 2007; Yin, 2003). Moreover, as the theory especially regarding the FFE in process development was limited, the replication logic was selected as this would allow for better and more valid analyses. Different industries have different contextual factors in terms of e.g. sources of innovation, environments and competition (Tidd, 2001; Utterback, 1994). To rule out as many differences as possible among the cases, they were selected within the same industry, the minerals and metals industry. As both product development and process development were focused upon in this thesis, the process industry was considered to be an appropriate context, since both innovation types are critical for competitiveness in these firms. The cases in the first research project were selected according to two criteria. The first criterion was that the product and process development projects should involve external partners where uncertainty was paramount, which most likely was reduced by some kind of control mechanism. A second criterion was that access to all the participating firms in both cases was considered to be important, as most studies of collaborative R&D projects often only present one view of the collaboration project. This would allow a more accurate view of the management and the different motives that the separate firms had for developing products and processes with external partners. In the second research project, good access was also an important selection criterion, as it is complex to study the FFE of innovation (Murphy and Kumar, 1997). Thus, to get a more accurate and realistic account of the difficulties of managing the FFE in product and process development, selecting firms that allowed us to get a broad perspective on the FFE was important, as well as the fact that the firms had great experience of conducting both product and process development. The case study firms in the second research project were Boliden AB, Höganäs AB, LKAB and SSAB (see Table 3 for an overview of the research sites). All of them are process firms and they belong to the minerals and metals industry. This sector is mature, and as a consequence the focus is primarily on conducting improvements rather than developing radically new products. Thus, the R&D in these firms mostly concerns development and few projects are strictly research-based. When the case study firms conduct research projects which can be classified as basic research, they most likely do these projects together with external partners such as universities or research centres. The customers within this sector are other industrial firms. For example, LKAB and Boliden AB are situated at the beginning of the value-chain,

32

33

whereas Höganäs AB and SSAB are their customers as their products are more processed. Another common characteristic is that all of the case study firms have a highly energy-demanding production process and it influences their product and process development. Environmental demands are sometimes the driver of new product and process development projects in these firms.

34

Tab

le 3

Res

earc

h sit

es

Cas

e st

udy

firm

s B

olid

en A

B

Hög

anäs

AB

L

KA

B

SSA

B

Mai

n b

usines

s M

etal

pro

duce

r Su

pplie

r of

met

al p

owde

r

Inte

rnat

iona

l hig

h-te

ch m

iner

als

grou

p G

loba

l pro

duce

r of

sev

eral

di

ffere

nt h

igh-

stre

ngth

ste

el

prod

ucts

M

ain a

pplica

tion

area

s/ p

roduct

s

The

mai

n m

etal

s ar

e co

pper

and

zin

c,

but

lead

, gol

d an

d sil

ver

are

othe

r im

port

ant

met

als.

A v

arie

ty o

f app

licat

ion

area

s, in

clud

ing

sinte

red

com

pone

nts,

soft

mag

netic

co

mpo

sites

, hot

pol

ymer

filtr

atio

n, a

nd

surf

ace

coat

ing.

Prod

uces

upg

rade

d ir

on o

re fo

r th

e st

eel i

ndus

try

and

indu

stri

al

min

eral

s pr

oduc

ts fo

r ot

her

indu

stri

es.

App

licat

ion

area

s ar

e pr

imar

ily

brid

ges,

build

ings

, shi

ps,

vari

ous

form

s of

veh

icle

s an

d lif

ting

devi

ces.

Pro

duct

dev

elopm

ent

Doe

s no

t de

velo

p pr

oduc

ts in

a

trad

ition

al w

ay, a

s th

e sm

elte

rs e

xtra

ct

and

prod

uce

pred

eter

min

ed b

asic

m

etal

s ou

t of

ore

. The

cha

lleng

e is

to

prod

uce

a pr

oduc

t of

con

siste

nt

qual

ity e

ven

thou

gh t

he q

ualit

y of

the

in

put

mat

eria

ls va

ries

.

Tec

hniq

ues

to p

rodu

ce d

iffer

entia

ted

prod

ucts

incl

ude

addi

ng c

oatin

g to

the

met

al

pow

der

or m

ixin

g th

e m

etal

pow

der

with

di

ffere

nt a

lloys

. Thu

s, pr

oduc

t de

velo

pmen

t ty

pica

lly in

clud

es s

mal

l-sc

ale

test

s, w

here

the

ai

m is

to

find

a m

etal

pow

der

with

the

op

timal

pro

pert

ies

for

the

cust

omer

s’ pr

oduc

tion

proc

esse

s.

The

pro

duct

dev

elop

men

t in

clud

es

both

impr

ovin

g ex

istin

g pr

oduc

ts

and

deve

lopi

ng n

ew p

rodu

cts

with

di

ffere

nt p

rope

rtie

s. O

ne t

ypic

al

prod

uct d

evel

opm

ent p

roje

ct is

to

chan

ge t

he p

rope

rtie

s of

the

iron

or

e, t

here

by c

usto

miz

ing

the

prod

uct.

Prod

uct

deve

lopm

ent

incl

udes

de

velo

pmen

t of

tot

ally

new

pr

oduc

ts a

s w

ell a

s im

prov

emen

ts o

f exi

stin

g pr

oduc

ts. A

larg

e pa

rt o

f all

prod

uct d

evel

opm

ent

is cu

stom

er-d

rive

n.

Pro

cess

dev

elopm

ent

The

focu

s of

pro

cess

dev

elop

men

t is

on c

ontin

uous

impr

ovem

ents

of t

he

prod

uctio

n pr

oces

s w

ithin

eac

h pl

ant.

The

pro

cess

dev

elop

men

t is

divi

ded

into

lo

cal a

nd g

loba

l dev

elop

men

t; in

crem

enta

l pr

oces

s de

velo

pmen

t is

cond

ucte

d lo

cally

at

the

prod

uctio

n pl

ants

, and

pro

cess

de

velo

pmen

t with

hig

her

degr

ee o

f new

ness

(e

.g. d

ecisi

ons

on r

educ

ing

the

num

ber

of

step

s in

the

pro

duct

ion

plan

or

to in

vest

in

new

pro

cess

equ

ipm

ent)

is c

entr

aliz

ed.

The

pro

cess

dev

elop

men

t pr

imar

ily

focu

ses

on i

mpr

ovin

g th

e ex

istin

g pr

oces

s te

chno

logy

, no

t on

im

plem

entin

g ra

dica

lly n

ew p

roce

ss

tech

nolo

gy

into

th

e pr

oduc

tion

proc

ess.

The

pro

cess

dev

elop

men

t is

mai

nly

cond

ucte

d at

the

bla

st

furn

aces

whe

re t

he s

teel

sla

bs

are

prod

uced

. The

ste

el s

labs

ar

e tr

ansp

orte

d to

the

pr

oduc

tion

plan

ts t

o be

pr

oces

sed

in t

he r

ollin

g m

ills.

Som

e pr

oces

s de

velo

pmen

t is

perf

orm

ed in

the

pro

duct

ion

plan

t as

wel

l. Pr

oces

s de

velo

pmen

t in

clud

es a

ran

ge

of a

ctiv

ities

from

con

tinuo

us

impr

ovem

ents

to

inve

stm

ent

in m

ore

mod

ern

proc

ess

equi

pmen

t.

Appro

xim

ate

num

ber

of

emplo

yees

4 40

0 em

ploy

ees

1

440

empl

oyee

s 3

700

empl

oyee

s 9

000

empl

oyee

s

35

There are several different units of analysis in both research projects, which corresponds to Yin’s (2003) notion of embedded units of analysis. In the first research project, the main unit of analysis was either the product development process or the process development process. Additional analyses were also made on the project level and firm level. For example, the motives for engaging in collaborative product and process development projects were analysed on the firm level and the selection of control mechanisms were analysed on the project level. In the second research project, the main unit of analysis was the FFE of product and process development and the sub-unit of analysis was the firm level (Yin, 2003). In Table 4 the specific research questions or purposes, data collection methods, analytical unit and the focus of the study applied in each appended paper are summarized.

36

Tab

le 4

Ove

rvie

w o

f res

earc

h m

etho

ds

Pap

ers

Res

earc

h pu

rpos

e or

que

stio

ns

Dat

a co

llect

ion

Ana

lyti

cal un

it a

nd foc

us o

f th

e st

udy

Pap

er I

C

ontr

ol m

echa

nism

s in

co

llabo

rativ

e R

&D

pro

ject

s: th

e ca

se o

f sup

plie

r an

d cu

stom

er r

elat

ions

hips

in

proc

ess

indu

stry

The

aim

of t

his

artic

le is

to

iden

tify

and

anal

yse

wha

t m

otiv

es c

olla

bora

ting

firm

s ha

ve w

hen

enga

ging

in

colla

bora

tive

R&

D p

roje

cts

and

how

the

se m

otiv

es

subs

eque

ntly

affe

ct t

he s

elec

tion

of c

ontr

ol

mec

hani

sms.

Inte

rvie

ws,

inte

rnal

and

ex

tern

al d

ocum

ents

T

he p

rodu

ct a

nd p

roce

ss d

evel

opm

ent

proc

esse

s

Firm

and

pro

ject

leve

l

Pap

er I

I A

ntec

eden

ts a

nd C

onse

quen

ces

of F

irm

s’ Pr

oces

s In

nova

tion

Cap

abili

ty: A

Lite

ratu

re

Rev

iew

and

Con

cept

ual

Mod

el

The

obj

ectiv

e in

thi

s pa

per

is to

add

ress

a p

robl

em

shar

ed b

y bo

th p

ract

ising

man

ager

s an

d th

e re

sear

ch

com

mun

ity: t

he is

sue

of h

ow t

o m

anag

e pr

oces

s in

nova

tion

so t

hat

its p

ositi

ve b

enef

its c

an b

e be

tter

re

aliz

ed.

Con

cept

ual

The

who

le p

roce

ss in

nova

tion

proc

ess

Firm

leve

l and

pro

ject

leve

l

Pap

er I

II

The

Fuz

zy F

ront

End

in N

on-

asse

mbl

ed P

rodu

ct

Dev

elop

men

t: A

Mul

tiple

Cas

e St

udy

of M

iner

al a

nd M

etal

Fi

rms

The

pur

pose

of t

his

pape

r is

to

incr

ease

kno

wle

dge

abou

t th

e FF

E in

non

-ass

embl

ed p

rodu

ct

deve

lopm

ent.

Inte

rvie

ws,

grou

p m

eetin

gs

with

R&

D m

anag

ers

and

engi

neer

s, in

tern

al a

nd

exte

rnal

doc

umen

ts

The

FFE

of t

he p

rodu

ct d

evel

opm

ent

proc

ess

Proj

ect

and

firm

leve

l

Pap

er I

V

Whe

re p

roce

ss d

evel

opm

ent

begi

ns: A

mul

tiple

cas

e st

udy

of

fuzz

y fr

ont

end

activ

ities

in

proc

ess

firm

s

The

pur

pose

of t

his

pape

r is

to

incr

ease

kno

wle

dge

abou

t th

e FF

E in

pro

cess

dev

elop

men

t.

Inte

rvie

ws,

inte

rnal

and

ex

tern

al d

ocum

ents

T

he F

FE o

f the

pro

cess

dev

elop

men

t pr

oces

s Pr

ojec

t an

d fir

m le

vel

Pap

er V

M

anag

ing

the

fuzz

y fr

ont

end:

In

sight

s fr

om p

roce

ss fi

rms

The

aim

of

this

pape

r is

to i

nfor

m r

esea

rche

rs a

nd

prac

titio

ners

ab

out

the

FFE

of

th

e in

nova

tion

proc

ess

in p

roce

ss fi

rms.

Inte

rvie

ws,

grou

p m

eetin

gs

with

R&

D m

anag

ers

and

engi

neer

s, in

tern

al a

nd

exte

rnal

doc

umen

ts

The

FFE

of t

he p

rodu

ct a

nd p

roce

ss

deve

lopm

ent

proc

ess

Proj

ect

and

firm

leve

l

The main data collection methods in the first research project were in-depth interviews and documents. A total of 14 interviews were collected: 6 interviews focusing on management control in a collaborative product development project, and 8 interviews within the context of process development. All project members were interviewed, and several internal documents were used in the analysis to find out what control mechanisms were used in collaborative development projects. The analysed documents were the project plan, status reports, and the final project report. Publicly available information, such as annual reports and home pages of the participating companies, were also compared with the data collected, in both case studies, to triangulate the empirical material (Yin, 2003). Interview guides were used in both case studies to ensure that important questions were discussed (see Appendices A and B). Most of the questions were open-ended, which made it possible to capture rich and informative data about the specific research questions. In the second research project, four exploratory interviews were conducted, with the overall aim of providing general background information on innovative activities as well as identifying suitable informants for the subsequent investigation. After the informants were identified, a total of 60 semi-structured interviews were done. Twenty-eight of the interviews focused on process development and the remaining 32 on product development. Interview guides were used in this research project as well (see Appendices C and D). Most of the questions were open-ended, which allowed us to capture rich and informative data about the specific research questions (Miles and Huberman, 1994). The interview guides were modified slightly during the collection of data as we gained more knowledge of what the firms did in the FFE of the innovation process. The data analyses in both research projects were conducted in two different stages: within case analysis and cross-case analysis (Eisenhardt, 1989; Miles and Huberman, 1994). First each case was analysed separately to get a good understanding of the important themes in each case. A case study history was written for each case, typically based on the interviews, field notes, observations, and secondary data. When important information was missing, several shorter telephone interviews were conducted at this stage to retrieve missing information. Each interview was tape-recorded, transcribed, and content-analysed in accordance with Miles and Huberman’s (1994) guidelines. Further, Miles and Huberman (1994) divide qualitative data analysis into three main

37

stages: data reduction and coding, data display, and conclusion drawing and verification. The first stage was to reduce the amount of data by clustering related questions together according to the general theme that was explored. At this stage, data were coded and thematically organized into themes. In the first research project, management control in collaborative product and process development projects, the data were coded in accordance with the theoretical framework. In the second research project, managing the fuzzy front end during product and process development, the empirical data were coded and thematically organized around four main themes, which emerged. Since the research problem of managing the front end when developing non-assembled products is a relatively new topic, especially the front end of process development, patterns were identified rather than confirmed (Edmondson and McManus, 2007; Eisenhardt, 1989). Existing theory was not used to retrieve codes; instead the author team read the transcripts and then we decided jointly on which codes to employ, based on the empirical data. The same codes were used when analysing the FFE in product development as well as process development. The first code addressed questions related to the definition and nature of the FFE, whereas the second code dealt with activities done in the FFE. The third code focused on managerial problems in the FFE, and the fourth code addressed potential success factors of FFE activities. The next step was to perform cross-case analyses, in which we compared the generated codes in the single cases across the cases and looked for similarities (Yin, 2003). Thus, we replicated the finding from one case to the next and searched for similar patterns. The first round of codes in each case was mainly descriptive, and then we went back to the existing literature to interpret the codes. The analyses were thus an iterative process that required repeated reading of the interview files, the notes, the secondary data, and the theoretical framework. We also searched for causal links to discover emerging relationships between the categories (Miles and Huberman, 1994). For example, as we conceptualized the FFE and identified different sub-phases, we related the key activities to these sub-phases. The analysis of the empirical data in this thesis was conducted in a similar way, although information of how the data analyses were conducted is described less in some papers whereas in other papers it is described in more detail. This was

38

done because different journals appear to have different views of what should be emphasized or not.

4.5 Some reflections on the frame of reference

The research conducted in this thesis builds on several different research disciplines. A large part of the literature used in the thesis has its roots in (assembled) product development, but other literature such as general literature on process development is used as well to get an understanding of the specific characteristics of the FFE in non-assembled product and process development. Literature from the field of management control is also used to create an understanding of how process firms can improve the management of product and process development. The underlying assumption in all literatures is that the innovation process can be managed in some way, but different situations require different tactics. In the introduction, three streams of research have been presented that have different underlying assumptions. The first stream in the literature that advocates a structured approach for managing innovation is closely linked to the rational plan perspective (Brown and Eisenhardt, 1995). Within this stream, the innovation process is seen as a predictable series of well-defined steps (Eisenhardt and Tabrizi, 1995). Therefore, a product that is well planned, implemented and supported will become successful. In the second stream of research, the innovation process is seen as an uncertain path through foggy, shifting markets and technology (Eisenhardt and Tabrizi, 1995). Therefore other tactics are deemed more suitable within this stream. This second stream of research resembles the disciplined problem-solving perspective in Brown and Eisenhardt’s article (1995) which evolved from studies of Japanese product development practices. Khurana and Rosenthal (1998) make a similar separation of a formalized and a culture-driven approach to manage the FFE in their article.

4.6 Assessing the quality of the research

In any research, assessing its quality is of major importance and can be done in several ways. Reliability, internal validity, construct validity, and external validity are often used to evaluate the quality of research (Yin, 2003).

39

Reliability is often associated with replication, i.e. that the same results can be obtained if another study is conducted. Replication of a qualitative study will normally not yield the same results, as there can be many different ways of interpreting data (Merriam, 2009). Instead, reliability in a qualitative study should be assessed against whether the results are consistent with the data gathered or not. Validity in a qualitative study is sometimes referred to as trustworthiness (Lincoln and Guba, 1985). Assessing whether the study is trustworthy or not involves answering the question of whether the findings are really about what they appear to be about. Internal validity is described as the extent to which the researcher can establish a causal relationship, whereby certain conditions are shown to lead to other conditions (Yin, 2003). Finally, external validity, also referred to as theoretical generalization, is the term for assessing whether a study’s findings can be generalized beyond the conducted case study. To strengthen reliability in this study, and limit potential biases in the data collection process, several things have been done. First, highly knowledgeable informants were selected, all with different roles and perspectives (e.g. project engineers, project managers, and top managers), to mitigate potential biases in the data collection process. Second, when writing co-authored papers, discussions among the author team (in papers III and IV) were conducted to create overlap between data collection and analysis. Third, a case study protocol was constructed, in the second research project, to increase reliability of the study. This was especially important as data were collected by three persons. Pattern matching is an important tactic to increase internal validity in explanatory or causal studies (Yin, 2003). The case studies in this thesis could be categorized as exploratory and/or descriptive, and thus pattern matching was not used. Patterns needed to be explored rather than confirmed (Eisenhardt and Graebner, 2007). Triangulation among different data sources was conducted to increase the (construct) validity (Yin, 2003). In the first research project focusing on management control in innovative projects, several different sources were collected. Interviews, internal documents, and external information mainly from newspapers and the firms’ websites were triangulated. In the second research project, all the above data sources were used as well as presenting the research results to representatives from all firms at several workshops. To further strengthen the construct validity of the study, a clear research framework was

40

designed and extensively discussed among the author team, prior to the data collection process in the second research project focusing on the FFE of innovation. To ensure consistency among the informants’ taped responses, several shorter telephone calls were made to follow up points that were unclear (Miles and Huberman 1994; Yin, 2003). Finally, multiple case studies strengthen the external validity, as this approach is suitable when the aim is to build new theory (Eisenhardt and Graebner, 2007). The fact that all cases were selected within the same industry is beneficial as it is easier to see clear patterns and make valid conclusions, although this also implies some limitations (see section 6.4).

4.7 The research process and the background to the appended papers

The process of writing a PhD thesis is similar to the product development process in many ways, as it often involves several different iterative activities. In the beginning the uncertainty is high about which ideas to go forward with and which ideas should be dropped. I started my PhD studies in the spring of 2004 in Accounting and Control. The main focus of interest at that time was on management control in collaborative development projects. In 2006 I finished the first half of my PhD studies and wrote a licentiate thesis (Kurkkio, 2006). When writing the licentiate thesis I found that the early stages of the innovation process were the most challenging to manage, and therefore they became the focus of attention in the second part of my studies. I started the second part in the autumn of 2007 and became a member of Promote – The Centre for Management of Innovation and Technology in Process Industries. The second research project was carried out within Promote and therefore I began to look more into the field of innovation management and combine it with my previous knowledge about management control. Four of the papers are written in the second part of my studies and the idea behind each individual paper is now presented. The background to paper II was that I was interested in finding out more about process development. The initial idea was to focus on more radical process development, referred to as process innovation, because the literature about process development was scattered and it was difficult to know whether this literature should be included or not. Initially I was also interested in limiting the search to the FFE of process innovation, but as I started the search for relevant

41

literature I found out relatively quickly that few papers existed on the front end of process innovation. Thus, we decided to focus on the whole “process innovation process” in the second paper. The focus here was on informing an understanding of antecedents and consequences of firms’ process innovation capability. This was an important step before I could turn the focus towards the FFE. The idea behind the third paper was also grounded in the existing literature. Even though several articles have been published on the FFE of product development, few papers address the special conditions that process firms face when they develop non-assembled products. This paper is based on empirical data from Höganäs AB, LKAB and SSAB. The fourth paper was written at the same time as the third paper, and the ideas for papers III, IV and V were all based on the existing literature and the gaps that were identified. In the fourth paper the focus was directed towards process development again. The limited attention towards the FFE of process development made it important to start with a conceptualization of the FFE. To find out more about when the FFE begins, when it ends and what key activities that are carried out. Managerial problems and suggestions of how to mitigate them have also been provided. Finally, the idea in the last paper was to compare the FFE of product and process development. Several researchers argue that more research is needed on other types of innovations than product development to increase our understanding of the innovation process. Therefore in paper V the FFE of (non-assembled) product development and the FFE of process development have been compared with each other to see similarities and differences between the two different contexts. Table 5 summarizes all interviews collected and it shows in which papers the different case studies are presented.

Table 5 An overview of the case studies and in which papers they are presented Number of interviews Main focus Used in paper

Case study 1 6 The whole product development process

Paper I

Case study 2 8 The whole process of process development

Paper I

Case study 3 36 The FFE of product development

Papers III and V

Case study 4 32 The FFE of process development

Papers IV and V

42

5. SUMMARY OF THE APPENDED PAPERS

This section gives a short overview of the papers in part II in this thesis. The core essence of the five appended papers is summarized in this section. The purpose is first presented and the methodological approach is described, followed by the main empirical findings of each paper, and finally the main conclusions are presented.

5.1 Paper I: Control mechanisms in collaborative R&D projects: the case of supplier and customer relationships in process industry

This first paper addresses management of collaborative product and process development projects. The aim of the paper was to identify and analyse what motives collaborating firms have when they engage in collaborative R&D projects (i.e. product and process development projects) and how these motives subsequently affect the selection of control mechanisms. The study had a qualitative research approach and two case studies were conducted. The first case study was a collaborative product development project between one customer and a supplier. The second case study was a collaborative process development project with several different partners involved. In both case studies, interviews were the main data source. In the product development project, internal documents such as project plans, status reports and the final project report were also collected and analysed. As no such data were available in the second case study, only interviews were collected and analysed. Each case was analysed separately and then a cross-case analysis was made. The main findings can be summarized by the fact that all case study firms had a variety of motives which affect the selection of (formal and informal) control mechanisms. The firms’ different motives were categorized into financial or relational, where the former was exemplified by the possibility of reducing development costs and sharing risks, and the latter was exemplified by building customer loyalty and sharing knowledge between the firms. In both cases a mix of formal and informal control mechanisms was used to manage collaborative product and process development projects. This paper contributes to the Management Control literature as it extends Dekker’s (2004) study and shows how the different motives affect the selection of control mechanisms used in collaborative product and process development

43

projects. Furthermore, this paper adds insights to the debate in the literature regarding the role of management control in the innovation process. In this paper management control is described as a system consisting of both formal and informal control mechanisms, and thus management control is seen as a supporting tool when developing new products and processes.

5.2 Paper II: Antecedents and Consequences of Firms’ Process Innovation Capability: A literature review and conceptual framework

Similar to product innovation, process innovation is a key source of long-term competitive advantage, yet few empirical studies address process innovation compared to the domain of product innovation. The literature on managing process innovation has not until now been systematically reviewed in the academic literature. The objective of this paper is to address a problem shared by both practising managers and the research community: the issue of how to manage process innovation so that its positive benefits can be better realized. The paper is a conceptual study, and a capability-based perspective is used to synthesize and organize the literature on managing process innovation. Inspired by research into absorptive capacity and open innovation, we make a distinction between a firm’s potential and realized process innovation capability. Here it is argued that proficiency in realization is critical to achieve process innovation efficiency. This has been disregarded in the existing literature, which mostly describes how firms can increase their potential process innovation capability. Antecedents and consequences of a firm’s process innovation capability are identified and discussed, which allows more proficient management of process innovation. By identifying the antecedent factors we highlight the key domains which tend to determine the potential process innovation capability of firms. The antecedent factors are: investment in new process technology, alignment between process innovation and strategy, top management support, product and process innovation integration, project selection and portfolio balancing, collaboration with external partners, collaboration among functions and departments, organizational climate and culture, organizational learning, and quality of relationship between labour and management.

44

By synthesizing the literature on managing process innovation, we address a knowledge gap in the literature regarding process innovation. By highlighting the difference between firms’ potential and realized process innovation capability, we argue that proficiency in realizing process innovation is essential for attaining the desired outcomes.

5.3 Paper III: The Fuzzy Front End in Non-assembled Product Development: A multiple case study of mineral and metal firms

Existing research on the FFE in product development has mainly focused on assembled products, whereas this paper explores the FFE in a new context, the FFE in non-assembled product development. Firms that develop and produce non-assembled products prevail within the food, minerals and metals, pulp and paper, and textile industries among others. Developing non-assembled products requires different characteristics than when assembled products are developed. This makes existing research on the FFE difficult to transfer without adjustments. The purpose of this paper is to increase the understanding of the FFE in non-assembled product development. Three case studies within the minerals and metals industry were conducted. Semi-structured interviews were the main data source, and internal and external documents were also used in the analyses of the case study firms. The analyses show that the fuzzy front end activities and managerial challenges in non-assembled product development diverge significantly from previously reported findings in the Technology and innovation management literature. First, we conceptualized the front end of non-assembled product development and found it to be an iterative trial-and-error process consisting of three sub-phases, dominated by activities such as analysis of input materials, identification of process constraints, laboratory tests, and batch tests with customers. Further, we identified six key managerial challenges for increasing both the speed and quality of the FFE process. These were integration of product and process development, cross-functional collaboration, senior management support, project management, a creative climate and culture, and a holistic perspective on development activities.

45

This study extends previous research on the FFE in product development, by showing that it is important to address idiosyncratic conditions to get a better understanding of the FFE. This paper contributes to the existing literature by adding increased understanding of the FFE in non-assembled product development. In more detail, several key activities in the FFE of non-assembled product development are identified, which have not been reported before. Anticipating the needs of the customer’s production process, analysis of input materials, and identification of the “process window” are all new insights important for process firms to address in the FFE of non-assembled product development. Other central activities in the FFE were conducting proficient literature reviews, laboratory tests, pilot plant tests, and batch tests. None of these activities has been reported in the extant literature. To summarize, we show that the FFE idiosyncrasies cannot fully be explained by prior research, and hence additional research about the FFE in non-assembled product development and other contexts is needed.

5.4 Paper IV: Where process development begins: a multiple case study of fuzzy front end activities in process firms

Despite the positive effects of process development on manufacturing performance and overall firm performance, knowledge about the front end activities which precede formal process development is limited. The fuzzy front end is known to be critical to overall product development success, but few if any studies have examined the front end in process development. Hence, the overall purpose of this paper is to increase knowledge about the FFE in process development in process firms. We conducted a multiple case study of four process firms within the minerals and metals industry. Semi-structured interviews were conducted and internal as well as external documents were collected. Employees working with process development from different perspectives and with different roles were selected to get a diverse view of how the FFE of process development was carried out and managed. Our results show that substantial differences in front end activities exist between the product and process development domains. We conceptualize the front end in process development to be an iterative trial-and-error process, dominated by

46

activities such as idea generation and refinement, literature reviews, anticipation of end-product changes, and various forms of experiments in bench scale, lab scale, and full-scale production. In addition, we highlight key problems in the FFE process, and managerial remedies for how to reduce them. The FFE was typically divided into four different sub-phases: informal start-up, formal idea-study, formal pre-study and the formal pre-project and each sub-phase included several different key activities. This paper provides new insights into the work process of process development. A first contribution is that we offer a descriptive framework that managers and researchers can use to understand the innovation activities carried out in the FFE of process development. Some of the key activities have not been reported in the existing literature such as anticipation of end-product changes and construction or modification of process equipment. Thus, our paper provides details of what actually happens in the fuzzy front end of process development. Another contribution is that we present empirical evidence of managerial problems in the FFE of process development. These were the reactive nature of process development, ad-hoc documentation of tests, early production involvement, and lack of time and resources. Contingency factors, such as degree of newness, state of existing knowledge, sources of ideas for process development, scope of the process development project, and centre of gravity of the process development project, may affect how the FFE is organized. A third contribution is that process development practices seem to be essential for achieving high product development performance, as product and process development are complementary.

5.5 Paper V: Managing the fuzzy front end: Insights from Process Firms

This final paper in the thesis compares the FFE of non-assembled product development and process development. More research that explores different types of industries as well as types of innovations has been called for. This paper addresses two research gaps in the existing literature which both refer to the one-sided focus on (assembled) product development. The paper compares the FFE of non-assembled product development against process development and similarities and differences are searched for. The aim of this paper is to inform researchers and practitioners about the fuzzy front end (FFE) of the innovation process in process firms.

47

A multiple case study was conducted at four process firms. The main data source was a total of 64 interviews with product and process employees. The conceptualizations showed that the FFE starts similarly in both cases, but the perhaps most central difference was the following stage in the innovation process. When products were developed, the next stage was typically formal development which included more tests and validations of the product concept (Cooper, 2008), whereas implementation in the production line was the stage after the FFE in process development (e.g. Lager et al., 2010). The process concept was instead implemented in the production process. This explains to some extent why the FFE in process development includes one extra sub-phase, compared to non-assembled product development. By comparing the FFE of non-assembled product development with process development, new insights emerge about what constitutes the FFE in process development and non-assembled product development. Some similarities in terms of key activities and managerial problems were found. However, the differences in terms of objectives, key activities and underlying logics suggest that extant knowledge of the FFE in product development is not applicable to the FFE of process development without adjustments. Further, there is a strong relationship between product development and process development that needs to be addressed already in the FFE.

48

6. DISCUSSION

This thesis aims to contribute to the fields of innovation management and management control. In this final chapter theoretical as well as managerial implications of the thesis are elaborated upon, followed by conclusions and limitations. Finally, directions for future research are suggested.

6.1. Theoretical implications

The first research question – What control mechanisms are important to consider when managing collaborative product and process development in process firms? – is addressed in paper I in the thesis. This paper adds new insights into how firms can manage collaborative product and process development across firm boundaries. Paper I shows that a combination of formal and informal control mechanisms was used to manage the collaborative development projects. A central contribution of this paper is that it explores how the firms’ motives affect the selection of formal and informal control mechanisms. For example, financial short-term motives (such as sharing development costs) were typically controlled by formal control mechanisms. On the other hand, informal control mechanisms were the main form of control if the motive was to build long-term relations to their customers. This paper extends prior research by Dekker (2004) as it clarifies how the motives affect the selection of control mechanisms. The literature on collaborative projects is immense. However, only a couple of researchers have studied accounting and control issues in collaborative projects, and more research has been called for (Dekker, 2004; Kamminga and Van der Meer-Kooistra, 2007). Further, two different development projects have been studied (a product development project and a process development project), and in both projects the firms used a mix of formal and informal control mechanisms. The second research question was: What are the critical factors for managing process innovation? This question was addressed in paper II. The main contribution in this paper is the identification of 10 antecedent factors (investment in new process technology, alignment between process innovation and strategy, top management support, product and process innovation integration, project selection and portfolio balancing, collaboration with external partners, collaboration among functions and departments, organizational climate and culture, organizational learning, and quality of relationship between labour and management) of firms’ potential process innovation capability. Then we make a distinction between

49

firms’ potential and realized process innovation capability, in order to give a more realistic account of how firms can achieve their desired outcomes and become more efficient. The existing literature on managing process innovation is scattered, and the existing models of process innovation are often very abstract and not very detailed (e.g. Lim et al., 2006; Pisano, 1997). This paper contributes to the literature on process innovation by synthesizing existing research on managing process innovation and conceptualizing how process innovation efficiency is enabled. The theoretical model offers new and more detailed insights into how firms can manage process innovation. The third research question was: How can the FFE of product and process development be conceptualized in process firms? The answer to this question is found in papers III, IV and V, all addressing the FFE of non-assembled product and process development. The two models presented in papers III and IV take a relatively detailed look at the FFE of product and process development in process firms. The FFE of product and process development is divided into different sub-phases. The FFE of product development entails three sub-phases: informal start-up, formal idea study and pre-study. The FFE of process development is divided into four sub-phases: informal start-up, formal idea-study, formal pre-study and formal pre-project. The most central difference was the ending of the FFE. The following stage when products were developed was formal development, which included more tests and validations of the product concept (Cooper, 2008). The next stage when conducting process development was implementation in the production process (e.g. Lager et al., 2010), and no more pre-tests were done. Instead the process concept was tested in the actual production process, and this explains to some extent why the FFE in process development includes one extra sub-phase, compared to the FFE in product development. In the existing literature, several conceptualizations of the FFE are offered (e.g. Griffiths-Hemans and Grover, 2006; Khurana and Rosenthal, 1998; Koen et al., 2001). However, all of them are based on assembled product development. Nobelius and Trygg (2002) have criticised a large part of the FFE research for presenting one “optimal” FFE stage, suitable in all different contexts. This thesis provides empirical evidence of what constitutes the FFE in two other contexts: non-assembled products and process development. A general conclusion is that the FFE of different types of innovations cannot be conceptualized in a similar way.

50

The fourth research question was: How can process firms manage the FFE of product and process development? This question was addressed in papers III, IV and V. To answer the question, key activities, managerial problems and success factors in the FFE have been identified. The focus in the papers has been on both the operative level (key activities) and the strategic level (managerial problems and success factors). Previous research has mostly focused on management from the strategic level, for example the research done by Cooper and others (Cooper, 1988; Cooper and Kleinschmidt, 1987: Khurana and Rosenthal, 1997). Another critical aspect is that most of the research on the front end is based on assembled product development (e.g. Murphy and Kumar, 1997; Seidel, 2007). This thesis contributes by exploring other contexts such as non-assembled product and process development. The findings in this thesis (in papers III and IV) show how firms can integrate strategic and operative activities and achieve what Khurana and Rosenthal (1998) call a holistic front end. On a more general level, a contribution made in this thesis is that it combines the management control literature and the innovation management literature. There is an ongoing discussion in both literatures regarding whether management control is suitable for managing innovation projects. In the introduction three streams of research with different views on the benefits and risks of controlling innovative projects have been outlined. The findings in this thesis are in line with the third stream of research, which argues for a balanced approach involving both control and creativity for managing innovation (e.g. Chiesa et al., 2009; Simons, 1994). A question posed in the introduction was: what are the tactics that balance control and creativity? In the FFE of process development, the tactics for increasing speed and quality and simultaneously granting conditions for creativity were: early involvement of production staff and cross-functional teamwork, while integration of product and process development, cross-functional teamwork, senior management support, an effective project manager, and a holistic perspective were the tactics in non-assembled product development. Many of the managerial problems found in this thesis suggest that the FFE would benefit from becoming more structured. This is in line with the main stream of research in the FFE literature (e.g. Cooper, 1988; Khurana and Rosenthal, 1998). Although the empirical findings suggest a more structured process, it is important to find the appropriate amount, as too much seems to cause rigidity and inflexibility while too little makes the FFE of innovation both ineffective and inefficient.

51

6.2 Managerial implications

Several recommendations for how firms can improve the management of product and process development are derivable, based on the results of this thesis. A first recommendation is that managers should encourage and control the conduct of specific issues and FFE activities. This does not mean that the employees are told how they should develop new ideas and how they should turn them into product or process concepts. Instead, managers should communicate with the development employees so that they know what is expected of them, and inform them so that the decision-making responsibilities are clear (Khurana and Rosenthal, 1998; Simons, 1995).This thesis has identified several problems in the FFE of product and process development. Some typical problems were that the rigour and proficiency by which the FFE activities were conducted varied a lot, and documentation of test results was often conducted in an ad-hoc manner. These problems could be reduced if the firms introduced a more structured work process for developing new products and processes. This is in line with one of the approaches Khurana and Rosenthal (1998) suggested to create a holistic front end. They found that a formalized work process can bring order to chaotic activities present in the FFE. On this basis, managers may reduce the uncertainty and equivocality among organizational participants. While it would be impossible and undesirable to reduce all fuzziness from the FFE, stronger formalization appears to be beneficial for the studied development projects. Another recommendation is that managers should encourage informal activities, since many new ideas for products and processes were developed outside the formalized project models. Although the description of the FFE in this thesis is sometimes very detailed and clear, it is important to remember that the development often is a fuzzy, trial-and-error process. Many of the activities in this stage are pursued more or less informally. An additional recommendation is that managers should align FFE activities with the firm’s overall strategy. Managers should communicate the firm’s values and beliefs to the development team in the FFE, thereby empowering engineers, encouraging new ideas and product and process concepts, and showing a tolerance of failures. This is in line with previous research showing that it is important that managers communicate the values and missions of the firm to accomplish strategic renewal (Simons, 1995).

52

Moreover, two more specific recommendations that are important both for improving the management of product and process development were to facilitate cross-functional collaboration and to encourage an entrepreneurial organization culture (also see papers III and IV). Finally, managers are encouraged to strengthen the relationship between product development and process development. Previous research has shown that this is important for improving both product and process development in process firms (Lager, 2002a; Lim et al. 2006; Linton and Walsh, 2008), as have several of the appended papers (III, IV and V). Most process firms are geographically dispersed, which makes it difficult to establish close integration between product development and process development – especially in the FFE, as the product and process engineers seldom worked together. Managers should focus on strengthening the interpersonal relationships between product and process engineers to overcome rigidity caused by formalized structures. Then problems and mistakes could be addressed early on in the innovation process. One possibility is to create new departments with employees from both product and process development. The recommendations above are mostly directed towards managers at the strategic level. Some recommendations important for other professions working with developing new products and processes within process firms, such as project managers, product engineers, process engineers, and production managers, have been found as well. A recommendation directed at the operative level is that it is important to clarify which activities need to be addressed in the FFE of product and process development. The conceptualization and identification of key activities in the FFE for both product and process development can be used as checklists. These conceptualizations help to make the FFE in product and process development more explicit, which is a first important step for improving the management of the FFE of innovation. Finally, the empirical findings in this thesis suggest that having an appropriate project leader in the FFE was considered to be important for success in the FFE. In the FFE of product development, this was described as decisive for both speed and quality of FFE activities. An effective project manager was described as being

53

well connected internally, possessing good communication skills, able to secure resources for development, and able to motivate and inspire other team members. At the same time, the project leader must have the ability to make tough decisions which are not always to the satisfaction of persons nearby.

6.3. Conclusions

Both research and practice will benefit from obtaining an increased understanding of the stated purpose. The general conclusions reached in this thesis are based on the main findings found in the appended papers. Summarizing the results from the appended papers, two conclusions can be drawn. The first one is that both product development and process development would benefit from more structured work processes in the FFE. This is found to be crucial in all of the appended papers. However, the difficulty is to find the appropriate level of control, as too much can inhibit innovation and too little results in inefficient use of the firm’s resources. The second conclusion is that while the results in this thesis show some similarities to FFE activities in product development, they stress that the FFE of process development cannot be managed and organized solely according to the principles for product development.

6.4. Overall limitations

Some limitations of the research conducted within this thesis are worth discussing. First, all empirical data are collected from the minerals and metals industry, and therefore care should be taken in generalizing the findings to other contexts. As all cases are situated within the same industry, the findings as a whole should be considered in their contextual setting, but some implications may very well be extended to similar process firms and other manufacturing firms. The selection of firms from one industry makes the analysis of the empirical data more valid for process firms, but it limits theoretical generalization to other contexts (Eisenhardt and Graebner, 2007). For example, the integration of product and process development, which is a key concern in process firms, may be critical in other manufacturing firms as well. However, managers in these manufacturing firms must interpret the findings in this thesis and apply them to their specific context. Second, the sampling strategy pursued in both research projects is theoretical sampling with replication logic (Yin, 2003). This is a limitation and, to further validate and strengthen the findings found in this thesis, other sampling strategies

54

such as contrary replication or sampling according to polar types (Eisenhardt, 1989) are recommended. Third, the qualitative approach is often criticised for relying on retrospective interviews, which can be problematic if informants forget and miss interpret information (Eisenhardt and Graebner, 2007). However, despite the critiques against the qualitative approach, this method was chosen as it ensured methodological fit among research questions, data collection, data analysis and status of current theory (Edmondson and McManus, 2007). Triangulation and selection of several informants with different perspectives and positions were carried out to increase the quality of the research.

6.5. Suggestions for future research

The suggestions for future research have been identified in several different areas of research. First, the FFE of process development should be explored further, as the work in this thesis needs to be validated in other contexts. Other manufacturing firms than process firms could be explored, and polar cases could be used as a sampling strategy to strengthen the findings in this thesis. The FFE of product development is still in its infancy, compared to research on the whole product development process, and as a result many of the papers are conceptual or exploratory case studies. Only recently have researchers begun to consider the importance of addressing different contingencies such as degree of newness, type of innovation project, and industry etc. Still, more empirical research with different contingencies is needed to understand more about how the FFE phase of the innovation process can be managed. Third, more research is needed to understand the role of management control when managing product and process development. A few studies have addressed this area (Chiesa et al., 2009; Poskela and Martinsuo, 2009), but it is suggested that more attention should be directed towards finding out how firms can achieve a balance between control and creativity. Finally, this thesis has focused on how the overall FFE is managed and organized in process firms, and the theoretical framework is mainly drawn from the management control literature and innovation management literature. However, other aspects such as organizational knowledge creation and organizational learning in the FFE probably also affect the performance of the FFE (e.g. Schulze and Hoegl, 2008). Thus a suggestion for future studies is to look into how

55

knowledge creation and organizational learning can be supported in the FFE of the innovation process to increase speed and quality in the FFE.

56

7. REFERENCES

Abernathy, M. and Brownell, P. (1997). Management control systems in research and development organizations: the role of accounting, behaviour and personnel control. Accounting, Organization and Society, 22(3-4): 233-248.

Alam, I. (2006). Removing the fuzziness from the fuzzy front-end of service innovations through customer interactions. Industrial Marketing Management, 35(4): 468-480.

Amabile, T.M. (1998). How to kill creativity. Harvard Business Review, 76(5): 77-87.

Anthony, R.N. (1965). Planning and Control Systems: A Framework for Analysis. Division of Research, Boston, Harvard Business School.

Bacon, G., Beckman, S., Mowery, D. and Wilson, E. (1994). Managing product definition high-technology industries: A pilot study. California Management Review, 36(3): 32-56.

Baer, M. and Frese, M. (2003). Innovation is not enough: climates for initiative and psychological safety, process innovations, and firm performance. Journal of Organizational Behaviour, 24 (1): 45-68.

Barnett, B.D. and Clark, K.B. (1996). Technological newness: an empirical study in the process industries. Journal of Engineering and Technology Management, 13(3-4): 263-282.

Björk, J. and Magnusson, M. (2009). Where Do Good Innovation Ideas Come From? Exploring the Influence of Network Connectivity on Innovation Idea Quality. The Journal of Product Innovation Management, 26(6): 662-670.

Boddy, D. Macbeth, D. and Wagner, B. (2000). Implementing collaboration between organizations: An empirical study of Supply Chain Partnering. Journal of Management Studies, 37(7): 1003-1017.

Bonner, J.M., Ruekert, R.W. and Walker, O.C. (2002). Upper management control of new product development projects and project performance. Journal of Product Innovation Management, 19(3): 233-245.

Brem, A. and Voigt, K-I. (2009). Integration of market pull and technology push in the corporate front end and innovation management-Insights from the German software industry. Technovation, 29(5): 351-367.

Brown, S.L. and Eisenhardt, K.M. (1995). Product development: Past research, present findings, and future directions. Academy of Management Review, 20(2): 343-378.

Bröring, S. (2005). The Front End of Innovation in Converging Industries: The case of Nutraceuticals and Functional Foods. Wiesbaden, Germany: Deutscher Universitäts-Verlag.

Bröring, S., Cloutier, M.L. and Leker, J. (2006). The front end of innovation in an era of industry convergence: evidence from nutraceuticals and functional foods. R&D Management, 36(5): 487-498.

57

Burrell, G. and Morgan, G. (1979). Sociological paradigms and organizational analysis. London, UK: Heinemann Educational Books.

Chang, S-L., Chen, C-Y. and Wey, S-C.(2007). Conceptualizing, assessing, and managing front-end fuzziness in innovation/NPD projects. R&D Management, 37(5): 469-478.

Chenhall, R.H. (2003). Management control systems design within its organizational context: findings from contingency-based research and directions for the future. Accounting, Organization and Society, 28(2-3): 127-168.

Chiesa, V., Frattini, F., Lamberti, L. and Noci, G. (2009). Exploring management control in radical innovation projects. European Journal of Innovation Management, 12(4): 416-443.

Collins, P.D., Hage, J. and Hull, F.M. (1988). Oragnizational and Technological Predictors of Change in Automaticity. Academy of Management Journal, 31(3): 512-543.

Cooper, R.G. (1988). Predevelopment activities determine new product success. Industrial Marketing Management, 17(3): 237-247.

Cooper, R.G. and Kleinschmidt, E.J. (1987). New products: What separates winners from losers? Journal of Product Innovation Management, 4(3): 169-184.

Cooper, R.G. (2008). Perspective: The stage-gate idea-to-launch process. Update, what’s new, and Nextgen systems. Journal of Product Innovation Management, 25(3): 213-232.

Daft, R.L. (1978). A dual core model of organizational innovation. Academy of Management Journal, 21(2): 193-210.

Davila, T. (2000). An empirical study on the drivers of management control systems’ design in new product development. Accounting, Organization and Society, 25(4-5): 383-409

Dekker, H. (2004). Control of inter-organizational relationships: evidence on appropriation concerns and coordination requirements. Accounting, Organization and Society, 29(1): 27-49.

Dennis, D.R. and Meredith, J.R. (2000). An Empirical Analysis of Process Industry Transformation Systems. Management Science, 46(8): 1085-1099.

Drucker, P.F. (1998). The Coming of the New Organization. Harvard Business Review, 66(1): 45-53.

Dubin, R. (1982). Management: Meanings, Methods, and Moxie. The Academy of Management Review, 7(3): 372-379.

Edmondson, A.C. and McManus, S.E. (2007). Methodological fit in management research. Academy of Management Review, 32(4): 1155-1179.

Elmquist, M. and Segrestin, B. (2007). Towards a New Logic for Front End Management: From Drug Discovery to Drug Design in Pharmaceutical R&D. Creativity and Innovation Management, 16(2): 106-120.

Eisenhardt, K.M. (1989). Building Theories from Case Study Research. Academy of Management Journal, 14(4): 532-550.

58

Eisenhardt, K.M. and Graebner, M.E. (2007). Theory building from cases: opportunities and challenges. Academy of Management Journal, 50(1): 25-35.

Eisenhardt, K.M. and Tabrizi, B.N. (1995). Accelerating Adaptive Processes: Product Innovation in Global Computer Industry. Administrative Science Quarterly, 40(1): 84-110.

Ettlie, J.E., Reza, E.M. (1992). Organizational integration and process innovation. Academy of Management Journal, 35(4): 795-827.

Frishammar, J. and Hörte, S.Å. (2005). Managing External Information in Manufacturing Firms: The Impact on Innovation Performance. Journal of Product Innovation Management, 22(3): 251-266.

Garcia, R. and Calantone, R. (2002). A critical look at technological innovation typology and innovativeness terminology: a literature review. The Journal of Product Innovation Management, 19(2): 110-132.

Goffin, K. and Mitchell, R. (2005). Innovation Management: Strategy and Implementation Using the Pentathlon Framework. Basingstoke, Palgrave Macmillan

Gopalakrishnan, S. and Damanpour, F. (1997). A review of innovation research in economics, sociology and technology management. Omega, 25(1): 15-28.

Gopalakrishnan, S., Bierly, P. and Kessler, E.H. (1999). A reexamination of product and process innovations using a knowledge-based view. Journal of High Technology Management, 10(1): 147-166.

Griffith-Hemans, J. and Grover, R. (2006). Setting the stage for Creative new products: Investigating the idea fruition process. Journal of the Academy of Marketing Science, 44(1): 27-39.

Guba, E.G. and Lincoln, Y.S. (2000). Competing paradigms in Qualitative Research. Handbook of Qualitative Research (Eds.) Denzin N.K. and Lincoln, Y.S., London. Sage Publications.

Hutcheson, P., Pearson, A.W. and Ball, D.F. (1995). Innovation in process plant: a case study of ethylene. Journal of Product Innovation Management, 12(5): 415-430.

IVA (2006). Increased competitiveness in the Swedish processing industry. Stockholm, Royal Swedish Academy of Engineering Sciences.

Jönsson, S. (2004). Product Development – Work for Premium Values. Kristianstad, Liber & Copenhagen Business School Press.

Kamminga, P.E. and Van der Meer-Kooistra, J. (2007). Management control patterns in joint venture relationships: A model and an exploratory study. Accounting, Organizations and Society, 32(1-2): 131:154.

Khazanchi, S., Lewis, M. and Boyer, K. (2007). Innovation-supportive culture: the impact of organizational values on process innovation. Journal of Operations Management, 25(4): 871-884.

Khurana, A. and Rosenthal, S.R. (1997). Integrating the fuzzy front end of new product development. Sloan Management Review, 38(2): 103-120.

Khurana, A. and Rosenthal, S.R. (1998). Towards holistic "front ends" in new product development. Journal of Product Innovation Management, 5(1): 57-74.

59

Kijkuit, B. and van den Ende, J. (2007). The Organizational Life of an Idea: Integrating Social Networks, Creativity and Decision-Making Perspectives. Journal of Management Studies, 44(6): 863-882.

Kim, J. and Wilemon, D. (2002). Focusing the Fuzzy Front-end in new product development. R&D Management, 32(4): 269-279.

Kirsch, L.J. (1996). The Management of Complex Tasks in Organizations: Controlling the Systems Development Process. Organization Science, 7(1): 1-21.

Koen, P., Ajamian, G., Bukart, R., Clamen, A., Davidson, J., D’Amore, R. Elkins, C., Herald, K., Incorvia, M. Johnson, A., Karol, R., Seibert, R., Slavejkov, A. and Wagner, K. (2001). Providing clarity and a common language to the “fuzzy front end”. Research Technology Management, 44(2): 46-55.

Kurkkio, M. (2006). Styrning i ett tvärfunktionell FoU-projekt. Licentiat uppsats, Luleå tekniska universitet.

Lager, T. (2000). A new conceptual model for the development of process technology in process industry. International Journal of Innovation Management, 4(3): 319-346.

Lager, T. (2002a). Product and process development intensity in process industry: A conceptual and empirical analysis of the allocation of company resources for the development of process technology. International Journal of Innovation Management, 6(2): 105-130.

Lager, T. (2002b). A structural analysis of process development in process industry – A new classification system for strategic project selection and portfolio balancing. R&D Management, 32(1): 87-95.

Lager, T., Hallberg, D. and Ericsson, P. (2010). Developing a process innovation work process: The LKAB experience. International Journal of Innovation Management, 14(2): 285-306.

Lim, L., Garnsey, E., and Gregory, M. (2006). Product and process innovation in biopharmaceuticals: A new perspective on development. R&D Management, 36(1): 27-36.

Linton, J.D. and Walsh, S.T. (2008). A theory of innovation for process-based innovations such as nanotechnology. Technological Forecasting & Social Change, 75(5): 583-594.

Macher, J.T. and Mowery, D.C. (2003). Managing” learning by doing: an empirical study in semiconductor manufacturing. Journal of Product Innovation Management, 20 (5): 391-410.

Merriam, S.B. (2009). Qualitative research: a guide to design and implementation. United Kingdom, John Wiley and Sons Ltd.

Miles, M. and Huberman, M. (1994). Qualitative data analysis. Thousand Oaks: Sage Publications.

Moenart, R.K., De Meyer, A., Souder, W.E. and Deschoolmeester, D. (1995). R&D/Marketing Communication During the Fuzzy Front-End. IEEE Transactions on Engineering Management, 42(3): 243-258.

60

Montoya-Weiss, M. and Calantone, R. (1994). Determinants of new product performance: A review and meta-analysis. Journal of Product Innovation Management, 11(5): 397-417.

Montoya-Weiss, M. and O'Driscoll, T. (2000). From experience: Applying performance support technology in the fuzzy front end. Journal of Product Innovation Management, 17(2): 143-161.

Murphy, S. and Kumar, V. (1997). The front end of new product development: A Canadian survey. R&D Management, 27(1): 5-15.

Nixon, B. (1998). Research and development performance measurement: a case study. Management Accounting Research, 9(3): 329-355.

Nobelius, D. and Trygg, L. (2002). Stop chasing the Front End process – management of the early phases in product development projects. International Journal of Project Management, 20(5): 331- 340.

Otley, D., Broadbent, J. and Berry, A. (1995). Research in Management Control: An Overview of its Development. British Journal of Management, 6(1): 31-44.

Ouchi, W. G. (1979). A Conceptual Framework for the Design of Organizational Control Mechanisms. Management Science, 25(9): 833-848.

Pearce, C.L. and Ensley, M. D. (2004). A reciprocal and longitudinal investigation of the innovation process: the central role of shared visions in product and process innovation teams (PPITs). Journal of Organizational Behaviour, 25 (2): 259-278.

Perks, H. (2000). Marketing information exchange mechanisms in collaborative new product development: the influence of resource balance and competitiveness. Industrial Marketing Management, 29(2): 179-189.

Pisano, G. (1997). The development factory: unlocking the potential of process innovation. Boston, HBS Press.

Poskela, J. and Martinsuo, M. (2009). Management Control and Strategic Renewal in the Front End of Innovation. Journal of Product Innovation Management, 26(6): 671-384.

Prajogo, D.I. and Sohal, A.S. (2006). The relationship between TQM practices, quality performance, and innovation performance: an empirical examination. International Journal of Quality and Reliability Management, 20(8): 901-918.

Reichstein, T. and Salter, A. (2006). Investigation the sources of process innovation among UK manufacturing firms. Industrial and Corporate Change, 15(4): 653-682.

Reid, S.E. and de Brentani, U. (2004). The fuzzy front end of new product development for discontinuous innovations: A theoretical model. Journal of Product Innovation Management, 21(3): 170-184.

Richtnér, A. and Åhlström, P. (2010).Top management control and knowledge creation in new product development. International Journal of Operations and Production Management, 30(10): 1006-1031.

61

Schulze, A. and Hoegl, M. (2008). Organizational knowledge creation and the generation of new product ideas: A behavioral approach. Research Policy, 37(10): 1742-1750.

Seidel, V.P. (2007). Concept shifting and the radical product development process, Journal of Product Innovation Management, 24(6): 522-533.

SGU. (2008). Investigation on behalf of the government: Mapping Sweden’s ore and mineral resources with the purpose of developing a knowledge base. Dnr 0-1329/2008.

Simons, R. (1994). How new top managers use control systems as levers of strategic renewal. Strategic Management Journal, 15(3): 169-189.

Simons, R. (1995). Levers of control: how managers use innovative control systems to drive strategic renewal, Boston MA, Harvard Business School Press

Smith, P.G. and Reinertsen, D.G. (1998). Developing Products in Half the Time – New rules, New York, John Wiley and Sons Inc.

Song. M.X. and Montoya-Weiss, M.M. (1998). Critical development activities for really new versus incremental products. Journal of Product Innovation Management, 15(2): 124-135.

Sundgren, M. (2004). New thinking, management control & instrumental rationality: Managing organizational creativity in pharmaceutical R&D. Dissertation, Chalmers University of Technology.

Tidd, J. (2001). Innovation management in context: Environment, organization and performance. International Journal of Management Reviews, 3(3): 169-183.

Trott, P. (2005). Innovation Management and New Product Development, Harlow, Financial Times Prentice Hall.

Tushman, M.L. and Nadler, D. (1986). Organizing for Innovation. California Management Review, 28(3): 74-92.

Utterback, J. (1994). Mastering the dynamics of innovation, Harvard Business School Press, Boston, M.A.

Utterback, J. and Abernathy, W. (1975). A dynamic model of process and product innovation. Omega, 3(6): 639-656.

Van de Ven, A.H. (1986). Central problems in the management of innovation. Management Science, 32(5): 590-607.

van Donk, D.P. and Fransoo, J.C. (2006). Operations management research in process industries. Journal of Operations Management, 24(3): 211-214.

Verganti, R. (1997). Leveraging on systemic learning to manage the early phases of product innovation projects. R&D Management, 27(4): 377-392.

Verworn, B. (2006). How German measurement and control firms integrate market and technological knowledge into the front end of new product development. International Journal of Technology Management, 34(3-4): 379-389.

Verworn, B., Herstatt, C. and Nagagira, A. (2008). The fuzzy front end of Japanese new product development projects: impact on success and differences between incremental and radical projects. R&D Management, 38(1): 1-19.

62

Yin, R. (2003). Case study research: Design and methods. Thousand Oaks, Sage Publications.

63

Appendix A: Interview questions on managing collaborative product development

Background information: Date, time, name, title, position within the firm. 1.Can you start with describing the project? (What did you develop? What was

the objective, how long did you collaborate etc.?) 2.What would you say was characteristic for this project? (E.g. was it well

planned, or did it develop over time, was it conducted within the appointed deadline and budget?)

3. What was it that you should make money on in this project? (Customer focus, technology focus, time to market or cost reductions?)

4.Can you divide the project into different stages? (How was the project structured, what incidents were important for the different stages, how were they managed?)

5.How should you describe the management of the project? (E.g. were there any formal guidelines or was it other aspects that influenced how the project was managed, such as informal rules, routines etc?)

6.Was the project managed in a similar way, or were the different stages managed in a different way?

7. In what way were the steering group’s decisions important? (Can you mention one example where the decisions of the steering group were decisive for the outcome of the project?)

8.Can you describe one or a few central incidents in the project? (Can you give an example of an incident that you regarded as critical for the outcome of the project?)

9. What were the possibilities of changing the direction of the project? (Was it planned in detail or were there some opportunities to be creative?)

10. What is your opinion of the result of the project? (Was it a success or not, and what is the reason for that?)

11. Do you think that there were any problems with the project (both the final result and the process)?

12. Can you tell me more about your role in the project? 13. Can you describe how the responsibility was divided among the project

members? 14. What is your experience of working in cross-professionally organized

projects?

64

Appendix B: Interview questions on managing collaborative process development

1. What is your position within the firm and what are the main activities that you work with?

2. Can you start with a general description of the project? (What did you develop? What was the objective of the collaborative project, etc.?)

3. What is your role in the collaborative project, and what are your responsibilities?

4. Did you have a formal contract for how the collaborative project would be realized and managed?

5. Is there a special group appointed to manage the collaborative project? 6. How did you make decisions within the management group? 7. Is the management of the collaborative project formal (top-down) or is the

management based on social control? 8. What is your opinion of the result of the project and how it was managed? 9. Are there any risks involved in these collaborative projects? (What are they?) 10.Do you think that the different partners trusted each other? (If yes, what did it

mean for your firm?) 11.Can you give an example of something that worked very well in this project? 12.Can you exemplify something that did not work well in this project? 13.On a general level, what do you consider as the three most important

advantages and disadvantages with collaborative projects?

65

Appendix C: Interview questions on managing and organizing the fuzzy front end in product development

Background information 1. What is your formal position within the firm? 2. What are the main activities that you work with and what is your area of

responsibility? 3. What are your previous work experience and education, and how long have

you been working for your current employer? General questions about product development 4. How do you define product development, and what activities constitute

product development? 5. What is the definition of product development at your firm? 6. What is the amount of time that you work with product development? 7. What are the main objectives/purposes of product development at your firm? 8. What is the overall mindset at the firm concerning product development? 9. How is product development organized at your firm? 10. What do the stages of product development look like, i.e. when are things

done and in what order? 11. Does your firm have a formal product development work process or method? 12. How are activities coordinated, which span over functional interfaces and

departments? 13. Do you have a holistic view of product development at your firm? 14. Can you describe how product development and process development

interact? 15. Do process development and product development have the same status at

your firm? 16. Do new products drive new processes or is it the other way around, i.e. is it

process development that enables development of new products? Questions of the fuzzy front end in product development 17.What is your experience of fuzzy front end activities in product development? 18.Who has ideas for new product development projects and where do the ideas

come from? 19. How does your firm screen new ideas for product development projects? 20. How does your firm choose which ideas eventually will be formal projects? 21. Against which criteria are ideas evaluated? 22. Does your firm balance and mix new ideas for product development

projects? 23. How are ideas for new products realized? That is, what do you do to create

formal projects?

66

24. What does your firm do to make the ideas work in practice? 25. How is the “fuzzy front end stage” structured/organized at your firm? 26. Who is involved in the early phases of product development at your firm? 27. How does the work in the early stage of product development function in

practice at your firm? 28. What are the problems, conflicts, and solutions? What is working well or not

so well? 29. What is the most important thing that will make your firm succeed with the

early stage in product development? 30. What do you need to change – if you were allowed to wish freely? 31. What are the obstacles in the early stage in product development? Questions on what drives success in the fuzzy front end of product development 32.What are the critical factors and activities that will make you succeed with the

front end? 33.Please explain, based on your previous experience, how to make FFE

activities work in practice? 34.Can you explain how the different activities and factors that you elaborated

on previously are related to each other, and what firms can do to link the individual factors into a coherent whole? Start with the factor that you deemed most important and elaborate on how it is related to the other factors that you judge as important.

35. Can you tell us, in general terms, how a product concept is developed?

67

Appendix D: Interview questions on managing and organizing the fuzzy front end in process development

Background information 1. What is your formal position within the firm? 2. What are the main activities that you work with, and what is your area of

responsibility? 3. What are your previous work experience and education, and how long have

you been working for your current employer? General questions about process development 4. How do you define process development and what activities constitute

process development? 5. What is the definition of process development at your firm? 6. What is the amount of time that you work with process development? 7. What are the main objectives/purposes of process development at your firm? 8. What is the overall mindset at the firm concerning process development? 9. How is process development organized at your firm? 10.What do the stages of process development look like, i.e. when are things

done and in what order? 11. Does your firm have a formal process development work process or method? 12. How are activities coordinated, which span over functional interfaces and

departments? 13. Do you have a holistic view of process development at your firm? 14. Can you describe how product development and process development

interact? 15. Do process development and product development have the same status at

your firm? 16. Do new products drive new processes or is it the other way around, i.e. is it

process development that enables development of new products? Questions of the fuzzy front end in process development 17.What is your experience of fuzzy front end activities in process development? 18.Who has ideas for new process development projects and where do the ideas

come from? 19. How do your firm screen new ideas for process development projects? 20. How does your firm choose which idea eventually will be formal projects? 21. Against which criteria are ideas evaluated? 22. Does your firm balance and mix new ideas for process development projects? 23. How are ideas realized in the context of production? That is, what do you do

to implement the ideas in the context of production? 24. What does your firm do to make the ideas work in practice?

68

25. How is the “fuzzy front end stage” structured/organized at your firm? 26. Who is involved in the early phases of process development at your firm? 27. How does the work in the early stage of process development function in

practice at your firm? 28. What are the problems, conflicts, and solutions? What is working well or not

so well? 29. What is the most important thing that will make your firm succeed with the

early stage in process development? 30. What do you need to change – if you were allowed to wish freely? 31. What are the obstacles in the early stage in process development? Questions on what drives success in the fuzzy front end of process development 32.What are the critical factors and activities that will make you succeed with the

front end? 33.Please explain, based on your previous experience, how to make FFE

activities work in practice? 34. Can you explain how the different activities and factors that you elaborated

on previously are related to each other, and what firms can do to link the individual factors into a coherent whole? Start with the factor that you deemed most important and elaborate on how it is related to the other factors that you judge as important.

35. Can you tell us, in general terms, how a process concept is developed?

69

Part II

Control mechanisms in collaborative R&D projects: the case of supplier and customer relationships in process industry Monika Kurkkio, 2009 International Journal of Technology Intelligence and Planning, Vol.5, No. 1, pp. 73-89

Paper A

Antecedents and Consequences of Firm’s Process Innovation Capability: A literature review and conceptual framework Monika Kurkkio, Johan Frishammar, and Lena Abrahamsson, 2010 Submitted to IEEE Transactions on Engineering Management

Paper B

The Fuzzy Front End in Non-assembled Product Development: A multiple case study of process firms Johan Frishammar, Ulrich Lichtenthaler, and Monika Kurkkio, 2010

A previous version was presented at the PDMA conference: Managing dualities in the innovation journey, University of Twente, 2009. In review with Journal of Engineering and Technology Management

Paper C

Where Process Development begins: A multiple case study of fuzzy front-end activities in process firms Monika Kurkkio, Johan Frishammar, and Ulrich Lichtenthaler, 2010

A previous version was presented at R&D Management Conference – Information, imagination and intelligence Manchester, 30 June-2 July 2010. In review with Technovation

Paper D

Managing the fuzzy front-end: Insights from process firms Monika Kurkkio, 2010 Accepted for publication in European Journal of Innovation Management

Paper E

Int. J. Technology Intelligence and Planning, Vol. 5, No. 1, 2009 73

Copyright © 2009 Inderscience Enterprises Ltd.

Control mechanisms in collaborative R&D projects: the case of supplier and customer relationships in process industry

Monika Kurkkio Luleå University of Technology, Centre for Management of Innovation and Technology in Process Industry (Promote), 971 87 Luleå, Sweden E-mail: [email protected]

Abstract: Collaborative R&D projects are often associated with high degrees of complexity and uncertainty therefore they are difficult to manage. The purpose of this paper is to identify and analyse which motives collaborating companies have to participate in inter-organisational R&D projects, and how these motives affect the selection of control mechanisms. Empirically, the paper draws on two case studies of joint R&D projects. Findings indicate that companies have a variety of motives which subsequently guides the mix of formal and informal control mechanisms. The paper adds to the literature on collaborative R&D projects and gives managerial advice.

Keywords: collaborative R&D projects; control mechanisms; process industry; product and process development.

Reference to this paper should be made as follows: Kurkkio, M. (2009) ‘Control mechanisms in collaborative R&D projects: the case of supplier and customer relationships in process industry’, Int. J. Technology Intelligence and Planning, Vol. 5, No. 1, pp.73–89.

Biographical notes: Monika Kurkkio is a PhD student at Luleå University of Technology. She is active at the Centre for Management of Innovation and Technology in Process Industry (Promote). Her research interests include organisational and management aspects of product and process innovation, primarily in the Process Industry. She is also interested in managing the ‘fuzzy front end’ of both product and process innovation.

1 Introduction

As technology and business processes become more sophisticated, companies are driven to collaborate with external partners and use an open strategy to acquire all the necessary information and competences instead of using the traditional business strategy, which encourages construction of barriers to competition (Chesbrough and Appleyard, 2007). Collaboration among companies to develop new technology is a strategy, which is becoming more and more important. The development of new technology is often associated with insufficient resources, high investment costs and high technical uncertainty. Collaboration with external partners has become a way to facilitate the

74 M. Kurkkio

development of new technology while remaining competitive. Therefore, many of the Research and Development (R&D) projects in today’s organisations take place in collaboration with external partners.

The collaboration among companies can be organised in many different ways, e.g., through strategic alliances, joint ventures, and supplier and customer collaboration (Emden et al., 2006; Madhok and Tallman, 1998). This paper focuses on one of the most common forms of collaboration between companies, joint R&D projects. A joint R&D project is not usually an own legal entity, which tends to make this type of collaboration more flexible than, for example, a joint venture. The number of companies in a joint R&D project can vary from two companies, e.g., a supplier and customer collaboration, to a handful of companies, and if the R&D project consists of several companies, then it is more suitable to speak of a network or an R&D consortium (Trott, 2005).

Collaborative projects among companies often occur because companies hope to acquire benefits that otherwise would not have been obtainable. Proclaimed benefits with collaborative R&D projects are e.g., shorter time to market, lower costs, knowledge transfer and more innovative solutions (Littler and Leverick, 1995; Perks, 2000). Although a number of scholars have acknowledged several advantages with collaborative R&D projects, there is furthermore evidence supporting the view that these projects are related to several problems, e.g., loss of competitive knowledge and difficulties of collaborating when firms have different values and norms (Boddy et al., 2000; Emden et al., 2006). Loss of competitive knowledge may mean that the collaborating companies lose their competitiveness, which ultimately can lead to losing their position relative to competitors. Difficulties connected with the management of the joint project may lead to a lack of commitment to collaboration between the companies concerned, and if these problems cannot be resolved they may eventually lead to failure. Previous studies indicate difficulties in managing these relationships due to the high level of complexity, risk, and uncertainties associated with collaborative R&D projects, and these are some important reasons behind the fact that many collaborative R&D projects result in failure. As appropriate control mechanisms are one option for firms wishing to reduce complexity and uncertainties, and therefore a means to decrease failure (Dekker, 2004; Kamminga and Van der Meer-Kooistra, 2007), they constitute the prime focus of this paper.

The problems related to management of collaborative R&D projects are not well researched (Boddy et al., 2000) and there is a need for more empirical studies covering various theoretical perspectives, different industries and types of collaborative R&D projects to understand the nature of these projects and how they can be managed. It thus seems important to further analyse why companies choose to engage in collaborative R&D projects and how these projects can be managed in a successful way. The aim of this paper is to identify and analyse which motives collaborating companies have when engaging in collaborative R&D projects and how these motives subsequently affect the selection of control mechanisms. The paper intends to contribute to the fields of technology management, industrial R&D, project management and general management theory, and deepens our understanding of how companies’ motives affect the selection of control mechanisms. Finally, the paper gives managerial advice mainly to project leaders and senior management on how joint R&D projects can be more effectively managed while at the same time maintaining the relationship between collaborating companies.

The rest of the paper is structured as follows. In the next section, the theoretical framework is presented and the third section contains the applied method and research

Control mechanisms in collaborative R&D projects 75

approach used in this paper. The following section contains empirical data from two case studies, which are illustrated and analysed in accordance with theory. Finally, the last section draws conclusions and managerial implications and gives recommendations for future research.

2 Theoretical framework

Research on collaborative R&D projects has appeared in a variety of research fields, for example strategic management, technology management, industrial R&D and project management. Another emerging stream of research that is considered to be useful for understanding management of collaborative R&D projects is the study of inter-organisational relationships, i.e., collaboration projects between companies. This emerging stream encompasses literature from a variety of disciplines like the ones presented above, and other relevant disciplines are marketing and management control. All these perspectives are used to understand what motives companies have to engage in joint R&D projects and how their motives affect the subsequent selection of control mechanisms.

2.1 Motives to engage in collaborative R&D projects

According to Smith et al. (1995), companies are willing to collaborate when the benefits of collaboration exceed the costs. An obvious first motive with collaborative R&D projects is the economic one of reducing costs and share risks by collaborating with external partners (Littler and Leverick, 1995; Perks, 2000). R&D projects are often associated with high costs and risks, and are viewed as somewhat fuzzy – involving high uncertainty with unclear rates of return (Nobelius, 2004). By collaborating with external partners, these costs and risks can be shared among the collaborating companies.

A second common motive for engaging in collaborative R&D projects is access to markets. According to Narula and Dunning (1998), companies engage in strategic alliances to improve market entry and presence; several R&D alliances are undertaken to protect or enhance the created product of the participating companies. A third motive presented in the extant literature is related to getting hold of relevant information about technological trends in the environment of the company (Lichtenthaler, 2003), which then can be used to learn from the collaborating companies; consequently, knowledge transfer can take place among the companies (Knudsen, 2007). For example, in product development projects involving a supplier company and their customers, learning and knowledge transfer are common expectations (Stjernström and Bengtsson, 2004). A supplier, for instance, can learn more about customers’ wishes and the customers can increase their understanding of the supplier’s production processes. Furthermore, companies can have relational motives, e.g., the intention of building long-term relationships that hopefully will foster loyalty (Knudsen, 2007). Finally, enhancing innovation is also mentioned as a motive for engaging in collaborative R&D projects (Narula and Dunning, 1998).

To summarise, there seem to be several different motives for engaging in collaborative R&D projects. All motives seem to have in common the aim of achieving competitiveness. This can be accomplished in the short term by lowering costs or in the long term by establishing lasting relationships and loyalty. The motives

76 M. Kurkkio

of the collaborating companies to engage in a joint R&D project often differ, however, which makes management of these projects difficult. This is furthermore mentioned as one reason for the high failure rates of collaborative projects (Boddy et al., 2000). Hence, research on how to manage and perhaps align the divergent motives in joint R&D projects is apparently of substantial importance, a fact that justifies this study.

2.2 Research on control mechanisms in joint R&D projects

There seems to be general agreement that all organisational control systems consist of both formal and informal control mechanisms (Langfield-Smith and Smith, 2003). Formal control mechanisms are often described as explicitly designed controls, while informal control mechanisms (also referred to as social control) are related to unwritten rules and norms that cannot be designed directly. Formal control has been divided into outcome control and behaviour control (Ouchi, 1979), where outcome control refers to measuring and monitoring the outputs of operations or behaviour, while behaviour control focus on specifying and measuring individuals’ behaviour (Langfield-Smith and Smith, 2003). Formal control in collaborative R&D projects consists of contractual obligations and formal organisational mechanisms for collaboration, e.g., in the form of technology planning, rules and regulations, performance measurements, reward structures and goal setting. In the early literature of technology management and project management, formal control mechanisms are furthermore viewed as important for successful management of R&D projects (see e.g., Lichtenthaler, 2003; Söderlund, 2004). Informal or social control mechanisms, on the other hand, are often developed from shared norms, values and beliefs, and are shaped e.g., by self-regulating mechanisms and informal cultures (Dekker, 2004; Langfield-Smith and Smith, 2003). Some examples of informal control mechanism in collaborative projects are trust and partner selection. In recent years, these social control mechanisms have received more attention in the literature, especially the importance of trust. Trust is a well-established concept both in project management and in the general management literature. Trust seems to be one of the most important informal control mechanisms in collaborative projects, and therefore it has furthermore been addressed in many previous studies on collaborative projects (Smith et al., 1995). Partner selection is another important informal control mechanism, which has been addressed in previous studies (see e.g., Emden et al., 2006), but will not be discussed further in this paper. This paper starts after the selection of an appropriate partner has been completed.

Trust is a difficult concept to study and, consequently, it has been defined and described in many ways (Langfield-Smith and Smith, 2003). In this paper, trust is classified in accordance with Sako (1992), who distinguishes three types of trust: contractual trust, competence and goodwill trust. Contractual trust is based on moral standards of honesty and requires that the companies keep their promises, whether there are written agreements or not. Competence trust is based on expectations that the collaborating partner has the necessary technological and management competences. In the case of competence trust, a company is expected to fulfil its obligations to the other party. Goodwill trust is referred to as the highest level of trust, which expects that the companies have an open commitment to each other; i.e., trust goes beyond what is explicitly written and the trustee is trusted at the highest level of discretion, further than contractual and competence trust. This classification has been used in previous studies on different types of collaborative R&D projects (Seal et al., 1999) and it furthermore seems


to be suitable in the context of joint R&D projects. Another distinction of the concept of trust is the one between personal trust and organisational trust. Barns et al. (2006) point out that personal relationships are important for successful projects, in collaborative R&D projects between industry and academia. This is probably central for a joint R&D project within the same industry, as personal relations among the project members lay the ground for a well-functioning project group. Personal relations among project members are not seen as an informal control mechanism in this paper, but are considered to be important because they affect the informal culture and the building of trust. Relations are expected to develop over time, and the nature of the collaboration is thus also expected to change and grow deeper.

Dekker (2004) has classified formal and informal control mechanisms based on the extant literature on inter-organisational management (see Table 1). When starting a collaborative R&D project, there are several aspects to consider in order for the joint project to be successful. The ex-ante mechanisms can help to mitigate control problems and align the partners’ different motives and aims with the collaboration project. Joint goal setting is one example of an aspect that will reduce uncertainty, if the collaborating partners decide the goals together before the implementation of the joint project. Examples of other relevant formal mechanisms related to behaviour controls are technology planning and regulations. The importance of these has been discussed previously in the paper. Finally, informal mechanisms like trust and social networks have been identified as central before starting a joint project. Dekker (2004) gives advice on which control mechanisms can be used to manage collaborative projects in progress. These are a mix of formal and informal control mechanisms as well.

Table 1 Classification of formal control mechanisms (outcome and behaviour control) and informal control mechanisms (social control) in inter-organisational relationships

Outcome control Behaviour control Social control Ex-ante mechanisms Ex-ante mechanisms Ex-ante mechanisms Goal setting Structural specifications: Partner selection Incentive systems/reward structures

Planning

Procedures

Rules and regulations

Trust (goodwill/capability):

Interaction

Reputation

Social networks

Ex-post mechanisms Ex-post mechanisms Ex-post mechanisms Performance monitoring and rewarding

Behaviour monitoring and rewarding

Trust building:

Risk taking

Joint decision-making and problem solving

Partner development

Source: Dekker (2004, p.32)

If we extend Dekker’s (2004) model of formal and informal control mechanisms, and relate the control mechanisms to the motives that companies can have to engage in, for example, joint R&D projects, then we can see how the motives affect the management of the projects. On the basis of extant literature presented earlier, it seems reasonable

78 M. Kurkkio

to believe that different companies have various motives to engage in these types of projects, which will influence what control mechanisms will be used to manage the joint project. If the focus of the R&D project is short-term-oriented and aims at e.g., gaining economic value in terms of lower costs or faster time to market, then formal control mechanisms are expected to be preferred over informal control mechanisms. On the other hand, if the primary motive with the joint R&D project is to build long-term relations, then informal control mechanisms are more likely to be used before formal control mechanisms. Hence, different motives with a joint R&D project demand a diverse mix of control mechanisms to manage the project in an efficient and successful way. This is further discussed in the fourth section of the paper.

3 Research method

In-depth case studies were judged to be the most appropriate research strategy for four main reasons. First, case studies are useful when exploring complex social phenomena in real-life contexts (Yin, 2003). Second, the limited amount of previous research on control mechanisms in collaborative R&D projects indicates that themes and patterns need to be identified rather than confirmed (Eisenhardt, 1989). Third, they enable the investigation of formal as well as informal processes within a joint R&D project (essential because of the social nature of a collaboration project). Finally, multiple sources of data can be used within the structure of case study research (Yin, 2003).

The cases were selected by means of theoretical sampling, i.e., on the assumption that they would fit the theoretical framework (Yin, 2003). One of the criteria for choosing the two cases was that the R&D projects should involve external partners where uncertainty was paramount, which most likely was reduced by some kind of control mechanism. Another criterion was that access to all the participating companies in both cases was considered to be important, as most studies of collaborative projects often only present one view of the collaboration project, e.g., the supplier or customer perspective (Stjernström and Bengtsson, 2004). It was assumed that the larger the joint R&D project was, the more complex it should be to manage and evaluate. This was also assumed in terms of technological complexity. The more complex technology used in the joint R&D project was assumed to make it more uncertain and therefore harder to manage and evaluate. However, the selected case should not be too large, as this would imply problems of getting an overview of the selected project. Both cases were conducted in the same industry, but the cases studied are quite different from one another in terms of how many actors were involved in the joint R&D project and also in the applied technology. To ensure anonymity, the two cases are labelled Case 1 and 2.

Case 1 was basically a new product development project between a supplier and a customer, characterised by incremental improvements. In Case 2, the same companies were present, but several other organisations were involved as well, such as a research centre, a university, and a competitor (to the customer in Case 1). The objective with the joint R&D project reported in Case 2 was to establish an international knowledge centre in Steelmaking.

Altogether, 14 in-depth, semi-structured interviews were conducted, lasting between 30 min and 1.5 h, with respondents from all involved parties in both cases. Six interviews were conducted in Case 1 and eight in Case 2 (see Table 2). A snowball (or chain) sampling strategy (Miles and Huberman, 1994) was used for selecting


the interviewees, and in both cases the interviewees were selected through a discussion with the project leaders, who helped to identify people with a central role in the project. The aim was to find knowledgeable, reliable and accurate respondents (Sigglekow, 2007). Project members from all levels (ranging from the strategic level to the operative level) were interviewed in both cases to get a richer and more accurate view of the selected projects.

Table 2 Specifications of the interviewed respondents in both cases

Respondents Company/organisation Gender CaseDuration of interview

Project leader Supplier Female 1 1 h Project leader Supplier Female 1 75 min “Member of the Steering committee” Supplier Male 1 90 min Steering committee/reference group Supplier Male 1 45 min Operative production manager Supplier Male 1 80 min Operative project manager Customer A Male 1 50 min Centre director Research centre Male 2 60 min Project leader Research centre Male 2 60 min Researcher Research centre Female 2 45 min “Member of the Steering committee” Supplier Male 2 45 min Researcher Customer A Male 2 30 min Researcher Customer A Male 2 30 min Project leader Customer B Male 2 50 min PhD student University Female 2 30 min

In Case 1, in addition to the interview data, internal documents were used in the analysis, which consisted of project planes, status reports and the final project report. The project plan and the final project report were two central documents in the empirical analysis. The project plan covered e.g., expected goals, time frames, and how risks should be divided between the companies within the joint R&D project, while the final project report was a summary of the results and stated e.g., how the expected goals actually were achieved. In both cases, data triangulation was furthermore utilised; publicly available information, such as annual reports and home pages of the participating companies, were compared with the data collected (Yin, 2003). As Case 2 was not completed, no such data were available at the time of the study. The empirical data for Case 1 were collected from October 2005 until January 2006, and in Case 2 all the empirical data were collected during December 2007.

The interviews were tape-recorded, transcribed into protocols, and content-analysed in accordance with Miles and Huberman’s (1994) guidelines. The analysis of the data began by organising each case study in chronological order and then each case was separately reorganised thematically in accordance with the theoretical framework. First, the different motives that companies had for engaging in collaborative R&D projects were identified, and thereafter the control mechanisms used in both projects were investigated. Finally, the selection of control mechanisms was analysed depending on the different motives that companies had with the joint R&D project. The project documents

80 M. Kurkkio

in Case 1 helped to identify and analyse the formal control mechanisms used in the joint R&D project. The project level is the primary unit of analysis, but some analysis is made on the company level, a fact that corresponds to Yin’s (2003) notion of embedded units of analysis. For example, the motives for engaging in joint R&D projects are analysed on the company level and the selection of control mechanisms is analysed on the project level. The analyses were conducted by the author of the paper and then discussed with several colleagues to strengthen the analysis. Aspects of trustworthiness were considered, e.g., through selecting cases with access to all involved companies at different levels within the studied projects, and the ensuring of consistency among the actors’ taped responses and the themes developed (Miles and Huberman, 1994; Yin, 2003). The theoretical representativeness of the study is strengthened through the selection of cases within the same industry, which makes the cases more comparable than if cases were selected from different contexts.

4 The cases of joint R&D projects in the process industry

In this section, the two case studies are presented and later analysed in accordance with theory. The section starts with a presentation of the general context for both case studies, namely the process industry. Thereafter, the general background of each case study is presented separately, followed by the motives that companies had for engaging in the joint R&D project. Finally, the control mechanisms in the two joint R&D projects are presented and subsequently analysed against the theoretical framework.

4.1 Joint R&D projects in the process industry

Process industries are characterised by a large quantity of materials that flow through the entire process, and this is related to costly production and heavy investments. The high investments in machines and equipment have promoted high-volume products and economy of scale for process-based companies to be successful and competitive. Changing technology, shortened product life cycles and increasing global competition are some aspects that have made product development a critical concern for companies in the process industry (Chronéer, 2003). In process industries, there is a strong relation between product and process development (Lager, 2002; Linton and Walsh, 2007). For example, product development has often been a result of process development.

The case studies reported in this paper are two joint R&D projects, with elements of both product and process development. The first project aims at developing one of the supplier’s products, i.e., this project is more of a product development project, and the other case aims at developing the production processes, i.e., a process development project. Both cases are joint R&D projects and have several similarities, but also significant differences. The similarities are for example that the joint R&D projects studied are not own legal partners, which affects how the projects are managed; both cases involve companies that had worked together previously, and two of the companies were involved in both cases. One of the differences is that the cases studied have different sizes (in terms of actors involved and budgets); this is interesting because more partners involved tend to make the project more complex to organise and manage. In Case 1, there are only two partners, and in Case 2 the same companies are involved as well as three


other companies. Figure 1 provides a schematic overview of the organisations involved in the two studied joint R&D projects.

Figure 1 Participating companies in the two joint R&D projects

Source: Self-constructed model

4.2 Case 1: The incremental product development project between a supplier and a customer

Case 1 is a small R&D project, which only involves two partners, a supplier company and one of its foreign customers. The project team can also be described as a small group consisting only of 6–8 people from both companies. The supplier company is a large mining company and the customer is a large steel-producing company, and both are located in northern Europe but in different countries. The overall objective of the collaborative R&D project was to accomplish product development, characterised by incremental improvements in one of the supplier company’s existing products. An iron pellet was the product, which was developed to better suit the customer’s production process. The joint R&D project was part of an ongoing collaboration between the two companies, and the project was carried out during approximately one year (ranging from the fall of 2005 until the end of 2006).

4.2.1 The motives that companies had for engaging in the joint R&D project

The joint R&D project started after a discussion between the customer company and the sales manager at the supplier company. The customer had a wish that the supplier’s products would fit their production processes better, and at the same time the supplier company had been working with developing its products to become more customer-oriented. They decided to start a joint project and planned how it was to be conducted. The project was described as the main assistance for exploiting customer relations, and the supplier saw the project as a chance to work closely with the customer and learn more about their needs and wishes. The project was also an opportunity to determine whether the product had a potential market. The customer described the joint project as an opportunity to learn more about the supplier’s products and how they were developed. Another objective with the joint R&D project in Case 1 was to improve the profitability for both companies and share development costs.

82 M. Kurkkio

4.2.2 Selection of control mechanisms in the joint R&D project

In the initial start-up phase of the joint R&D project, the companies discussed the terms for the project and how it should be conducted. Thereafter, a contract was written, which stated how the risks and costs would be divided between the companies, what each part should do, and what was expected by both partners. The contract was described as a formal document, which set the rules for the joint project. Further, the contract was a way of reducing uncertainty, which was related to the differences in e.g., culture and values between the two collaborating companies. R&D projects are often associated with high levels of uncertainty, as the outcomes in these projects are hard to measure and anticipate. A joint R&D project is even more uncertain when different companies are involved, and therefore both companies tried to reduce as much uncertainty as possible before the project was started. The sales manager from the supplier company said that

“it was important to make a contract which both companies agreed upon before the project started so that we did not destroy the relationship with the customer.”

Another way of reducing the uncertainty in the joint project was to make a plan of how to conduct the project as realistic and accurate as possible. The project leader said that much time was devoted to developing the plan; this was especially important when the project was conducted with an external partner. The planning document was used as a steering document by the project leader and it described e.g., how the project was organised, what the different companies should do in different phases of the project, expected goals of both companies and a joint goal for the entire project, how risks should be divided between the companies, and the time frame that specified when the different actions were planned to occur. The project was decomposed into three phases: a starting phase, an implementation phase, and a final phase. The work was divided between the companies so that the supplier did most of the work in the first two phases and the customer was involved and in charge of the final phase. In the first phase, the supplier company builds the test plant and coated the product, an iron pellet, before it was shipped off to the customer. In the third phase, the new product was tested at the factory in the customer company. There was little interaction between the companies in the second phase, while both companies worked on their respective parts of the joint project, but the project leader still held systematic project meetings. The project leader said that this was a way of assuring the customer that the joint R&D project was important for the supplier, who therefore wanted to inform the customer even if not much had been done yet at the time. In the final phase, the companies interacted more because the results of the project were interpreted by both companies and conclusions of the projects were drawn.

When the companies in Case 1 worked on their respective parts of the project, they did not impose any formal control on one another; instead, they trusted that each party should do as they had agreed upon. The previous collaboration between the companies had helped to establish trust between them. This was mentioned by one respondent as the most significant aspect of conducting a successful joint R&D project.

4.3 Case 2: The joint process development project

As noted previously, Case 2 involved the same supplier and customer as in Case 1 but also included several other organisations. These were a research centre, another customer (customer B) and a university (see Figure 1). The research centre conducts research in the


metallurgical area, and the supplier company is one of its largest customers. Customer B produces steel products of high quality such as steel sheets and steel plates. The university has a technical profile and one of the research profiles is materials engineering, where high-performance steel is one central area. The supplier company and customer A have already been presented in Case 1. The project team in this joint R&D project consists of approximately 20 persons, with at least two persons from each participant. The objective with the joint R&D project was to establish an international knowledge centre of steelmaking. In more detail, the project aimed at developing models and systems to optimise the entire processes from iron to complete steel, i.e., the project can be classified as a process development project. The project entails four different units or research areas. The first area is called future processes and production systems, the second area encompasses optimum raw material design, the third focuses on sustainable energy strategies, and the fourth area deals with residual products and recycling systems. In each of these areas, the overall aim is to develop and render the entire process of steelmaking more effective.

The research centre took the initiative to start the joint R&D project and invited all the organisations, who accepted the offer. The participating organisations had a workshop where they discussed the aim of the joint R&D project, how it was going to be conducted, and what each participant should contribute. The research centre, which was the initiative taker, put together a project description and secured funding from an external financier. The project application was accepted and this was the start of the project in 2006. It is expected to be finished after three years, in 2009.

4.3.1 The motives that companies had for engaging in the joint R&D project

The respondents in Case 2 mentioned several different motives for engaging in the joint R&D project. Saving costs by making the entire process from iron to steel more effective was described as the primary motive, which all companies agreed upon. Besides saving costs, one of the companies mentioned that the project aimed at developing more environmentally friendly processes, which can in the long run be a competitive advantage primarily for the participating industrial companies. Several of the actors said that they hoped to learn new things and to develop a social network by participating in the joint R&D project. The social network can be described as a company network based on the personal contacts among the project members.

4.3.2 Selection of control mechanisms in the joint R&D project

In the first phase of the joint R&D project, several contracts were written. First, a general contract was established between the financier and the research centre. As the research centre was the initiative taker, it took care of most of the administrative works, e.g., in terms of writing the project application, putting together the evaluation reports to the financier and so forth. Much time and effort were spent in the first phase by the research centre on planning the joint R&D project. Individual contracts were written between the research centre and each participating organisation. These regulated what the different organisations should do and how much of the resources they contributed. The conditions, e.g., the level of funding, differed for each participating organisation.

The joint project was organised as a company with a classical hierarchical organisational structure. The project had a steering committee, where all parties were

84 M. Kurkkio

involved; a vice director, from the research centre; a project group of operational management with participants from each organisation; four project units/research areas. Despite this structure, several of the respondents describe the project as a flat organisation. One of the respondents said that the communication in the joint project was good. He gave an example: there were several regular information meetings in the project and all organisations attended these meetings. Another respondent said that “there is always someone you can speak to if you need help”. Most of the full-time project members sat together in one building, located near three of the organisations. The intention of gathering most of the project members who work operationally in the project was to create a good climate and to build personal relations within the project team. Several of the companies had collaborated together previously in other constellations, but not all organisations had worked together. For example, the research centre and the supplier company and customer B had a very close relationship, which had resulted in several previous collaborations, but customers A and B were not used to collaborate. They were primary competitors but, as the joint R&D project was primarily a process development project, this was described as unproblematic. One of the respondents said that “if the joint R&D project had been a product development project instead, the rivalry between customers A and B could have been a problem”. This project had a process development focus and one expected result was that less energy would be needed to produce new products; this was said to be of more general interest instead of an organisation-specific interest.

One of the respondents said that it was relatively easy to establish the formal structure of the project, when each organisation could have a member in the steering committee and take part in all the decisions that were made. The operational management team in the joint R&D project had several meetings, every 6–8 weeks. In these meetings, information was shared on e.g., the status of the project thus far. In the end of the first year, the financier also requested a status report, where the companies presented what results had been achieved until then in their part of the joint R&D project. There were several status reports and some evaluation points, which many of the respondents mentioned as a positive thing.

One of the operative project leaders said that “trust is the most important thing in these kinds of projects”, but several other respondents pointed out the importance of formal control mechanisms such as the use of contracts, monitoring and continuous evaluation for managing the joint R&D project. As the project was still in the first phase, informal control mechanisms were not yet suitable and therefore formal mechanisms were stressed as the most important ones by the project members. In this case study, the organisations primarily used formal control mechanisms to manage the joint R&D project.

4.4 How do motives affect the selection of control mechanisms?

Findings from the present case studies have shown that companies can have diverse motives for engaging in joint R&D projects. In Case 1, both companies emphasised long-term goals, such as building relations to foster loyalty and learning from each other, but the joint R&D project also had a short-term motive of collaborating, e.g., by sharing costs and risks. The respondents in Case 2 mentioned several different motives. Saving costs by making the entire process from iron to steel more effective was described as the primary motive, which all organisations agreed upon. Hence, in both cases,


cost reduction was pointed out as an important motive for engaging in joint R&D projects. These findings are in line with Lager (2002) who points to cost reduction as one of the most common motives for process development. This does not come as a surprise, as most companies in process industries often have high investment costs (Lager, 2002).

The long-term goals such as building durable relations and knowledge transfer, which were the primary motives in Case 1 for engaging in the joint R&D project, are in line with Knudsen’s (2007) findings. She identifies these motives as important in collaborative projects. The results also agree with Stjernström and Bengtsson’s (2004) study, which states that learning and knowledge transfer are common expectations in supplier and customer collaborations. Further, the creation of a social network was described as the main motive for engaging in the joint R&D project by one of the organisations in Case 2. Building social networks per se is perhaps not the final objective for participating in collaborative R&D projects, but is seen as a means to achieve future profitability. Judging from the empirical observations, it seems that companies’ motives to engage in joint R&D projects can be separated into two types; financial and relational motives. Financial motives often have a short-term perspective by e.g., reducing development costs or sharing risks, while relational motives have the aim of achieving economic benefits over a longer-time period for example by building up customer loyalty.

In Case 2, the primary motive of the joint R&D project was to achieve cost-effectiveness (primarily financial motives), which resulted in the use of formal control mechanisms such as contractual obligations and planning. If, on the other hand, relational motives were the primary focus of the joint R&D project, as in Case 1, then informal control mechanisms, such as trust and joint decision-making, tended to be used. Hence, the results show that depending on the type of motives the companies had with the joint R&D project, different control mechanisms were used to manage the project.

In the extant literature, both formal and informal control mechanisms are expected in all companies or projects (Langfield-Smith and Smith, 2003). This seems to be the case in this study as well, but it is the mix of formal and informal control mechanisms that is interesting. In Case 2, the organisations managed the joint R&D project primarily with formal control mechanisms. One possible reason is that all of the organisations involved did not know each other very well, i.e., they had no previous experience of working together. In this joint R&D project, several organisations participated, which made it more complex and uncertain compared with Case 1; therefore, formal control mechanisms were used to manage the project. The project leaders tried to establish a good climate that hopefully would build up trust among the project members, but as the joint R&D project was in an early phase this had not yet been attained. The organisations had set up rules and regulations in a contract. Previous studies on collaborative projects show that contracts are usual in these projects (Seal et al., 1999; Langfield-Smith and Smith, 2003). The contract was a way of reducing uncertainty among the organisations. The joint R&D project was further evaluated by the financier, which several of the respondents mentioned as positive. This is in line with Dekker’s (2004) findings and supports the view of performance monitoring, as an important control mechanism to manage collaborative projects.

Informal control mechanisms seemed to be dominant in Case 1, although formal control mechanisms were also used to reduce the uncertainties in the project. Trust was the main informal control mechanism used in Case 1 and the type of trust can be related

86 M. Kurkkio

to competence trust, which means that the collaborating partners trusted each other and thought that the partner was competent (Sako, 1992). The results of Case 1 show that the companies used a mix of formal and informal control mechanisms in the joint R&D project. In the first phase of the joint project, joint decision-making and goal setting were the foundation for how the project was going to be conducted. This is in line with Dekker’s (2004) results. He has classified joint decision-making as an informal control mechanism and goal setting as a formal control mechanism. The supplier who initiated the project was very concerned that the project should not damage the relationship with the customer, and this was the primary reason for using a contract and joint decision-making. The operative work in the joint project was primarily conducted by one company at a time, and the control mechanisms used were mostly informal ones such as trust.

If the appropriate mix is identified and used, then the joint R&D project can be managed efficiently while the relations between the companies are preserved. In the next section, these important and somewhat opposite goals, which are difficult to obtain simultaneously, are discussed and managerial ideas of how to achieve this aim are proposed.

5 Discussion and conclusions

This paper examines companies’ motives for engaging in collaborative R&D projects and how these motives subsequently affect the selection of control mechanisms in the joint R&D project. The empirical findings suggest that companies have various such motives and that these do affect the selection of control mechanisms used in the project. The motives were separated into those of financial or relational character. The financial motives, such as reducing development costs and sharing risks by collaborating with external partners, had primarily a short-term perspective, while relational motives often have the aim of achieving economic benefits over a longer-time period for example by building up customer loyalty or sharing knowledge between the companies.

According to the findings, formal control mechanisms such as contracts, planning and performance monitoring are used in projects with cost reduction as the chief objective. Formal control mechanisms seemed to be important in the early stages of a joint R&D project, when the persons involved did not know each other and mutual competence or goodwill trust was not established. In joint R&D projects where the companies have positive experience from previous collaboration, informal control mechanisms seemed to be the appropriate control mechanisms. Formal control mechanisms can be used to reduce uncertainty and help to maintain the relation between the companies. Further formal control mechanisms, e.g., planning and control techniques, were important to manage the more complex collaborative R&D project. Previous studies show that collaborative R&D projects have a high failure rate (Boddy et al., 2000), which can be improved if the appropriate control mechanisms are used in the project. It is important to note that contextual factors such as size of the joint R&D project, degree of formalisation of the operative work, and degree of technology affect how the mix of control mechanisms should be developed and implemented.


This study contributes to research on collaborative R&D projects in three ways. First, there are only a few empirical studies of specific control mechanisms used in inter-organisational relationships (Dekker, 2004). This study adds to the knowledge of how companies’ motives affect the selection of control mechanisms. The study has further provided empirical evidence on what type of motives companies have when they decide to engage in collaborative R&D projects. Second, the study has combined the perspectives of technology management, project management and general management theory to exemplify a comprehensive approach to understanding management in collaborative R&D projects. Third, the study suggests that if the primary aim of a joint R&D project is of financial character and has a short-term orientation, then formal control mechanisms are used before informal ones. The mechanisms that are being used consist to a great extent of management and technology practices such as planning, cost calculation, and open book accounting.

For project leaders and other managers at the strategic level of a company, this study suggests that it is important to understand what motives the different participants have for engaging in a joint R&D project. Therefore, it is important to conduct a discussion before the project starts, with persons from all of the participating companies, so that every party can present its expectations, motives, ambitions, and goals with the joint project. Another aspect that seemed to be central was the selection of the project members. The persons involved in the joint R&D project should be independent individuals who can be held accountable and have the ability to collaborate with other persons, so that personal relations can develop. The right persons can help to build a climate with a trusting atmosphere, which seems to be important in these types of projects. Once the project is in progress, the project leader should be able to use a mix of formal and informal control mechanisms so that the project is managed effectively and at the same time the relationship is maintained between the companies. These somewhat opposite goals are both important to pursue simultaneously. The project leader may use formal control mechanisms to define boundaries for the joint R&D project, and this is especially important at the beginning of a new collaboration where the companies do not have established mutual trust. Informal control mechanisms, on the other hand, are suitable when companies have worked together previously or later on in the project after the project members have gotten to know each other.

As in most case studies, the sample size is limited, and therefore no attempts have been made to generalise beyond the sample, but theoretical generalisations are useful. The results presented above provide insights into how the different motives of companies can and should affect the selection of control mechanisms in a joint R&D project. The explorative nature of the study, combined with the low sample size, suggests that more empirical work is desirable to strengthen the external validity of the results (Yin, 2003). Further, it would be interesting to examine how the control mechanisms in a joint project evolve over time. In other words, there is a need for more longitudinal studies of joint R&D projects, which can deepen our understanding of the mixture of control mechanisms and their ways of changing. This can be seen as a search for higher-level mechanisms of control, whether implicit or explicit, over the control mechanisms themselves.

88 M. Kurkkio

References Barns, T.A., Pashby, I.R. and Gibbons, A.M. (2006) ‘Managing collaborative R&D projects

development of a practical management tool’, Int. J. Project Management, Vol. 24, No. 5, pp.395–404.

Boddy, D., Macbeth, D. and Wagner, B. (2000) ‘Implementing collaboration between organizations: an empirical study of supply chain partnering’, Journal of Management Studies, Vol. 37, No. 7, pp.1003–1017.

Chesbrough, H.W. and Appleyard, M.M. (2007) ‘Open innovation and strategy’, California Management Review, Vol. 50, No. 1, pp.57–76.

Chronéer, D. (2003) ‘Have companies shifted their centre of gravity during the 90s?’, Int. J. Innovation Management, Vol. 7, No. 1, pp.95–129.

Dekker, H. (2004) ‘Control of inter-organizational relationships: evidence on appropriation concerns and coordination requirements’, Accounting, Organization and Society, Vol. 29, No. 1, pp.27–49.

Eisenhardt, K. (1989) ‘Building theories from case study research’, Academy of Management Review, Vol. 14, No. 4, pp.532–550.

Emden, Z., Calantone, R.J. and Droge, C. (2006) ‘Collaborating for new product development: selecting the partner with maximum potential to create value’, Journal of Product Innovation Management, Vol. 23, No. 4, pp.330–341.

Kamminga, P.E. and Van der Meer-Kooistra, J. (2007) ‘Management control patterns in joint venture relationships: a model and an exploratory study’, Accounting, Organization and Society, Vol. 32, Nos. 1–2, pp.131–154.

Knudsen, M.P. (2007) ‘The relative importance of interfirm relationships and knowledge transfer for new product development success’, The Journal of Product Innovation Management, Vol. 24, No. 2, pp.117–138.

Lager, T. (2002) ‘A structural analysis of process development in process industry’, R&D Management, Vol. 32, No. 1, pp.87–95.

Langfield-Smith, K. and Smith, D. (2003) ‘Management control systems and trust in outsourcing relationships’, Management Accounting Research, Vol. 14, No. 3, pp.281–307.

Lichtenthaler, E. (2003) ‘Third generation management of technology intelligence processes’, R&D Management, Vol. 33, No. 4, pp.361–375.

Linton, J.D. and Walsh, S.T. (2007) ‘A theory of innovation for process-based innovations such as nanotechnology’, Technological Forecasting and Social Change, Vol. 75, No. 5, pp.583–594.

Littler, D. and Leverick, F. (1995) ‘Joint ventures for product development: learning from experience’, Long Range Planning, Vol. 28, No. 3, pp.58–67.

Madhok, A. and Tallman, S.B. (1998) ‘Resources, transactions and rents: managing value through interfirm collaborative relationships’, Organization Science, Vol. 9, No. 3, pp.326–339.

Miles, M.B. and Huberman, A.M. (1994) Qualitative Data Analysis – An Expanded Sourcebook, Sage Publications, Inc., Thousand Oaks.

Narula, R. and Dunning, J.H. (1998) ‘Explaining international R&D alliances and the role of governments’, International Business Review, Vol. 7, No. 4, pp.377–397.

Nobelius, D. (2004) ‘Towards the sixth generation of R&D management’, Int. J. Project Management, Vol. 22, No. 5, pp.369–375.

Ouchi, W.G. (1979) ‘A conceptual framework for the design of organizational control mechanisms’, Management Science, Vol. 25, No. 9, pp.833–848.

Perks, H. (2000) ‘Marketing information exchange mechanisms in collaborative new product: the influence of resource balance and competitiveness’, Industrial Marketing Management, Vol. 29, No. 2, pp.179–189.


Sako, M. (1992) Prices, Quality and Trust: Inter-firm Relations in Britain and Japan, Cambridge University Press, Cambridge.

Seal, W., Cullen, J., Dunlop, A., Berry, T. and Ahmed, M. (1999) ‘Enacting a European supply chain: a case study on the role of management accounting’, Management Accounting Research, Vol. 10, No. 3, pp.303–322.

Sigglecow, N. (2007) ‘Persuasion with case studies’, Academy of Management Journal, Vol. 50, No. 1, pp.20–24.

Stjernström, S. and Bengtsson, L. (2004) ‘Supplier perspective on business relationships: experience from six small suppliers’, Journal of Purchasing and Supply Management, Vol. 10, No. 3, p.137–146.

Smith, K.G., Carroll, S.J. and Ashford, S.J. (1995) ‘Intra – and interorganizational cooperation: towards a research agenda’, Academy of Management, Vol. 38, No. 1, pp.7–23.

Söderlund, J. (2004) ‘On the broadening scope of the research on projects: a review and a model for analysis’, Int. J. Project Management, Vol. 22, No. 8, pp.655–667.

Trott, P. (2005) Innovation Management and New Product Development, Pearson Education, Harlow.

Yin, R.K. (2003) Case Study Research: Design and Methods, Sage Publications, Inc., Thousand Oaks.


Paper A


Paper B



Paper C



Paper D


Paper E

Antecedents and Consequences of Firms’ Process InnovationCapability: A literature review and a conceptual framework

Monika KurkkioCenter for Management of Innovation and Technology in Process Industry

(Promote)Luleå University of Technology, SE 971 87 Luleå, Sweden

[email protected]

Johan FrishammarCenter for Management of Innovation and Technology in Process Industry

(Promote)Luleå University of Technology, SE 971 87 Luleå, Sweden

Tel: 0046 920 [email protected]

Lena AbrahamssonDivision of Industrial Work Environment

Luleå University of Technology, SE 971 87 Luleå, [email protected]

AbstractThis article reviews and analyzes the extant literature on managing processinnovation. Similar to product innovation, process innovation is a key source oflong term competitive advantage, yet few studies address process innovationcompared to the well researched domain of product innovation. In addition,the literature on managing process innovation has not yet been systematicallyreviewed in the scholarly literature. The aim of this paper is to provide a reviewof the process innovation literature in order to inform the understanding ofantecedents and consequences of firms’ process innovation capability. Thepaper offers several important contributions. First, it makes a principaldistinction between a firm’s potential and realized process innovationcapability, and argues that high quality realization mechanisms are critical forachieving desired process innovation outcomes. Second, antecedents andconsequences of a firm’s process innovation capability are identified anddiscussed, which allows more proficient management of process innovation.Finally, by drawing on a capability based perspective, important implicationsfor management practice and recommendations for future research arederived.

Keywords: Process innovation, Capabilities, literature review

1

I. INTRODUCTIONProcess innovation is crucial for the renewal, survival, and growth ofmanufacturing firms [1], [2]. The intended outcomes of process innovation aretypically cost reductions, increased production volumes, improved productquality, and faster development. All these outcomes can clearly advance afirm’s competitive position [3], [4], [5], [6].

The literature on managing process innovation is multidisciplinary, andspans technology and innovation management [7], [8], [9] operationsmanagement [10], organizational behavior [11], [12] and the generalmanagement literature [13], [14]. All these fields of inquiry have theorizedabout which factors affect the success of process innovation [7], [15], [16], butfew if any attempts have been made to systematically structure and presentprior research findings in a cumulative way. Such cumulative studies exist forproduct innovation [17], [18], [19], [20] but are so far lacking in the processinnovation domain.

We realize that the terms “process” and “innovation” carry multiplemeanings. In accordance with prior literature, we define process innovation asdeliberate and systemic development related mainly to internal productionobjectives, implying the introduction of significantly new elements into theproduction process with the purpose of creating or improving methods ofproduction [3], [7], [9], [10], [11], [21], [22]. Thus, we use the term “processinnovation” to refer to innovation in production processes.

A. Research gaps and contributionGrounded in an extensive literature review, this article offers severalcontributions. First, we present a novel approach to the literature on managingprocess innovation by utilizing a capability based perspective. By drawing onresearch into absorptive capacity, we discriminate between firms’ potentialand realized process innovation capabilities, which appears important forbetter understanding how the desired outcomes of process innovation areachieved. A capability based perspective to product innovation has beenutilized in several studies [23], [24], [25], but little is still known about the skillsand procedures which underline a firm’s process innovation capability [26],[27] and about how such capabilities underpin desired outcomes [1].

Second, we identify the antecedents and consequences of a firm’s processinnovation capability, which allows more proficient management of processinnovation. In sum, we address a problem shared by both practicing managersand the research community: the issue of how to manage process innovationso that its positive benefits can be better realized.

2

II. RESEARCH APPROACH AND ASSORTMENT OF LITERATUREOur review of the process innovation literature started with an unprejudiced‘read through’ of books on process innovation [1], current textbooks ininnovation management [8], and frequently quoted articles on processinnovation [14], [28]. This allowed us to establish the most relevant key wordswhich were subsequently employed in the formal searches for literature. Thenext step was to construct a classification matrix consisting of differentanalytical dimensions for classifying the literature. This matrix enhanced rigorand quality in subsequent steps when classification and analysis of theliterature were made with three researchers involved.

During the formal literature search, we looked for articles on ‘process ANDinnovation’, ‘process AND development’, ‘process technology’, and‘production’ or ‘manufacturing’ in combination with ‘development’ or‘innovation’. Initially, our literature search was broad and encompassedsearches in the databases of Blackwell Synergy, EBSCOhost, IEEE Explore, JStore, Sage, Science Direct, Springer and Wiley. Together, these databasescontain all major journals publishing articles on process innovation. To improvereliability, all databases were searched twice by two researchers workingindependently. The authors then met to compare their findings, discussingwhether the papers that had been found would be included or excluded fromthe review.

This generated a first, tentative list of literature. Next, two authors againread the abstracts for all papers and then ‘negotiated’ about excluding orincluding the articles that emerged as uncertain cases.

Despite the initial broad scope, the review was limited in several ways. Firstof all, the definition of process innovation employed was an initial limitation,which meant that literature outside the production and manufacturing domainwas disregarded. Second, the article is about management of processinnovation at the firm level. Therefore, articles focusing solely on the industrylevel were excluded, e.g. [29]. Articles focusing partly on the industry level, butwith implications on the firm level, were included [21]. Articles written fromthe aggregate perspective of economics were also excluded, because sucharticles tend to examine differences in e.g. patterns of process innovationacross countries and industrial sectors rather than firm level issues [30]. Mostdifficult to judge were articles concerning process technology. Although accessto process technology, i.e. in the form of a firm’s knowledge base, is necessaryfor successful process innovation, it is not synonymous with process innovationas such. Therefore, articles on the role of process technology in processinnovation were included, whereas those on process technology outside theprocess innovation context were excluded. For reasons of both scope andquality, our literature review is grounded mainly in peer reviewed journalarticles, although a few seminal books on the topic were included as well.

Finally, we also excluded several related bodies of literature. Articles onconcurrent engineering and business process re engineering were all

3

disregarded. We also excluded articles focusing on business process innovation,e.g. [31]. Articles on process management practices, which include a series ofquality related initiatives such as total quality management, ISO 9000, and sixsigma programs (e.g. [32]) were disregarded for multiple reasons. First, ourdefinition of process innovation implies more bold efforts than what thisliterature encompasses. Second, the literature on process managementpractices is massive, making its inclusion difficult for reasons of scope. Weconsidered but did not include literature on simultaneous engineering anddesign for manufacturability, as much of the focus there has been oncoordinating product and process design rather than on managing processinnovation as such.

This procedure refined our first list of literature into a more comprehensiveone. In addition, the use of a ‘snowball technique’, i.e. reading through thereference lists of the reviewed papers, provided some additional insights intoliterature missed in the formal searches. All in all, our final list of literatureincluded 51 peer reviewed research articles and books which constitute thebasis of the review.

III. A CONCEPTUAL FRAMEWORK OF FIRMS’ PROCESS INNOVATIONCAPABILITY

In this and coming sections, the literature on managing process innovation issynthesized and organized according to a framework grounded in a capabilitybased perspective [33], [34], [35]. A fundamental assumption is thus that afirm’s process innovation capability is paramount for process innovationoutcomes [1].

Inspired by research into absorptive capacity [34] and open innovation [36],[37], we make a distinction between firms’ potential and realized processinnovation capabilities. A firm’s potential process innovation capability involvesacquisition and assimilation activities, whereas the realized process innovationcapability focuses on exploitation and transformation activities [34], [37]. Forexample, the float glass technology pioneered by Pilkington requiredacquisition and assimilation of new process technology and collaboration withexternal partners, and in subsequent steps exploitation of the technology andradical transformation of Pilkington’s production processes [38].

Potential and realized process innovation capabilities are both essential forachieving desired outcomes, but they have complementary roles. The potentialprocess innovation capability involves assimilating and acquiring tangible andintangible resources to increase a firm’s potential capacity [34]. In the contextof managing process innovation, these resources have been identified in priorliterature which has presented them as antecedents, enablers or so calledsuccess factors for process innovation. In this paper, we refer to these asantecedent factors to a firm’s potential process innovation capability. The

4

realized process innovation capability refers to a firm’s exploitation andtransformation capacity. That is, it implies executing the antecedent factors sothat process innovation is enabled. Arguably, firms need high qualityrealization mechanisms to manage the transition between a potential and arealized process innovation capability [39], [40] rather than performing processinnovation work in an ad hoc manner.

We argue that a firm’s process innovation capabilities should be viewed ascapabilities rather than dynamic capabilities. Winter defines a capability as “ahigh level routine that, together with its implementing input flows, confersupon an organization’s management a set of decision options for producingsignificant outputs of a particular type” [35, p. 983]. Dynamic capabilities areset to affect overall organizational change and define a firm’s path of evolution[33]. Dynamic capabilities could thus configure other capabilities, for examplein the context of production [34]. Figure 1 presents a conceptual framework offirms’ process innovation capability.

Figure 1: A conceptual framework of firms’ process innovation capability

Potential

PROCESS INNOVATION CAPABILITY

Realized

Proficiency in realizing process innovation solutions

ANTECEDENTS TO FIRMS’ POTENTIALPROCESS INNOVATION CAPABILITY

Investment in new process technology

Alignment between process innovation and strategy

Top management support

Product and process innovation integration

Project selection and portfolio balancing

Collaboration with external partners

Collaboration among functions and departments

Organizational climate and culture

Organizational learning

Quality of relationship between labor and management

PROCESS INNOVATION OUTCOMES

Cost reductions

Increased production volumes

Increased quality and reliability

More environmentally friendly production

Reduced time-to-market

Facilitated production ramp-up

A. Antecedents to a firm’s potential process innovation capabilityA number of key antecedent factors have been identified which all have apositive influence on a firm’s potential process innovation capability. Table 1provides a comprehensive overview of the antecedent factors, and theliterature from which they were obtained.

5

Table 1: Factors antecedent to firms’ potential process innovation capability

Antecedent factors Authors

Investment in new process technology [44], [51], [49], [52], [50], [53], [14], [42],[41], [46], [45], [43], [47], [1], [48]

Alignment between processinnovation and strategy

[44], [4], [7], [54], [16], [21]

Top management support [56], [55], [16]

Product and process innovationintegration

[58], [52], [57], [60], [9], [4], [45], [59], [47],[7], [38], [21]

Project selection and portfoliobalancing

[68], [9], [22], [61], [62]

Collaboration with external partners [63], [64], [7], [65], [48], [66]

Collaboration among functions anddepartments

[63], [53], [3], [68], [64], [12], [55], [67], [66]

Organizational climate and culture [70], [11], [53], [10], [69], [28], [12], [5], [65]

Organizational learning [70], [63], [11], [51], [49], [71], [53], [3], [10],[28], [59], [61], [12], [5], [62], [65]

Quality of relationship between laborand management

[3], [10], [61], [62]

Investment in new process technology. The ability of firms to develop,introduce, and expand production with new process solutions is one of themost important firm specific factors for competition [41]. The literature onprocess innovation describes two seemingly different approaches to newprocess renewal: creating and developing new process technology internally, oradopting new process technology developed elsewhere [42]. The borderbetween the two is not perfectly clear – as when modifying technologydeveloped by somebody else [43] – and they seem closely interconnected andequally important.

Process technology constitutes a significant part of a firm’s knowledge base,and has to be in place to enable process innovation [44], [45]. A firm thatinvests in new process technology and adopts it quickly and effectively will besuperior in introducing new products to the market, sustain lowerdevelopment risks, and protect its position from would be imitators [1], [46],[47]. Acquiring external technology is often dealt with in an ad hoc fashion, and

6

firms often lack the skills needed to do this in a proficient way [48]. Thus, therisk is that firms increase nothing but their costs.

Consequently, skills and prior experience are critical for investments intonew process technology [14], [43], [49], [50]. For example, new technologywhich requires low adjustments to local use will be more easily adoptedbecause the firm has relevant experience and skills [42] [49] [51]. Anotherexample of the importance of experience is found in Damanpour andGopalakrishnan [52], who found that high performance firms adopt productand process innovations more evenly than low performance firms. Finally,flexibility in both technology and organization was essential for improving theadoption process [53]. Thus, it is argued that investment in new processtechnology is an important antecedent to a firm’s potential process innovationcapability.

Alignment between process innovation and strategy. Prior research hasemphasized the importance of strategic decision making at the firm level inshaping the chances that a firm will be able to enact and sustain processinnovation [7]. This seems especially important in later or ‘systemic’ phases ofan industry’s development, where a strong focus on process innovation oftenprevails. In fact, Utterback and Abernathy [21] claim that firms’ capabilities toprocess innovate are a matter of overall strategy. An obvious example of theimportance of aligning process innovation work with strategy is the processindustries, where process innovation is the real key to competitive advantage[16].

Process innovation has a positive effect on strategy especially when firmstake a proactive stance towards process innovation [44]. A proactive stance canfor example allow a firm to move from low price commodity products to anemphasis on low cost special/customized products [44]. In a similar vein,Schroeder [54] finds that a firm’s strategic posture plays a major role in shapingprocess innovation. Finally, research shows that process innovation projectsplay an important role also in emerging strategies, by highlighting the need forchange and leading the way forward [4].

Top management support is essential to successful process innovation efforts.Top managers are the ones who decide how the firm’s resources are spent andin what strategic direction the firm is heading. Therefore, it is important tohave support and recognition from top management when process innovationprojects are conducted [55]. The skills and mind set of top managers aredecisive for achieving competitive advantages from process innovation [16],[56]. For example, the executives should be specialists in the firm’smanufacturing process and at the same time be technology literate andgenerally competent. If they possess all of the mentioned traits, then theyknow how each competitor is competing process wise and are better preparedto prevail.

7

Product and process innovation integration was defined by Ettlie as ‘…that setof focused, disciplined, rigorous practices designed to concentrate efforts onevolving concepts to market introduction’ [57, p. 1226]. Integration of thesetwo R&D activities is important, as new products often require new or modifiedprocesses, whereas new processes in turn sometimes allow product conceptsthat were previously impossible to develop. In firms developing and producingnon assembled products specifically, process innovation not only allows orrequires but rather implies or imposes immediate changes to the final product[45], [58], [59]. Therefore, for non assembled products, a change to themanufacturing process can result in significant changes to end product features[4], [9], [45]. Integration is therefore good for performance [47], [57].

Integration of product and process innovation has been examined at boththe industry level, e.g. [21], [45], and the firm level [52]. Prior research at thefirm level shows that firms which innovate in products are also likely toinnovate in processes [7], [52].

At the industry level, the seminal work of Utterback and Abernathy [21] hasprofoundly shaped the thinking on product and process innovation integration.In essence, as an industry matures, the focus of innovative activity graduallyshifts from product to process [60]. These patterns will not necessarily hold fornon assembled products, however. Utterback [38] suggests the same basicpattern for non assembled products, with the difference that these kinds ofproducts have a lower relative intensity of product innovation and a higherrelative intensity of process innovation. In either case, the fact that the locus ofinnovation shifts with the stage of development provides clues on how toallocate the R&D budget, as the type of innovation likely to succeed varies overthe three stages [21].

Project selection and portfolio balancing. Most if not all firms have scarceresources and cannot do everything, and therefore they need proficiency inselecting the ‘right’ process innovation projects to invest in [61]. Furthermore,it is important to have an appropriate mix of different innovation projects, i.e.portfolio balancing [9]. Specifically, firms need to balance ‘production’ and‘development’ because too much focus on short sighted productivity issues canreduce the space for development, learning and innovation [62]. Conversely,radical and long term projects cannot dominate solely either, and it seemsindeed complicated to find the appropriate balance.

To try to select the project that provides the best return on investment isone way to evaluate what should be developed or not [61]. A firm’s strategy isanother evaluation criterion and it is often the key determinant for theallocation of resources [9]. Consequently, Lager [22] presents a classificationmatrix for process innovation projects which can be used as a tool for thecreation of a firm’s R&D strategy and selection of a well balanced portfolio ofprocess innovation projects, which appears essential to performance.

8

Collaboration with external partners. The literature on inter organizationalcollaboration for process innovation seems to be dominated by the use ofsuppliers as partners [7]. For example, the key to becoming a successfulsupplier (seller) of process innovation solutions is to engage in relationshipactivities and actively involve potential buyers in the commercialization process[63]. Similarly, Reichstein and Salter [7] concluded that suppliers are the bestsource of knowledge if process innovation is the goal. However, Linton [64] listsa couple of other important partners, such as other competing firms, noncompeting firms, customers and consultants, all of which are important whencreating and implementing process innovations.

The ability to gain knowledge from external partners seems to be frequentlyaddressed as a key issue when engaging in inter organizational collaborationfor developing process innovations [48], [65], [66]. Sawy et al. [65] found thatpersons outside an organization can function as knowledge catalysts thattrigger and facilitate innovative ideas. Since the development process often isassociated with insufficient resources, high investment costs and hightechnological uncertainty, the use of inter organizational collaboration (e.g.technical alliances) can strengthen a firm’s potential process innovationcapability [67].

Collaboration among functions and departments is a way to ensure that theexpertise of each relevant department or function is incorporated into theprocess innovation team. Indeed, several authors identify cross functionalcollaboration as critical for managing process innovation proficiently [12], [53],[63], [64], [68].

Several factors have been found to facilitate intra organizationalcollaboration in the context of process innovation. For example, a shared visionamong team members and intra personnel flow (i.e. job rotation and mobility)are crucial in cross functional work teams [12], [53], [55]. The objective ofcross functional collaboration is to ensure the right mix of skills andcompetences when developing process innovation. However, cross functionalteams are not suitable in all technological change processes – it seems todepend on the novelty and type of uncertainty involved in the productionprocess [66]. Research by Tyre and Hauptman (1992) revealed that in highlytechnical and complex projects the use of cross functional teams was lessbeneficial because such teams could not manage the type of uncertaintyinvolved.

As process innovations are systemic by nature [3], several differentfunctions and departments within the firm are typically involved in both thedesign and implementation of process innovation work. Therefore,collaboration among functions and departments is a critical antecedent tofirms’ potential process innovation capability.

9

Organizational climate and culture. The literature on climate and culture isfairly homogeneous, and most of the reviewed articles highlight organizationalculture and climate as a prerequisite for successful process innovation. Some ofthe papers, like Lager and Hörte [69], simply mention the need for a companyto have a good and stimulating climate for process innovation work, whileother scholars discuss organizational climate and culture in depth.

Climate is defined as a broad class of organizational, rather thanpsychological, variables which describe the organizational context forindividuals’ actions. Culture, on the other hand, is on a higher level ofabstraction and refers to a deeper set of meanings, basic beliefs and values[11]. Climate and culture are not mutually exclusive, and both refer to thecreation, implementation and replication of organizational routines andorganizational knowledge [5], [28].

Khazanchi et al. [10], for example, stress organizational culture as a key tomanaging process innovation. A culture of experimentation, freedom,playfulness and empowerment is crucial to enhancing plant performance [12],[70]. Pearce and Ensley [12] specifically highlight employee empowerment,team creativity and shared vision of the process innovation team, becausethese enhance the innovation process and support idea generation. Accordingto Baer and Frese [11], many process innovations often lack the human andorganizational side of the socio technical system, which is one importantexplanation for why many process innovations fail to provide the expectedpositive outcomes.

Baer and Frese [11] studied specifically two dimensions of organizationalclimate: the climate for initiative and for psychological safety. Their resultsshow that firms with a high degree of process innovativeness but with lowlevels of climate for initiative and psychological safety came out worse than ifthey had not innovated at all. In other words: innovation is not enough and inorder to work it needs to be complemented by an adequate climate forinitiative and psychological safety. Therefore firms need to develop anorganizational climate where people feel safe in taking interpersonal risks, andare encouraged to propose new ideas, openly discuss problems, andproactively approach work. In essence, for process innovation to work, it isimportant to create shared beliefs, a knowledge sharing climate/culture,altruism and courtesy.

Organizational learning. Organizational learning appears critical to processinnovation performance [10], [11], [28].

Although all scholars agree on the importance of organizational learning,there are visible differences in perspectives. Some scholars have a ratherpragmatic definition of learning, where it is viewed as an easy to handlemethod for achieving positive results. For example, employee learning isessential in process innovation [61], and Athaide et al. [63] relate learning toeducation and training when discussing the seller buyer interaction during

10

commercialization of process innovations. Authors in this group generally seelearning as something that can be measured, e.g. Macher and Mowery [59]measure knowledge transfer and learning by doing (the rate of learning) bylooking at participating companies’ manufacturing yield. Cabral and Leiblein’s[51] argument on ‘learning by doing’ provides another example of thisperspective. They have studied semiconductor firms’ experiences with newtechnology adoption and found that the relationship between productionexperience and new technology adoption was relatively short lived. Theirfindings indicate that ‘knowledge’ derived from learning by doing rapidlybecomes obsolete.

In a related but somewhat more complex perspective, learning is seen assomething that can be planned, arranged and controlled for. This stance issometimes referred to as knowledge management. Gopalakrishnan andcolleagues [3] employed this perspective when they found that processinnovations are associated with knowledge that may be embedded and storedin equipment, tools, organizational systems, operating procedures, routines,and the individual operators. Although all knowledge is initially created byindividuals, it does not become ‘organizational knowledge’ until it is transferredthroughout the organization, where it can be stored in the organization’s‘memory’ and institutionalized. Within this perspective, learning is usually seenas something that is possible to transfer. For example, knowledge can betransferred from one plant to another, or it can be transferred within the sameplant [49], [51], [53], [59], [71].

A third stream of literature targets learning as a process, and the articlesfocus on continuous generation of knowledge and innovation within theorganization (and not just adoption to new technology). The people, the actualwork places, the organization, the communication and the work tasks are ‘thelearning environment’, the space where learning happens and is created.Repenning and Sterman [62], for example, suggest that the criticaldeterminants of success in efforts to learn and improve are the interactionsbetween managers’ explanations for poor organizational performance and thephysical structure of the workplace. Similarly, Sawy et al. [65] conclude thathaving an organizational environment that encourages knowledge creation andsharing will lead to more successful process innovation projects.

Regardless of which perspective is advocated, the literature is quiteunanimous about what types of organizational system and structures supportand stimulate learning. Findings suggest team based organizations,cooperation and integration of work tasks and competences, and there are alsostrong strains of experiment friendly and creative climate, dialogue,empowerment, discretion, decentralization and broad participation over a widerange of parts of the organization, and system thinking [5], [12], [53], [59], [61],[70]. A key conclusion that can be drawn is that for learning to increase thefirm’s potential process innovation capability, people must be given the

11

opportunity to develop and create ideas together and the organization mustfacilitate the flow of ideas and information.

Quality of relationship between labor and management. Good workingrelationships between management and the operating personnel arepresented as a key factor for managing process innovation [62]. A goodrelationship is characterized by mutual understanding of the other party’sperspective, shared beliefs and values, good communication, and a facilitatingsocial and cognitive environment [10], [61]. Further, it is the basic foundationof process innovation, and if the relationship is not working, it is highly likelythat the process innovation project will result in failure. Process innovationoften requires both technological and organizational changes, whichpresupposes high quality relationships [3].

According to a study by Repenning and Sterman [62], managers’ beliefsabout those who work for them are not always in line with the employees’perceptions, which can cause severe problems. O’Hara and colleagues [61] alsoaddress the need for a good relationship between managers and the operatingpersonnel. They suggest that it can be achieved by involving the operatingpersonnel early in a development process, as this will lead to more engagedteam members [61].

In a similar vein, a few authors have pointed out the importance of engagingand empowering the operating personnel [10], [61]. For example, Khazanchiand colleagues [10] have suggested that operating personnel should work withincremental process innovations, as they are the ones with the appropriateskills and knowledge. In essence, with involvement comes greater satisfactionand lower resistance to change.

B. The difference between potential and realized process innovation capabilitiesFor process innovation to occur successfully, firms must be proficient in the‘actual process’ of taking an idea or blueprint for a new process innovation, andsubsequently implementing it in the production process, i.e. ‘realizing it’. Thisrequires a ‘process for process innovation’, but also specific capabilities of thefirm. We argue that the antecedent factors previously elaborated determine afirm’s potential process innovation capability. A firm’s proficiency in realizationdetermines the extent to which the potential capability will resemble therealized one. Realization of process innovation therefore resembles LeonardBarton’s [28] notion of a core capability, as it can differentiate a companystrategically.

Generally, process innovation has less well developed work processescompared to product innovation [1], although some authors have addressedthis topic. Pisano [6], for example, splits the “process development process”into three phases: process research, pilot development, and commercial plantscale up. In the first stage, a description of the product concept is made, whichin most cases will be incomplete. The product concept is the formula for the

12

changes required to the end product, and it further includes a set of functionaltarget specifications. At this first stage, development work is generallyperformed in laboratory settings, and scientific literature is used as a source toenhance knowledge of key problems [6]. The next stage is referred to as pilotdevelopment. This stage is more “empirical” and focuses on scaling up theprocess to an intermediate level and anticipating key process parameters. Thethird and final stage involves scaling up the process created to full scale andtransferring it to the plant where the process will be adapted and used toproduce the product [5]. A similar process has been suggested by Lager [68],and a somewhat more iterative model by Lim and colleagues [4].

Within this ‘process innovation process’, several different activities areneeded for the firm to eventually succeed with realization. These includeproficiency in idea generation and project formulation, individuals with suitableskills, knowledge and experience, good networks for research and technicaldevelopment, and adequate incentives for process innovation [13], [69]. Allthese enable process innovation realization, and in addition, firms need toconsider the overall process design and resources to enable realization ofprocess innovation projects [4]. Another important capability is problemsolving when dealing with technological change [66]. In addition, informationon product performance needs to be gathered to detect bugs, education andtraining of production staff are required, and annual services and upgradesneed to be provided. Moreover, support from various parts of the organizationneeds to be ensured [63].

C. Process innovation outcomesFinally, we argue that a firm’s realized process innovation capability determinesthe extent to which desired outcomes actually are achieved. These outcomesvary according to which type of process innovation project is being pursued,and include for example cost reductions and increases in productivity [5], [6],increased production volumes [4], and increased product quality and reliability[3]. Other key indicators include more environmentally friendly production [9]and reduction of the time to market period [1].

IV. DISCUSSIONDespite the importance of process innovation for firm performance, relativelyfew studies address this topic in comparison with product innovation [1], [7].Although several authors have called for a more firmly grounded understandingof how process innovation capabilities are created and sustained [1], [15] suchknowledge remains scarce. We contribute to this literature by addressing theproblem of how to manage process innovation more proficiently so that itspositive benefits can be better realized.

13

Both potential and realized process innovation capabilities are essential forachieving competitive advantage, but they have complementary roles. Firmsneed to acquire and assimilate both tangible and intangible resources, whichare represented by the antecedent factors. Altogether, these factors determinea firm’s potential process innovation capability. In addition, the firm mustpossess the capabilities for realizing process innovations to achieve desiredoutcomes. However, proficiency in realization alone does not guarantee theexploitation of critical knowledge [34]. The firm must be good at both acquiringand assimilating resources and competences, as well as at transforming andexploiting them in the context of production.

Process innovation has generally less well developed work processes thanproduct innovation [1], and we argue that proficiency in realizing processinnovation is essential for attaining the desired outcomes. This idea extendsprior research into work processes for process innovation [4], [68].

Factors antecedent to a firm’s process innovation capability are recognizedand presented in the above framework, by means of an extensive review andanalysis of the literature. These factors reflect the fact that process innovationspans both organizational and technological issues. Technology related factorssuch as new tools and process equipment are critical components [49], [51] butthey should be complemented by organizational factors and strategicconsiderations. For example, firms need to possess the mechanisms necessaryto integrate process and product innovation, and align process innovationefforts with the firm’s strategy [44], [47], [57]. Knowledge and learningcapabilities of employees are another non technical issue of critical importance[14], [43], [50].

V. ADDITIONAL REFLECTIONS ON THE LITERATUREA main feature of the literature is that the articles included are heavily coloredby the time period in which they were written. Therefore, they reflect theprevailing spirit of place, time and discourse in management and organizationtheory as well as production problems. For example, articles from the early1990s often discuss or advocate a ‘Japanese way of doing things’, a trend whichgained momentum during the 1980s. Japanese modes of work organizationwere highlighted as the savior for European and U.S. manufacturing industry,with an emphasis on flexible production technology and lean production. Themain focus in the studied articles from this period was on work organization,but concepts such as team organization, integration of functions, crossfunctional teams, process orientation and customer focus were also central[14], [53], [61]. During the 1990s, it became fashionable to address the firm ‘asa whole’, with a focus on both production and organization. Central concepts inthe studied articles from this period were corporate culture, innovation,entrepreneurship and learning [10], [11], [12], [65]. Focus was often oncreating independent, multi skilled and committed employees in order to

14

maintain a flexible and learning organization. Far reaching thoughts that allpersonnel should work in teams – technical experts as well as productionworkers – can be spotted. In manufacturing industry today, the pendulumseems to have swung back to an emphasis on the assembly line,standardization and lean production. But these trends and eventualconsequences for the possibilities of managing successful process innovationare not reflected in any of these articles.

Regarding the empirical studies in the reviewed articles, they are quitebalanced in terms of quantitative and qualitative studies. A third of all thereviewed articles combine these two approaches. However, a weakness is thatthe majority of the reviewed articles focus on top or middle level managers,often with a background in research and development. Consequently, theyreflect their attitudes and perceptions on what is important and fundamentalfor managing process innovation. Few if any articles focus on the shop floorworkers who actually make most of process innovation happen. The literatureis thus biased towards a ‘top down’ rather than a ‘bottom up’ perspective, andthe end result is sometimes a scratch on the surface rather than acomprehensive picture of how process innovation really happens. It is likelythat managers follow the logic of appropriateness – answers to questions arephrased to be accepted by an audience residing within the same institutionalsetup. Instead of ‘staying in touch with reality’, such actions allow respondentsto stay ‘in touch with their identity’ [72, p. 161]. To increase our knowledge ofhow process innovation can be better managed, more diverse samples arecalled for in future research.

VI. CONCLUSIONS, LIMITATIONS, AND OUTLOOKThis article contributes to the literature on managing process innovation inseveral ways. First, the distinction between a firm’s potential and realizedprocess innovation capabilities provides new insights into how processinnovation is fundamentally enabled. Second, by identifying the antecedents toa firm’s potential process innovation capability, we highlight the key domainswhich tend to determine the potential process innovation capability of firms.Finally, we emphasize the importance of having high quality realizationmechanisms, which has been underplayed and overlooked in the extantliterature on managing process innovation.

In addition, this article provides a couple of important managerialimplications. The antecedent factors offer guidance on how firms can go aboutenhancing their potential process innovation capability. This is not sufficient,however. By highlighting the need for high quality realization mechanisms, westress that process innovation efficiency, similarly to product innovation, isultimately enabled through planned, structured and formalized workprocesses. Together, these findings allow managers to enhance the extent towhich process innovation efficiency is fulfilled.

15

Some important limitations are also worth mentioning. First, the frameworksuggested is a conceptual construction. Although grounded in the extantliterature, empirical testing is necessary. Second, the antecedent factorspresented do not constitute an exhaustive list, and certain key factors mayhave been overlooked. Third, the antecedent factors may not be mutuallyexclusive, and covariance among factors can be expected, which needs to beaddressed in subsequent empirical research. Another limitation is the exclusionof literature on more incremental and continuous process improvementefforts, such as total quality management, kaizen, lean production, and ISO9000. While we do recognize that such efforts and improvements may escalateinto significant ones over time, this literature was disregarded for reasons ofboth definition and scope, i.e. our intention to review works that treat moresignificant changes in the production process. Finally, we also recognize thatsome of the antecedent factors are more firmly grounded in the literature thanothers. The fact that some are more heavily researched than others is notnecessarily a sign of their being unimportant, but rather a sign of moreresearch being needed in this domain.

A first and obvious suggestion for further research is thus to test theconceptual model empirically, preferably through a survey. This impliesdeveloping valid measures for the constructs included, to examine covarianceamong the antecedent factors and establish their relative importance, tocarefully select one or a few dependent variables, and to assess theexplanatory power of the model. In addition, more research is needed into theissue of how to manage the realization process. While this issue has attractedconsiderable attention in product innovation by means of research into stagegate processes [73], [74], knowledge of formalized work processes to enableprocess innovation is limited. Qualitative, smaller sample studies are likely tobe more appropriate for understanding the realization process. In addition, weencourage more work on integration of product innovation and processinnovation. The distinction between these in manufacturing firms is to someextent artificial, as they are linked in the firm’s technological core. Both productand process innovation are often needed within the same innovation project,and it is somewhat surprising to see that this issue has attracted relatively littleattention.

Suggestions for further research are not only limited to issues present in theextant literature, however, but should also address topics not currentlyconfronted. First, pre development activities in process innovation seemneglected, although the importance of such activities has been noted [4] andresearch into the “fuzzy front end” of product innovation has revealed that thefoundation of failure often lies in the fuzzy front end [75] , [76], [77]. Furtherresearch into the process innovation domain would benefit from studying theearly, essentially creative parts of idea generation and concept formation in theprocess innovation context. Finally, as a counterweight to the current focus onfirm level constructs, we urge future studies on the microfoundations of

16

managing process innovation. Currently a key issue in both strategicmanagement and entrepreneurship [26], the distinct skills, procedures anddecision rules which underlie firm level sensing in process innovation remainlargely unexplored.

17

REFERENCES[1] G.P. Pisano, The development factory: unlocking the potential of process

innovation, Boston, MA: Harvard Business School Press, 1997.[2] M.L. Tushman and D. Nadler, “Organizing for Innovation”, California

Management Review, vol. 28, no. 3, pp. 74 92, 1986.[3] S. Gopalakrishnan, P. Bierly and E.H. Kessler, ”A reexamination of product

and process innovations using a knowledge based view”, Journal of HighTechnology Management, vol. 10, no. 1, pp. 147 166, 1999.

[4] L.P.L. Lim, E. Garnsey and M. Gregory, “Product and process innovation inbiopharmaceuticals: A new perspective on development”, R&DManagement, vol. 36, no. 1, pp. 27 36, 2006.

[5] G.P. Pisano, “Knowledge, integration, and the locus of learning: an empiricalanalysis of process development”, Strategic Management Journal, vol. 15,pp. 85 100, 1994.

[6] G.P. Pisano, “Learning before doing in the development of new processtechnology”, Research Policy, vol. 25, no.7, pp. 1097 1119, 1996.

[7] T. Reichstein and A. Salter, “Investigating the sources of process innovationamongUK manufacturing firms”, Industrial and Corporate Change, vol. 15, no. 4,pp. 653 682, 2006.

[8] K. Goffin and R. Mitchell, Innovation Management, Strategy andImplementation

Using the Pentathlon Framework. New York, Palgrave Macmillan, 2005.[9] T. Lager, “Product and process development intensity in process industry: A

conceptual and empirical analysis of the allocation of company resources forthe development of process technology”, International Journal of InnovationManagement, vol. 6, no. 2, pp. 105 130, 2002.

[10] S. Khazanchi, M.W. Lewis and K.K Boyer, “Innovation supportive culture:the impact of organizational values on process innovation”, Journal ofOperations Management, vol. 25, no. 4, pp. 871 884, 2007.

[11] M. Baer and M. Frese, “Innovation is not enough: climates for initiativeand psychological safety, process innovations, and firm performance”,Journal of Organizational Behaviour, vol. 24, no. 1, pp. 45 68, 2003.

[12] C.L. Pearce and M.D. Ensley, “A reciprocal and longitudinal investigation ofthe innovation process: The central role of shared vision in product andprocess innovation teams (PPITs)”, Journal of Organizational Behavior, vol.25, no. 2, pp. 259 278, 2004.

[13] P.D. Collins, J. Hage and F.M. Hull, “Organizational and TechnologicalPredictors ofChange in Automacity”, The Academy of Management Journal, vol. 31, no. 3,pp. 512 543, 1988.

18

[14] J.E. Ettlie and E.M. Reza, “Organizational integration and processinnovation”, Academy of Management Journal, vol. 35, no. 4, pp. 795 827,1992.

[15] Q. Lu and B. Botha, “Process development: a theoretical framework”,International Journal of Production Research, vol. 15, no. 1, pp. 2977 2996,2006.

[16] W. Skinner, “The shareholder’s delight: companies that achievecompetitive advantage from process innovation”, International Journal ofTechnology Management, vol. 7, no. 1/2/3, pp. 41 48, 1992.

[17] S.L. Brown and K.M. Eisenhardt, “Product development: past research,present findings, and future directions, Academy of Management Review,vol. 20, no. 2, pp. 343 378, 1995.

[18] R. Balachandra and J.H. Friar, “Factors for Success in R&D Projects andNew Product Innovation: A Contextual Framework”, IEEE Transactions onEngineering Management, vol. 44, no. 3, pp.276 287, 1997.

[19] V. Krishnan and K.T. Ulrich, “Product development decisions: a review ofthe Literature”,Management Science, vol. 47, no. 1, pp. 1 21, 2001.

[20] M.M. Montoya Weiss and R. Calantone, “Determinants of New ProductPerformance: A Review and Meta Analysis”, Journal of Product InnovationManagement, vol. 11, no.5, pp. 397 417, 1994.

[21] J.M. Utterback and W.J. Abernathy, “A dynamic model of process andproduct Innovation”, Omega, vol. 3, no.6, pp. 639 656, 1975.

[22] T. Lager, “A structural analysis of process development in processindustry”, R&D Management, vol. 32, no. 1, pp. 87 95, 2002.

[23] E. Danneels, “The Dynamics of Product Innovation and FirmCompetences”, Strategic Management Journal, vol. 23, no. 12, pp. 10951121, 2002.

[24] K.M. Eisenhardt and J.A. Martin, “Dynamic capabilities: what are they?”Strategic Management Journal, vol. 21 no. 10/11, pp. 1105 1121, 2000.

[25] J. Jayaram and R. Narasimhan, “The influence of New ProductDevelopment Competitive Capabilities on Project Performance”, IEEEtransactions on Engineering Management, vol. 54, no. 2, pp.241 256, 2007.

[26] D.J. Teece, “Explicating dynamic capabilities: The nature andmicrofoundations of (sustainable) enterprise performance”, StrategicManagement Journal, vol. 28, no. 13,pp. 1319 1350, 2007.

[27] D. Teece, Dynamic capabilities and strategic management – Organizing forinnovation and growth, Oxford University Press, 2009.

[28] D. Leonard Barton, “Core capabilities and core rigidities: a paradox inmanaging new product development”, Strategic Management Journal, vol.13, Special Issue: Strategy Processes: Managing Corporate Self Renewal, pp.111 125, 1992.

19

[29] A. Mantovani, “Complementarity between product and process innovationin a monopoly setting”, Economics of Innovation & New Technology, vol. 15,no. 3, pp. 219 234, 2006.

[30] E. Huergo and J. Jaumandreu, “Firms’ age, process innovation andproductivity Growth”, International Journal of Industrial Organization, vol.22, no. 4, pp. 541 559, 2004.

[31] K.W. Bender, J.E. Cedeño, J.F. Cirone, K.P. Klaus, L.C. Leahey and T.D.Menyhert, ”Process innovation – case studies of critical success factors”,Engineering Management Journal, vol. 12, no. 4, pp. 17 24, 2000.

[32] M.J. Benner and M. Tushman, “Process management and technologicalinnovation: a

longitudinal study of the photography and paint industries”, AdministrativeScience

Quarterly, vol. 47, no. 4, pp. 676 706, 2002.[33] D.J. Teece, G. Pisano, and A. Shuen, “Dynamic capabilities and strategic

management”, Strategic Management Journal, vol. 18, no. 7, pp. 509 533,1997.

[34] S.A. Zahra and G.A. George, “Absorptive capacity: A review,reconceptualization, and extension”, Academy of Management Review, vol.27, no. 2, pp. 185 203, 2002.

[35] S. Winter, “The satisfying principle in capability learning”, StrategicManagement Journal, vol. 21, no. 10/11, pp. 981 996, 2000.

[36] U. Lichtenthaler, “Relative capacity: Retaining knowledge outside a firm’sboundaries”, Journal of Engineering and Technology Management, vol. 25,no. 3, pp. 200 212, 2008.

[37] U. Lichtenthaler and E. Lichtenthaler, “A capability based framework foropen innovation: Complementing absorptive capacity”, Journal ofManagement Studies, vol. 46, no. 8, pp. 1315 1338, 2009.

[38] J.M. Utterback, “Mastering the Dynamics of Innovation”, Boston, MA:Harvard Business School Press, 1994.

[39] U. Lichtenthaler and J. Frishammar, “The impact of aligning productdevelopment andtechnology licensing: a contingency perspective”, Journal of ProductInnovation Management (forthcoming).

[40] X.M. Song and M.E. Parry, “A cross national comparative study of newproduct development processes: Japan and the United States”, Journal ofMarketing, vol. 61, no. 2, pp. 1 18, 1997.

[41] N.W. Hatch and D.C. Mowery, “Process innovation and learning by doing insemiconductor manufacturing”, Management Science, vol. 44, no. 11, pp.1461 1477, 1998.

20

[42] S. Gopalakrishnan and F. Damanpour, “Patterns of generation andadoption of innovation in organizations: Contingency models of innovationattributes”, Journal of Engineering and Technology Management, vol. 11 no.2, pp. 95 116, 1994.

[43] P.W Meyer, K. Sivakumar and C. Nakata, “Implementation of IndustrialProcess Innovations: Factors, Effects, and Marketing Implications”, Journalof Product Innovation Management, vol. 16, no. 3, pp. 295 311, 1999.

[44] S. Brown, ”Managing process technology – further empirical evidence frommanufacturing plants”, Technovation, vol. 21 no. 8, pp. 467 478, 2001.

[45] J.D.Linton and S.T. Walsh, “A theory of innovation for process basedinnovations such as nanotechnology”, Technological Forecasting & SocialChange, vol. 75, no. 5, pp. 583 594, 2008.

[46] M.R. Kelley and H. Brooks, “External Learning Opportunities and theDiffusion of Process Innovations to Small Firms”, Technological Forecastingand Social Change, vol. 39 no. 1 2, pp. 103 125, 1991.

[47] G.P. Pisano and S.C. Wheelwright, “The new logic of high tech R&D”,Harvard Business Review, vol. 73, no. 5, pp. 93 105, 1995.

[48] G.N. Stock and M.V. Tatikonda, “External technology integration in productand process development”, International Journal of Operations andProductions Management, vol. 24, no. 7, pp. 642 665, 2004.

[49] K H. Chai, M.J. Gregory and Y.S. Shi, “An exploratory study of intrafirmprocess innovations transfer in Asia”, IEEE Transactions on EngineeringManagement, vol. 51, no. 3, pp. 364 374, 2004.

[50] J.E. Ettlie, “Intrafirm Mobility and Manufacturing Modernization”, Journalof Engineering and Technology Management, vol. 6, no. 3 4, pp. 281 302,1990.

[51] R. Cabral and M.J. Leiblein, “Adoption of process innovation with learningby doing: Evidence from the semiconductor industry”, Journal of IndustrialEconomics, vol. XLIX, no. 3, pp. 269 280, 2001.

[52] F. Damanpour and S. Gopalakrishnan, “The dynamics of the adoption ofproduct and process innovations in organizations”, Journal of ManagementStudies, vol. 38, no. 1, pp. 45 65, 2001.

[53] J.E. Ettlie and J.D. Penner Hahn, ”Adoption complexity and economies ofscope for new process technology in manufacturing”, Journal of HighTechnology Management, vol. 5, no. 1, pp. 19 38, 1994.

[54] D.M. Schroeder, “A dynamic perspective on the impact of processinnovation upon competitive strategies”, Strategic Management Journal,vol. 11, no. 1, pp. 25 41, 1990.

[55] S.G. Scott, “Social identification effects in product and processdevelopment teams”, Journal of Engineering and Technology Management,vol. 14, no. 2, pp. 97 127, 1997.

21

[56] E. Lefebvre and L.A. Lefebvre, “Firm innovativeness and CEO characteristicsin small manufacturing firms”, Journal of Engineering and TechnologyManagement, vol. 9, no. 3 4, pp. 243 277, 1992.

[57] J.E. Ettlie, “Product process development integration in manufacturing”,Management Science, vol. 41, no. 7, pp. 1224 1236, 1995.

[58] B.D. Barnett and K.B. Clark, “Technological newness: an empirical study inthe process industries”, Journal of Engineering and TechnologyManagement, vol. 13, no. 3 4, pp. 263 282, 1996.

[59] J.T. Macher and D.C. Mowery, “Managing learning by doing: an empiricalstudy in semiconductor manufacturing”, Journal of Product InnovationManagement, vol. 20, no. 5, pp. 391 410, 2003.

[60] P. Hutcheson, A.W. Pearson and D.F. Ball, “Innovation in process plant: acase study of ethylene”, Journal of Product Innovation Management, vol. 12,no. 5, pp. 415 430, 1995.

[61] J.P. O'Hara, H.E. Evans and T.F. Hayden, “Developing new manufacturingprocesses: a case study and model”, Journal of Engineering and TechnologyManagement, vol. 10, no. 3, pp. 285 306, 1993.

[62] N.P. Repenning and J.D. Sterman, “Capability traps and self confirmingerrors in the dynamics of process improvement”, Administrative ScienceQuarterly, vol. 47, no. 2, pp. 265 295, 2002.

[63] G.A. Athaide, P.W. Meyers and D.L. Wilemon, “Seller buyer interactionsduring the commercialization of technology process innovations”, Journal ofProduct Innovation Management, vol. 13, no. 5, pp. 406 421, 1996.

[64] J.D. Linton, “The role of relationships and reciprocity in theimplementation of process innovation”, Engineering Management Journal,vol. 12, no. 3, pp. 34 38, 2000.

[65] O.A.E. Sawy, I. Eriksson, A. Raven and S. Carlsson, “Understanding sharedknowledge creation spaces around business processes: precursors toprocess innovation implementation”, International Journal of TechnologyManagement, vol. 22 No. 1/2/3, pp. 149 173, 2001.

[66] M.J.Tyre and O. Hauptmann, “Effectiveness of organizational responses totechnological change in the production process”, Organization Science, vol.3, no. 3, pp. 301 320, 1992.

[67] F.K. Sen and W.G. Egelhoff, “Innovation Capabilities of a Firm and the Useof Technical Alliances”, IEEE Transactions on Engineering Management, vol.47, no. 2, May 2000.

[68] T. Lager, “A new conceptual model for the development of processtechnology in process industry”, International Journal of InnovationManagement, vol. 4, no. 3, pp. 319 346, 2000.

22

[69] T. Lager and S Å. Hörte, “Success factors for improvement and innovationof process technology in process industry”, Integrated ManufacturingSystems, vol. 13, no. 3, pp. 158 164, 2002.

[70] A.E. Akgün, H. Keskin and J. Byrne, “Organizational emotional capability,product and process innovation, and firm performance: An empiricalanalysis”, Journal of Engineering and Technology Management, vol. 26,no.3, pp. 103 130, 2009.

[71] R.D. Dewar and J.E. Dutton, “The adoption of radical and incrementalinnovations: an empirical analysis”, Management Science, vol. 32, no. 11,pp. 1422 1433, 1986.

[72] J.G. March and J.P. Olsen, Rediscovering Institutions. The OrganizationalBasis of Politics, Free Press: New York, 1989

[73] R. Cooper and E.J. Kleinschmidt, “Benchmarking the firm’s critical successfactors in new product development”, Journal of Product InnovationManagement, vol. 12, no. 5, pp. 374 391, 1995.

[74] R. Cooper, “Perspective: The stage gate idea to launch process – update,what’s new, and nextgen systems”, Journal of Product InnovationManagement, vol. 25, no. 3, pp. 213 232, 2008.

[75] R. Cooper, “Predevelopment activities determine new productsuccess”, Industrial Marketing Management, vol. 17, no. 3, pp. 237 247,1988.

[76] J. Kim and D. Wilemon, “Focusing the fuzzy front end in new productdevelopment”, R&D Management, vol. 32, no. 4, pp. 269 279, 2002.

[77] A. Khurana and S.R. Rosenthal, “Towards holistic "front ends" in newproduct development”, Journal of Product Innovation Management, vol. 15,no. 1, pp. 57 74, 1998.

23


Paper A


Paper B



Paper C



Paper D


Paper E

The Fuzzy Front End in Non-assembled Product

Development: A multiple case study of mineral- and metal

firms

Johan Frishammar

Luleå University of Technology Center for Management of Innovation and Technology

in Process Industry SE-97187 Luleå; Sweden Phone: +46-920491407

Email: [email protected]

Ulrich Lichtenthaler WHU - Otto Beisheim School of Management

Innovation and Organization Burgplatz 2; D-56179 Vallendar; Germany

Phone: +49-(0)261-6509-411 Email: [email protected]

Monika Kurkkio Luleå University of Technology

Center for Management of Innovation and Technology in Process Industry

SE-97187 Luleå; Sweden Phone: +46-920491865


Abstract

This paper examines front-end practices under conditions of non-assembled product development, and show that front-end activities and managerial challenges diverge significantly from previously reported findings in the technology- and innovation management literature. We conceptualize the front end in non-assembled product development to be an iterative trial-and-error process, dominated by activities such as analysis of input materials, identification of process constraints, laboratory tests, and batch tests with customers. In addition, we identify the key managerial challenges for increasing both the speed and quality of the FFE process. Our findings have major implications for how R&D managers can increase the success and quality of emerging product concepts.

Keywords: Fuzzy front end, non-assembled products, process firms, case-study

1. Introduction

The fuzzy front end (FFE) is defined as the period between when an opportunity for a new product is first considered, and when the product idea is judged ready to enter formal development (Kim and Wilemon, 2002). Although new product projects often fail at the end of a development process, or during the subsequent commercialization phase, the foundations for failure seem often to be established at the very beginning (Cooper, 1998). Thus, proficiency in conducting the FFE is of utmost importance for the whole new product development process.

Khurana and Rosenthal (1997) have shown that the FFE is a crossroads of complex information processing, tacit knowledge, conflicting organizational pressures, and considerable uncertainty. In addition, the FFE is often ill-defined and characterized by ad-hoc decision-making in many firms (Montoya-Weiss and O’Driscoll, 2000). This makes mistakes hard to avoid, and the search for better predictive guidelines justifiable.

Although knowledge of the front end is scarce compared to knowledge of the subsequent “formal” product development phase, important insights have been made by previous scholars. The theoretical and managerial interest in the FFE has particularly grown in recent years because several works have underscored the importance of these activities. The FFE has been studied conceptually (Chang et al., 2007, Kim and Wilemon, 2002; Reid and De Brentani, 2004), and in the context of both services (Alam, 2006) and assembled products (e.g. Khurana and Rosenthal, 1997; 1998; Murphy and Kumar, 1997; Verganti, 1997; Verworn et al., 2008). Research on the FFE in non-assembled product development is non-existent, however.

There is a strong deficit of research into the FFE of non-assembled product development for a couple of reasons. First and foremost, the specific contingency characteristics of firms developing and producing non-assembled products make previous research results difficult to transfer. Firms developing non-assembled products prevail within the mineral- and metals, pulp and paper, chemicals, and food industries, among others, and differ in many ways from other manufacturing industries. In particular, they have the following typical characteristics. Incoming material is often raw, rather than consisting of components from suppliers; the production plants are typically very large and strongly integrated on one site; and the production process is often continuous with on-line control in real time (Hutcheson, Pearson and Ball, 1995).

Furthermore, non-assembled products are typically used in somebody else’s subsequent production, thus creating strong supply chain dependencies. In addition, product life-cycles are often very long; the production process is inert, inflexible and difficult to change, and also very capital-intensive (Hutcheson et al., 1995; Lager, 2002a; 2002b). Furthermore, high degrees of interdependence between product development and process development exist (Linton and Walsh, 2008). That is, changes of the production process frequently result in changes of the end product. Conversely, new product concepts may require significant changes of the production process.

2

Second, the literature on what drives success in the FFE is scattered because a variety of potential activities and success factors prevail in the extant literature, with only a vague understanding of their relative importance. Many of these factors also seem context-dependent (i.e. some apply to physical products, some apply to services only, and some may apply to both types).

Third, better knowledge of which factors are important would not only allow firms developing non-assembled products to enact better product concepts for subsequent development, but could also increase speed of development and cut costs (Smith and Reinertsen, 1992) as well as increasing the overall success of new product development efforts. Finally, since development of non-assembled products account for nearly one third of all R&D expenses of manufacturing firms (Lager, 2010), increased knowledge about front-end practices in such firms appears critical. There are no theoretical reasons to believe that the importance of front end activities should be contingent on the type of product being developed, i.e. assembled-, non-assembled products or services (Khurana and Rosenthal, 1998). In fact, all formal development projects first pass the important FFE phase (Cooper, 1988). However, there are reasons to believe that extant knowledge, based on studies of the FFE in assembled products, cannot be uncritically transferred to the domain of non-assembled product development, due to various industry idiosyncrasies.

The overall purpose of this article is therefore to increase knowledge about the FFE in non-assembled product development. Drawing on 36 deep interviews with R&D professionals and other data sources, we identify the key activities and issues which constitute the FFE in non-assembled product development. Moreover, we present a conceptualization of the FFE applicable to non-assembled products, and we highlight the central managerial challenges in the FFE for this type of development. Thus, this article provides insights that are equally relevant to academics and industry professionals.

The remainder of the article is organized as follows. First, we review previous literature to root our findings theoretically. Next, the research design is described, leading to a section which provides key results. The article ends with discussion, analysis, implications and an outlook towards future research avenues.

2. Theoretical background

High degrees of interdependence exist between product and process development in firms developing and manufacturing non-assembled products. In addition, product development and process development are “fuzzy sets” in such firms, meaning that a new R&D project typically spans both domains (Lim et al., 2006). For example, new product concepts frequently require changes to the production process, thus calling for process development. Conversely, process development often change end product features (Lager, 2010; Linton and Walsh,

3

2008). To acknowledge these interdependencies, our conceptual framework combines literature on managing the FFE with the literature on managing process development. To our knowledge, however, no research exists on the FFE in process development. Therefore, we draw on the general process development literature for studying FFE practices and activities. Thus, these two literatures combined serve as a feasible theoretical starting point for a study such as the current one.

This theoretical background is therefore split into three subsections. The first revisits the literature on the FFE in product development. The second reviews the literature on managing process development. Finally, the third subsection encompasses issues and ideas present in both the FFE and process development literatures.

2.1 The fuzzy front end literature

Previous literature on the FFE reveals that firms need proficiency in idea refinement and screening of ideas (Griffiths-Hemans and Grover, 2006; Khurana and Rosenthal, 1997). Deficiencies (e.g. poor ideas, too few ideas, or poor screening) often result in costly problems at later phases of the NPD process (Cooper, 1988). Screening must take place in two domains: business analysis screens a new product idea in terms of its viability as a business proposition, whereas feasibility analysis determines whether a firm can support a subsequent development project with sufficient resources (Murphy and Kumar, 1997).

Another important issue is preliminary technology assessment, i.e. an early consideration of the technology underpinning a new product. Preliminary technology assessment reduces uncertainty and highlights the technical viability of the product. Such assessment means addressing the questions of whether the product can be developed, what technical solutions will be required, and at what cost (Cooper and Kleinschmidt, 1987; Verworn, 2006).

Still another important issue is project priorities. These include making trade-offs among the three competing virtues of scope (product functionality), scheduling (timing), and resources (cost). Previous research has shown that unclear project priorities are a major cause of time delays and product over-engineering (Khurana and Rosenthal, 1997; Murphy and Kumar, 1997).

Proficient project management is another critical issue. While recognizing that FFE activities often take place without a formal project manager assigned, an effective project manager can, while present, lobby for support and resources and coordinate both technical and design issues (Khurana and Rosenthal, 1997). The fact that front-end projects vary in activities, sequencing and time duration also calls for effective project management (Nobelius and Trygg, 2002).

Moreover, a cross-functional executive review committee is likely to affect the performance of front-end activities positively, as cross-functional competence is needed not only when creating product concepts, but also when evaluating them (Khurana and Rosenthal, 1997).

In addition, firms need proficiency in environmental scanning and analysis. Significant issues here include attention to prospective and current product offerings of competitors and legally mandated issues in emerging product

4

definitions (Bacon et al., 1994). But development teams are also advised to have a broad outlook. One-sided attention to technical details often leads to product concepts which resemble what the firm is already offering (Börjesson et al., 2006).

Finally, according to prior work, an early and well-defined product definition is of utmost importance in the FFE. A product definition includes a product concept, which represents the goal for the development project (Montoya-Weiss and O’Driscoll, 2000; Seidel, 2007). In addition, it contains information about target markets, customer needs, and product specifications (Montoya-Weiss and Calantone, 1994) as well as product positioning and product requirements (Cooper and Kleinschmidt, 1987). Proficiency in creating product definitions is crucial as it deeply impacts the go/kill decision prior to formal product development (Cooper, 1988). 2.2 The literature on managing process development

The adoption of process development solutions could be a crucial factor during the FFE phase, as firms in mature industries often have to rely on external actors for process development solutions (Hutcheson et al., 1995; Utterback and Abernathy, 1975). In particular, engineering firms whose underlying competences are not product-specific play a key role in such adoption (Arora and Gambardella, 1997).

The emergence of new product concepts during the FFE also requires a firm to invest in new process technology, i.e. innovation in throughput technology (Linton and Walsh, 2008). Firms that do so proficiently are often superior in introducing new products (Pisano, 1997). It is therefore advised that firms view investments in new process technology as an investment, and not only as costs (Brown, 2001).

Critical to consider during the FFE is also product and process development integration. This factor seems especially important in process firms, as mutual interdependence exists between product and process development (Lim et al., 2006; Linton and Walsh, 2008). The fact that the focus of innovative activities gradually shifts from product to process development as industries mature (Utterback and Abernathy, 1975) also highlights the need for such integration.

Finally, an adequate relationship between labour and management appears to be crucial. Characteristics of such a relationship include mutual understanding of the other party’s perspective, shared beliefs and values, good communication, and a facilitating social and cognitive environment (Khazanchi et al., 2007; Repenning and Sterman, 2002). As process development often requires revised work procedures and changes in work routines, this issue is critical.

2.3 Critical issues and activities present in both the FFE literature and the literature on managing process development

The importance of senior management involvement has been highlighted in the literature on managing process development (e.g. Skinner, 1992), but it has been given even greater attention in the FFE literature. Senior management involvement can overcome resistance to change and facilitate concept

5

development if managers get personally involved and act as executive champions (Murphy and Kumar, 1997). Managers can also align individual activities which often cut across functional boundaries (Khurana and Rosenthal, 1998).

Firms also need proficiency in project selection and portfolio balancing. Khurana and Rosenthal (1997) found this to be missing completely in several firms that they studied, and it was only performed in an ad-hoc and sporadic way in several others. The importance of such activities has also been highlighted in the literature on managing process development, suggesting that new project ideas must match with the firm’s resources and expected return on investment (Lager, 2002a).

To increase success in the FFE, firms also need to encourage idea visionaries or product champions (Griffiths-Hemans and Grover, 2006). Such persons can overcome stability, inertia and status quo, but also facilitate the interpretation of emerging product concepts (Heller, 2000).

A creative organizational culture is furthermore believed to be a key to success. Culture refers to a deep set of meanings, basic beliefs and values (Baer and Frese, 2003). Such a culture is believed to stress experimentation, empowerment, team creativity and shared vision (Pearce and Ensley, 2004) as well as risk-taking, problem-solving and proactiveness (Baer and Frese, 2003; Pearce and Ensley, 2004). In addition, a creative culture allows a firm to utilize the creativity and talents of its employees in the FFE (Murphy and Kumar, 1997).

An additional issue is alignment between NPD and strategy. In essence, new product concepts must capitalize in some way on the firm’s core competences. Therefore, new projects must be aligned with both the overall business strategy of the firm (Bacon et al., 2004) and the product strategy (Khurana and Rosenthal, 1997). Otherwise, decision-making during the FFE is made on project-specific criteria rather than issues of strategic fit (Khurana and Rosenthal, 1998). Similar ideas prevail in the literature on managing process development (e.g. Brown, 2001; Reichstein and Salter, 2006).

An adequate degree of formalization is also emphasized. A “formal” front-end process should be explicit and widely known, have clear decision-making responsibilities, and contain specific performance measures (Khurana and Rosenthal, 1998). Several previous studies show that formalization impacts positively on FFE activities (de Brentani, 2001; Khurana and Rosenthal, 1998). Formalization imposes a risk in addition, however; both an absence of formalization and an excessive reliance on it can be harmful (Gassmann et al., 2006). The literature on managing process development also shows that formalization has positive effects, as it helps achieving objectives (Lager, 2000; Lim et al., 2006).

Cross-functionality, too, is beneficial in the FFE, as it promotes idea refinement (Verworn, 2006) and facilitates knowledge creation (Heller, 2000; Verganti, 1997). It also ensures task relevance for engineers and task feasibility on the marketing side (Gassmann et al., 2006) and reduces uncertainty (Moenaert et al., 1995). In essence, an emerging product concept needs exposure and criticism from all affected functions and departments. Cross-functionality is emphasized in the literature on managing process development as well (Athaide et al., 1996;

6

Lager, 2000; Pearce and Ensley, 2004) as it is needed in both the design and implementation phases of process development work (Gopalakrishnan et al., 1999).

In addition to cooperation among functions and departments, however, FFE activities benefit from inter-organizational collaboration (Khurana and Rosenthal, 1997; Lichtenthaler and Ernst, 2006). Collaboration with suppliers is critical because suppliers play an important role by refining existing technologies and improving equipment reliability and capabilities (Hutcheson et al., 1995; Reichstein and Salter, 2006). Such collaboration also allows firms to acquire information, competence or skills that otherwise could not have been accessed (Lichtenthaler and Ernst, 2009).

Finally, firms should consider early customer involvement during the FFE, although customer involvement has different meanings in the context of product development and process development respectively. In the FFE literature, customer involvement helps clarify customer needs (Verworn, 2006). Early customer involvement is also strongly linked to proficiency in idea generation (Langerak et al., 2004). It further facilitates the screening and evaluation of emerging concepts (Alam, 2006), and it allows for uncertainty reduction (de Brentani, 2001). Product development presupposes an external customer, however. In the context of process development, the “customer” is the firm’s own production function (Lager, 2002a). Therefore, it seems highly relevant to front-end practices in process firms to involve both types of customers – external and internal.

In sum, prior research points to a variety of different factors and activities that seem to determine the extent to which a firm is able to complete the FFE successfully. As pointed out in detail above, however, extant FFE research focuses quite exclusively on firms developing assembled products or services. At the same time, research into managing process development has hardly addressed the FFE in prior empirical research. As such, the brief overview of earlier research underscores the need for further work on the FFE in non-assembled product development, which is the focus of our study. In the following section, we describe the study’s methods.

3. Research design

A multiple case study was employed as a research strategy to capture a comprehensive and contextual assessment of the complex and iterative activities which constitute the FFE (Edmondson and McManus, 2007). In addition, the multiple case study approach was preferred because it enabled the collection of comparative data (Eisenhardt and Graebner, 2007; Yin, 1994). The study and data analysis followed the suggestions given in seminal works on case study research (Eisenhardt, 1989; Yin, 1994). Thus, the research approach of this study was similar to the designs of recent case study based works in the Journal of Engineering and Technology Management and other leading journals (e.g. Kester et al., 2009; Ren, 2009).

7

8

3.1 Research sites

Three firms operating in the mineral- and metal industry were selected by means of judgement sampling (Denzin and Lincoln, 1994). Höganäs AB develops and manufactures metal powders for the global market. Metal powder technology is used in a variety of application areas, including sintered components, soft magnetic composites, hot polymer filtration, and surface coating. The firm has 1,600 employees in 28 countries. LKAB develops and produces upgraded iron ore and industrial mineral products for the steel and other industry sectors, and employs 3,800 persons in 15 countries. SSAB develops and manufactures high-strength steel for the world market. Application areas include bridges, buildings, ships, and various forms of vehicles and lifting devices. The firm currently employs 9,200 employees in 45 countries. 3.2 Data collection and data sources

At these three firms, we conducted a total number of 36 in-depth interviews over the course of a 20 month period. Four of these interviews were exploratory and unstructured, and conducted with the aim of providing general information on innovative activities. The remaining 32 interviews were semi-structured with an interview guideline, which was based on prior research and the initial unstructured interviews.

Each interview was recorded and transcribed for further analysis, and all interviews were undertaken to specifically examine the topics that are addressed in this paper. Table 1 provides detailed information about the semi-structured interviews.

Posi

tion

Inte

rvie

w

date

Dur

atio

nD

egre

eY

ears

of

empl

oym

ent

Posi

tion

Inte

rvie

w d

ate

Dur

atio

n D

egre

eY

ears

of

empl

oym

ent

Posi

tion

Inte

rvie

w

date

Dur

atio

n D

egre

eY

eem

Mar

ketin

g m

anag

er (R

1)20

08-1

2-05

1h, 5

7 m

inM

Sc2

Res

earc

h en

gine

er

(R13

)20

09-1

0-13

31 m

inM

Sc4

Res

earc

h en

gine

er

(R22

)20

09-1

0-21

59 m

inM

Sc

Seni

or

deve

lopm

ent

engi

neer

(R2)

2009

-09-

241h

, 5 m

inPh

.D7

Man

ager

pro

duct

de

velo

pmen

t (R

14)

2009

-11-

121h

, 8 m

inM

Sc14

Gen

aral

man

ager

pr

oduc

t dev

elop

men

t (R

23)

2008

-09-

103h

, 39

min

Ph.D

Man

ager

glo

bal

deve

lopm

ent (

R3)

20

08-1

2-04

1h, 3

2 m

inLi

c. E

ng25

Res

earc

h en

gine

er

(R15

)20

09-0

9-24

41 m

inM

Sc2

Func

tiona

l man

ager

(R

24)

2009

-10-

191h

, 38

min

Lic.

Eng

Dev

elop

men

t en

gine

er (R

4)20

09-1

0-05

1h, 3

min

MSc

8R

esea

rch

engi

neer

(R

16)

2009

-09-

2559

min

Ph.D

3M

anag

er p

rodu

ct

deve

lopm

ent (

R25

)20

09-1

1-09

1h, 3

min

MSc

Dev

elop

men

t en

gine

er (R

5)20

09-0

9-28

55 m

in

MSc

10Se

nior

rese

arch

er

(R17

)20

09-1

0-13

39 m

inPh

.D7

Dev

elop

men

t en

gine

er (R

26)

2008

-10-

311h

, 59

min

MSc

Tech

nica

l dire

ctor

(R

6)20

08-1

0-10

1h, 2

9 m

inM

Sc35

Met

allu

rgic

al

deve

lopm

ent m

anag

er

(R18

)

2008

-03-

261h

, 13

min

MSc

13Se

nior

rese

arch

er

(R27

)20

08-1

0-30

1h, 5

1 m

inPh

.D

Dev

elop

men

t en

gine

er (R

7)

2008

-10-

091h

, 10

min

MSc

8Se

nior

rese

arch

er

(R19

)20

09-0

9-24

45 m

inM

Sc29

Res

earc

h en

gine

er

(R28

)20

09-1

0-21

58 m

inM

Sc

Seni

or

deve

lopm

ent

engi

neer

(R8)

2009

-09-

042h

, 55m

inPh

.D9

Man

ager

pro

duct

de

velo

pmen

t (R

20)

2009

-09-

2448

min

MSc

25Se

nior

join

ing

spec

ialis

t (R

29)

2008

-10-

311h

, 59

min

Ph.D

Func

tiona

l m

anag

er (

R9)

2008

-10-

101h

, 52

min

MSc

12R

esea

rch

engi

neer

(R

21)

2009

-10-

1338

min

MSc

7M

anag

er p

rodu

ct

deve

lopm

ent (

R30

)20

08-1

0-31

1h, 3

4 m

inPh

.D

Mat

eria

ls

deve

lopm

ent

man

ager

(R10

)

2008

-12-

051h

, 39

min

Ph.D

11M

anag

er p

rodu

ct

deve

lopm

ent (

R31

) 20

08-1

0-30

1h, 0

7 m

inM

Sc

Prod

uct m

anag

er

(R11

)20

09-0

9-29

59 m

inB

Sc20

Func

tiona

l man

ager

(R

32)

2008

-10-

281h

, 04

min

MSc

Mat

eria

ls

deve

lopm

ent

man

ager

(R12

)

2008

-10-

091h

, 36

min

MSc

20

Inte

rvie

ws a

t Hög

anäs

AB

In

terv

iew

s at L

KA

B

In

terv

iew

s at S

SAB

ars o

f pl

oym

ent

3 18 11 12 12 18 10 6 19 23 17

Tab

le 1

: Sam

ple

info

rmat

ion

for

the

sem

i-str

uctu

red

inte

rvie

ws

The interviews ranged from about 30 minutes up to 3½ hours, and covered four sets of questions. The first set provided background information about employment time, work duties and job experience. The second set of questions focused on the firm’s general NPD practices, and included a discussion about how NPD was organized and conducted, with an emphasis on the problems and opportunities associated with the current NPD processes and organization of NPD work. The third section focused explicitly on front-end practices, i.e. key activities, roles, responsibilities, problems and opportunities present in the FFE. The final set of questions allowed informants to elaborate upon what drives success in the FFE, i.e. which factors and activities were most essential for successful outcomes. Departures from the open-ended questions were permitted, and the format of the interviews was accordingly adapted and changed slightly to pursue interesting and particularly relevant new facets as they emerged.

The interviews were complemented with observations made through informal conversation with employees and several group meetings with R&D managers and engineers. The interviews were also supplemented by secondary data such as documented procedures and company brochures, thus allowing for empirical triangulation of the firm’s NPD practice.

3.3 Data analysis, validity, and reliability

The data analysis started by analyzing each case individually, thus creating a case study history based on the interview data, field notes, observations, and internal documents (Yin, 1994). To facilitate the analysis, related questions were conceptually clustered together according to the general theme that they were exploring (Miles and Huberman, 1994). We then performed a cross-case analysis, in which we compared the findings from the within-case analysis across the cases and looked for similar themes (Yin, 1994). The cross-case analysis was conducted after completing the data collection phase, to allow meaningful replication (Ozcan and Eisenhardt, 2010). Our main unit of analysis was the FFE phase of non-assembled product development. To this background, the cross-case analysis did not result in any major differences across the three cases. As in similar studied (Kleinsmann et al., 2009; Langner and Seidel, 2009; ), the analysis was a continuous and iterative process requiring repetitive reading of the interview accounts, the field notes, the secondary data, and the theoretical framework.

To increase reliability (transparency and future replication), a case study protocol was constructed together with a case study data base, containing case study notes, documents, and the narratives collected during the study, all with the aim of facilitating retrieval for future studies (Yin, 1994). To strengthen internal validity, a clear research framework was designed and extensively discussed within the research team prior to data collection. To mitigate bias in the data collection, we selected knowledgeable informants from different hierarchical levels who viewed the FFE from diverse perspectives (Eisenhardt and Graebner, 2007). As the current study was conducted in a previously unexplored context, rather than of replicated nature, pattern matching (a comparison between observed patterns and those established in previous studies) was sometimes difficult (Denzin and Lincoln, 1994). Therefore, construct

validity was addressed mainly by trying to establish a clear chain of evidence to allow readers to see how the initial research purpose matches with key conclusions (Yin, 1994).

4. Results 4.1 Informal start-up

The very first activities in the FFE of the firms studied referred to informal start-up tasks. Before a formal assessment of new product ideas was undertaken, important informal activities took place. One such activity was idea generation and refinement. Ideas typically came from customers or from internal development personnel. Independently of the source, much time and energy was spent on specifying the ideas as far as possible. To anticipate the requirements of the customers’ production processes was another key activity, because the performance level of the final product is largely determined by how the subsequent customers process it. One informant articulated this issue: ‘…our customers can destroy the performance level of our products completely if they do something wrong…so we have to understand how our product works in their production process. So their process, our process, and the materials input make an end-user product’ (R8).

Besides idea generation, the preliminary analysis of input material was a critical issue, especially when new raw materials were a part of the emerging product concept. In the firms in our sample, input materials were typically raw materials. As input material often varies in quality and holds unique chemical properties, their influence on subsequent development and manufacturing needs to be taken into consideration. Therefore, the analysis of input material constituted a key activity in the firms investigated.

Another key activity performed in the earliest stage was identifying the “process window”. The process window was a metaphor used by multiple informants to label the constraints inherent in the internal production process. Conceptually, a new product idea is always either inside or outside the process window. Being inside the process window means the idea can be developed within the frame of existing production processes, without subsequent investments in or modifications of process technology. Being outside the process window implies a more radically new product idea which requires changes of current production processes. The distinction is important because, as one informant articulated, ‘When we step outside the process window they [the process development department] have to be seriously involved’ (R22).

4.2 Formal idea study

The initial informal activities were followed by the first “formal” sub-phase, which could be named idea study – and which was typically initiated by a product manager. The formal idea study was described as theoretical rather than empirical. This phase typically started with a literature review conducted to map existing technical and chemical knowledge relevant to the emerging product idea. The findings of the literature review were subsequently discussed with

11

colleagues and, later on, with senior researchers within each respective firm. Based on the literature review, preliminary laboratory tests were performed to verify whether the product idea was feasible to develop. Documenting the results from the preliminary laboratory tests was deemed critical for both time and cost reasons, yet deficiencies were often spotted. Especially development engineers regarded such administrative matters as boring and bothersome. In particular, failed experiments were frequently not documented, often resulting in the firms having to reinvent the wheel in subsequent steps. The typical idea study ended with the enactment of a preliminary product definition, a definition of project goals and, eventually, a decision to proceed or not to proceed to the formal pre-development study. 4.3 Formal pre-development study

The decision to proceed to a pre-development study was typically made by a cross-functional management team, including managers from product development, production and marketing. The final sub-phase in the FFE, the pre-study, was described as empirical rather than theoretical. Common activities in the pre-study phase were more detailed and fine-grained laboratory tests, subsequent tests in pilot plants when applicable, and batch tests with customers.

Typically, at this phase, several related, alternative product concepts were created simultaneously. Many informants described these activities in a manner similar to baking a cake. While following the same basic recipe, the volume of the ingredients was varied (i.e. the exact chemical composition) as were the process parameters (e.g. heat, pressure and time). The best product concept was often determined in close collaboration with customers, thus making customer involvement critical to this phase. Finally, a product concept was selected for subsequent formal development, followed by the creation of a project plan and budgets. A formal go/no-go decision marked the end of the pre-study phase. Based on the empirical data, Figure 1 conceptualizes the FFE of non-assembled product development in the firms studied.

12

Figure 1: A conceptual model of the FFE in non-assembled product development.

Go/No go

decision

Time

� Idea generation and refinement

� Anticipation of customer’s process requirements

� Analysis of input materials

� Identification of process window

FORMAL IDEA-STUDY

� Literature review

� Preliminary laboratory tests incl. documentation

� Creation of preliminary product concepts

� Definition of project goals

FORMAL PRE-STUDY

� Laboratory tests

� Pilot plant tests (when applicable)

� Batch tests with customers

� Enactment of several product concepts

� Customer involvement

� Selection of one product concept

� Project planning

INFORMALSTART-UP

Start of the formal NPD

project

4.4 Additional remarks on the FFE in non-assembled product development

In addition to the activities conducted in the various phases, a couple of additional remarks apply when taking an integrative perspective on the FFE activities. First, idea studies and pre-studies were often loosely coupled and characterized by low degrees of formalization. Although such studies were common in all firms, they varied in rigour, quality and execution. Due to time and resource constraints, these activities were sometimes done in a less proficient way or even skipped. Second, in all firms, the activities conducted in the FFE varied depending on the novelty and sources of the idea. Customer-driven ideas had much clearer boundaries, whereas internal ideas were often technically driven and encompassed a higher degree of newness.

Furthermore, the FFE was described by a clear majority of informants as a trial-and-error process. Iteration among activities and phases was common, as both product and process parameters were difficult to anticipate exactly in advance.

Moreover, there was a strong need for integration between the areas of product and process development. While this issue was deemed critical for projects outside the “process window”, even projects within the process window frequently resulted in failure when process development competences were not sufficiently brought into a project. Product and process development were often

13

argued to be artificially separated, and sometimes it was even unclear to informants as to which of the two domains a specific activity belonged. As one informant put it, ‘you cannot work with product development without doing process development in this type of industry.’ (R8) Another key issue experienced was that of scaling problems. While small-scale trials in laboratories were described as the perhaps most important activity to increase knowledge and reduce uncertainty in emerging product concepts, the results of these trials were often difficult to transfer to full-scale production. In part, this was attributed to differences in process technology between laboratories and factories, but changes in process parameters such as pressure, humidity and temperature occurred in addition, whose influence was often very difficult to anticipate. As one informant articulated, ‘Scaling up is another important problem…and the main problem with scaling is that you lose properties. Everything is possible to mix…but the properties don’t always follow when you move from small scale to larger scale.’ (R2)

One of the firms studied, SSAB, was often incapable of doing small-scale trials. This was for a reason opposite to the above: difficulties in scaling down the process technology for making high-strength steel. Hence, pilot tests were often made in the actual production line. Failed tests could thus be extremely costly, resulting in over 100 tons of scrap metal. This underscores another important distinction between firms involved with non-assembled product development, and other manufacturing industries. A related issue is that new product concepts ultimately cannot be tested proficiently without affecting the normal production. That is, eventually a new concept has to be tried with the process equipment intended to be used in future manufacturing. As the production facilities are operated around the clock, this often resulted in disagreements between R&D and production on capacity utilization. In many ways, this problem resembles the classical exploration-exploitation dilemma, i.e. balancing today’s needs and requirements with those of tomorrow. An informant articulated this: ‘One thing that encumbers us is the production conditions. It is crowded in the production line and we do as many tests as we are allowed to. When you perform full-scale trials you always have an opportunity cost to keep in mind. That is, the normal production which is lost when we perform the experiments.’ (R27) 4.5 Central managerial challenges

In addition to the FFE activities and issues described in the previous sections, our data reveal that management can make deliberate efforts to improve both the speed and quality of the FFE process. In particular, six factors appeared critical for increasing the proficiency of FFE activities. First, integration of product development and process development was paramount. Process development thus needs to be involved early in the planning and conduct of FFE activities in non-assembled product development. Integrating these two innovation types was essential to all projects, but especially to those outside the “process window”.

14

Second and similarly, cross-functional collaboration was heavily emphasized. Multiple competences were needed in idea generation and concept creation, and the relevant competences span not only the knowledge of product and process issues, but also engineering, material science, and marketing. This means that a variety of competences are needed in a pre-development team.

Third, senior management support was judged to be critical. Such support includes not only concrete activities such as securing funds and resources and aligning FFE activities with firm strategies, but also the communication of values – including empowering engineers, encouraging new ideas and product concepts, and showing a tolerance of failures.

Fourth, project management was decisive for both speed and quality of FFE activities. An effective project manager was described as being well connected internally, possessing good communication skills, being able to secure resources for development, and able to motivate and inspire other team members. Simultaneously, such a person must often make tough decisions, not always to the satisfaction of persons nearby. Such persons were few in all firms studied, and one informant articulated that ‘Being an effective project manager is extremely awkward…you must be able to walk on other people’s feet and then collaborate with them afterwards…very, very few people can manage to do so’ (R6).

Fifth, a creative climate and culture were frequently emphasized as critical for success. According to the informants, such a culture allows experimentation, skunk-work among development staff, and tolerance of individual differences among employees.

Finally, a clear majority of informants emphasized a holistic perspective on development activities. In short, this means not only that individual development activities need to be aligned, but in addition that deficiencies in single activities run the risk of jeopardizing whole FFE projects. For example, one informant said: ‘My reflection is that these activities and factors…the ideas which are inherent in them…are all crucial and must be fulfilled. All pieces of the puzzle must be in place; otherwise you end up with nothing. If you don’t look at it as a whole, you run an overwhelming risk of sub optimization’. (R18)

5. Discussion Overall, our results highlight that the FFE of non-assembled product development possesses unique and context-specific characteristics. Given that the technology- and innovation management literatures already have made significant advances in recent years to better understand, describe and conceptualize the FFE, our results extend those of previous scholars and show that additional research into idiosyncratic industry domains is needed. Despite the extant knowledge of the FFE, many firms still experience major difficulties in FFE activities, which constitute a major managerial challenge in NPD. As such, our study deepens the understanding of the FFE in non-assembled product development. These findings are particularly relevant in light of the desire of

15

both scholars and managers to better understand the activities, issues, and managerial challenges needed to enhance proficiency of front end activities. 5.1 Theoretical implications

This study provides important theoretical implications for research into the FFE. First, the idea of informal activities preceding the somewhat more formalized FFE activities is not novel (Khurana and Rosenthal, 1998). In addition, the importance of idea generation and refinement has been noted previously (Cooper, 1988; Kim and Wilemon, 2002). However, to anticipate the needs of the customer’s production process, analysis of input materials and identification of the “process window” constitute novel findings, which have not yet been reported although the process development literature advocates the early consideration of production needs (Lager, 2002a). These findings, especially the identification of the so-called “process window”, may very well extend to other manufacturing firms.

Second, many of the activities which constitute the formal idea study, such as the enactment of a preliminary product definition and definition of project goals, clearly constitute standard FFE activities (Bacon et al., 1994; Kim & Wilemon, 2002; Khurana and Rosenthal, 1997). The importance of proficient literature reviews and laboratory tests have not, however, been reported in either the FFE or the process development literatures.

Third, while customer involvement is known to be critical for FFE success (Alam, 2006; Kim and Wilemon, 2002), activities such as detailed laboratory tests, pilot plant tests and batch tests are not reported in the extant literature. Moreover, the extant literature has envisaged that product concept shifting may be necessary in the FFE (Seidel, 2007), but the fact that multiple concepts are developed in parallel also constitutes a novel finding. In the firms studied, these are described as key activities of the formal pre-development phase. These findings also extend prior literature.

A fourth contribution is our showing that mainstream conceptualizations of the FFE do not directly apply to the context of non-assembled products. While some similarities are spotted, the conceptualization proposed diverges in large part from previous research conceptualizing the FFE (Cooper, 2006; Griffiths-Hemans and Grover, 2006; Khurana and Rosenthal, 1997; 1998; Koen et al., 2001).

Finally, we highlight a set of central managerial challenges. While activities such as cross-functionality (Kim and Wilemon, 2002; Verganti, 1997), senior management involvement (Koen et al., 2001), project management (Khurana and Rosenthal, 1997; Nobelius and Trygg, 2002) and culture (Langerak et al., 2004) have been described in the FFE literature, our analysis shows these to be paramount. Moreover, we confirm findings in the process development literature and show that product and process development integration is essential to FFE activities (Lim et al., 2006). Thus, we show that new product concepts may require new process technology, a fact that needs to be anticipated early in a firm’s development efforts. Moreover, our analyses underscore that tight

16

integration between product and process development is a key to being able to successfully produce the new product in subsequent phases. Again, this has not yet been reported in the field of FFE research. In addition, our data show the need for a wider and more comprehensive view of what constitutes a holistic perspective on FFE activities than what has been reported previously (Khurana and Rosenthal, 1998).

Altogether, these findings have valuable implications for research not only into firms developing and manufacturing non-assembled products but may very well extend to other manufacturing industries as well. In sum, by studying the FFE in non-assembled product development, we show that the FFE idiosyncrasies cannot fully be explained by prior research. Evidently, additional research on the FFE context of process firms and other specific contexts is needed.

5.2 Managerial implications

Our results provide direct managerial implications for enhancing the success of FFE activities and, ultimately, to the benefit of the whole new product development process, especially concerning non-assembled product development. First, managers can strongly enhance their firms’ proficiency in the FFE by paying particular attention to the specific issues and activities which constitute the FFE. As the rigour and proficiency with which these activities are conducted vary, managers can both encourage and control the conduct of these activities. Regarding the informal activities which precede the more formal idea-study and pre-development study, they involve ensuring that the customer’s process requirements are fully examined, that input materials are sufficiently analyzed, and that the “process window” for the development project is fully comprehended. In the idea-study, management can ensure that relevant literature reviews are sufficiently compiled, that preliminary laboratory tests are not only conducted but also documented, that the preliminary product concept is sufficiently clear to facilitate subsequent development efforts, and that project goals are clear and measurable. As for the pre-development study, management can ensure that laboratory tests are conducted and interpreted correctly, that customers are involved to a sufficient degree, that the evaluation and ultimate selection of product concepts are done proficiently, and that adequate project planning is performed. This implies a call for routines. We thus make a plea for formalization of current FFE practices. A formal FFE process is explicit, widely known, and characterized by clear decision-making responsibilities (Khurana and Rosenthal, 1998). Although too much discipline may endanger creativity (Gassmann et al., 2006), the fact that documentation of experiments and other activities often were conducted less than satisfactorily speaks for more rather than less formalization. The background where idea-studies and pre-development studies were often only loosely coupled also calls for formalization. That being said, order and predictability rather than rigidity are needed (Boeddrich, 2004; de Brentani, 2001). Accordingly, we do not recommend that managers should overemphasize

17

the formalization of the FFE activities. However, all firms studied were characterized by insufficient formalization of the FFE activities. Accordingly, stronger formalization is unlikely to lead to rigidity at the current state of organizing for the FFE in these firms. Thus, stronger formalization will help to reap the benefits from organizational routines to facilitate experimentation and creativity.

Moreover, our analyses highlight the fact that some factors and activities serve as configuration devices, in the sense that proficiency in these implies a form of meta-competence for making other factors and activities work. Specifically, these factors were deemed of particular importance to make the concrete activities which constitute the informal start-up, idea-study, and pre-development study activities work more proficiently. Consequently, managers need to embrace product and process development integration because development project often spans the product/process interface (Lim et al., 2006; Linton and Walsh, 2008). In addition, cross-functional work practices should be encouraged because multiple competences from a variety of different functional domains are needed in FFE projects.

Further, FFE activities would benefit from senior management support, which concerns not only funding and resource issues, but also strategic alignment and encouragement throughout the process. Moreover, project managers with adequate skills should be assigned to lead FFE activities, so as to ensure ultimate success. The importance of culture as a coordination mechanism has been noted previously, and found to impact positively on feasibility analysis, customer focus, and coordination (Khurana and Rosenthal, 1998). Our findings confirm these, and show the importance of a culture which stresses experimentation and allows skunk-work among employees. Clearly, managers can reinforce such cultural norms.

Finally, our results emphasize a holistic perspective on FFE activities. It follows that deficiencies in single activities run the risk of jeopardizing FFE projects. The idea of integrating factors and activities is not novel. The extant literature on the FFE has emphasized synergy among skills and resources in technology, marketing and manufacturing (Cooper, 1988), fit among product concept, product technology, and process design (1997), and fit among product idea, firm strategy and operating capabilities (Murphy and Kumar, 1997). In addition, Khurana and Rosenthal (1998) have shown that understanding the interrelationships among activities is as important as the activities themselves.

Excelling in individual factors and activities thus seems a necessary but not a sufficient condition for success. Firms need to show proficiency in a large number of dimensions, but also to integrate these into a coherent whole to achieve the synergy arising from the complementarity of multiple factors. Some of these factors belong to the product domain, others to the process domain, and some cut across the product/process interface. Managers are therefore advised to align individual activities, besides focusing on them as such, in order to establish a proficient FFE process.

18

5.3 Limitations and outlook Some limitations of this study are worth noting. Despite a large number of

interviews, the current study is restricted to only three firms within the mineral- and metals industry. Although case studies are not devoid of generalizations (Gibbert et al., 2008), external validity (generalizability) is problematic with a research design like the current one. Managers outside our sample must therefore draw their own inferences by analogy.

Furthermore, this study has limitations common to case-study research. The method used is descriptive rather than exploratory, which means that the behaviour of firms and respondents can be described rather than explained. Similarly, while the selected key informants provided rich information on current FFE activities, the method as such leaves room for important details to be omitted. In addition, the information collected is retrospective, and is therefore subject to the problems inherent in human memory.

That being said, the FFE in non-assembled product development would clearly benefit from continued research. This includes testing the current results in a wider sample of process firms in other industrial domains, e.g. pulp and paper, chemical and food industries. Another interesting research question would be how firms can go about creating a holistic perspective on FFE activities in practice – that is, how they can achieve a tight degree of integration of the many and often heterogeneous activities which constitute the FFE. Still another interesting issue would be to study strategies for increased formalization while simultaneously avoiding interfering with creativity, as well as testing the current results quantitatively through a survey. Our findings indicate that more research is needed to offer a better understanding to firms aiming at profiting from innovation. The results of future qualitative and quantitative studies may well provide results that are equally relevant to academics and practitioners.

19

References Alam, I., 2006. Removing the fuzziness from the fuzzy front-end of service

innovations through customer interactions. Industrial Marketing Management 35, 468-480.

Arora, A., Gambardella, A., 1997. Domestic markets and international competitiveness: generic and product-specific competencies in the engineering sector. Strategic Management Journal 18, 53-74.

Athaide, G., Meyers, P., Wilemon, D., 1996. Seller-buyer interactions during the commercialization of technology process innovations. Journal of Product Innovation Management 13, 406-421.

Bacon, G., Beckman, S., Mowery, D., Wilson, E., 1994. Managing product definition in high-technology industries: A pilot study. California Management Review 36, 32-56.

Baer, M., Frese, M., 2003. Innovation is not enough: climates for initiative and psychological safety, process innovations, and firm performance. Journal of Organizational Behaviour 24, 45-68.

Brown, S., 2001. Managing process technology – further empirical evidence from manufacturing plants. Technovation 21, 467-478.

Börjesson, S., Dahlsten, F., Williander, M., 2006. Innovative scanning experiences from an idea generation project at Volvo cars. Technovation 26, 775-783.

Chang, S-L., Chen, C-Y., Wey, S-C., 2007. Conceptualizing, assessing, and managing front-end fuzziness in innovation/NPD projects. R&D Management 37, 469-478.

Cooper, R.G., 1988. Predevelopment activities determine new product success. Industrial Marketing Management 17, 237-247.

Cooper, R.G., Kleinschmidt, E.J., 1987. New products: What separates winners from losers? Journal of Product Innovation Management 4, 169-184.

de Brentani, U., 2001. Innovative versus incremental new business services: Different keys for achieving success. Journal of Product Innovation Management 18, 169-187.

Denzin, N., Lincoln, Y., 1994. Handbook of qualitative research. Sage Publications, Thousand Oaks.

Edmondson, A.C., McManus, S.E., 2007. Methodological fit in management research. Academy of Management Review, 32 1155-1179.

Eisenhardt, K.M., 1989. Building theories from case study research. Academy of Management Review 14, 532-550.

Eisenhardt, K.M., Graebner, M.E., 2007. Theory building from cases: opportunities and challenges. Academy of Management Journal 50, 25-32.

Gassmann, O., Sandmeier, P., Wecht, C., 2006. Extreme customer innovation in the front-end: Learning from a new software paradigm. International Journal of Technology Management 33, 46-66.

Gibbert, M., Ruigrok, W., Wicki, B., 2008. What passes as a rigorous case study? Strategic Management Journal 29, 1465-1474.

20

Gopalakrishnan, S., Bierly, P., Kessler, E., 1999. A reexamination of product and process innovations using a knowledge-based view. Journal of High Technology Management, 10, 147-166.

Griffiths-Hemans, J., Grover, R., 2006. Setting the stage for creative new products: Investigating the idea fruition process. Journal of the Academy of Marketing Science 34, 27-39.

Heller, T., 2000. If only we'd known sooner: Developing knowledge of organizational changes earlier in the product development process. IEEE Transactions on Engineering Management 47, 335-359.

Hutcheson, P., Pearson, A.W., Ball, D.F., 1995. Innovation in process plant: a case study of ethylene. Journal of Product Innovation Management 12, 415-430.

Kester, L., Hultink, E.J., Lauche, K., 2009. Portfolio decision-making genres: A case study. Journal of Engineering and Technology Management 26, 327-341.

Khazanchi, S., Lewis, M., Boyer, K., 2007. Innovation-supportive culture: the impact of organizational values on process innovation. Journal of Operations Management 25, 871-884.

Khurana, A., Rosenthal, S.R., 1997. Integrating the fuzzy front end of new product development. Sloan Management Review 38, 103-120.

Khurana, A., Rosenthal, S.R., 1998. Towards holistic "front ends" in new product development. Journal of Product Innovation Management 15, 57-74.

Kim, J., Wilemon, D., 2002. Focusing the fuzzy front-end in new product development. R&D Management 32, 269-279.

Kleinsmann, M., Bujis, J., Valkenburg, R., 2009. Understanding the complexity of knowledge integration in collaborative new product development teams: A case study. Journal of Engineering and Technology Management 27, 20-32.

Koen, P., Ajamian, G., Burkart, R., Clamen, A., Davidson, J., D’Amore, R., Elkins, C., Herald, K., Incorvia, M., Johnson, A., Karol, R., Seibert, R., Slavejkov, A., Wagner, K., 2001. Providing clarity and a common language to the "fuzzy front end". Research Technology Management 44, 46-55.

Langerak, F., Hultink, E., Robben, H., 2004. The role of predevelopment activities in the relationship between market orientation and performance. R&D Management 34, 295-309.

Lager, T., 2000. A new conceptual model for the development of process technology in process industry. International Journal of Innovation Management 4, 319-346.

Lager, T., 2002a. Product and process development intensity in process industry: A conceptual and empirical analysis of the allocation of company resources for the development of process technology. International Journal of Innovation Management 6, 105-130.

Lager, T., 2002b. A structural analysis of process development in process industry. R&D Management 32, 87-95.

21

Lager, T., 2010. Managing process innovation. Imperial College Press, London, UK.

Langner, B., Seidel, V.P., 2009. Collaborative concept development using supplier competitions: Insights from the automotive industry. Journal of Engineering and Technology Management 26, 1-14.

Lichtenthaler, U., Ernst, H., 2009. Opening up the innovation process: The role of technology aggressiveness. R&D Management 39, 38-54.

Lichtenthaler, U., Ernst, H., 2006. Attitudes to externally organising knowledge management tasks: a review, reconsideration and extension of the NIH syndrome. R&D Management 36, 367-386.

Lim, L., Garnsey, E., Gregory, M., 2006. Product and process development in biopharmaceuticals: A new perspective on development. R&D Management

36, 27-36. Linton, J.D., Walsh, S.T., 2008. A theory of innovation for process-based

innovations such as nanotechnology. Technological Forecasting & Social Change 75, 583-594.

Miles, M., Huberman, M., 1994. Qualitative data analysis. Sage Publications, Thousand Oaks.

Montoya-Weiss, M., Calantone, R., 1994. Determinants of new product performance: A review and meta-analysis. Journal of Product innovation Management 11, 397-417.

Montoya-Weiss, M., O'Driscoll, T., 2000. From experience: Applying performance support technology in the fuzzy front end. Journal of Product Innovation Management 17, 143-161.

Moenaert, R., De Meyer, A., Souder, W., Deschoolmeester, D., 1995. R&D/marketing communication during the fuzzy front-end. IEEE Transactions on Engineering Management 42, 243-258.

Murphy, S., Kumar, V., 1997. The front end of new product development: A Canadian survey. R&D Management 27, 5-15.

Nobelius, D., Trygg, L., 2002. Stop chasing the front end process – Management of the early phases in product development projects. International Journal of Project Management 20, 331-340.

Pearce, C., Ensley, M, 2004. A reciprocal and longitudinal investigation of the innovation process: The central role of shared vision in product and process innovation teams (PPITs). Journal of Organizational Behavior 25, 259-278.

Pisano, G., 1997. The development factory: unlocking the potential of process innovation. HBS Press, Boston.

Reichstein, T., Salter, A., 2006. Investigation the sources of process development among UK manufacturing firms. Industrial and Corporate Change 15, 653-682.

Reid, S.E., de Brentani, U., 2004. The fuzzy front end of new product development for discontinuous innovations: A theoretical model. Journal of Product Innovation Management 21, 170-184.

22

Ren, T., 2009. Barriers and drivers for process innovation in the petrochemical industry: A case study. Journal of Engineering and Technology Management 26, 285-304.

Repenning, N., Sterman, J., 2002. Capability traps and self-confirming errors in the dynamics of process improvement. Administrative Science Quarterly 47, 265-295.

Seidel, V.P., 2007. Concept shifting and the radical product development process. Journal of Product Innovation Management 24, 522-533.

Skinner, W., 1992. The shareholder’s delight: Companies that achieve competitive advantage from process innovation. International Journal of Technology Management 7, 41-48.

Smith, P., Reinertsen, D., 1998. Developing products in half the time. John Wiley & Sons, New York.

Utterback, J., Abernathy, W., 1975. A dynamic model of process and product innovation. Omega 3, 639-656.

Verganti, R., 1997. Leveraging on systemic learning to manage the early phases of product innovation projects. R&D Management 27, 377-392.

Verworn, B., Herstatt, C., Nagagira, A., 2008. The fuzzy front end of Japanese new product development projects: Impact on success and differences between incremental and radical projects. R&D Management 38, 1-19.

Verworn, B., 2006. How German measurement and control firms integrate market and technological knowledge into the front end of new product development. International Journal of Technology Management 34, 379-389.

Yin, R., 1994. Case study research: Design and methods. Sage Publications, Thousand Oaks.

23


Paper A


Paper B



Paper C



Paper D


Paper E

Where process development begins: A multiple case study of fuzzy front-end activities in process firms

Monika Kurkkio

Luleå University of Technology Center for Management of Innovation and Technology

in Process Industry SE-97187 Luleå; Sweden Phone: +46-920491865


Johan Frishammar Luleå University of Technology




Ulrich Lichtenthaler WHU - Otto Beisheim School of Management

Innovation and Organization Burgplatz 2; D-56179 Vallendar; Germany

Phone: +49-(0)261-6509-411 Email: [email protected]

Abstract The fuzzy front end stage is known to be critical to overall product development success, but few if any studies have examined the front end in process development. By means of a multiple case study of process firms, this article aims to bridge this knowledge gap. Our results show that substantial differences in front end activities exist between the product development and process development domains. We conceptualize the front end in process development to be an iterative trial-and-error process, dominated by activities such as idea generation and refinement, literature reviews, anticipation of end-product changes, and various forms of experiments in bench scale, lab scale, and full-scale production. In addition, we highlight key problems in the front-end process, and managerial remedies for how to mitigate them. While these findings provide theoretical implications for research into product development, process development and production management, the findings are particularly relevant to process development managers, plant managers, and development engineers interested in increasing the efficiency of production processes. Keywords: Fuzzy front end; Process development; case study

1

1. Introduction Product development and process development constitute two types of

innovative activities critical to competitive advantage of manufacturing firms (Chen et al., 2010; Khachanzi et al., 2007; Koc and Ceylan, 2007). While product development is driven by the desire to create a new product, process development relates primarily to internal production objectives, such as cost reductions, enhancement of production volumes, or sustaining more environment-friendly production (Lager, 2002a; Pisano, 1997). Accordingly, process development is a key determinant of successful technological innovation (Linton and Walsh, 2008; Reichstein and Salter, 2006). With a background where front end activities – the activities which precede subsequent formal development efforts – are stressed as the key to effective product development (Cooper et al., 2004; Verworn, 2009), this article aims to shed light on front end activities in the context of process development, i.e. the activities conducted before a new process concept is implemented into the production process (Lager, 2002; Pisano, 1996).

Over the past decades, the conduct of front end activities in product development (but not process development) has been a recurrent theme in the literature (e.g. Cooper, 1988; Khurana and Rosenthal, 1997, 1998; Verworn, 2009). The fuzzy front end (FFE) stage of the product development process starts when a firm has an idea for a new product, and it ends when the firm decides to launch a formal development project or, alternatively, decides not to do so (Kim and Wilemon, 2002). Consequently, the FFE precedes the typical “formal” stage-gate approach to product development (Cooper, 2008). Previous research pictures the FFE stage as often being ill-defined, unclear, uncertain, and equivocal (Chang et al., 2007). In addition, the FFE is often characterized by ad-hoc decision-making and conflicting organizational pressures (Khurana and Rosenthal, 1998; Montoya-Weiss and O’Driscoll, 2000). This makes mistakes hard to avoid and the search for better predictive guidelines justifiable.

Both quantitative (e.g. Murphy and Kumar, 1997; Verworn, 2009) and qualitative studies (e.g. Brem and Voigt, 2009; Khurana and Rosenthal, 1997; Seidel, 2007) have demonstrated the critical role of front end activities for final product development success. In contrast, the process development literature has not specifically studied front end activities. This literature has merely highlighted the assumed importance of front end activities, but has not examined such activities in detail (Lager, 2000; Lim, Garnsey and Gregory, 2006; Pisano, 1996). This limited attention is remarkable because effective improvements to manufacturing processes and throughput technology are a key source of

2

competitive advantage in manufacturing firms (Pisano, 1997). Furthermore, product development and process development are often tightly linked in the technological core of manufacturing firms with the aim of offering high-quality products to customers (Damanpour and Gopalakrishnan, 2001; Ettlie, 1995; Pisano, 1997; Reichstein and Salter, 2006).

Accordingly, proficient process development may be a requirement rather than an option to facilitate successful product development (Brown, 2001). These conditions underscore the strong deficit of research into the FFE for process development (Lim et al., 2006). This research deficit is further highlighted by the importance of front end activities in product development (Verworn, 2009) and the managerial challenges inherent in the FFE (Khurana and Rosenthal, 1998). Accordingly, the topic of front end activities in process development appears equally relevant to research and management practice. To this background, the overall purpose of this article is to increase knowledge about the FFE stage in process development. Specifically, we explore the stages which constitute the FFE stage in process development and identify the key activities therein. We also elaborate key problems and identify a number of key success factors applicable to the FFE in process development.

As such, this article offers several contributions. First, we contribute to research into the FFE and process development by conceptualizing the FFE of process development and by identifying the key activities which constitute this stage. Thus, we shed light on an essential yet under-researched topic in technology- and innovation management (Brown and Maylor, 2005). Second, we identify both key problems and success factors which are critical to ultimate success in process development. While our results highlight some similarities to front end activities in product development, they also show that extant knowledge on the FFE in product development cannot be directly transferred to the process domain because of differences in scope and key activities. Accordingly, our findings may be essential to successful process development in firms. Third, we identify important interdependences of process development and product development, which indicate that process development practices may be essential to achieve high product development performance. Altogether, these issues have been pointed out as areas ripe for further study in recent work on process development (Khazanchi et al., 2007; Lager, 2000; Lim et al, 2006; Samson and Whybark, 1998).

The remainder of the paper is organized as follows. In section 2, we present the theoretical background to our study. In section 3, we describe the research design. In section 4, we elaborate the empirical findings and provide analyses of

3

these. In section 5, we discuss the study’s theoretical and managerial implications before addressing limitations and providing an outlook.

2. Theoretical Background

The theoretical background section is divided into three sub-sections. First, we define process development and elaborate its typical outcomes. Second, we provide a general overview of the “process of process development”. Third, in a context where front end activities in process development have hardly been examined in the extant literature, we review the FFE literature in product development, with the aim of using this literature as a theoretical starting point for studying FFE activities in the process domain. 2.1. Defining process development

Process development concerns improving how work is done (Utterback and Abernathy, 1975), in contrast to product development which focuses on what is done, i.e. the final output sold and delivered to external customers (Gopalakrishnan et al., 1999). No all-encompassing definition of process development exists in the extant literature, but a variety of somewhat similar definitions prevail (Utterback and Abernathy, 1975; Baer and Frese, 2003; Ettlie and Reza, 1992; Khazanchi et al., 2007). Thus, most authors seem to agree on the general characteristics of what constitute process development. Consistent with earlier research, we therefore define process development as deliberate and systemic development related mainly to production objectives, implying the introduction of new elements into the production process with the purpose of creating or improving methods of production (Baer and Frese, 2003; Gopalakrishnan et al., 1999, Lager, 2002a; Lager, 2002b: Khazanchi et al., 2007; Reichstein and Salter, 2006).

Based on this definition, process development first of all implies deliberate organizational attempts to change or modify the production process. That is, improvement efforts are conscious and planned. While process development can occur by serendipity, that is rarely the case (Baer and Frese, 2003). Second, process development does not occur in isolation. Process development is organizationally complex and spans multiple functions which make it systemic. Changes in the production process often affect other processes within a firm (Gopalakrishnan et al., 1999), for example product development, manufacturing strategy, and operations strategy (Pisano, 1997). Thus, improvements and modifications of the production process are likely to result in changes in other processes close by. Third, because process development is performed in the context of production, the objectives are typically internal to the firm and often

4

center on cost reductions and improved product quality (Lager, 2002a). Fourth, process development implies the introduction of new elements into the production process (Khazanchi et al., 2007). These elements may include new input materials, task specifications, work and information flow mechanisms, or new equipment used to produce a product (Gopalakrishnan et al., 1999; Reichstein and Salter, 2006).

Finally, process development occurs on a degree-of-newness continuum. Scholars often distinguish between process innovation and process improvement as two ideal types of process development (Gopalakrishnan et al., 1999; Reichstein and Salter, 2006; Utterback and Abernathy, 1975). As process development itself is organizationally complex and spans multiple functions, from laboratory settings to full-scale trials, the degree of newness is difficult to anticipate in advance (Pisano, 1997). Process improvement typically refers to incremental development efforts such as improvements of existing processes which may allow slightly increased efficiency or effectiveness (Utterback and Abernathy, 1975). Process innovation, on the contrary, presupposes wider and more all-encompassing changes to the manufacturing process such as the launch of a next-generation process for a new product (Reichstein and Salter, 2006; Pisano, 1997). Thus, process development is a matter of a continuum rather than clear-cut categories ranging between incremental and more radical development efforts (Pisano, 1997).

The outcomes of process development are multiple, and vary according to the goals of each specific process development project. Cost reductions, increased production volumes, and yields from production are key examples of desired outcomes (Lim et al., 2006; Pisano, 1994; 1996). Additional outcomes include increased product quality and reliability, reduced time to market, and sustaining more environment-friendly production (Gopalakrishnan et al., 1999; Lager, 2002a; Pisano, 1997). A firm’s proficiency in process development determines the extent to which these outcomes materialize (Pisano, 1997). 2.2. Conceptualizing the process development process

In comparison with the stage-gate literature on the product development process (Cooper, 1988; Cooper et al., 2004), much less attention has been devoted to the “process development process”. The few studies that do exist typically focus on this phenomenon in process firms (Lager, et al., 2010; Lim et al., 2006; Pisano, 1996). Process development is critical to process firms whose manufacturing processes are inert, capital-intensive, and critical to competitive advantage (Hutcheson et al., 1995). In addition, process development often

5

requires large investments in R&D spending in process firms, compared to other manufacturing industries (Lager, 2002b).

One of the first models of the process development work process was proposed by Pisano (1994). This model splits process activities into three stages: process research, pilot development, and commercial plant scale-up. In the first stage, a description of the product concept is made, which in most cases will be incomplete (Pisano, 1996). The product concept is the formula for the changes required to the end product, and it further includes a set of functional target specifications, e.g. unit cost, capacity required, or quality levels. At this first stage, development work is generally performed in laboratory settings, and scientific literature is used as a source to enhance knowledge of key problems (Pisano, 1996). The next stage is referred to as pilot development. This stage is more “empirical” and focuses on scaling up the process to an intermediate level and to anticipate key process parameters, e.g. time and temperature. The third and final stage involves scaling up the process created to full-scale and transferring it to the plant where the process will be adapted and used to produce the product (Pisano, 1994).

In a similar vein, Lager (2000) presents a conceptual model for the “process development process”, which comprises three stages: (1) identifying internal production needs, (2) process development work in laboratories, including pilot plant and production plant tests, and (3) transferring development results to production. His findings indicate that practitioners often conduct the first and last stages with low proficiency, although these are indeed critical to accomplish a well-functioning manufacturing process. In addition, Lim et al. (2006) suggest an iterative, multi-phased development model of the process development, enacted by studying biopharmaceutical development. The biopharmaceutical context is characterized by radical innovation, thus making this model less useful in the context of this paper.

Common to all three models is the importance given to the early stages of process development. Despite the assumed importance of the early stages, however, all of these models lack sufficient details about which activities actually constitute the early stage (Lager, 2000; Lim et al., 2006; Pisano, 1994). As the extant knowledge on the FFE in process development is so scarce, the next section revisits the literature on the FFE in product development which serves as a guiding device for our empirical study. 2.3. The fuzzy front end in product development

6

The FFE in product development has been described by previous scholars as critical to ultimate new product success (e.g. Cooper, 1988; Khurana and Rosenthal, 1997; 1998; Griffiths-Hemans and Grover, 2006). The starting point for the FFE is when a firm has an idea for a new product (Kim and Wilemon, 2002). This implies that the product idea needs to be shared in a social structure, as opposed to residing within the head of an individual (Khurana and Rosenthal, 1997). The FFE ends with a decision to either approve or disapprove a formal product development project (Khurana and Rosenthal, 1998; Verworn, 2006). In between the idea and the decision to proceed to formal development or not, a variety of key activities constitute the FFE. Such activities include, but are not limited to, preliminary opportunity identification, idea refinement, market analysis, preliminary technology assessment, product and portfolio strategy, and the enactment of a product concept (Khurana and Rosenthal, 1998; Montoya-Weiss and O’Driscoll, 2000; Seidel, 2007).

The FFE thus typically starts with a more precise opportunity, and if the opportunity is considered worth exploring, a small team is assigned to investigate it further (Kim and Wilemon, 2002). Proficiency in idea refinement and screening of ideas thus constitute key activities (Griffiths-Hemans and Grover, 2006; Khurana and Rosenthal, 1997) as deficiencies here often result in costly problems at later stages of the product development process (Cooper, 1988). Another key activity performed early in the FFE is market analysis, which typically contains information about target markets and the definition of market requirements (Cooper and Kleinschmidt, 1987; Montoya-Weiss and O’Driscoll, 2000). In addition, firms typically conduct preliminary technology assessment which involves identifying potential technologies and suitable applications for developing the new product as well as an estimation of related costs (Cooper and Kleinschmidt, 1987; Verworn, 2006).

Furthermore, it is crucial to determine whether the proposed product fits with existing business plans and to assess synergy with existing products, i.e. product and portfolio strategy considerations (Khurana and Rosenthal, 1998). The evaluation of the competitive situation – environmental scanning and analysis – is yet another activity which needs to be addressed during the FFE. In particular, competitors’ prospective and current product offerings need to be taken into account when exploring the new product idea (Bacon et al., 1994).

Subsequently, a product concept is enacted, which constitutes the “output” of the previously performed key activities. A product concept represents the goals for the development project (Seidel, 2007). In addition, it includes a statement of customer benefits, information about target markets, product specifications,

7

product requirements, and the technology required to produce the product (Cooper and Kleinschmidt, 1987; Montoya-Weiss and O’Driscoll, 2000). Finally, feasibility analysis and project planning are conducted which include tests of the product concept and specifications of the resources that are needed to complete the project (Khurana and Rosenthal, 1997). Altogether, these activities form the basis for the decision to proceed to a formal product development project or not (Khurana and Rosenthal, 1998).

Conceptually, some of these activities are likely to be critical for the FFE in process development as well (e.g. preliminary opportunity identification, idea refinement, and preliminary technology assessment), while others are clearly product-specific (market analysis, identification of market segment, and an evaluation of the competitive situation). The empirical study to explore these issues is described in the following section.

3. Research Design After a detailed literature analysis, our research design was selected to

achieve high methodological fit (Edmondson and McManus, 2007). For multiple reasons, a case study was selected as the research strategy for this study. First, the FFE in development projects is pictured as complex, consisting of multiple iterative activities (Chang et al., 2007; Khurana and Rosenthal, 1997; Montoya-Weiss and O’Driscoll, 2000), which make the case study a feasible approach. Second, given the background that FFE research in process development is severely limited, a case study allows a more contextual assessment of social phenomena in real-life contexts (Meredith, 1998; Yin, 1994). Third, the limited amount of previous research means that themes and patterns need to be identified rather than confirmed (Edmondson and McManus, 2007).

The case study allowed for responding to the need for deep understanding and local contextualization (Miles and Huberman, 1994). In designing and conducting the case study, we followed the guidelines in seminal work on case study research (Yin, 1994). A multiple case study approach was preferred because it enabled the collection of comparative data, which is more likely to yield accurate, generalizable theory than single cases (Eisenhardt and Graebner, 2007; Yin, 1994). The cases were selected using literal sampling, with the aim of sampling cases that would replicate each other and consequently extend emergent theory (Eisenhardt and Grabner, 2007; Yin, 2003). 3.1. Research sites

8

Aligned with this sampling strategy, firms in the process industries were deemed relevant for the empirical study. While process firms are a subset of manufacturing firms, they are defined by several typical characteristics. In particular, the input material is typically raw rather than components from suppliers, the production process is capital-intensive and rigid, and process development is critical to competitive advantage in such firms (Barnett and Clark, 1996; Lager, 2000; Linton and Walsh, 2008). Furthermore, production plants are typically very large and strongly integrated on one site, and the production process is continuous with on-line control in real time (Hutcheson et al., 1995). Process firms do not constitute a completely homogeneous group of firms, however, because of differences not only in the production processes but also in manufacturing strategy (Dennis and Meredith, 2000; Van Donk and Fransoo, 2006). Nevertheless, the activities which constitute the FFE in such firms may still be similar (Lager, 2002; Lim et al., 2006; Linton and Walsh, 2008). Overall, the process industry can be considered as an ideal setting for a study such as ours, because of the key role played by process development (Barnett and Clark, 1996).

Four firms in the metal and mineral industry were selected, as they all had great experience of conducting process development, and good access was given to all firms. Boliden AB is a mining and smelting firm producing mainly zinc and copper. Other important metals produced are lead, gold, and silver. This firm has about 4600 employees in 7 countries. Höganäs AB develops and manufactures metal powders for the global market. Metal powder technology is used in a variety of application areas, including sintered components, soft magnetic composites, hot polymer filtration, and surface coating. The firm has about 1600 employees in 28 countries. LKAB develops and produces upgraded iron ore and industrial mineral products for the steel and other industry sectors, and employs about 3800 persons in 15 countries. SSAB develops and manufactures high-strength steel for the world market. Application areas include bridges, buildings, ships, various forms of vehicles and lifting devices. The firm currently employs around 9200 employees in 45 countries. 3.2. Data Collection

The primary data source consisted of 32 in-depth interviews, which was motivated by the fact that interviews are an efficient approach to collect rich empirical data (Eisenhardt and Graebner, 2007). Four initial interviews were unstructured and exploratory, and they were conducted with the aim of providing general information on innovative activities as well as identifying

9

suitable informants for the subsequent investigation. The remaining 28 interviews were semi-structured and specifically addressed the research purpose. We used a “snowball/chain sampling” approach to select the 28 informants, and this approach allowed us to identify the key persons working with process development (Miles and Huberman, 1994). The sampling approach led to differences in how many informants were selected at each firm. In all firms, however, informants were selected from different hierarchical levels to get diverse perspectives of the activities and problems in the FFE stage, which also mitigate respondent bias (Eisenhardt and Graebner, 2007). Table 1 provides descriptive information about the semi-structured interviews.

10

11

Int

ervi

ews

at H

ögan

äs A

B

Inte

rvie

ws

at L

KA

B

Chi

ef M

etal

lurg

ist O

ffic

ier

2008

-10-

102h

, 40

min

Ph.

D9

Sen

oir

Res

earc

her

2008

-09-

241h

, 36

min

M.S

c.24

Glo

bal P

roce

ss

Dev

elop

men

t Man

ager

2009

-09-

241h

, 5 m

inM

.Sc.

21D

epar

tmen

t M

anag

er P

roce

ss

Tec

hnol

ogy

2008

-10-

061h

, 12

min

Ph.D

25

Plan

t Man

ager

2008

-10-

101h

, 29

min

M.S

c.11

Man

ager

of

Pro

cess

Tec

hnol

ogy

2008

-10-

031h

, 58

min

M.S

c.19

Dev

elop

men

t Eng

inee

r20

09-0

9-28

48 m

inM

.Sc.

8C

oord

inat

or

2008

-10-

231h

, 31

min

Non

aca

dem

ic

trai

ning

18

Pro

ject

Man

ager

, Con

stru

ctio

n 20

08-1

0-03

2h, 0

2 m

inPh

.D13

Pro

ject

Man

ager

, Con

stru

ctio

n 20

08-0

9-23

1h, 3

8 m

inM

.Sc.

29P

roje

ct M

anag

er, C

onst

ruct

ion

2008

-09-

231h

, 38

min

M.S

c.17

Inte

rvie

ws

at B

olid

en A

BIn

terv

iew

s at

SSA

B

Chi

ef S

trat

egy

Off

icie

r20

08-1

1-06

2h, 4

0 m

inM

.Sc.

15C

hief

Met

allu

rgic

al D

evel

opm

ent

2008

-10-

011h

, 18

min

M.S

c.20

Seni

or M

etal

lurg

ist

2008

-11-

061h

, 27

min

M.S

c.13

Dev

elop

men

t Eng

inee

r20

08-1

1-13

1h, 3

2 m

inM

.Sc.

4

Seni

or M

etal

lurg

ist

2008

-11-

071h

, 54

min

Ph.

D29

Man

ager

of

R&

D a

nd Q

uali

ty20

08-1

1-13

1h, 0

4 m

inM

.Sc.

25

Proc

ess

Tec

hnol

ogy

Dev

elop

men

t Man

ager

2008

-11-

071h

, 02

min

M.S

c.11

Pro

ject

Man

ager

, CI

2008

-10-

241h

, 19

min

M.S

c.3

Min

ing

Tec

hnol

gy

Man

ager

2008

-11-

0652

min

M.S

c.1

Proj

ect

Eng

inee

r20

08-1

1-07

47 m

inM

.Sc.

24

Seni

or M

etal

lurg

ist

2008

-11-

071h

, 48

min

Non

aca

dem

ic

trai

ning

23

Exp

lora

tion

Dir

ecto

r20

08-1

1-06

1h, 0

3 m

inM

.Sc.

9T

echn

olog

y M

anag

er20

08-1

1-06

1h, 3

2 m

inM

.Sc.

30Pr

oces

s M

etal

lurg

ist

2009

-10-

2356

min

M.S

c.4

Proc

ess

Met

allu

rgis

t20

09-1

0-23

56 m

inM

.Sc.

22Pr

oces

s D

evel

opm

ent

Dev

elop

men

t Man

ager

2009

-10-

2341

min

M.S

c.12

Proc

ess

Met

allu

rgis

t20

09-1

0-23

41 m

inM

.Sc.

4

Deg

ree

Yea

rs o

f em

plo

ymen

t

Deg

ree

Yea

rs o

f em

plo

ymen

t

Pos

itio

nIn

terv

iew

d

ate

Dur

atio

nD

egre

eY

ears

of

empl

oym

ent

Pos

itio

nIn

terv

iew

d

ate

Du

rati

on

Yea

rs o

f em

ploy

men

tP

osit

ion

Inte

rvie

w

dat

eD

ura

tion

P

osit

ion

Inte

rvie

w

dat

eD

urat

ion

Deg

ree

Tab

le 1

Info

rman

ts for

the

sem

i-st

ruct

ure

d inte

rvie

ws

The semi-structured interviews were based on an interview guideline, which was developed based on prior work into the FFE in product development and general research into the process development work process (see the appendix). The interview guideline comprised four sections. The first section provided background information about the informants’ age, education, and work duties. The second section encompassed general questions about the firm’s process development practices and a discussion about how process development work was organized and conducted, with an emphasis on the problems and opportunities associated with the current process development processes. The third section focused explicitly on front end practices, i.e. key activities, roles, responsibilities, problems and opportunities present in the FFE of process development. The final set of questions allowed informants to elaborate upon what drives success in the FFE stage, i.e. which factors and activities were most critical to successful outcomes. Due to the exploratory nature of the research, the format of the interviews was adapted and changed slightly over the course of the data collection period, to pursue interesting and particularly relevant new facets as they emerged.

All interviews were recorded, transcribed, and included in a case study protocol (Yin, 1994). The interviews ranged from 40 up to 160 minutes and were conducted over the course of approximately two years, from August 2008 to October 2009. To extend the insights from the interviews, informal discussions with managers and engineers were held and observations were made. In addition, we collected secondary data on documented procedures, annual reports, and other company information, thus allowing for empirical triangulation of the firms’ process development practices. 3.3. Data analysis

We started the data analysis by analyzing each case separately, thus compiling a case study history based on the interviews, field notes, observations, and secondary data (Yin, 1994). Each interview was transferred into a spreadsheet for further analysis. To facilitate the analysis, related questions were conceptually clustered together according to the general theme that they explored (Miles and Huberman, 1994). The main unit of analysis is the FFE stage of process development, whereas the identification of key success factors at the firm level constitutes an embedded unit of analysis (Yin, 2003).

At this stage, data were coded and thematically organized around four main themes, which emerged and thus constituted our first-order (informant) findings. The author team jointly decided on which codes to employ, based on an

12

unprejudiced reading of the transcripts. Because we were interested to explore rather than to prejudge or ex-ante determine the FFE stage of process development, we sought to let patterns emerge from the interview data. The first code addressed questions related to the definition and nature of the FFE stage, whereas the second code dealt with activities done in the FFE. The third code focused on managerial problems in the FFE, and the fourth code addressed potential success factors of FFE activities. Each transcript was first read and coded, and then the findings were summarized for each case. If a dimension was identified by a majority of our informants, it became part of our first-order findings. When important information was missing, several shorter telephone interviews were conducted at this stage to retrieve such missing information.

We then performed a cross-case analysis, in which we compared the first-order findings in each case across the cases and looked for similar themes (Yin, 1994). We performed the cross-case analysis after completing the data collection, so that we could replicate the cases against one another. As the first-order coding was descriptive, we made a second-order coding where we interpreted the descriptive data and compared the cases against each other and searched for causal links (Miles and Huberman, 1994). Thus, the empirical data was compared to the literature which constituted the conceptual set up of this paper. Thus, the second-order coding process moved beyond the first-order findings. For example, as we conceptualized the FFE stage and identified different sub-phases we related the key activities to these sub-phases. As in similar studies (Galunic and Eisenhardt, 2001; Smith et al., 2009), this analysis was a continuous process that required repeated reading of the interview files, the notes, the secondary data, and the theoretical framework. 3.4. Validity and reliability issues

To strengthen internal validity, a clear research framework was designed and extensively discussed among the author team prior to the data collection process. We also used multiple theoretical perspective that were evaluated when trying to explain and find counterarguments to the main theses in this paper, drawing on literature from technology- and innovation management, organization theory, product development and operations management. Because the current study was conducted in a previously unexplored context, pattern matching was difficult. Consequently, internal validity could have been affected negatively (Denzin and Lincoln, 1994). Construct validity was addressed mainly by triangulation and by establishing a clear chain of evidence which would allow readers to see how the research questions matched with key conclusions (Yin, 1994). For example,

13

secondary data on formalized work processes for process development were compared to insights from the interviews. On this basis, our multiple case study approach also strengthened external validity and helped to mitigate observer bias (Eisenhardt and Graebner, 2007).

In order to limit potential biases in the data collection process, we selected highly knowledgeable informants who had diverse perspectives concerning the FFE in process development. To create overlap between data collection and data analysis, frequent discussions and preliminary analysis of emergent findings were shared among the authors of this paper. To further increase reliability (transparency and future replication), a case study protocol was constructed together with a case study data base, containing case study notes, documents, and the narratives collected during the study, all with the aim of facilitating retrieval for future studies (Yin, 1994).

4. Empirical Findings This section reports the results from the cross-case analysis, but key

information on each case is provided in addition. We organize our findings into three main sections. First, we conceptualize the FFE stage of process development and describe the critical activities performed therein. We link activities to a conceptual model to provide a comprehensive overview of “what happens” in each specific phase. Subsequently, an integrated perspective on the FFE in process development is presented. Second, typical problems with FFE practices in process development are highlighted. Finally, we elaborate what managers could do to facilitate front end activities performed by individuals or a project team. 4.1. The FFE stage in process development

Overall, the FFE stage was pictured by informants as complex yet critical to end results. In all firms, this stage involved both formal and informal activities.. Common to all firms in our study was a systematic project model applicable to the early stages of process development. Although differences in details were spotted among the firms studied, it basically consisted of four sub-phases which we label informal start-up, formal idea-study, formal pre-study, and formal pre-project. However, this model was not always applied to the FFE projects pursued. Often, the model was more rigorously used for projects with higher degrees of newness, such as those requiring major investments in new process technology, and less rigorously applied to incremental development projects.

14

Below, the sub-phases of the FFE are described in detail and the activities which typically belong to each sub-phase are illustrated. 4.1.1. The informal start-up phase

In the beginning of the FFE, ideas for process development were often fuzzy and unclear, containing multiple loose ends. Therefore, a key activity in the start-up phase was idea generation and refinement. Typically, ideas either came from top management or middle management, or they originated among development personnel on own initiatives without any explicit order. If the idea came from management and it was in line with the firm’s manufacturing strategy, it was considered legitimated and supported. Although a need for refinement was often spotted, the idea often moved quickly into a formal idea-study. If, on the other hand, the idea originated from development personnel, additional informal activities took place to gain legitimacy. These activities were typically performed outside the formal project model.

Because process development efforts span from minor changes in work methods to significant investments in new process technology, ideas came from a variety of different sources. In principle, any person involved with the production process or with development in general could come up with ideas for new process development projects. Most of the ideas in all firms came from the production personnel, but ideas also originated in the R&D department, product development, or central process development departments. Although the triggers to new ideas were multiple, cost reductions and quality problems with existing products were the most common starting points.

To further refine ideas, informal conversations with colleagues frequently took place, e.g. at coffee breaks. In addition, process development ideas were discussed and refined when persons from different functional departments gathered and discussed problems in the production process. Hence, the second key activity in the informal start-up phase was informal discussions. An idea which gained momentum and legitimacy among development personnel was subsequently explored in a formal idea study. 4.1.2. The formal idea-study

The idea-study was pictured as the first “formal” phase of the FFE, and it was typically initiated by a metallurgical or process development manager. At this stage, the idea was further specified and a clearer conception of the problem was enacted, to which the idea was thought to present the solution. Specifying the ideas was typically done by means of group discussions and/or by conducting a

15

formal literature review, depending on the type of process innovation project carried out. For example, if the anticipated problem and the idea related to quality issues in the end product, the idea for a solution was often frequently discussed among process operators and development engineers. If the idea for the process development project instead came from senior management, thus implying a more long-term objective such as to evaluate future production processes, opportunities and constraints, then a formal literature review was complied.

When the problem and the idea were sufficiently understood, they were typically extensively discussed in subsequent steps with other colleagues and, later on, with senior researchers within each respective firm. A key question asked in the idea study was whether the process change that was required would affect the properties of the final product and, if so, in what way. Thus, there was a strong interdependence between process development and product development. Because changes in the production process often affect the end-product, the second key activity was the anticipation of end-product changes. Typically, the next key activities performed in the idea-study were to test and further validate the idea by means of small-scale trials. Preliminary tests in bench scale or preliminary laboratory tests were typically conducted at this stage.

The idea-study was described as highly theoretical, meaning that most of the work was conceptual and previous research and knowledge were regarded as key inputs. Few practical tests were performed at this stage, and if tests occurred, these were conducted in a small-scale laboratory setting. For incremental process development projects, this stage was often deemed to be of less importance. For more radical projects, implying more all-encompassing changes to the production process, this stage was considered paramount. Another key activity was the creation of a preliminary “process concept”. Similar to the conception of a product concept, respondents envisioned this to include a set of functional specifications of the desired process objectives. The “output” of the idea study was typically a report describing the preliminary objectives for the project, together with a specification of the problem to be solved and the hypothesized gains to be reaped. Hence, the final key activity in the idea-study can be summarized as definition of project objectives. Finally, a decision to proceed or not to proceed to a pre-study was taken, typically by a process development department manager. 4.1.3. The formal pre-study

The idea-study served as the input material to the pre-study, which was further explored by doing more detailed, fine-graded, and rigorous empirical

16

tests. The project idea was thus further specified. These specifications often triggered investments in or reconstruction of process equipment which would further enhance efficiency in the production processes. Questions such as what we should develop, how we should do it, and what are the costs and consequences of the proposed process change were typical in the beginning of the pre-study. Risk analyses of different types, e.g. internal and external environmental analysis, were a typical key activity in the pre-study phase. Questions such as how will the process changes affect the work environment and what external effects the proposed change may have on the environment outside the firm were critical, e.g. environmental pollution issues.

In most cases, a project group with persons from different functional backgrounds was assigned to conduct the pre-study. A key task of the project group was to verify ideas by performing a series of experiments at several different scales. Hence, key activities performed in the pre-study were more fine-grained bench-scale tests, laboratory tests, and pilot plant tests. The type of tests that were performed depended on the type of process development project. Tests in the bench scale and laboratory had the smallest scale, and they were often conducted with the purpose of anticipating changes in the chemistry of the end product. Process engineers and researchers altered process parameters, e.g. temperature and pressure, to see whether preliminary goals for the project could be reached. Subsequently, the process was scaled up to some intermediate level and the appropriate level of key process parameters was selected.

As more knowledge was gained about the effects and consequences of the planned process change, the preliminary objectives were often changed slightly, which created a need for iteration among key activities. For example, the results of experiments frequently improved the emerging process concept. Consequently, important project objectives were revised, and these revisions triggered new tests. The refinement of the preliminary process concept was therefore the last key activity in the pre-study, as the final evaluation of the process concept typically took place in the pre-project phase. 4.1.4. The formal pre-project

The concluding phase of the FFE in process development was the pre-project phase. In this phase, the first key activity was specification and selection of final process concept. If the project involved building or buying new process equipment, i.e. construction/modification of process equipment, this equipment was tested and evaluated at this final phase. Full-scale experiments were another activity conducted to further verify the process change. The full-scale experiment was the

17

most accurate estimation of all, as the idea was tested in the actual production environment. All other experiments such as bench-scale tests, laboratory tests and pilot plant tests involved many restrictions related to scaling problems. Moreover, a feasibility analysis and a project plan were viewed as key activities. By conducting a feasibility analysis, the current capabilities and requirements were explored as well as whether the proposed process change was in line with the firm’s overall strategy. In the project plan, resources, costs and a time plan were estimated for the overall development project. Based on the data analysis, Figure 1 provides a visual representation of the FFE stage of process development.

Go/Nogo

decision

Time

Idea generation andrefinement

Informaldiscussions

FORMAL IDEA-STUDY

Literature review

Anticipation of end-product changes

Preliminary bench tests

Preliminary laboratory test

Creation of a preliminary process concept

Definition of project objectives

FORMAL PRE-STUDY FORMAL PRE-PROJECT

More fine-graded bench tests

Laboratory test

Pilot plant tests

Risk analysis

Refinement of preliminary process concept

Specification and selection of final process concept

Construction/ modification of process equipment

Full-scale test

Feasibility analysis

Project planning

INFORMALSTART-UP

Figure 1 The FFE stage of process development

4.1.5. An integrated perspective on the fuzzy front end in process development

An important observation is that the activities performed within the FFE in process development differ depending on the type of process development project that was pursued. Because of the divergence among different types of projects, it is somewhat difficult to present one general and all-encompassing model that would be applicable to all sorts of projects. Key activities performed differed both among and within the firms studied, and Figure 1 thus represents an ideal type, which is a good proxy for most process development projects.

18

In particular, our study allowed identifying several key factors that influence firms’ management of front end activities. First, the degree of newness of the process development project is a critical determinant of the activities in the FFE. All else being equal, a higher degree of newness made the project more complex and uncertain. Accordingly, higher newness resulted not only in more activities, but also in longer time-frames for individual activities, e.g. to verify ideas and problems. For example, if the process change implied radical changes, such as the introduction of the next-generation production process, it strongly affected not only the number of activities performed but also the rigor of execution.

Second, a firm’s prior expertise about the current problem strongly affected the project design and outcome of the FFE stage. If the firm had previous knowledge of a given problem, the idea-study was often skipped and the process engineers started directly by doing experiments. For example, in cases concerning minor process changes, the process engineers already had the required knowledge to solve the problem upfront, which meant that activities involving acquiring relevant knowledge could be skipped. On the other hand, if the problem was novel and knowledge was lacking, then a key activity was to acquire solid information about the anticipated problem through a thorough literature review and small-scale experiments.

Third, the source of the process change was deemed critical. Long-term oriented process development projects typically came from senior management, and they required a more formalized structure. In comparison, more short-term oriented projects initiated at lower levels were often conducted informally, often with the characteristics of “skunk-work”. For example, if a project proposal was ordered by senior management, activities were typically performed in a more formal and proficient way. Literature reviews and documentation were example of activities which varied in conduct depending on the source of the process change.

Fourth, the time frame for the project was another key issue. FFE projects were often given an approximate timeframe at the outset, which greatly affected the conduct of key activities. Under conditions of a short time frame, some activities were often skipped or performed in an ad-hoc manner to keep the project on time. Typically, a complete literature review, documentation of experiments, and exact anticipation of the process concept often suffered in projects with shorter time-frames.

Finally, process development projects often had a center of gravity in either technical or organizational change. Although most projects spanned the technology/organization interface (e.g. the installation of new process equipment

19

which requires new routines and work practices on the part of labor), the center of gravity affected which project activities were conducted. In particular, technical changes were rarely effective without being accompanied by organizational changes. For example, if the process change implied investments in new process technology, technology assessment was critical. But the establishment of new work methods among operating personnel was at least as critical.

4.2. Key problems in the FFE stage of process development

Several key problems and deficiencies in FFE projects emerged from analyzing the data. While some of these challenges were common to all firms studied, others were firm-specific. A key problem shared by all firms, which also constitutes a paradox, was early production involvement. In all firms studied, the production department was the main source of process development ideas. While this ensured that new projects were relevant, it also favored incremental adjustment of existing operations and processes at the expense of more radical ones. That is, most projects addressed the daily problems in the production process whereas projects of high future importance often were difficult to pursue. The focus on incremental adjustments depended mainly on the restriction of new investments and the fact that current production problems were typically prioritized at the expense of future-oriented ideas. This problem was reinforced when production plants were cost-centers and plant managers prioritized among different ideas, thus delimiting the influence of central process development departments, who often were the sources of more radical ideas.

Second and similarly, most FFE projects in process development were reactive, implying that process engineers mainly worked with existing and acute problems in the production plants. An analogy of the work performed by the process engineers was made to the work performed by a fire brigade. This analogy typically demonstrates that process development is mainly reactive (i.e. to put out existing fires) rather than proactive (i.e. to plan ahead and prepare for future challenges).

A third common problem shared by all firms was the lack of time and resources to conduct proficient process development. The reactive nature of process development made it difficult to plan ahead and estimate the appropriate time and resources to conduct process development. For example, due to time and resource constraints, evaluations of experiments and calculations in the FFE stage were often based on personal judgments and gut feeling rather than on acquired facts.

20

Finally, documentation of results was a general problem within all firms. Especially, failed experiments were seldom documented, resulting in the same mistakes being repeated over and over again. In addition to these problems, the firms reported specific problems in the FFE stage when conducting process development projects. For example, idea generation was considered a weak link at LKAB. Boliden AB pointed out screening and evaluation of ideas as one of its primary problems, whereas the main problem at Höganäs was described as getting acceptance for ideas that were outside the traditional way of conduct. At SSAB the main problem was pictured as an overall resistance to the change of routines and work methods. 4.3. Key success factors in the FFE activities of process development

The informants in our cases listed early involvement of production staff, cross-functional teamwork, and a creative culture as the key aspects to support activities in the FFE. Early production involvement was judged decisive for a proficient FFE stage. Frequent and close communication with operating personnel provided a clearer understanding of the current production problems and could provide insights about future challenges. Inviting the operating personnel to participate in the FFE was an alternative way to gain direct access to critical knowledge of the production process, and it was also believed to reduce the resistance to change, which thus facilitates the subsequent implementation of the process development project.

Moreover, cross-functional teams were further stressed as critical to achieve enhanced performance in the FFE stage. The systemic nature of process development implies that a number of different functions and departments within the firm should be involved in the design and implementation of process development work, which calls for close collaboration among functions and departments. Further, the systemic nature of process development made it critical to integrate technical choices with operating conditions, because process performance is affected by interactions among actual operating conditions, technical choices and capabilities of the future production process. The process engineers thus needed to collaborate with several different functional groups, e.g. operating personnel and product engineers. To facilitate cross-functional collaboration, frequent meetings and common goals were deemed important.

Finally, a creative culture was considered to be critical to enhance performance in the FFE stage. A creative culture, according to the informants, involved freedom, playfulness, tolerance for individual differences among employees, and trust and empowerment on the part of management. According

21

to the informants, to create a creative culture involved addressing all of the above aspects as well as enabling slack, where the employees were encouraged to pursue informal discussions, tests, and experiments to evaluate emerging ideas.

5. Discussion and Implications Despite the positive effects of process development on manufacturing

performance (Reichstein and Salter, 2006) and overall firm performance (Gopalakrishnan et al., 1999; Utterback and Abernathy, 1975), knowledge about the process development activities which precede formal projects has remained scarce. To a background where the technology- and innovation management literature has established these activities as critical to success in new product development (Cooper, et al., 2004; Verworn, 2009), the absence of studies in the process domain is somewhat surprising. While previous scholars have pointed out the need for additional research into the FFE in process development (Lager, 2000; Lim et al., 2006; Pisano, 1997), extant knowledge still remains limited.

Our research has helped to close this knowledge gap. While our results show some similarities to front end activities in product development, they stress that extant knowledge on the FFE in product development cannot be directly transferred to the process domain because of differences in key activities. As front end activities are critical to ultimate success in process development (Pisano, 1997), our study deepens the understanding of the FFE in a previously unexplored context. These findings are particularly relevant in light of the desire of both managers and academics to better understand the activities, issues, and managerial challenges needed to enhance proficiency in process development (Hatch and Mowery, 1998; Khazanchi et al., 2007; Reichstein and Salter, 2006). 5.1. Theoretical implications

Our empirical study has important implications for research into process development, new product development, and production management. First, regarding process development research, our study provides new insights into the “process development process”. In particular, we conceptualize the FFE stage of process development and identify the key activities which constitute this stage. Our findings thus extend those of previous scholars (Lager, 2000; Lim et al., 2006; Pisano, 1996; 1997) by addressing in detail what constitutes the FFE stage in process development. These findings are important because existing models of the process development process are relatively abstract and lack sufficient details of what actually happens in the early phases, which appear critical for success. By

22

conceptualizing the FFE stage in process development, an important step is taken to increase the understanding of how process development efficiency is ultimately enabled.

Specifically, we have found that the FFE stage in process development typically consists of four different sub-phases. These were labeled informal start-up, formal idea-study, formal pre-study, and formal pre-project. To our knowledge, none of these sub-phases have been reported in the literature on process development before, but similar models prevail in the literature on the FFE in product development (Khurana and Rosenthal, 1997; 1998). Further, our study deepens the understanding of the key activities that take place throughout the various sub-phases of the FFE. Some of the activities performed are equivalent to the ones conducted in the FFE stage of product development, such as idea generation, development of a product/process concept, definition of project goals and project planning (Khurana and Rosenthal, 1998; Seidel, 2007; Griffiths-Hemans and Grover, 2006). In contrast, other critical activities such as anticipation of end-product changes and construction/modification of process equipment have not been reported previously. Thus, our study provides important new insights into the critical activities within the FFE stage in process development.

In addition, we present empirical evidence of managerial problems in the FFE stage of process development. Our study has shown that the reactive nature of process development, ad-hoc documentation of tests and early production involvement are typical problems in the FFE in process development. These problems have not been reported in the extant literature on the FFE in product development, while problems such as a lack of time and resources have been reported previously (Khurana and Rosenthal, 1998). Accordingly, our study contributes to a clearer picture of the managerial challenges of process development activities.

Moreover, we identified several factors which may impact the organization of FFE activities in process development. In particular, these factors were the degree of newness, state of existing knowledge, sources of ideas to process development, scope of the process development project, and center of gravity of the process development project. According to these factors, the FFE stage in process development may vary substantially. Consistent with contingency theory (Perrow, 1970), our findings suggest that there is no single best way to organize the FFE in process development. Rather, these factors should be considered before structuring the FFE. Consequently, the FFE stage in process development looks different with respect to which activities are performed and prioritized.

23

Accordingly, researchers should not oversimplify the organization of the relevant tasks because the organization depends on the specific situation and environment of the process development project.

Our analyses additionally revealed that both informal and formal activities are present in the FFE stage of process development, and that many incremental process development projects are conducted without starting a formal project, whereas in the cases of more radical process development the FFE tends to be relatively formalized. Thus, our results are contrary to product development research on the FFE suggesting that incremental product development is more formalized than radical product development (Khurana and Rosenthal, 1997; Reid and de Brentani, 2004). This appears to be an important difference between the FFE in process and product development. This finding calls for breaking new ground because any attempts to transfer knowledge from product development research to process development seem to be questionable with regard to the formalization of the FFE activities. Accordingly, this issue needs to be highlighted as a particularly fruitful avenue for further research.

Second, regarding new product development research, we have identified important interdependences between process development and product development. These interdependences have received little attention in previous research, but our study indicates that process development practices may be essential to achieve high product development performance. Accordingly, our findings suggest that product and process development are complementary, i.e., more of any one of them increases the returns to the other (Ennen and Richter, 2010; Lichtenthaler, 2009). Complementarity theory helps explain the super-additive value of particular resource configurations (Milgrom and Roberts, 1995; Song et al., 2005). Consistent with this theory, our findings suggest that strong process development enhances a firm’s benefits from product development. As product and process development are interlinked in a firm’s overall innovation activities, changes in the process most likely affect the end-product (Barnett and Clark, 1996) and this makes it important to anticipate changes in the end-product as early as possible.

Third, concerning research into production management, our study has highlighted the important role of production management in the early stages of a firm’s process development activities. Early production involvement is described as really critical and therefore it is highlighted as a major success factor in the FFE of process development – and at the same time, it is difficult to implement, which makes it a major problem. Thus, our findings suggest that early production involvement constitutes a paradox which needs to be better understood. Poole

24

and van de Ven (1989) argued that addressing social-scientific paradoxes could further advance management practice as well as organization theory, because paradoxes are about tensions and oppositions between incompatible positions of real-world activities, which are subject to temporal and spatial constraints. Therefore, the paradox of early production involvement suggests that future studies should pay particular attention to how the benefits of such integration may be reaped, while at the same time avoiding the risks.

In light of the extant process development literature, our findings suggest that the role of the production department in process development has been acknowledged insufficiently (Linton and Walsh, 2004). While earlier work has pointed out the importance of production management for product development, our findings emphasize its key role in process development. Beyond the importance of integrating production management in the later stages of the new product development process, our interviews have underscored the need for integrating production management expertise from the very beginning of the process development process (Nihtila, 1999). 5.2. Managerial implications

Our study provides direct managerial implications for enhancing the success of FFE activities in process development and, consequently, for improving the overall process development process. Above all, our findings clarify which activities need to be addressed in the FFE stage in process development. These results provide implications relevant to top managers, project managers, plant managers, and development engineers of process development projects. In essence, we present a template for how the FFE stage can be structured and managed. Managers can thus not only control the conduct of formal activities, but also encourage the more informal ones in order to enhance the performance in the FFE stage.

In particular, the FFE is a relatively unclear and ill-defined stage, and it comprises substantial ad-hoc decision-making (Chang et al., 2007). To some extent, these problems are difficult to avoid because of the inherently fuzzy nature of this initial stage of process development. Despite this fuzziness, however, there are some specific recommendations for managers which will most likely contribute to enhancing a firm’s FFE in process development. The findings of our study further reveal other problems in the FFE stage of process development, such as low levels of innovativeness due to early production involvement, a lack of time and resources, ad-hoc documentation of tests, and that most process development efforts were reactive rather than proactive. Many

25

of these problems could, arguably, be diminished through a more formalized work process.

Accordingly, our findings imply a call for routines in the FFE stage of process development. A formalized work process is explicit, widely known, and characterized by clear decision-making responsibilities (Khurana and Rosenthal, 1998). On this basis, managers may reduce the uncertainty and equivocality among organizational participants. While it will be impossible to reduce all fuzziness from the FFE, a stronger formalization appears to be beneficial for all process development projects that we have studied. More formalization in the FFE stage will most likely help firms to reap the benefits from organizational routines in process development.

Moreover, early production involvement was considered a critical driver of improving a firm’s performance in the FFE stage. By involving the production experts early, important knowledge is acquired and the process concept can be refined in more detail. Early production involvement further reduces resistance to change. Moreover, it ensures that the process change will be implemented in production, which leads to more efficient overall process development. While management clearly can reinforce early production involvement, particular emphasis needs to be paid to the need for innovation despite early production involvement, because otherwise early production involvement may result in low levels of innovativeness.

Beyond the importance of the production department, cross-functional integration constitutes a major success factor in the FFE of process development. In particular, cross-functional collaboration is critical to facilitate efficient product and process development and sufficient alignment between product development and process development (e.g. Chen, et al., 2010; Pearce and Ensley, 2004). In the FFE stage, the involvement of persons from different functions and backgrounds helps define the objectives and requirements of the new process change before excessive resources are invested. A shared vision of the process concept further facilitates mutual understanding among the functional groups performing specialized roles in the FFE stage, and it reduces resistance to change. Accordingly, managers need to pay particular attention to the cross-functional activities and interfaces in the FFE of process development.

Finally, an entrepreneurial organizational culture provides a major foundation for successful FFE in process development. In particular, an entrepreneurial culture allows employees to test their ideas and to conduct experiments in small scale without excessive bureaucracy. Moreover, trust, freedom, and playfulness tend to play important roles in entrepreneurial cultures,

26

and these aspects constitute important components to actively establish a creative culture in firms (Akün, Keskin and Byrne, 2009; Pearce and Ensley, 2004). Managers thus need to trust employees and reinforce a good relationship among managers and employees to foster a creative culture. Clearly, management can encourage and reinforce such cultural norms (Khazanchi et al., 2007; Pagelle and LePine, 2002). 5.3. Limitations and further research

Some key limitations associated with this study are worth mentioning. First, the findings in this study are based on an in-depth study of four firms, all active within the metal and minerals industry. While sampling cases within one single industry strengthens external validity, it still delimits generalizability outside this context (Yin, 1994). As such, the firms in our study may have idiosyncratic characteristics and, therefore, the suggested conceptualization of the FFE and the identified activities need to be validated against other cases and methods. Another limitation is the sole focus on continuous processes. Thus, an important suggestion for future research is to investigate the FFE in other types of processes, e.g. job shop processes, batch processes, and flow processes. In addition, retrospective sense-making is always an issue to consider when dealing with qualitative interview data. There is thus a risk of relying on retrospective interviews, as informants can forget and misinterpret important facts (Eisenhardt and Graebner, 2007).

On this basis, the FFE stage in process development would clearly benefit from additional research. First, it would be beneficial if the results emerging from this study were more fine-tuned and tested in other cases with different conditions (i.e. polar cases), to test whether the results hold. A second avenue is to extend the framework to other industries such as assembled consumer products. A critical question is whether the same components and activities of the FFE stage in process development are relevant for consumer products. A third avenue is to test the current results in a wider sample of multiple industrial domains. Thus, there is a need for large-scale empirical efforts, which may contribute to further deepening our understanding of the FFE in process development.

Specifically, we encourage mixed-methods studies, which combine qualitative and quantitative data collection. This type of studies may be an essential direction for future work because it enables researchers to gain a thorough understanding while still providing quantitative evidence of complex phenomena, such as the FFE in process firms. As much remains to be explored,

27

there are great opportunities for further research into process development, especially concerning the role of FFE activities. In particular, additional case studies and quantitative analyses may provide results that are equally relevant to academics and practitioners.

28

References Akgün, A.E., Keskin, H., Byrne, J., 2009. Organizational emotional capability,

product and process innovation, and firm performance: An empirical analysis. Journal ofEngineering and Technology Management 26 (3), 103-130.

Bacon, G., Beckman, S., Mowery, D. and Wilson, E., 1994. Managing product definition high-technology industries: A pilot study. California Management Review 36 (3), 32-56.

Baer, M. and Frese, M., 2003. Innovation is not enough: climates for initiative and psychological safety, process innovations, and firm performance. Journal of Organizational Behaviour 24 (1), 45-68.

Barnett, B.D., Clark, K.B., 1996. Technological newness: an empirical study in the process industries. Journal of Engineering and Technology Management 13 (3-4), 263-282.

Brem, A. and Voigt, K-I., 2009. Integration of market pull and technology push in the corporate front end and innovation management - Insights from the German software industry. Technovation 29 (5), 351-367.

Brown, S. 2001. Managing process technology – further empirical evidence from manufacturing plants. Technovation 21 (8), 467-478.

Brown, S. and Maylor, H., 2005. Strategic resonant firms, mass producers, big fish & flat liners: a study of policies, practices & performance in innovation. Technovation 25 (4), 307-319.

Chang, S-L., Chen, C-Y. and Wey, S-C., 2007. Conceptualizing, assessing, and managing front-end fuzziness in innovation/NPD projects. R&D Management 37 (5), 469-478.

Chen, J., Damanpour, F., Reilly, R., 2010. Understanding antecedents of new product development speed: A meta-analysis. Journal of Operations Management 28 (1), 17-33.

Cooper, R.G., 1988. Predevelopment activities determine new product success. Industrial Marketing Management 17 (3), 237-247.

Cooper, R.G. and Kleinschmidt, E.J., 1987. New products: What separates winners from losers? Journal of Product Innovation Management 4 (3), 169-184.

Cooper, R.G., Edgett, S.J. and Kleinschmidt, E.J., 2004. Benchmarking best NPD practices- 1. Research Technology Management 47 (1), 31-43.

Cooper, R.G., 2008. Perspective: The stage-gate idea-to-launch process. Update, what’s new, and Nextgen systems. Journal of Product Innovation Management 25 (3), 213-232.

Damanpour, F. and Gopalakrishnan, S., 2001. The dynamics of the adoption of product and process innovations in organizations. Journal of Management Studies 38 (1), 45-65.

Dennis, D.R. and Meredith, J.R., 2000. An analysis of process industry production and inventory management systems. Journal of Operations Management 18 (6), 683-699.

29

Denzin, N. and Lincoln, Y., 1994. Handbook of Qualitative Research. Thousand Oaks, Sage Publications.

Edmondson, A.C. and McManus, S.E., 2007. Methodological fit in management research Academy of Management Review 32 (4), 1155-1179.

Eisenhardt, K.M. and Graebner, M.E., 2007. Theory building from cases: opportunities and challenges. Academy of Management Journal 50 (1), 25-35.

Ennen, E. and Richter, A., 2010. The Whole Is More Than the Sum of Its Parts – Or Is It? A Review of the Empirical Literature on Complementarities in Organizations. Journal of Management 36 (1), 207-233.

Ettlie, J.E., 1995. Product-process Development Integration in Manufacturing, Management Science 41 (7), 1224-1237.

Ettlie, J.E. and Reza, E.M., 1992. Organizational integration and process innovation. Academy of Management Journal 35 (4), 795-827.

Galunic, D.C. and Eisenhardt, K.M., 2001. Architectural innovation and modular corporate forms. Academy of Management Journal 44 (6), 1229-1249.

Gopalakrishnan, S., Bierly, P. and Kessler, E., 1999. A reexamination of product and process innovations using a knowledge-based view. Journal of High Technology Management 10 (1), 147-166.

Griffith-Hemans, J. and Grover, R., 2006. Setting the stage for Creative new products: Investigating the idea fruition process. Journal of the Academy of Marketing Science, 44 (1), 27-39.

Hatch, N.W., Mowery, D.C., 1998. Process innovation and learning by doing in semiconductor manufacturing. Management Science 44 (11), 1461-1477.

Hutcheson, P., Pearson, A.W. and Ball, D.F., 1995. Innovation in process plant: a case study of ethylene. Journal of Product Innovation Management 12 (5), 415-430.

Khazanchi, S., Lewis, M. and Boyer, K., 2007. Innovation-supportive culture: the impact of organizational values on process innovation. Journal of Operations Management 25(4), 871-884.

Khurana, A. and Rosenthal, S.R., 1997. Integrating the fuzzy front end of new product development. Sloan Management Review 38 (2), 103-120.

Khurana, A. and Rosenthal, S.R., 1998. Towards holistic "front ends" in new product development. Journal of Product Innovation Management 15 (1), 57-74.

Kim, J. and Wilemon, D., 2002. Focusing the Fuzzy Front-end in new product development. R&D Management 32 (4), 269-279.

Koc, T. and Ceylan, C., 2007. Factors impacting the innovation capacity in large-scale companies. Technovation 27 (3), 105-114.

Lager, T., 2000. A new conceptual model for the development of process technology in process industry. International Journal of Innovation Management 4(3), 319-346.

30

Lager, T., 2002a. Product and process development intensity in process industry: A conceptual and empirical analysis of the allocation of company resources for the development of process technology. International Journal of Innovation Management 6 (2), 105-130.

Lager, T., 2002b. A structural analysis of process development in process industry. R&D Management 32 (1), 87-95.

Lager, T., Hallberg, D. and Ericsson, P. (2010). Developing a process innovation work process: The LKAB experience. International Journal of Innovation Management, 14 (2), 285-306.

Lichtenthaler, U. (2009). Absorptive capacity, environmental turbulence, and the complementarity of organizational learning processes. Academy of Management Journal, 52 (4), 822-846.

Lim, L., Garnsey, E., and Gregory, M., 2006. Product and process innovation in biopharmaceuticals: A new perspective on development. R&D Management 36 (1), 27-36.

Linton, J.D., Walsh, S.T., 2004. Integrating innovation and learning curve theory: an enabler for moving nanotechnologies and other emerging process technologies into production. R&D Management 34 (5), 517-526.

Linton, J.D., Walsh, S.T., 2008. A theory of innovation for process-based innovations such as nanotechnology. Technological Forecasting & Social Change 75 (5), 583-594.

Meredith, J., 1998. Building operations management theory through case and field research. Journal of Operations Management 16 (4), 441-454.

Miles, M. and Huberman, M., 1994. Qualitative data analysis. Thousand Oaks: Sage Publications.

Milgrom, P. and Roberts, J., 1995. Complementarities and fit: Strategy, structure, and organizational change in manufacturing. Journal of Accounting and Economics, 19, 179-208.

Montoya-Weiss, M. and O'Driscoll, T., 2000. From experience: Applying performance support technology in the fuzzy front end. Journal of Product Innovation Management 17 (2), 143-161.

Murphy, S. and Kumar, V., 1997. The front end of new product development: A Canadian survey. R&D Management 27 (1), 5-15.

Nihtila, J., 1999. R&D-production integration in the early phases of new product development projects. Journal of Engineering and Technology Management 16 (1), 55-83.

Pagell, M. and LePine, J.A., 2002. Multiple case studies of team effectiveness in manufacturing organizations. Journal of Operations Management 20 (5), 619-639.

Pearce, C.L., Ensley, M.D., 2004. A reciprocal and longitudinal investigation of the innovation process: The central role of shared vision in product and process innovation teams (PPITs). Journal of Organizational Behavior 25 (2), 259-278.

31

Perrow, C., 1970. Organizational Analysis: a Sociological View, London, Tavistock.

Pisano, G.P., 1994. Knowledge, integration, and the locus of learning: an empirical analysis of process development. Strategic Management Journal 15 (Special Issue: Competitive organizational behaviour), 85-100.

Pisano, G.P., 1996. Learning-before-doing in the development of new process technology. Research Policy 25 (7), 1097-1119.

Pisano, G., 1997. The development factory: unlocking the potential of process innovation. Boston: HBS Press.

Poole, M.S. and van de Ven, A.H., 1989. Using Paradox to Build Management and Organization Theories. The Academy of Management Review 14 (4), 562-578.

Reichstein, T. and Salter, A., 2006. Investigating the sources of process innovation among UK manufacturing firms. Industrial and Corporate Change 15 (4), 653-682.

Reid, S.E. and de Brentani, U., 2004. The Fuzzy Front End of New Product Development for Discontinuous Innovations: A Theoretical Model. Journal of Product Innovation Management 21 (3) 170-184.

Samson, D. and Whyback, D.C., 1998. Tackling the ever so important ‘soft’ issues in operations management. Journal of Operations Management 17 (1), 3-7.

Seidel, V.P., 2007. Concept shifting and the radical product development process. Journal of Product Innovation Management 24 (6), 522-533.

Smith, A.D., Plowman, D.A., Duchon, D., Quinn, A.M., 2009. A qualitative study of high- reputation plant managers: Political skills and successful outcomes. Journal of Operations Management 27 (6), 428-443.

Song, M., Droge, C., Hanvanich, S., and Calantone, R. 2005. Marketing and technology resource complementarity: An analysis of their interaction effect in two environmental contexts. Strategic Management Journal, 26 (3), 259-276.

Utterback, J. and Abernathy, W., 1975. A dynamic model of process and product innovation. Omega 3 (6), 639-656.

Van Donk, D.P. and Fransoo, J.C., 2006. Operations management research in process industries. Journal of Operations Management 24 (3), 211-214.

Verworn, B., 2009. A structural equation model of the impact of the “fuzzy front end” on the success of new product development. Research Policy 38 (10), 1571-1581.

Verworn, B., 2006. How German measurement and control firms integrate market and technological knowledge into the front end of new product development. International Journal of Technology Management 34 (3-4), 379-389.

Yin, R., 1994. Case study research: Design and methods. Thousand Oaks: Sage Publications.

32

Appendix: Interview questions on managing and organizing the fuzzy front end in process development Background information 1. What is your formal position within the firm? 2. What are the main activities that you work with and what is your area of responsibility? 3. What are your previous work experience and education, and how long have you been working for your current employer?

General questions about process development 4. How do you define process development and what activities constitute process development? 5. What is the definition of process development at your firm? 6. What is the amount of time that you work with process development? 7. What are the main objectives/purposes of process development at your firm? 8. What is the overall mindset at the firm concerning process development? 9. How is process development organized at your firm? 10. What do the stages of process development look like, i.e. when are things done and in what order? 11. Does your firm have a formal process development work process or method? 12. How are activities coordinated, which span over functional interfaces and departments? 13. Do you have a holistic view of process development at your firm? 14. Can you describe how product development and process development interact? 15. Do process development and product development have the same status at your firm? 16. Do new products drive new processes or is it the other way around, i.e. is it process development that enables development of new products?

Questions of the fuzzy front end in process development 17. What is your experience of fuzzy front end activities in process development? 18. Who has ideas for new process development projects and where do the ideas come from? 19. How does your firm screen new ideas for process development projects? 20. How does your firm choose which ideas eventually will be developed into formal projects? 21. Against which criteria are ideas evaluated? 22. How does your firm balance and mix new ideas for process development projects? 23. How are ideas realized in the context of production? That is, what do you do to implement the ideas in the context of production? 24. What does your firm do to make the ideas work in practice?

33

25. How is the “fuzzy front end stage” structured/organized at your firm? 26. Who is involved in the early phases of process development at your firm? 27. How does the work in the early stage of process development operate in practice at your firm? 28. What are the problems, conflicts, and solutions? What is working well or, perhaps, not so well? 29. What is the most important thing that will make your firm succeed with the early stage in process development? 30. What do you need to change – if you were allowed to wish freely? 31. What are the obstacles in the early stage in process development?

Questions on what drives success in the fuzzy front end of process development 32. What are the critical factors and activities that will make you succeed with the front end? 33. Please explain, based on your previous experience, how to make FFE activities work in practice? 34. Can you explain how the different activities and factors that you elaborated on previously are related to each other, and what firms can do to link the individual factors into a coherent whole? Start with the factor that you deemed most important and elaborate on how it is related to the other factors that you judge as important. 35. Can you tell us, in general terms, how a process concept is developed?

34


Paper A


Paper B



Paper C



Paper D


Paper E

Managing the Fuzzy Front-end: Insights from Process Firms

Monika Kurkkio Luleå University of Technology




Abstract Purpose – The aim of this paper is to inform researchers and practitioners about the fuzzy front-end (FFE) of the innovation process in process firms. Design/methodology/approach – A multiple case study of four process firms was conducted with a total of 64 semi-structured interviews. Findings – The paper gives new insights into the FFE in non-assembled product and process development in process firms. The FFE of non-assembled product and process development is first conceptualized and key activities are identified. Further, it is discussed how the strong relationship between product and process development can be managed in the FFE. Research limitations/implications – All four firms are from the mineral and metals industry, prompting caution when generalizing the results to other contexts. This research offers insights about the FFE in process firms. Theoretical implications are added to the existing literature on the FFE and general process development literature, and the paper increases our understanding of innovation management in general. Practical implications – From a practical point of view, the paper gives advice on how managers in process firms can increase speed and clarity in the FFE. The conceptualizations and the identified front-end key activities are suggested as checklists for improving the FFE stage. Originality/value –This study compares how the FFE within two different types of innovations is conceptualized and managed. Thus, the FFE in non-assembled product and process development is explored. The FFE of process development is an unexplored context. Keywords – Product development, Process development, fuzzy front end, Process industry, case study Paper type – Case study

1

1. Introduction

The importance of innovation is clear—either firms take advantage of new opportunities, or they are replaced by other firms that do (e.g. Cooper and Kleinschmidt, 1987; Utterback, 1994)—but the critical challenge is how to achieve innovation. In the past decade, the fuzzy front-end (FFE) of innovation has been given increased attention (e.g. Björk and Magnusson, 2009; Chang et al., 2007; Seidel, 2007), because the foundation of successful innovation is found in the FFE stage (e.g. Cooper and Kleinschmidt, 1987; Cooper, 1988; Koen et al., 2001). The FFE starts when a firm has a new idea that is shared in a social structure, as opposed to residing within the head of an individual (Khurana and Rosenthal, 1998), and ends when the firm makes a decision to either launch a formal development project or, alternatively, decides not to do so (Kim and Wilemon, 2002).

The FFE is recognized as the most difficult stage to manage in the innovation process (Kim and Wilemon, 2002), as it involves a significant degree of uncertainty and equivocality (Chang et al., 2007). In addition, the FFE is often characterized by ad-hoc decision-making, and conflicting organizational pressures (Khurana and Rosenthal, 1998; Montoya-Weiss and O’Driscoll, 2000). For example, the FFE is difficult to plan and conduct as there are several sources of uncertainty and many ideas arise outside the formal systems (Griffith-Hemans and Grover, 2006). The potential market and the underlying technology are the two main sources of uncertainty in the FFE (Montoya-Weiss and O’Driscoll, 2000; Verworn, 2006).

The FFE of product development has been conceptualized several times in the literature (e.g. Cooper, 2008; Griffiths-Hemans and Grover, 2006; Koen et al., 2001) and key activities have been identified therein as well (Khurana and Rosenthal, 1998; Kim and Wilemon, 2002). Some studies focus on one specific activity in the FFE, e.g. ideation (Björk and Magnusson, 2009; Griffith-Hemans and Grover, 2006), screening of ideas (Khurana and Rosenthal, 1998), and the creation of a product concept (Seidel, 2007), while others take a broader perspective and explore the whole FFE and identify general success factors at the front-end (e.g. Koen et al., 2002; Kim and Wilemon, 2002). However, the FFE literature is criticized for presenting one generic front-end for various kinds of innovation types, in whatever industry (Nobelius and Trygg, 2002). Since many of the early research endeavours do not take into account factors such as innovation type, degree of newness, and industry, it is difficult for firms to actually apply this research to their specific innovation process.

Most of the existing research on the FFE of the innovation process has focused on assembled product development (e.g., Khurana and Rosenthal, 1998; Seidel, 2007; Verganti, 1997). Assembled products are put together by several different components which constitute the end product (Utterback, 1994). Non-assembled products, on the other hand, are composed by only one or a few materials and they are typically used as input in somebody else’s production (Barnett and Clark, 1996). Textiles, chemicals, minerals, metals, and food are examples of non-assembled products. Firms producing non-assembled products are commonly referred to as process firms.

2

Process firms are suppliers of input materials to a variety of different manufacturing industries, for example the automotive industry, the construction industry, the consumer food industry etc. Steel, for example, is an important metal which has become a cornerstone in every home, household, and in the entire society (SGU, 2008). Process firms have historically been important in building up the prosperity and civilization of today’s society. Still, process firms have received modest attention in the innovation management literature compared to other manufacturing industries (Barnett and Clark, 1996; Lager, 2002, Lager et al., 2010; Pisano, 1997). It is of great importance for the prosperity and civilization in the whole society to increase the knowledge of how process firms can improve their innovation processes.

In process firms and several of the traditional manufacturing firms, process development is the key innovation type, rather than product development (Utterback and Abernathy, 1975; Hutcheson et al., 1995). Compared to product development, process development has received limited attention in the innovation management literature (Reichstein and Salter, 2006). However, process development is important for building long-term competitive advantages in manufacturing firms (Pisano, 1997).

This paper explores the FFE in non-assembled product and process development. It is motivated by the narrow focus on product development in the FFE literature, which fails to account for the intricacies of the innovation process as it applies to other innovation types (e.g. processes, non-assembled products, and services), and by the difficulty that firms have in applying the existing research to their specific innovation process. Limited attention has been paid to the FFE of non-assembled products (Bröring et al., 2006; Elmquist and Segrestin, 2007) and more research is needed to understand the FFE in this context. No studies have addressed the FFE of process development, but a few studies have found that the FFE of process development is important for the overall success of the innovation project (Lim et al., 2006; Pisano, 1997). The aim of this paper is to inform researchers and practitioners about the FFE of the innovation process in process firms. Specifically, the research questions are the following:

1. How is the FFE of the innovation process conceptualized in process firms? 2. What are the differences and similarities in terms of key activities in the FFE of product and process development? 3. How is the relationship between product and process development managed in the FFE of the innovation process?

The rest of the paper is organized as follows. Since the FFE in process

development has hardly been studied in the extant literature, the theoretical background in the next section is based on literature about the FFE of the innovation process (mostly referring to assembled products) and general literature on process development (primarily conducted within process firms). In the subsequent section, the method is presented. The empirical findings based on the

3

multiple case studies are presented in the fourth section of the paper, and finally a discussion of theoretical and managerial implications is given as well as suggestions for future research.

2. Theoretical background

Technical innovations can basically be divided into product development (products can be either physical or in the form of services) and process development (Utterback, 1994; Reichstein and Salter, 2006). These two innovation types have different focuses and objectives, and they work according to different logics. For example, while product development is driven by the desire to create a new product, process development relates primarily to internal production objectives, such as cost reductions, enhancement of production volumes, or sustaining more environment-friendly production (Lager, 2002a; Pisano, 1997). Thus, the market for process development is internal to the firm, whereas the market for product development is external to the firm.

Discovery

Stage 1

Gate 1

Idea Screen

Second Screen

ScopingBuilding Business

Case

Go ToDevelopment

Development

Go ToTesting

Go ToLaunch

Testing & Validation

Launch

Post-LaunchReview

Stage 2 Stage 3Stage 4 Stage 5

Gate 2

Gate 3

Gate 4

Gate 5

Figure 1 A typical Stage-Gate process (Cooper, 2008)

The Stage-Gate model is by far one of the most applied models for managing the innovation process (Cooper, 2008). Cooper’s model consists of a series of stages including a number of activities ranging from idea to launch, which are complemented by gates where go/kill decisions are made on whether the firm should invest further in the project or not. The model describes the product development process, beginning with an ideation stage called discovery and ending with product launch (see Figure 1). The front-end in product development involves all the activities which preceded subsequent formal development efforts, and thus the stages of idea screening, scoping, and building a business case. This is a common way to conceptualize the FFE in product

4

development, although several different conceptualizations of the FFE exist in the literature (e.g. Griffiths-Hemans and Grover, 2006; Khurana and Rosenthal, 1998; Koen et al., 2001).

Several scholars picture the FFE in product development as a sequential process, consisting of several sub-phases with iterations among and within them (e.g. Cooper, 2008; Griffith-Hemans and Grover, 2006; Khurana and Rosenthal, 1998). For example, if an idea is rejected at the first gate, it can be refined and given a second chance. Khurana and Rosenthal (1998) have divided the FFE into three sub-phases: pre-phase zero, phase zero and phase one in their model. In a similar vein, Griffith-Hemans and Grover (2006) divide the FFE into idea creation, idea concretization and idea commitment. Their conceptualization is done on the individual level and highlights informal activities. In the idea creation phase, an employee comes up with a product idea. Facets of the potential new product are specified in the following phase, idea concretization, for the purpose of convincing the stakeholders. Then in the idea commitment phase, resources are committed to the concretized product idea and the organization formally accepts it.

Koen et al. (2001) describe the FFE as a five-part circle, instead of a sequential process. The elements are opportunity identification, opportunity analysis, idea genesis, idea selection, and concept and technology development. Ideas are expected to flow and iterate among the five elements in the model. All these previous conceptualizations are based on assembled products, however, which raises the question whether they are applicable to the FFE of non-assembled products as well.

Compared to the product development literature, few attempts have been made in the process development literature to conceptualize the FFE. In this literature, the FFE is pointed out as one of the most critical stages for achieving successful process development (e.g. Lim et al., 2006; Pisano, 1997). The FFE in process development can be defined as the activities conducted before a new process concept is implemented into the production process (Lager, 2002; Pisano, 1997). It seems logical to assume that the FFE is conceptualized differently, since the objectives and activities differ in product and process development. However, as no conceptualizations of the FFE in process development have been made in the existing literature, the first research question addresses this knowledge gap.

2.2 Key activities in the front-end of the innovation process One of the first activities generally undertaken in the FFE is

preliminary opportunity identification (Khurana and Rosenthal, 1998; Kim and Wilemon, 2002). For example, one individual identifies flaws or gaps in the current state of thinking. The sources of new innovative products or processes can be either inside or outside the boundaries of the firm. Idea generation is not always done explicitly, but when the idea is shared in a social structure (Khurana and Rosenthal, 1998) it becomes explicit. The next step is typically to refine the idea (Griffiths-Hemans and Grover, 2006); the idea is further considered,

5

specified and screened. Screening typically takes place in two dimensions: business analysis and feasibility analysis (Khurana and Rosenthal, 1998). First, an idea is evaluated in terms of its viability as a business proposition. Then, the feasibility of the product idea is determined. Product ideas can be screened and evaluated accordingly, if the firm has the required technological and knowledge resources.

Preliminary market and technology analysis constitute another set of key activities in the FFE (Khurana and Rosenthal, 1998). Information is gathered about the feasibility of the idea in terms of market and technology assessment. Customer needs, target market and definition of market requirements are examples of information collected (Cooper and Kleinschmidt, 1987; Montoya-Weiss and O’Driscoll, 2000). Other information includes identifying the technical solution required to develop the product, as well as an estimation of related costs (Cooper and Kleinschmidt, 1987; Verworn, 2006). Moreover, the proposed product requires evaluation against the existing business plans and products within the firm’s product and portfolio strategy (Khurana and Rosenthal, 1998). The idea should further be evaluated against the competitive situation, with competitors’ prospective and current product offerings, and environmental scanning and analysis should be conducted.

The next step is typically to create a product definition. This includes a product concept, information about target markets, customer needs, product specifications, and product positioning and requirements (Cooper and Kleinschmidt, 1987). The product concept is a representation of the development goals of the project (Montoya-Weiss and O’Driscoll, 2000; Seidel, 2007), and is one of the most essential parts of the product definition. A well-defined product definition, created early on, will facilitate the important go/kill decision that determines whether the product idea will enter formal development or not (Cooper, 2008).

Finally, project priorities among scope (product functionality), scheduling (time), and resources (costs) are required. Previous research has shown that unclear priorities are a major cause of delays (Murphy and Kumar, 1997). Thus, product definitions are often evaluated by a cross-functional executive committee (Khurana and Rosenthal, 1998).

Due to the limited amount of previous research on the FFE in process development, most of the key activities in this section are derived from (assembled) product development. Nevertheless, some of these activities are usually also critical when developing new processes and non-assembled products (e.g. preliminary opportunity identification, idea refinement, and preliminary technology assessment). It seems reasonable to assume that idea generation and refinement are conducted similarly, regardless of whether products or processes are being developed. Preliminary technology assessment is another key activity that is important to consider in the FFE of process development, i.e., when new technology is introduced into the production process.

In the general process development literature, early key activities include: managing and converting raw materials, identifying internal production needs, pre-testing, and transferring development results to full-scale production

6

(Reichstein and Salter, 2006). The importance of FFE activities (e.g. identifying internal production needs and performing pre-testing) has been noted (Lager, 2002; Lim et al., 2006; Pisano, 1997), yet research on the FFE in process development has not been conducted to the best of our knowledge. The fact that product and process development have different objectives suggests that some key activities in the FFE are product-specific whereas others are process-specific. The second research question addresses the differences and similarities in key activities in the FFE of product and process development.

2.3 The relationship between product and process development in process firms As stated in the beginning, there is a strong relationship between

product and process development in process firms. New products often require new or modified processes, whereas new processes sometime allow concepts that were previously impossible to develop (Barnett and Clark, 1996; Lim et al., 2006). Linton and Walsh (2008) show that for non-assembled products, any change of the manufacturing process results in significant changes of the end product. This makes it important to consider both product and process development simultaneously.

In the process industry, product and process development are often required within the same innovation project. Thus, an activity can span both product and process domains at the same time. The purpose of an activity often determines whether it is classified as a product- or a process-development project (Lager, 2002; Lim et al., 2006). For example, product development in the process industry often entails changes in chemistry and adjustments in process parameters, which together result in a modified or new product (Barnett and Clark, 1996). Hence, it is difficult to separate product development from process development since many projects entail a little of both.

However, it is also a challenge to integrate them, as product and process development have different objectives (i.e. efficiency and effectiveness), key activities, and customers (production versus external customers). The third research question explores how the relationship between product and process development is managed in the FFE of the innovation process.

3. Method A case study approach was selected for several reasons. The iterative

and complex nature of the FFE favoured the case study methodology (Chang et al., 2007; Montoya-Weiss and O’Driscoll, 2000; Murphy and Kumar, 1997). Another motive was to ensure methodological fit among the research questions, data collection, data analysis, and the status of the current theory (Edmondson and McManus, 2007; Eisenhardt and Graebner, 2007).

A multiple-case study approach is more likely to yield accurate, generalizable theory than a single-case study (Eisenhardt and Graebner, 2007). The cases in the study were selected according to the literature. The cases were expected to replicate each other and extend the emerging theory of the FFE of

7

8

the innovation process. Each case was treated as an independent analytical unit (Eisenhardt and Grabner, 2007; Yin, 1994).

Four process firms from the mineral and metals industry were selected. Since process development is the primary innovation type in process firms, the process industry was considered an ideal setting for this study (Barnett and Clark, 1996). Although the firms produced different (non-assembled) products, it was assumed that the FFE of the innovation process most likely was managed and organized in a similar way. The four case study firms all had great experience of conducting both product and process development, and good access was given to all firms. See Table 1 for a summary of the research sites.

9

Tab

le 1

. Res

earc

h sit

es

Cas

e st

udy

firm

s A

lpha

B

eta

G

amm

a D

elta

M

ain b

usines

s M

etal

pro

duce

r Su

pplie

r of

met

al p

owde

r

Inte

rnat

iona

l hig

h-te

ch m

iner

als

grou

p G

loba

l pro

duce

r of

sev

eral

di

ffere

nt h

igh-

stre

ngth

ste

el

prod

ucts

M

ain a

pplica

tion

area

s/ p

roduct

s

The

mai

n m

etal

s ar

e co

pper

and

zi

nc, b

ut le

ad, g

old

and

silve

r ar

e ot

her

impo

rtan

t m

etal

s.

A v

arie

ty o

f app

licat

ion

area

s, in

clud

ing

sinte

red

com

pone

nts,

soft

mag

netic

com

posit

es, h

ot p

olym

er

filtr

atio

n, a

nd s

urfa

ce c

oatin

g.

Prod

uces

upg

rade

d ir

on o

re fo

r th

e st

eel i

ndus

try

and

indu

stri

al

min

eral

s pr

oduc

ts fo

r ot

her

indu

stri

es.

App

licat

ion

area

s ar

e pr

imar

ily

brid

ges,

build

ings

, shi

ps, v

ario

us

form

s of

veh

icle

s an

d lif

ting

devi

ces.

Pro

duct

dev

elopm

ent

Doe

s no

t de

velo

p pr

oduc

ts in

a

trad

ition

al w

ay, a

s th

e sm

elte

rs

extr

act

and

prod

uce

pred

eter

min

ed

basic

met

als

out

of o

re. T

he

chal

leng

e is

to p

rodu

ce a

pro

duct

of

con

siste

nt q

ualit

y ev

en t

houg

h th

e qu

ality

of t

he in

put

mat

eria

ls va

ries

.

Tec

hniq

ues

to p

rodu

ce d

iffer

entia

ted

prod

ucts

incl

ude

addi

ng c

oatin

g to

th

e m

etal

pow

der

or m

ixin

g th

e m

etal

pow

der

with

diff

eren

t al

loys

. T

hus,

prod

uct d

evel

opm

ent

typi

cally

in

clud

es s

mal

l-sc

ale

test

s, w

here

the

ai

m is

to

find

a m

etal

pow

der

with

th

e op

timal

pro

pert

ies

for

the

cust

omer

s’ pr

oduc

tion

proc

esse

s.

The

pro

duct

dev

elop

men

t in

clud

es b

oth

impr

ovin

g ex

istin

g pr

oduc

ts a

nd d

evel

opin

g ne

w

prod

ucts

with

diff

eren

t pr

oper

ties.

One

typ

ical

pro

duct

de

velo

pmen

t pro

ject

is t

o ch

ange

th

e pr

oper

ties

of t

he ir

on o

re,

ther

eby

cust

omiz

ing

the

prod

uct.

Prod

uct

deve

lopm

ent

incl

udes

de

velo

pmen

t of

tot

ally

new

pr

oduc

ts a

s w

ell a

s im

prov

emen

ts

of e

xist

ing

prod

ucts

. A la

rge

part

of

all

prod

uct

deve

lopm

ent

is cu

stom

er-d

rive

n.

Pro

cess

dev

elopm

ent

The

focu

s of

pro

cess

dev

elop

men

t is

on c

ontin

uous

impr

ovem

ents

of

the

prod

uctio

n pr

oces

s w

ithin

eac

h pl

ant.

The

pro

cess

dev

elop

men

t is

divi

ded

into

loca

l and

glo

bal d

evel

opm

ent;

incr

emen

tal p

roce

ss d

evel

opm

ent

is co

nduc

ted

loca

lly a

t th

e pr

oduc

tion

plan

ts, a

nd p

roce

ss d

evel

opm

ent

with

hi

gher

deg

ree

of n

ewne

ss (

e.g.

de

cisio

ns o

n re

duci

ng t

he n

umbe

r of

st

eps

in t

he p

rodu

ctio

n pl

an o

r to

in

vest

in n

ew p

roce

ss e

quip

men

t) is

ce

ntra

lized

.

The

pr

oces

s de

velo

pmen

t pr

imar

ily

focu

ses

on

impr

ovin

g th

e ex

istin

g pr

oces

s te

chno

logy

, no

t on

im

plem

entin

g ra

dica

lly

new

pro

cess

tec

hnol

ogy

into

the

pr

oduc

tion

proc

ess.

The

pro

cess

dev

elop

men

t is

mai

nly

cond

ucte

d at

the

bla

st

furn

aces

whe

re t

he s

teel

sla

bs a

re

prod

uced

. The

ste

el s

labs

are

tr

ansp

orte

d to

the

pro

duct

ion

plan

ts t

o be

pro

cess

ed in

the

ro

lling

mill

s. So

me

proc

ess

deve

lopm

ent

is pe

rfor

med

in t

he

prod

uctio

n pl

ant

as w

ell.

Proc

ess

deve

lopm

ent

incl

udes

a r

ange

of

activ

ities

from

con

tinuo

us

impr

ovem

ents

to

inve

stm

ent

in

mor

e m

oder

n pr

oces

s eq

uipm

ent.

A

ppro

xim

ate

num

ber

of

emplo

yees

4 40

0 em

ploy

ees

1

440

empl

oyee

s 3

700

empl

oyee

s 9

000

empl

oyee

s

Num

ber

of

inte

rvie

ws

14 in

terv

iew

s to

tally

, all

focu

sed

on

proc

ess

deve

lopm

ent

17 in

terv

iew

s to

tally

, whe

re 5

focu

sed

on p

roce

ss d

evel

opm

ent

and

12 o

n pr

oduc

t dev

elop

men

t

17 in

terv

iew

s to

tally

, whe

re 8

fo

cuse

d on

pro

cess

dev

elop

men

t an

d 9

on p

rodu

ct d

evel

opm

ent

16 in

terv

iew

s to

tally

, whe

re 5

fo

cuse

d on

pro

cess

dev

elop

men

t an

d 12

on

prod

uct

deve

lopm

ent

3.1 Data Collection Interviews were the primary data collection method, and this is

generally the most efficient approach to collect rich empirical data (Eisenhardt and Graebner, 2007). A total of 64 interviews were conducted. A first round of interviews (with one informant for each case) aimed to achieve an overall understanding of general information on innovative activities and identify suitable informants for the subsequent investigation. In the later interviews, more specific questions were asked to address the research purpose explicitly. Thirty-two interviews were conducted with employees primarily developing new products, while 28 interviews were conducted with employees mainly working with process development (see Table 1 for more information).

The informants were selected from different hierarchical levels to get diverse perspectives of the management of the FFE of the innovation process, and it also mitigated respondent bias (Eisenhardt and Graebner, 2007). Moreover, we selected highly knowledgeable informants who had diverse perspectives concerning the FFE in both product and process development in order to limit bias. A “snowball/chain sampling” approach was used to select the informants, and this approach allowed us to identify the key persons working with product and process development (Miles and Huberman, 1994).

Two semi-structured interview guides were designed. One focused on the FFE in product development while the other concerned the FFE in process development. The interview guides were based on prior research of the FFE, and general literature on product and process development. The interview guides included four sections. The first section provided background information of the informant. The second section covered general questions about the firm’s product and process development practices and organization of the development work. The following section focused explicitly on front-end practices, i.e. key activities, roles, responsibilities, problems and opportunities present in the FFE. The final set of questions included the informants’ view of what factors and activities in the FFE were most critical to successful outcomes. A clear research framework was designed and discussed among the research team prior to the data collection process, which strengthened internal validity (Yin, 1994).

To create overlap between data collection and data analysis, frequent discussions and preliminary analysis of emergent findings were shared among the research team. To further increase reliability (transparency and future replication), a case study protocol was constructed containing case study notes, documents, and the narratives collected during the study, all with the aim of facilitating retrieval for future studies (Yin, 1994). The interviews ranged from 30 minutes to 3.5 hours and were conducted over the course of approximately two years, ranging from August 2008 to October 2009. The interviews were tape-recorded and transcribed. Informal discussions, observations, and group meetings and seminars with R&D managers and engineers were used to complement the interviews. Moreover, secondary data were collected on documented procedures, annual reports and other company information. This permitted empirical triangulation of each firm’s product and process development practices.

10

3.2 Data analysis The analysis of the empirical data started with a separate analysis for

each case of the FFE in product and process development. A total of seven case study histories were created, based on interviews, field notes, observations, informal discussions and secondary data. As Alpha does not develop products in the traditional way, only its process development was analysed. Several shorter telephone interviews were conducted at this stage to retrieve missing information. The main unit of analysis is the FFE stage in product and process development, whereas the relationship between product and process development, at the firm level, constitutes an embedded unit of analysis (Yin, 2003). This affected how the empirical data have been structured in the paper. Since the main unit of analysis is the FFE and not the firm level, the empirical findings are displayed according to the research questions which do not take into account differences on the firm level.

All interviews were transferred into a spreadsheet for further analysis. Relevant questions were conceptually clustered together according to one of three general themes, which facilitated the analysis (Miles and Huberman, 1994). The first theme included questions related to the conceptualization of the FFE, the second theme focused on activities done in the FFE, and the third theme brought together questions related to the relationship between product and process development in the FFE. This was an effort to establish a clear chain of evidence which would allow the readers to see how the research questions matched with key conclusions (Yin, 1994).

The next step was to compare and contrast the findings for product and process development across the cases studied (Yin, 1994) within each theme. Several times, the empirical findings were evaluated against the theoretical framework. This was important, as the empirical data were collected from another context and the theoretical framework was largely based on assembled product development and to some extent on general process development literature. The analyses thus required repeated reading of the interview files, notes, secondary data, and the theoretical framework.

4. Empirical findings and analysis The empirical findings are displayed according to the research

questions. Subsequently, the FFE of the innovation process is conceptualized in the next section. Thereafter front-end key activities are analysed, and in the final section the relationship between product and process development is explored.

4.1 Conceptualizing the FFE of the innovation process in process firms All firms had two different sequential project models—one for the

overall product development process, and the other for the process development process. The FFE was present in both models, although the models were not applied in each and every case. The case study firms organized the early stages

11

slightly differently, depending on whether a new product or process was being developed (see Figures 2 and 3). The figures are based mostly on the informants’ descriptions of what they did in the FFE of product and process development, and the formal project models for the overall processes were used as well.

INFORMALSTART-UP

FORMALIDEA-STUDY

FORMALPRE-DEVELOPMENT

STUDY

Figure 2 The fuzzy front-end in product development

INFORMALSTART-UP

FORMALIDEA-STUDY

FORMALPRE-STUDY

FORMALPRE-PROJECT

Figure 3 The fuzzy front-end in process development

The FFE in non-assembled product development was divided into

three sub-phases: informal start-up, formal idea-study and formal pre-development study. The typical starting point for the FFE in product development was an idea which typically came from the customers, or from internal development personnel. In the last sub-phase a product concept was created and a decision to proceed to formal development or not was made, which marked the end of the FFE. The stage after the FFE was typically formal development, where more tests were done to further develop the product concept and to validate it at larger scale.

The FFE of process development included four sub-phases: informal start-up, formal idea-study, formal pre-study and formal pre-project. Thus, one extra sub-phase constituted the FFE in process development. Similar to the FFE in product development, the starting point was an idea for an improvement in the production process. The ideas typically came from inside the firm and most ideas came from the production department or from other departments nearby, such as the R&D department, product development, or the central process development department. Process development mainly has an internal focus; accordingly the ideas typically involved cost reductions and quality problems with existing products. The end of the FFE was called a pre-project. In this final sub-phase the process concept was specified, and the decision was made on

12

whether the new idea should be implemented or not in the production process. After the FFE in process development, no more tests were done; instead the process concept was implemented.

Several of the informants described the FFE as a trial-and-error process with a number of iterations among the different sub-phases, regardless of whether products or processes were developed. The formal description of the two FFE stages were described as linear processes, but it is important to stress that in some cases, the sub-phases may overlap, run in parallel or be repeated, which is represented by the dotted arrows in both figures. Figures 2 and 3 represent ideal types of the FFE in product and process development. Thus, these figures capture most ideas and events in the FFE in process firms, but not necessarily all of them.

A development engineer at Delta expressed the iterations like this: ‘It is an iterative process; sometimes you must go back and fix a number of parameters, so all phases are connected, and it is not a linear process.’ This quotation is not peculiar to Delta; iterations were common in all cases and they were carried out in all sub-phases, especially when preliminary tests were done.

The FFE differed depending on whether products or processes were developed. For example, the nature of the development project typically influenced how the FFE was organized. Because process development is systemic in nature, it affects other sub-processes throughout the firm, not just the intended change in the production process. This, combined with the large range of process development projects (process development ranges from incremental adjustments in existing operations to creating or buying totally new process equipment), clearly made it difficult to conduct all process development within the formal project model. Hence, it was difficult to have a generic work process for process development. One of the plant managers at Beta described this: ‘One deficiency is that we don’t have a work process when performing process development projects and we move forward too fast. We are in fact at the prototype stage but it is already implemented in the production process.’ Several interviewees from all cases described process development as complex and unsystematic, primarily because of the systemic nature and the diversity in different types of process improvements. When developing new products, the complexity was instead related to technological and market uncertainty.

The degree of newness also affected how the FFE was organized and conducted. For larger improvements in the production process, there was a high degree of formalization in the FFE, due to high capital costs and risks. When conducting large improvements, it was very important to consider every possible aspect in the FFE, so that failures could be avoided early and not when the process change was implemented in the production process. Ideas with a lower degree of newness (incremental process development projects) were accordingly often conducted informally, outside the formal project model. The same pattern was found in the FFE of product development as well.

13

4.2 Front-end key activities in the FFE of the innovation process The front-end key activities were divided into general activities (i.e.

conducted in both of the FFE stages) and specific activities, which were carried out when either products or processes were developed. Table 2 gives an overview of the different key activities in the FFE of the innovation process.

Idea generation was a general key activity. In product development, customers were the most important source for new product ideas, and other sources were the R&D or marketing departments. Ideas for process changes came from a variety of internal sources. Ideas were mostly informally created. Although formal systems existed to capture ideas from employees, they were seldom used.

Once an idea was generated, it was typically explored further by one person or a small team. The idea was first discussed with colleagues, and then senior researchers helped to refine the idea. These activities normally took place outside the formal project model. Informal discussions were important in order to further refine the ideas throughout the whole FFE. As ideas often were fuzzy, unclear and containing multiple loose ends, idea refinement at the beginning of the FFE was an important key activity for developing both new products and processes.

The next general key activity was to conduct a literature review. The aim was to map the existing knowledge of that particular area and to reduce uncertainty. The literature review usually kicked off the next sub-phase, the formal idea-study. The results of the literature review were discussed with colleagues and senior researchers, and the results were evaluated once again before conducting empirical tests. This sub-phase is more theoretical than empirical, as most of the knowledge was acquired by exploring existing theoretical knowledge in the literature.

The literature review was typically followed by more empirically grounded work, such as analysis of input materials and preliminary tests. This was especially critical if new materials were suggested in product development. Raw materials were used in the case study firms, and these often vary in quality and have unique chemical properties which made analysis of input materials essential. At this stage, the product or process engineers typically conducted several different empirical tests (e.g. laboratory tests and tests at bench scale) to get a feel for the problem and the optimal solution. When feasible, preliminary tests at bench scale and laboratory tests were done when developing both products and processes. More fine-graded tests, larger-scale tests such as pilot plant tests, and ultimately full-scale tests were also performed. The next sub-phase generally began once these tests achieved more accuracy.

The final output of the FFE is a product or process concept where several different objectives are summarized. The creation of a product or process concept was a general key activity in the FFE for all case study firms. Creating a project plan and a budget were also important, in which costs, resources and timeframes were estimated. Together these formed the basis for deciding whether the project idea should be developed further or terminated.

14

Table 2. Key activities in the FFE of the innovation process within process firms

General key activities in the FFE of the innovation process

Specific key activities in the FFE of the innovation process

Idea generation Primary product specific key activities

Idea refinement Anticipation of customers’ process requirements Literature review Identification of a process window Analysis of input materials Customer involvement Preliminary test at small scale (e.g. bench scale, laboratory test etc.) Primary process specific key activities

Pilot plan tests Anticipation of end-product changes Full-scale tests Risk analysis Creation of a product/process concept Construction/modification of process equipment Creating a project plan and a budget

Besides the general key activities, a number of specific key activities were also found. Key activities specific for product development in the FFE were anticipation of customers’ process requirements, identification of a process window and customer involvement.

Anticipation of customers’ process requirements was crucial for achieving the best performance, since the performance level of the final product was largely determined by how the customers used it in their production. Identification of a process window was a metaphor used by several informants to describe the opportunities and constraints of the production process when developing new products. This is further elaborated upon in section 4.3 where the relationship between product and process development is described. Finally, customer involvement in the FFE was described as a key activity, as some of the best product concepts often were developed in close collaboration with customers. Two of the three key specific activities focused on achieving satisfied customers, which corresponds to the general objective of product development—to produce products that provide benefits for the firm’s customers. As customers were the most important source of new product ideas, a large part of the product development at the case study firms was customer-driven. The third specific key activity, identifying a process window, shows that the interdependence between product and process development was strong already in the FFE of product development. This will be discussed in 4.3.

In the FFE of process development, specific key activities were anticipation of end-product changes, risk analyses, and construction/modification of process equipment. Changes in the production process were often done to decrease the production cost or to increase quality in the existing products. However, due to the close relationship between product and process development, these changes could occasionally lead to deteriorations in the end-product, which made anticipation of end-product changes a key activity in the FFE. The systemic nature of process development could lead to unexpected changes in other processes close by. Depending on the proposed idea, the process concept sometimes included construction/modification of process equipment. When conducting larger changes in the production process, construction/modification of process equipment was

15

often an essential activity. The final specific activity in the FFE of process development was risk analysis. Since most process changes involve changes in routines and work methods, risk analysis was central in the FFE. A number of different risk analyses were conducted to verify that the intended change had no negative effects, on either the internal or the external environment. For example, work environment and safety issues were typically analysed before deciding whether a process change could be implemented or not. Risk analysis is important when developing new products as well, although it was not a key activity in the FFE of product development. In product development, risk analyses are usually done in the formal development stage after the FFE. Finally, since process development efforts range from minor changes in work methods to significant investments in new process technology, the key activities varied according to the type of project.

To summarize, many of the general key activities presented in Table 1 are also conducted in assembled product development, such as idea generation, refinement, and creation of a product concept (e.g. see Khurana and Rosenthal, 1998; Reid and de Bretani, 2004). Whereas analysis of input materials, tests at smaller scales, and pilot plant test are specific for process firms producing non-assembled products. A number of specific key activities were identified, whether the goal of the innovation project was to produce a new product or to improve the production process. Thus, the FFE of the innovation process involves different activities depending on the type of innovation that is developed.

4.3 The relationship between product and process development in the front end There is generally a high degree of interdependence between product

and process development in process firms. For example, changes in the production process often change the properties of the final product. Therefore, process engineers must anticipate, as early as possible, how the final product is affected when a process development project is planned. From the perspective of the product engineers, the production process holds opportunities and constraints for what products can be developed. An efficient production process can speed up the firm’s ability to develop new products. Constraints need to be considered as well, so that product concepts that are impossible to produce in the existing production process are terminated early on.

A product idea can be either outside or inside the process window. Being inside the process window means that the process technology is known and the product idea is possible to develop without new production equipment. Whereas being outside the process window represents radically new product ideas that require investments in new equipment, and the process technology is unknown. Integration is more important when the product idea is outside the process window. Thus, identifying the process window in the FFE is essential for process firms.

For all case study firms, several product and process engineers and senior researchers confirmed that the interdependence between product and process development was strong. For example, process development projects

16

could sometimes be initiated by the development of a new product. As one of Beta’s interviewees said: ‘I got a feeling that almost all changes in the production process are derived from new products and not from the process itself.’

Many development projects entailed both product and process development, and there were no clear borders between them. The process development manager at Delta stated: ‘It is a problem when you think that you can isolate one part from the other. Unfortunately, it happens all the time. An example is when changes in the production process are made, which they don’t think are significant, but which considerably affect the end product.’

However, integrating product and process development was often difficult. Product and process development were divided into different departments in all case study firms, and they were often located at different places. In the FFE, before any formal project was created, personal relationships were critical in order to manage the strong relationship between product and process development.

5. Discussion The FFE is identified as the weakest stage in the whole innovation

process, but the decisions made in the beginning largely determine the outcomes of the innovation process (Koen et al., 2001). Existing research on the FFE has focused on assembled product development, while research on other innovation types is still limited. Specific characteristics of process firms make it difficult to apply previous knowledge on the FFE. An absence of studies on the FFE in process development further entitles the focus on process firms. Therefore, the aim of this paper was to inform both researchers and practitioners about the FFE of the innovation process by exploring non-assembled product and process development within process firms. Three research questions were created: How is the FFE of the innovation process conceptualized in process firms? What are the differences and similarities in terms of key activities in the FFE of product and process development? And how is the relationship between product and process development managed in the FFE of the innovation process?

The first research question explores how the FFE is conceptualized in non-assembled product development and process development, respectively. Figures 2 and 3 display the FFE in product and process development in process firms. The conceptualizations showed that the FFE starts similarly in both cases. Typically a new idea was generated which originated either outside the firm or within the firm. Perhaps the most central difference was the following stage in the innovation process. When products were developed, the next stage was typically formal development which included more tests and validations of the product concept (Cooper, 2008), while implementation in the production line was the stage after the FFE in process development (e.g. Lager et al., 2010). The process concept was instead implemented in the production process. This explains to some extent why the FFE in process development includes one extra sub-phase, compared to the FFE in product development.

By conceptualizing the FFE in process firms, it can be concluded that the FFE is contingent on whether new products or processes are being

17

developed. Previous research has addressed the deficiencies of early stages in the process development process (Pisano, 1997) as well as the lack of contingency factors when studying the FFE (Bröring et al., 2006; Nobelius and Trygg, 2002). This paper adds new insights into what constitutes the FFE in process development and non-assembled product development.

The second research question compares key activities in the FFE of product and process development. Table 2 shows that some general key activities are conducted in the FFE of both product and process development, despite contradictory objectives. Idea generation, screening, and the creation of a product or process concept (Khuruna and Rosentahl, 1998; Kim and Wilemon, 2002) are examples of fundamental activities in the FFE regardless of innovation type or context. The specific key activities for each innovation type suggest that extant knowledge of the FFE in product development cannot be directly transferred to process development, as they work according to two different logics (product development is mainly effectiveness-driven and process development is mainly efficiency-driven) and with different objectives.

Moreover, it was found that both informal and formal activities were performed in the FFE regardless of the innovation type. Informal and unstructured activities were described as very important in the literature for generating new ideas in the context of discontinuous product development (Reid and de Bretani, 2004). This study shows that this was the case even in more incremental product and process development.

Regarding the final research question, how the relationship between product and process development is managed in the FFE of the innovation process, the relationship was described as strong in all case study firms. In the literature, the relationship has been explored on the firm level and the industry level (Lim et al., 2006; Linton and Walsh, 2008; Utterback and Abernathy, 1975). However, there have been few empirical examples of how the relationship looks and how it can be managed. This paper shows that the relationship needs to be considered as early as in the FFE of the innovation process. This has not been reported previously, either in the FFE literature or in the literature on process-based innovations.

From a practical point of view, these findings can increase the speed and conduct of the FFE. The conceptualizations of the FFE are a first attempt to clarify and structure what is done in the FFE of non-assembled product development and process development, and what needs to be done. The conceptualizations and the identified front-end key activities can be used as checklists, and they are a first step toward making the FFE less fuzzy. This is an initial and important measure for improving the overall innovation process.

The FFE in process development was described as complex and unsystematic in all case study firms, and most of them lacked a formal process development work process. Thus, the FFE of process development would benefit from increased formalization. A formalized work process in the FFE should guide the employees so that they know what is expected of them and the decision-making responsibilities are clear (Khurana and Rosenthal, 1998). However,

18

formalization has a back side as it can cause rigidity, and it is important to find the appropriate level and not impose too much formalization.

To manage the relationship between product and process development, managers are advised to encourage informal activities and strengthen interpersonal relationships between product and process engineers. Then problems and mistakes can be addressed early on in the innovation process.

6. Limitations and suggestions for future research As with any research, there are limitations associated with this study.

The empirical data are collected within one specific industry (i.e. the mineral and metals industry) and therefore care should be taken when generalizing the findings to other contexts. The findings should be considered in their contextual setting, but some implications may be extended to other manufacturing firms. Analysis of the empirical data is more valid for process firms due to the sampling strategy; however, it limits theoretical generalization to other contexts (Eisenhardt and Graebner, 2007). Retrospective sense-making must be considered, as well as keeping in mind that interviews are the primary data source, since informants can forget or misinterpret important facts (Eisenhardt and Graebner, 2007).

The FFE of product development differs in many respects from process development, and therefore only parts of existing research can be used to understand the FFE in process development. This suggests that the FFE in process development warrants its own realm of study. More research is suggested to better understand the specific conditions of organizing and managing the FFE of innovation. The product- and process-specific key activities found in this exploratory study need to be validated further. Finally, to increase the understanding of the FFE, more research is needed on innovations other than assembled product development, as the existing knowledge is still limited.

19

References Alam, I. (2006), “Removing the fuzziness from the fuzzy front-end of service

innovations through customer interactions”, Industrial Marketing Management, Vol. 35 No. 4, pp. 468-480.

Barnett, B.D. and Clark, K.B. (1996), “Technological newness: an empirical study in the process industries”, Journal of Engineering and Technology Management, Vol. 13 No. 3-4, pp. 263-282.

Björk, J. and Magnusson, M. (2009), “Where Do Good Innovation Ideas Come From? Exploring the Influence of Network Connectivity on Innovation Idea Quality”, Journal of Product Innovation Management, Vol. 26 No. 6, pp. 662-670.

Bröring, S., Cloutier, M.L. and Leker, J. (2006), “The front end of innovation in an era of industry convergence: evidence from nutraceuticals and functional foods”, R&D Management, Vol. 36 No. 5, pp.487- 498.

Chang, S-L., Chen, C-Y. and Wey, S-C. (2007), “Conceptualizing, assessing, and managing front-end fuzziness in innovation/NPD projects”. R&D Management, Vol. 37 No. 5, pp. 469-478.

Cooper, R.G. and Kleinschmidt, E.J. (1987), “New products: What separates winners from losers?” Journal of Product Innovation Management, Vol. 4 No. 3, pp. 169-184.

Cooper, R.G. (2008), “Perspective: “The stage-gate idea-to-launch process. Update, what’s new, and Nextgen systems”, Journal of Product Innovation Management, Vol. 25 No. 3, pp. 213-232.

Edmondson, A.C. and McManus, S.E. (2007), “Methodological fit in management research”. Academy of Management Review, Vol. 32 No. 4, pp. 1155-1179.

Eisenhardt, K.M. and Graebner, M.E. (2007), “Theory building from cases: opportunities and challenges”, Academy of Management Journal, Vol. 50 No. 1, pp. 25-35.

Elmquist, M. and Segrestin, B. (2007). “Towards a New Logic for Front End Management: From Drug Discovery to Drug Design in Pharmaceutical R&D”, Creativity and Innovation Management, Vol. 16 No. 2, pp. 106-120.

Gopalakrishnan, S., Damanpour, F. (1997), “A review of innovation research in economics, sociology and technology management”. Omega, Vol. 25 No. 1, pp. 15-28.

Griffith-Hemans, J. and Grover, R. (2006), “Setting the stage for Creative new products: Investigating the idea fruition process”, Journal of the Academy of Marketing Science, Vol. 44 No. 1, pp. 27-39.

Hutcheson, P., Pearson, A.W. and Ball, D.F. (1995), “Innovation in process plant: a case study of ethylene”, Journal of Product Innovation Management, Vol. 12 No. 5, pp. 415-430.

Khurana, A. and Rosenthal, S.R. (1998), “Towards holistic "front ends" in new product development”, Journal of Product Innovation Management, Vol. 15 No. 1, pp. 57-74.

20

Kim, J. and Wilemon, D. (2002), “Focusing the Fuzzy Front-end in new product development”, R&D Management, Vol. 32 No. 4, pp. 269-279.

Koen, P., Ajamian, G., Bukart, R., Clamen, A., Davidson, J., D’Amore, R. Elkins, C., Herald, K., Incorvia, M. Johnson, A., Karol, R., Seibert, R., Slavejkov, A. and Wagner, K. (2001), “Providing clarity and a common language to the “fuzzy front end”. Research Technology Management, Vol. 44 No. 2, pp. 46-55.

Lager, T. (2002), “Product and process development intensity in process industry: A conceptual and empirical analysis of the allocation of company resources for the development of process technology”, International Journal of Innovation Management, Vol. 6 No. 2, pp. 105-130.

Lager, T., Hallberg, D. and Ericsson, P. (2010), “Developing a process innovation work process: The LKAB experience”, International Journal of Innovation Management, Vol. 14 No. 2, pp. 285-306.

Lim, L., Garnsey, E., and Gregory, M. (2006), “Product and process innovation in biopharmaceuticals: A new perspective on development”, R&D Management, Vol. 36 No. 1, pp. 27-36.

Linton, J.D. and Walsh, S.T. (2008), “A theory of innovation for process-based innovations such as nanotechnology”, Technological Forecasting and Social Change, Vol. 75 No. 5, pp. 583-594.

Miles, M. and Huberman, M. (1994), Qualitative data analysis. Thousand Oaks: Sage Publications.

Montoya-Weiss, M. and O'Driscoll, T. (2000), “From experience: Applying performance support technology in the fuzzy front end”, Journal of Product Innovation Management, Vol. 17 No. 2, pp. 143-161.

Murphy, S. and Kumar, V. (1997), “The front end of new product development: A Canadian survey”, R&D Management, Vol. 27 No. 1, pp. 5-15.

Nobelius, D. and Trygg, L. (2002), “Stop chasing the Front End process – management of the early phases in product development projects”, International Journal of Project Management, Vol. 20 No. 5, pp. 331- 340.

Pisano, G. (1997), The development factory: unlocking the potential of process innovation, Boston, HBS Press.

Reichstein, T. and Salter, A. (2006), “Investigation the sources of process innovation among UK manufacturing firms”, Industrial and Corporate Change, Vol. 15 No. 4, pp. 653-682.

Reid, S.E. and de Brentani, U. (2004), “The fuzzy front end of new product development for discontinuous innovations: A theoretical model”, Journal of Product Innovation Management, Vol. 21 No. 3, pp. 170-184.

Seidel, V.P. (2007), “Concept shifting and the radical product development process”, Journal of Product Innovation Management, Vol. 24 No. 6, pp. 522-533.

SGU. (2008), “Investigation on behalf of the government: Mapping Sweden’s ore and mineral resources with the purpose of developing a knowledge base”, Dnr 0-1329/2008.

21

Smith, P.G. and Reinertsen, D.G. (1998), Developing Products in Half the Time – New rules, New York, John Wiley and Sons Inc.

Utterback, J. (1994) Mastering the dynamics of innovation, Harvard Business School Press, Boston, M.A.

Utterback, J. and Arbernathy, W. (1975) “A dynamic model of process and product innovation”, Omega, Vol. 3 No. 6, pp. 639-656.

Verganti, R. (1997) “Leveraging on systemic learning to manage the early phases of product innovation projects”, R&D Management, Vol. 27 No. 4, pp. 377-392.

Verworn, B. (2006), “How German measurement and control firms integrate market and technological knowledge into the front end of new product development”, International Journal of Technology Management, Vol. 34 No. 3-4, pp. 379-389.

Yin, R. (1994), Case study research: Design and methods, Thousand Oaks, Sage Publications.

22

managing the fuzzy front end of product and process

Documents