MASTER THESIS
The Innovative Performance of Firms and its
effects on Partner Selection for Strategic Alliances
An empirical investigation of Dutch ICT companies
ROBERT KOOPS, 1425773 First Supervisor: Dr. G. de Jong Second Assessor: Dr. G. Gemser University of Groningen
Faculty of Management and Organization Cluster Strategy and Innovation
The Innovative Performance of Firms and its effects on
Partner Selection for Strategic Alliances
An empirical investigation of Dutch ICT companies
Robert Koops
ACKNOWLEDGEMENTS
This study has been carried out at the faculty of Management and Organization of the University of Groningen and is the final part of the Master Strategy and Innovation. The research started in April 2005 and finished May 2006. I would like to thank the people that have supported me in writing this thesis and finishing my study.
The faculty of Economics of the Groningen University played a crucial role in process of conducting the study by granting the funds needed to conduct the survey. Without this financial help this study would not have been impossible.
I owe many thanks to my supervisor, Dr. Gjalt de Jong, who‘s theoretical insights and support were very valuable to the research. His feedback proved to be a very helpful contribution for finishing the thesis. He also played a crucial role in raising the funds that were needed to mail the questionnaire.
Thanks to Dr. Gerda Gemser for her input as second assessor. Her feedback on the first version of this thesis proved very valuable and is incorporated in the final version. Minke van der Horst is thanked for improving my English.
I also would like to thank my sister Thera, who has been a big support especially during the last years of my study. Finally I would like to express a special word of gratitude to my parents. They have been very supporting during the whole of my study. Groningen, May 2006
SUMMARY
This study attempts to link the innovative nature of firms with the characteristics that are used in the selection process of alliance partners. The main aim of this study is to learn whether specific selection characteristics become more (or less) important when firms are more innovative. The study aims at analyzing how the innovative nature of firms influences the selection process of partners for strategic alliances. Five selection characteristics play a central role in the study. These are successively ‘special skills that you can learn from your partner’, ‘willingness to share expertise’, ‘complementarity of capabilities’, ‘technical capabilities’ and ‘costs of alternatives’.
A survey is deployed in order to answer the research questions. The questionnaire is mainly constructed around fourteen selection characteristics that have been used in prior research and several innovation parameters. Furthermore, control variables are added, for example size of the focal firm and alliance experience.
The questionnaire is sent to 334 strategic managers active in the Dutch ICT sector. All strategic managers that are contacted are employed at companies that are known to have at least one strategic alliance. The achieved response is 51 usable surveys (response rate 15.2%).
The results of the analysis suggest that the innovative nature of firms hardly influences the importance that selecting managers place on partner selection variables. Solely the cost of alternatives is (negatively) correlated with the innovative nature of the selecting firm. Firms that are more innovative tend to place less importance on the costs of the alliance they establish.
Size of firms and alliance experience seem to play a more significant role in selection processes. The larger the selecting firm, the more importance is placed on the financial situation of the partner. Furthermore, the larger the selecting firm, the less important the technical capabilities of partner firms become. Firms that are experienced in allying place significant more importance on financial assets and intangible assets of their partners, while less experienced firms are more interested in the industry in which the partner is active.
CONTENTS
1 Introduction 1
1.1 Research Topic 1
1.2 Research Questions 1
1.3 Research Design 3
1.4 The Dutch ICT Sector 4
1.5 Contents of the thesis 5
2 Theoretical Background 6
2.1 Strategic Alliances 6
2.1.1 A multi-disciplinary approach to partner selection 7
2.1.2 Empirical research on partner selection 10
2.2 Innovation 11
2.2.1 Firm Innovativeness Measures 12
2.2.2 Firm Innovativeness and Alliances 13
2.2.3 Final Overview 14
2.3 Hypotheses 14
2.4 Control Variables 16
2.4.1 Firm Specific Variables 17
2.4.2 Alliance Specific Variables 17
2.5 Conceptual Model 17 3 Methodology 19 3.1 Questionnaire 19 3.2 Sample 19 3.3 Data Collection 20 3.4 Non-Response Analysis 21 3.5 Data Analysis 21 4 Results 23
4.1 General sample characteristics 23
4.2 Innovativeness and Partner Selection 25
4.2.1 Partner selection in the Dutch ICT sector 25
4.2.2 Innovative Performance 26
4.3 Correlation Coefficients 27
4.4 Intergroup Differences 29
4.5 Regression Analysis 30
5 Conclusion and Discussion 33
5.1 Conclusion 33
5.2 Limitations 35
5.3 Future Research 35
References 38
Appendices 41
Appendix I The 14 Selection Characteristics 41
Appendix II Dutch Questionnaire 42
Appendix III Correlation Matrix 46
Appendix IV Tests of Difference 47
1 INTRODUCTION1
Alliances are an important subject within the field of strategic management. An increasing number of firms are teaming up for a variety of reasons (De Man & Duysters, 2003). Some of these alliances prove successful, but a great number of them fail in the end. This study investigates the relation between innovative nature of the firm and the selection of partners for alliances.
1.1 Research Topic
Alliances are subject to a broad range of studies2. Therefore it may seem rather challenging to come up with a study that offers new features. This study attempts to link the innovative nature of firms with the characteristics that are used in the selection process of alliance partners. Until now, studies have investigated the effects of alliances on the innovative performance (cf. De Man & Duysters, 2003). This study is unique because it is one of the first that investigates this causal relation in the opposite way. The main aim of this study is to learn whether selection characteristics of partners become more (or less) important when firms are more innovative. For firms that prefer to ally with innovative firms the results indicate on which characteristics they are selected, which makes it possible to influence the selection process in a positive manner. Furthermore, this study investigates other firm specific variables (such as size of the focal firm and alliance experience), and in addition alliance specific variables (such as the scope of the alliance) in relation with the used selection characteristics.
1.2 Research Questions
The focal point of this study is the selection process of partners for strategic alliances. Each time firms decide to cooperate, managers within these firms have to select the best partnering firm. In these decision making processes various considerations may play a role. For example, are managers who choose partner firms more interested in the possibilities of learning from their future partner, or is the quality of the products that these potential partners provide leading in decision making processes? An important element within these decisions might be the innovative performance of firms. It seems likely that managers of innovative firms choose partner firms with different motives than their counterparts within less innovative firms. The main research question analyzed in this thesis is:
RESEARCH QUESTION: How does the innovative nature of firms influences the selection process of partners for strategic alliances?
The aim of the research is to study whether innovative firms select their partners for alliances in a different way than less innovative firms. The study is based on three schools of thought which are the Organizational Learning (OL), the Research Based View (RBV) and Transaction Costs Economics (TCE). Most studies in the field of Management and Organization apply a similar kind of multidisciplinary approach, but they do this rather implicit. In this thesis the underlying paradigms are explicitly pointed out, for this provides a better understanding of the range and the limitations of this study.
The three schools of thought correspond to the three sub-questions which are presented below. In the theoretical section the differences and similarities of these schools are discussed in-depth. This section, however, provides a first introduction of the schools of thought in relation to alliances and innovation.
SUB-QUESTION 1: Do innovative firms select primarily on the base of learning objectives?
The first sub-question aims at finding learning motives in partner selection for strategic alliances. This aspect of partner selection is derived from the OL school. Several scholars have studied learning in alliances. Inkpen (1998) for example, observes a tight relation between alliances and learning by assuming that the formation of an alliance is an acknowledgement that the alliance partner has useful information. Dacin, Hitt, and Levitas (1997) and Hitt, Levitas, Arregle, and Borza (2000) give Organisational Learning objectives an important place within their fourteen selection characteristics of potential partners3. Three of these selection characteristics are directly related to the Organisation Learning School. These characteristics are ‘willingness to share expertise’, ‘the partner's ability to acquire your firms special skills’ and ‘special skills that you can learn from your partner’.
The aim of sub-question 1 is finding whether innovative firms are more interested in learning from their partners than less innovative firms. It is likely that firms that try to be innovative are more interested in learning from their partners than less innovative firms, since learning from alliance partners is a fairly easy method of increasing the innovativeness of firms (cf. van Gils and Zwart, 2004).
3 The fourteen selection characteristics which are developed in Dacin et al. (1997) and Hitt et al. (2000),
SUB-QUESTION 2: What resource characteristics of the potential partners are most important for innovative firms?
The second sub-question is related to the RBV School. The central paradigm of the RBV is that firms are able to generate a sustainable competitive advantage by procuring or creating resources that are valuable, scarce, inimitable and non-substitutable (Barney, 2001). Partnering can be a method of realizing competitive advantage (e.g. Norman, 2004; Eisenhardt & Schoonhoven, 1996). The selection characteristics mentioned in Dacin et. al. (1997) which are regarded as being related to the RBV school are: ‘financial assets’, ‘complementarity of capabilities’, ‘unique competencies’, ‘market knowledge/access’, ‘intangible assets’, ‘managerial capabilities’, ‘capabilities to provide quality product/service’ and ‘technical capabilities’. The main aim is to find whether the used selection characteristics differ among innovative and less innovative firms.
SUB-QUESTION 3: Do less innovative firms select on the basis of cost alternatives?
The last sub-question aims at finding the significance of constructs related to TCE. TCE is based on the paradigm that transactions are organized in the most cost efficient way. Within TCE three ways of organizing transactions are possible. These are successively market, hierarchy and hybrid transactions (Powell 1990). Alliances can be regarded as a type of network transaction (e.g. Osbon & Hagedoorn, 1997). TCE suggests that before starting an alliance, a firm normally concludes that it is more cost efficient to collaborate than to do it by itself or buy specific products/services. The potential partner has to provide in the needs of the firm in such a way that it is economically viable to cooperate. One of the fourteen characteristics of Hitt et al. (2000) relates to TCE, this characteristic is ‘cost of alternatives’. TCE in itself does not directly say a lot about innovativeness of firms. Nevertheless, the reason to use the school of thought is because it provides a lot of insights in the formation processes of alliances. Sub-question 3 is proposed using the assumption that whenever cost alternatives play a too important role, innovation efforts of firms might suffer.
The theoretical section presents five hypotheses4. All of them originate from the main research question, the three sub questions, and relevant theories. The first two (1a and 1b) are founded in the OL. The next two hypotheses (2a and 2b) are linked to the RBV, and the last question is linked to TCE (3).
1.3 Research Design
As mentioned above, a survey is deployed in order to answer the research questions. For the most part, the questionnaire is constructed around the fourteen selection characteristics of Hitt et al. (2000) and Dacin et al. (1997), and several innovation parameters. The questionnaire was sent in September 2005 to 334 respondents who are active in the Dutch ICT sector. All respondents are strategic managers in organizations that have at least one strategic alliance. The respondents are selected from the website www.persberichten.com. This website provides press reports from companies that operate in the ICT sector. Firms that contributed a press report which mentions partnering and/or alliance activities are selected and added to the database. All respondents are addressed with a Dutch survey which is added to this thesis in Appendix II. The data that is collected is analyzed by using the statistical package SPSS 14.0 for Windows. The methodology section discusses the research design more thoroughly.
1.4 The Dutch ICT Sector
All firms in the survey are active in the Dutch ICT sector. Because of the fact that situational conditions can play a role in this study (cf. Hitt et al., 2000), this section discusses the main issues within this specific sector.
The Dutch ICT Sector is usually segmented into two sub-sectors; ICT Industry and ICT Services (cf. CBS, 2005; CPB, 2003). The ICT Industry produces IT related commodities, for example chips, semi-conductors and mobile-phones. The ICT Service sector provides two types of services. The first type is telecommunication services, like services of mobile telephone providers. The second type is other computer services, like building websites and development of new software.
The ICT sector is of significance in the Netherlands. One out of every 30 Dutch companies is active in this sector. The sector has witnessed a period of significant growth in de last decade: the number of ICT companies doubled. The main growth took place in the period between 1996 and 2000. This period of growth stagnated after the so called ‘ICT bubble’ evaporated just after 2000. Nevertheless, the total number of companies still increases.
The contribution of the sector to R&D in the Netherlands is disproportionate large; it encapsulates about 20 percent of the total national R&D expenses. Especially the ICT Industry, which realizes 86 percent of the total R&D expenses in the ICT sector, is highly involved in research and development. In this sector a geographical separation between R&D and production activities is observed. Often companies seem to conduct their R&D activities in the Netherlands whereas the actual production takes place elsewhere.
1.5 Contents of the thesis
2 THEORETICAL BACKGROUND
This section presents an overview of existing theories and literature on (the formation of) strategic alliances. Furthermore, it discusses relevant empirical research that is already conducted in the field of alliance formation. Apart from strategic alliances, the section also discusses different theories on innovation. The emphasis on innovation is on performance measures because of its relevance for the study. The reader should be aware that empirical research and existing literature are intertwined to present a readable overview instead of a plain list of relevant papers.
2.1 Strategic Alliances
Strategic alliances are defined by Das & Teng (2003) as interfirm cooperative arrangements aimed at pursuing mutual strategic objectives. Inkpen (1998) defines strategic alliances as “relatively enduring interfirm cooperative arrangements that utilizes
resources and/or governance structures from autonomous organizations” (p. 279). De
Man & Duysters (2003) classify strategic alliances as “cooperative agreements in which
two or more separate organizations team up in order to share reciprocal inputs maintaining their own corporate identities”(p. 3). What is unambiguous from these
definitions is the notion that strategic alliances are arrangements which last over a longer period of time between two organizations that remain autonomous. The general intent of strategic alliance partners is to establish and maintain a long-term cooperative relationship in order to compete more effectively with firms outside the relationship (Hitt
et al. 2000). Firms collaborate for a variety of reasons (Tidd, Besant & Pavitt, 2001).
Some of the multiple motives for starting a strategic alliance are for example reducing the cost of development or market entry, reducing the risk of development of market entry, achieving economies of scale, and reducing the time of developing new products
information which is shared is not critical to the competitive position of the firm. The difference between these two types of alliances is too ambiguous and too vague to use in empirical study, so this difference is not incorporated in this study.
As mentioned above there are different types of alliances. Pekar & Allio (1997) sum up the following list of possible forms of alliances: Collaborative advertising, R&D partnerships, Lease service agreements, Shared distribution, Technology transfer, Co-operative bidding, Cross-manufacturing, Resource venturing, Government and Industry partnering, Internal spin-offs, Cross-licensing. Das & Teng (1998) add to this list Joint Ventures, Equity Investment, Licensing, Buyer-Supplier Relationships and Joint Marketing. The combination of these two lists seems to provide a complete list of strategic alliances which fit in the broad view of alliances. All forms of cooperation mentioned above qualify for strategic alliance, in the definition which is used in this study. Appendix V contains the definitions of the mentioned forms of alliances.
Many strategic alliances fail in the end. The average success rate found by several scholars is 40-50 percent (e.g. Duysters, Kok, & Vaanderdrager, 1999). Scholars have found several reasons for this relatively high percentage of failure. There can be a lack of trust between the partners which makes collaborating difficult if not impossible (Barney & Hansen, 1994). Another possible reason is that one of the partners may enter the alliance with a hidden agenda, which results in actions that do not benefit both partners (Duysters et al., 1999). Sometimes the partners have different expectations with regards to the value that will be captured by the alliance, which can result in a conflict (Inkpen & Li, 1999). This is just a small selection of reasons why alliances fail. Nonetheless it can prove useful for many firms to recognize the high failure rate and be more restrictive when they select potential partners (Overby, 2005).
2.1.1 A multi-disciplinary approach to partner selection
approaches to provide a multidimensional view on the subject. This multidimensional approach aims at incorporating all relevant characteristics of potential partners, which are used in partner assessment processes.
This section presents the core views of three important schools and their contribution to the study of alliance formation. The first school is Transaction Cost Economics (henceforth TCE). This school focuses on firms and their decision between markets and hierarchies in organizing their economic activities (Williamson, 1975). Secondly the impact of the Organizational Learning (henceforth OL) School will be discussed. This approach aims at analyzing how individuals, groups and organizations perceive and interpret information and how this is used to create a better fit with their environment (Aldrich, 1999). Thirdly the Resource Based View (henceforth RBV) will be subject of analysis. The prime objective of the RBV is to study the relationship between assets which firms dispose and the resulting capability to create and maintain a competitive advantage (Mol & Wijnberg, 2004).
Transaction Cost Economics
One of the most applied schools of thought when theorizing on partnering and alliances is TCE (Eisenhardt & Schoonhoven, 1996). This approach focuses on the economic reasons for allying. The TCE is based on the paradigm that transactions are organized in the most cost efficient way. In the classical approach two types of transactions governance were identified: market transactions and hierarchical transactions (Williamson, 1975). Alliances can be seen as a hybrid government form between hierarchy and market transactions (Williamson, 1985).
The transaction cost perspective argues that the transaction costs of exchange are the most important determinant’s of the governance form. Before entering into an alliance a firm normally would conclude that it would be more cost efficient to collaborate than to do it by itself or buy specific products/services. The potential partner has to provide in the needs of the firm in such a way that it is economically viable to cooperate.
The TCE provides several partner characteristics that can be used when assessing potential partners. An important selection characteristic is the assessment of the partner in regard to cost alternatives. Potential partners can be compared with each other by using the cost alternatives. The TCE paradigm states that the firms are likely to choice the partner who contributes most and costs less.
Organizational Learning
can help to achieve this advantage by sharing knowledge. Inkpen (1998) notices a tight relation between alliances and learning by assuming that the formation of an alliance is an acknowledgment that the alliance partner has useful information. Sometimes this information can be sought after and acquired in a very conscious fashion. Other firms seek this knowledge less conscious but still they are learning through interaction with their alliance partners.
Alliance partners provide learning possibilities in several areas. Some firms are interested to learn new managerial skills and capabilities from their partners. Other firms are more interested in market knowledge and technical knowledge. The specific learning interest depends on several situational conditions like resource position of firms and the market situation (Hitt et al., 2000).
The OL perspective provides some characteristics by which potential partners are selected (Hitt et al., 2000). Firms can consider whether they can learn special skills from their potential partners. When comparing partners that have such skills and the ones that do not, it is likely that firms will choose a firm that can provide new skills. Firms may also consider the managerial capabilities of the potential partners to analyze which firm provides most new insights in managing skills. Another aspect is to analyze the potential willingness to share expertise by their potential partners. Potential partners that appear to be willing to share capabilities are likely to be favored more than the firms that do not show this willingness. On the other hand, firms can also fear appropriation of knowledge by their partners (Norman, 2002). The ability to acquire the firm’s special skills by the potential partners can then be taken in consideration when selecting.
The learning efforts of alliance partners are likely to have their impact on the innovative performance of the firms. Firms that learn from their partners are likely to be more innovative than the ones that do not.
Resource Based View
Next to the OL and TCE approach is the RBV, one of the most frequently used schools when it comes to analyzing alliances. The central paradigm of the RBV is that firms are able to generate a sustainable competitive advantage by procuring or creating resources that are valuable, scarce, inimitable, and non-substitutable5 (Barney, 2001). One of the methods of realizing this competitive advantage is by cooperating with partners (e.g. Norman, 2004; Eisenhardt & Schoonhoven 1996). Das & Teng (2003) present two RBV related items that can be used in partner selection processes, namely resource
5 Between the OL and RBV exists overlap. However OL solely aims at knowledge as source of competitive
characteristics and resource alignment. The resource characteristics are the aggregated resources that will be available through the alliance. Resource alignment is the type of fit between the partners’ resources that will be contributed to the alliance. Thus firms will not only analyze which resources will be contributed by their potential partners, they will also analyze the interfirm compatibility of these resources. Hitt et al. (2000) provides six RBV related characteristics that firms use when assessing potential partners. These characteristics are successively: financial assets, intangible assets, unique competences, technical capabilities, complementary of capabilities and the capability to provide quality products/services. Within these groups there are resource characteristics as well as resource alignment.
2.1.2 Empirical research on partner selection
Despite the fact that much research is conducted in the field of strategic alliances, the aspect of selecting partners needs more research (cf. Hitt et al., 2000; Overby 2005). Especially the basis on which partners are selected needs more thorough investigation. This study focuses on partner analysis, which is defined by Das & Teng (2003) as “the
examination of overall match between the partner firms in an alliance in terms of their inter partner market commonality, resource characteristics and resource alignments” (p.
280). In this context, resource alignment is the pattern that integrates the resources of the partner firms. The interpartner market commonality is the degree of presence that partner firms manifest in the market targeted by the alliance. This definition clearly indicates the fact that Das & Teng (2003) have formulated their definition with two schools of thought strongly in mind, namely the Resource Based View and Industrial Organization.
The choice of the right partner is regarded to be a major requirement for success of strategic alliances. Considering the high failure rates among strategic alliances, selecting the right partner is a critical task in the process of partnering (Overby, 2005). Varis & Conn (2002) provide an overview of the most important studies which are conducted in this specific field of research. They argue, similar to Das & Teng (2003), in favor of a multi-disciplinary approach towards studying partner selection, like is deployed in this thesis.
The most groundbreaking study in this field of research is conducted by Geringer (1991). This study was the first to provide sharp defined partner selection criteria that can be used in empirical research (Varis & Conn, 2002). Geringer (1991) distinguishes two types of criteria, namely task-related and partner-related criteria. Task-related criteria are “associated with operational skills and resources which a venture requires for its
competitive success” (p. 45). Examples of these criteria are patents, management
efficiency and effectiveness of partners’ cooperation” (p. 46). Examples are trust in the
top management teams of the partner, size of the partner, and the partners’ corporate culture. Furthermore, this survey study includes fifteen categories of selection chareteristics. These categories and criteria still feature most studies of partner selection criteria, for they prove to be a set of context-irrelevant partner selection criteria (Varis & Conn, 2002). This thesis focuses on individual selection characteristics. The theoretical division between partner and task related criteria is not important to this study.
Studies following Geringer (1991) investigate the importance of partner selecting criteria in various situations. Some examples - most of them mentioned by Varis & Conn (2002) - are shortly provided in this section6. Glaister (1996) for example, concluded that knowledge of local markets, distribution channels and knowledge of local culture are the most important task-related criteria for UK firms that have alliances with Western European organizations. Hoffman & Schlosser (2001) suggest that small and medium sized company managers base their partner selection respectively on four factors: ‘The presence of trust’; ‘complementary contribution to resources and strategy’; ‘partner excellence in cooperation’ and ‘agreement of fundamental values and cultures’. Other studies examined the differences in the applied partner selection criteria in different countries and cultures (e.g. Dacin et al., 1997). Nielsen (2002) provides some evidence that the importance of various selection characteristics vary when dealing with partners from different geographical/cultural regions. Despite this growing research base there is still a need for further research in this particular subject (Hitt et al., 2002).
2.2 Innovation
Innovation is seen as essential to create competitive advantage (e.g. Dress, Lumpkin & Taylor, 2005). Changes in customers needs, the discovery of new technologies and changes in the competitive landscape are some of the (many) factors that force firms to continually innovate. Especially in the light of ever faster changing business environment it seems essential for organizations to continuously change themselves, their products, and services. To survive it is essential to not only focus on exploitation of old certainties, but also to explore new possibilities (March, 1991). Innovation can be the result of this exploration of new possibilities.
An innovation is defined by Wijnberg (2004) as “something new which is presented
in such a way that the value will be determined by the relevant selectors” (p. 1472). This
implies that firms that are innovative not only have to produce a new product (or service), but they also have to market the product in such a way that the firm appropriates (part) of
6 Glaister (1996), Hoffman & Schlosser (2001) and Dacin et al. (1997) are reviewed in Varis & Conn (2002).
the value from the product (Teece, 1987). This notion clarifies the difference between an invention and an innovation. An invention is something that becomes an innovation at the moment it is put on the market (or to the relevant selectors).
Scholars have developed various classifications within the concept innovation. An often used distinction is between product and process innovation (e.g. Dress et al., 2005). Sometimes the category organizational innovation is added to this classification method (Wijnberg, 2004). Another often used classification method is the distinction between
radical innovations and incremental innovation (e.g. Dress et al., 2005). These two
variables are regarded the two extremes of a continuum. Radical innovations entail fundamental changes to existing practices. They usually change the way the innovating company does business dramatically. It may even occur that the whole industry fundamentally changes. Incremental innovation on the other hand involves small changes in existing products and/or services.
2.2.1 Firm Innovativeness Measures
marketing press releases of firms, which can be regarded not a real objective measuring method. Nevertheless, the study of Hagedoorn & Cloodt (2002) shows relatively high correlation between the different indicators. Thus using a single method of measuring innovative performance does not necessarily imply biased outcomes. Nonetheless it seems rather useful to incorporate two ore more indicators when measuring the innovative performance of firms
Apart from the four indicators already mentioned, scholars apply several other measuring methods. De Man & Duysters (2003) add financial indicators, stock market reactions to technology alliances, and the number of researchers assigned to an alliance. These measurement methods appear to be less direct but are employed in several empirical studies into the subject of firm innovative performance.
2.2.2 Firm Innovativeness and Alliances
De Man and Duysters (2003) provide an overview of relevant papers on the subject of firm innovative performance and alliances. This paper, that analyzes 30 survey studies, provides a relatively up-to-date representation of current research in this specific field of study. De Man and Duysters (2003) argue that alliances may have a positive as well as a negative impact on innovative performance. Nevertheless, in only four out of thirty papers that are analyzed in the paper, a negative correlation between alliances and innovative performance of firms is found. The negative impact only occurs in a-typical situations or can be related to ambiguous interpretation of derived data. An example of this ambiguous interpretation is already mentioned above and relates to measuring innovative performance solely through R&D expenses. Whenever R&D expenses decrease after starting an alliance, one can interpret this in two distinctly different ways. The first is that firms are getting less innovative through alliances. The other is the possibility that innovation can be organized more efficient through cooperation, which results in reduced R&D expenses. This example clearly shows the already mentioned usefulness of applying multiple methods (Hagedoorn & Cloodt, 2002) when measuring innovativeness.
De Man and Duysters (2003) conclude that firm innovative performance in general is positively influenced by alliances. There are specific conditions which have an effect on the relative influence of alliances on innovativeness like different types of alliances, sector differences, and differences between countries. In this section some examples of papers that are somewhat more specific are presented. Stuart (2000)7 demonstrated that technology alliances with large and innovative partners improve innovation and growth
7 Stuart (2000) is reviewed in De Man & Duysters (2003). Chang (2003) is not reviewed in De Man &
rates. On the other hand, collaborations with small and technologically unsophisticated partners have an immaterial effect on performance. Chang (2003) also examined the relationship between inter-organizational cooperation and innovative performance at the firm level. That study consisted of a postal questionnaire that was send to 400 firms in the Biotechnology sector in the UK and Taiwan. The results of the study show that types of inter-organizational cooperation enhancing a firm’s innovative performance vary across sectors and countries.
Most studies that have examined the relationship between alliances and innovativeness indicate that, at least to some degree, this relation exists. This study will take research a step further by focusing on the selection characteristics that firms apply and the innovative performance of the firm. Furthermore, the relation of the two variables, alliances and innovation, is researched in an opposite manner. The effect of alliances on the innovative performance of firms is not the subject. Instead the question is whether innovativity of firms is reflected in the choices they make in the alliance partnering process.
2.2.3 Final overview
The relation between innovative performance and strategic alliances is confirmed in various studies (cf. De Man & Duysters, 2003). Nevertheless the innovative nature of firms in relation with the preferred selection characteristics of partners for strategic alliances remains hardly investigated. It seems likely that firms that are innovative have other preferences towards their partners than less innovative firms. This study attempts to provide evidence for such a relation.
Various schools of thought provide insight in the formation process for strategic alliances. This study uses three schools of thought, because a multidimensional approach provides a more comprehensive insight in subject (Osborn & Hagedoorn, 1997). The selection of the partner, that provides an optimal match, is an important issue within the formation process of alliances (Overby, 2005). This study attempts to make a contribution to the knowledge base on the selection process of strategic alliances.
2.3 Hypotheses
HYPOTHESIS 1a: The more the innovative performance of the focal firm, the more the
focus on their partners knowledge resources will be
Innovative firms are likely to be more focused on learning from their potential partners than firms with less innovative performance. Firms’ innovative performance can be increased by co-operating with partners (De Man & Duysters, 2003). Chang (2003) finds that research co-operations often increase the innovative performance of the partners involved. In these kinds of co-operational arrangements the exchange of knowledge resources is one of the key components. Especially larger firms are interesting partners, for they often posses leading edge technological resources (Stuart, 2000). One of the methods of acquiring these resources is selecting partners which have potentially interesting knowledge resources. It is likely that innovative firms are more explicitly focused on their partners’ knowledge resource than less innovative firms.
HYPOTHESIS 1b: The more the innovative performance of the focal firm, the more the
focus on their partners willingness to share information
Innovative firms are to be expected not only to consider the question whether their potential partners actually possess knowledge resources, they are also likely to select firms that are willing to share their knowledge. More innovative firms are likely to select partners that are willing to share their knowledge. Significant examples of this behavior are the early alliances between Japanese firms with Western partners. In those alliances the Japanese firms selected their Western counterparts by evaluating their knowledge and the learning ability that could be provided (Varis & Conn, 2002). Nevertheless, by sharing knowledge firms critical skills and knowledge are exposed to their partners which can lead to unwanted appropriation or imitation (Norman, 2002). In particular competitor firms have to be careful in sharing information because sensitive information must not be passed on (Inkpen & Li, 1999).
HYPOTHESIS 2a: The more the innovative performance of the focal firm, the more the
focus on complementarity of the resources of their partner will be
co-operation with government institutions, which are completely different than themselves. In addition it is being theorized that by bringing together different firms with unique skills and capabilities, alliances can create powerful learning opportunities (Inkpen, 1998). The complementarity of resources is important to innovation efforts of firms. In case the partners have similar resources, innovation is not likely to be stimulated. In situations where resources differ too much, both partners will not be able to make a ‘fit’ (Das & Teng, 2002). Thus complementarity of resources is likely to be more important for firms that are innovative than for firms that are not.
HYPOTHESIS 2b: The more the innovative performance of the focal firm, the more the
focus on technical competences of their partner will be
Technical competences of firms are often crucial for innovation as it is a crucial productive resource. Das & Teng (1998) recognize the fact that technology is not easily transferred between firms. Patent systems, for example, prevent the use of technology owned by other firms. Strategic alliances can prove very useful to access technologies of other firms. Of course not all strategic alliances partners will fully share all their technical knowledge and resources. There will be issues like trust and unwanted appropriation. Nonetheless, it seems likely that firms with obvious innovative motives are eager to cooperate with firms that have specific technical competences.
HYPOTHESIS 3: The more the innovative performance of the focal firm, the less the
focus on cost alternatives will be
Despite TCE does not explicitly incorporate innovation issues, a hypothesis with a direct link to it is taken in this thesis. The main reason is the importance of the TCE within this field of study. The central paradigm within the TCE is that firms will choose the alternatives that are most cost efficient (Williamson, 1975). Whenever firms use this principle in process of partnering it is likely that innovation issues are less important. Firms that mainly focus on cost alternatives when selecting their strategic alliance partners are therefore expected to be less innovative than firms that do not.
2.4 Control Variables
selection variables are more important than for firms in emerging countries. This section discusses the variables that might play such a confounding role.
In this thesis a distinction is made between firm specific variables and alliance specific variables. Firm specific variables are factors within the selecting firm that are likely to influence the selection process. Next to the innovative nature of the firm, the variables in this study are size of the firm and alliance experience of the firm. Alliance
specific variables are variables that are unique for the alliance. Variables that are
regarded specific for the alliance are for example the nationality of the partner and the scope of the alliance.
2.4.1 Firm Specific Variables
Next to the main variable in this study, the innovative nature of the firm, other firm specific variables might play a role in the selection process of strategic alliances. For instance, the size of the focal firm might be of significance. The size of the firms in this study is measured by the number of employees and the annual turnover in 2004.
Alliance experience is accumulated through establishment of prior strategic alliances by the focal firm. It is likely that through gaining experience the relative importance of partner selection criteria changes. For example, it is likely that firms less experienced in partnering will choose a partner that already has some experience. The importance of the partners alliance experience is therefore likely to be less for more experienced firms. – The alliance experience is measured through the number of alliances established by the focal firm.
2.4.2 Alliance Specific Variables
Alliance specific variables are variables that are specific for the alliance situation. The ultimate goal of the alliance might be of importance when selecting partners for strategic alliances. For firms that select a partner for an alliance aimed at innovation, for example a R&D partnership, it is likely that technical capabilities are more important than alliances that have a commercial objective. Strategic alliances might have multiple goals and therefore the nature of it can be diverse. Next to the nature of the alliance the scope of the alliance (cf. Norman, 2004) might also play a role in the selection process. The scope of the alliance is measured by cumulating the nature of the alliances8; the more goals the broader the scope of the alliance.
2.5 Conceptual Model
In the previous sections various variables have been discussed that might influence the importance of selection characteristics. This section provides a conceptual model that arises from the theoretical discussion above. This model is the basis of the research conducted in this thesis. Figure 2.1 shows the conceptual model that is constructed.
The main elements of the model are the centrally placed items. The key relation in the model is between the innovative nature and the importance of partner characteristics. This thesis investigates the assumption that the innovative nature of the firm is reflected in the choices that are made when choosing a partner. Other variables that might influence the weight of the characteristic are the size of the focal firm, alliance experience of the focal firm, nationality of the partner, and scope of the alliance. These are the control variables discussed above. Finally the model shows that the importance of specific partner selection characteristics will affect the choice of a partner for an alliance. For example, it is likely that firms which consider it important to learn from their alliance partners, subsequently select partners that provide opportunities to learn.
Figure 2.1 Conceptual Model
Firm Specific
Variables Alliance Specific Variable
Scope of the alliance Innovative Nature of the focal firm Size of the
focal firm Importance of
3 METHODOLOGY
Because most of the data required to answer the research question is not available from archival sources, the research is conducted by deploying a survey. In this section the methodology is discussed in-depth to provide insight in the process of collecting and analyzing the data.
3.1 Questionnaire
The survey is designed by using constructs which are already used in other research. For the selection characteristics constructs are used that also have been used in (international) studies (Dacin et al., 1997; Hitt et al., 2000). All items related to the importance of selection criteria are tested using a 5-point Likert scale. For measuring innovativeness, three variables are deployed: the number of new products in 2004, the number of new patents in 2004, and the percentage of R&D budget in 2004. Despite the fact study (Hagedoorn & Cloodt, 2003) shows that using one measure for determining innovativeness normally proves enough, there is chosen for multiple measures for firm innovative performance. To simplify completing the survey, a scale division is used as much as possible. Next to the variables directly linked to the research questions and hypotheses, some extra items were inserted to control and analyze the data more in-depth. Besides examination of the position and seniority of the respondents themselves, some general questions on alliances and the organisations involved are added. The final questionnaire consists of 35 items.
The questionnaire is pre-tested by a respondent from the field. The pre-test consists of completing the survey by the respondent, followed by a one hour during interview on the survey and alliance related subjects. Next to pre-testing in the field, the survey was tested by an academic in order to increase viability. After these tests and their evaluation, some adjustments were made to the survey. Apart from adjusting the sequence of the subjects, some improvements to constructs related to the selection criteria were made. The survey is also shortened to four pages to increase the potential response rate. Appendix II presents the final Dutch survey which is sent to the respondents.
3.2 Sample
partner9 involved, not only the company address data could be retrieved easily, also the names of the strategic managers involved10. After collecting the company names and names of strategic managers, the company address and telephone data were added by using company webpages and on-line address databanks. Using this method of data collection a database containing 334 Dutch respondents is created.
The survey is not send to foreign partners because it is only drawn up in Dutch to avoid unnecessary biases. The used method of research is in line with research on strategic alliances and organizational learning conducted by Norman (2002). In that study a similar kind of survey method is deployed for data collection.
3.3 Data collection
The survey, accompanied by a letter, was send to all 334 respondents in September 2005. To increase the response rate, a free port response envelope was enclosed. The questionnaires were numbered to ensure optimal administrative processing as well as providing opportunity for (non)-response analysis. After one week, 26 usable surveys were returned (response rate 8.08%). Another 21 surveys were returned blank because of a variety of reasons11. In October 2005 the telephonic follow-up started. Respondents who did not have responded were asked whether they had received the survey and if they could return their survey as soon as possible. When respondents did not receive the survey it was e-mailed to them. 50 E-mails were sent this way. After conducting 412 telephone (re)calls in seven days, another 25 usable surveys returned (response rate 7.49%). The total response is 51 usable surveys. Besides these 51 usable surveys, 3 surveys returned but could not be used. Two surveys dealt with mergers instead of strategic alliances, the other was not sufficiently filled in.
The overall response rate is 15.2%. A comparable study by Norman (2002) had a slightly lower response rate (13.6%, 61 usable questionnaires). Compared to the alliance selection study by Nielsen (2002) among Danish firms the response rate is significant higher (6.5%, 120 usable questionnaires). On the other hand, the survey of Chang (2003) among 400 international CEO’s and top-managers, analyzing co-operation and innovative performance, achieved a higher response rate (41%, 162 usable questionnaires). Compared to Dutch surveys, among strategic decision makers, the response rate can be considered fairly high. A survey under Dutch CEO’s conducted by van Ees, Postma and van der Laan (2005) for example, achieved a response rate of 8.5% (135 usable
9 When an internationally operating company has a Dutch field office and the alliance activities were clearly
conducted in the Netherlands, the International Partner is also considered Dutch and is contacted for the survey.
questionnaires). Factors influencing this discrepancy are, amongst others, the intensity of the follow-up, and the length of the survey or the country/culture in which the survey is conducted (Dillman 1978, 2000).
3.4 Non-Response Analysis
To check whether non-response differed significant from the response, analysis was conducted using the Reach database12. This is a database that, amongst others, captures information from the Dutch Chambers of Commerce. Testing the difference between response and non-response the variable number of employees in 2004 is being used. Of the total of 334 companies in the sample, data of 222 companies could be retrieved. From the 51 respondents, 36 could were found in the Reach database. The fact that the data of not all firms in the sample could be retrieved has various reasons.. Some firms, for example, do not exist anymore; others operate under another name than their name registered in the Reach database. The Mann-Whitney test is deployed to check for difference. No significant difference was found testing the two samples (p = 0.328). Thus both samples appear not to have a significantly different distribution.
3.5 Data Analysis
The analysis of the collected data is completed using the statistical package SPPS 14.0 for windows. Before using some of the variables some editing was necessary. This section discusses the steps that were taken to make the selection variables and innovative performance variables ready to use in the research.
The importance of the 14 partner selection characteristics (Hitt et al. 2000) is measured by using 21 items. Seven partner selection characteristics are measured using 2 items. The other seven are measured using one item. The multi-item measures are averaged to create a new comparable measure. Table 3.1 shows the Cronbach’s alphas for the seven multi-item measures. For two items the alpha is below 0.6. Norman (2002) argues that alphas in the 0.5 – 0.6 range are acceptable in the early stages of research. Furthermore, this range is regarded acceptable for constructs that are two-item scales, for alpha increases as the number of items in a scale increases.
Variable Alpha New Complementarity of capabilities 0.57 4.41 Market knowledge/access 0.69 3.92 Quality product/service 0.60 3.84 Unique competencies 0.53 3.80 Intangible assets 0.76 3.69
Partner's ability to acquire skills 0.78 3.44 Technical capabilities 0.65 3.21
Table 3.1 Cronbach’s alpha of the multi-item constructs
The innovative nature of the firms in the sample is measured using three variables: R&D budget in 2004, the number of new product in 2004, and patents registered in 2004. Because the number of patents registered in 2004 is 0 for all respondents, this variable proves unusable. In section 4.2.2 this outcome is discussed more thoroughly. The other two variables prove to be of use.
4 RESULTS
This section presents the research results. First it provides general sample characteristics. Next it presents the univariate analysis of the selection characteristics and innovation parameters. Thirdly it presents the correlation analysis, followed by four tests of difference. The final section discusses the regression results.
4.1 General sample characteristics
The survey examines the latest created strategic alliance the respondents were involved in. To asses the potential historical bias, the respondents were asked when the alliance was created. Table 4.1 provides an overview of the period in which the strategic alliances came into being. Most strategic alliances (35 = 68.6%) existed less than two years, at the time the survey was conducted. Nearly all strategic alliances (50 = 98%) in the study still existed at the time the questionnaire was returned. This percentage seems to be inconsistent with the 40-50% survival rate of strategic alliances that is normally observed (e.g. Duysters, Kok, & Vaanderdrager, 1999). It is likely that this difference is due to the fact that most strategic alliances in this study are relatively recent. Most firms in the study are experienced in partner companies. The average number of alliances is 5.86 (with standard deviation of 6.171, skewness of 2.861 and curtosis of 10.014).
Start Period of the alliance Number Percentage
First Half of 2003 9 17.6 % Second Half of 2003 4 7.8 % First Half of 2004 9 17.6 % Second Half of 2004 15 29.4 % First Half of 2005 10 19.6 % Second Half of 2005 1 2.0 % Unkown 3 5.9 %
Table 4.1 When did the strategic alliances in the sample start?
To control for country specific characteristics all firms involved in the study are Dutch ICT companies. This is also a direct result of the fact that the survey was solely conducted in Dutch to prevent a lingual bias. Nonetheless, not all alliances involved are strictly national. Most alliances in the study are between two Dutch companies (32 = 62.7%). The other 19 (37.3%) alliances are between a Dutch and a foreign partner. In these cases the respondent originates from the Dutch firm.
Turnover in 2004 Number Percentage Less than 200,000 6 11.8 % Between 200,000 and 500,000 7 13.7 % Between 500,000 and 1,000,000 6 11.8 % Between 1,000,000 and 5,000,000 15 29.4 % Between 5,000,000 tot 10,000,000 8 15.7 % More than 10,000,000 9 17.6 %
Table 4.2 Size of the firms in the sample
To assess the viability of the response, the manager with overall responsibility for alliance management was targeted. 80.4 percent of the questionnaires is indeed completed by the manager with overall alliance responsibility. This 80.4% correspondents with the position of the respondents within their firms (see table 4.3). Most respondents are owners or top-managers within their organization. To assess the viability of the data, the respondents were asked how long they had spent working in the industry. On average this is 13.6 years (with s.d. of 6.7). In addition, the respondents were asked how long they had spent working at their current firm (mean = 6.2 years, s.d. = 4.2). Both figures prove reasonable seniority of the respondents. Norman (2002) also included these variables and came with comparable mean years of experience (With averages of respectively 18.7 years within the industry and 8.9 years within the organization,).
Position of the Respondent Number Percentage
Operational 1 2.0 %
Middle Management 6 11.8 %
Staff 3 5.9 %
Top Management 11 21.6 %
Owner 30 58.8 %
Table 4.3 Position of the respondents
The sample includes different types of strategic alliances. The respondents were asked to characterize the nature of the strategic alliance. The result can be found in table 4.4. Respondents were free to pick more than one option13. From the combined list of types of strategic alliances provided by Pekar & Allio (1997) and Das & Teng (1998), ones which are directly linked to innovation, like technology transfer (15 = 29.4%) and R&D partnerships (13 = 25.5%) were reasonably mentioned often. This is consistent with the assumption that strategic alliances in the ICT sectors are often aimed at innovation (cf.
13 Respondents had the option to typify the alliance themselves. The following types are therefore also
Varis & Conn, 2002). Another significant feature is the fact that more than half (28 = 54.9%) of all strategic alliances clearly is aimed at a commercial objective.
Nature of Alliance Number Percentage
Joint Marketing 28 54.9 % Technology transfer 15 29.4 % Equity Investment 14 27.5 % R&D partnership 13 25.5 % Licensing / Cross-licensing 11 21.6 % Shared distribution 10 19.2 % Buyer-Supplier Relationships 9 17.6 % Cross-manufacturing 6 11.8 % Joint Ventures 4 7.8 % Co-operative bidding 2 3.9 %
Lease service agreements 0 0 %
Table 4.4 Nature of strategic alliances in the sample
4.2 Innovativeness and partner selection
This section discusses the main variables of the study. The partner selection variables are discussed first. Secondly the innovation variables are discussed.
4.2.1 Partner selection in the Dutch ICT sector
The fourteen selection characteristics are based on the studies of Dacin et al. (1997) and Hitt et al. (2000). Table 4.5 provides an overview of the relative importance of the partner selection characteristics for strategic alliances in the Dutch ICT sector. Appendix I provides definitions of the various variables as used in this study as well as the two studies mentioned above.
Variable average s.d.
1. Complementarity of capabilities 4.41 0.64 2. Industry attractiveness 4.20 0.87 3. Willingness to share expertise 3.98 0.68 4. Market knowledge/access 3.92 0.89 5. Quality product/service 3.84 0.75 6. Unique competencies 3.80 0.59 7. Intangible assets 3.69 0.97 8. Financial assets 3.67 0.93 9. Managerial capabilities 3.59 0.83 10. Partner's ability to acquire skills 3.44 0.80 11. Technical capabilities 3.21 0.97 12. Special skills that you can learn 3.00 0.96 13. Previous alliance experience 2.88 0.99
14. Cost of Alternatives 2.80 0.94
Table 4.5 Partner Selection Criteria in order of importance
the focal firm. For managers in the Dutch ICT sector it is important that the resources of their partners provide new possibilities (unique competences is placed sixth). However, it is considered to be more important that the resources of both parties are compatible.
Furthermore, Dutch executives place relatively high importance on the marketing resources of their partner. The expertise and/or ability of a partner to effectively operate in a market and the degree to which an industry presents a favourable environment are both in the top 5 of most important selection characteristics. This seems to be in line with the earlier notion that the alliances in the Dutch ICT sector are primarily aiming at achieving marketing goals (see table 4.4). On the other hand, it is worth noticing that the question whether the partner has technical capability that might be of importance of the focal firm is seen as relatively unimportant. This seems to be contrary to the fact that many alliances are aimed at innovation goals.
Learning objectives seem to play a less important role. Dutch managers consider the fact that the partner is willing to share knowledge important. However, the question whether the partner has the ability to acquire skills of their partner seems to be less important. It also appears to be of less importance for Dutch executives that their partners might possibly be able to acquire skills which their firm possesses. Appropriation of knowledge by the partner is regarded less important than acquiring knowledge themselves.
Cost of alternatives is the most unimportant partner selection characteristic. Firms hardly place any significance in alternatives that might prove more cost efficient. Firms also place little significance on the alliance experience of their potential partner. This learning opportunity has the penultimate place in the ranking.
Comparing this to the relative importance of selection variables in other situations, many differences come about. Table 4.6 and 4.7 display the selection criteria in other studies in order of importance. Both tables display many differences with the results in this study. Therefore, the conclusion is justified that the importance of partner selection is likely influenced by situational variables. However comparing the rankings below with the ranking in current analysis is difficult. Current analysis is aimed at the Dutch ICT sector and the rankings below are international studies that did not take into account the sectors in which the respondents were active in, only their nationality.
Korean Executives U.S. Executives
1. Technical Capabilities 1. Financial Assets 2. Industry Attractiveness 2. Managerial Capabilities 3. Special skills that you can learn 3. Quality product/service
4. Willingness to Share Expertise 4. Complementarity of capabilities 5. Capabilities to Provide Quality 5. Unique competencies
Variable
1. Market knowledge/access (4) 3.43
2. Financial assets (8) 3.20
3. Quality product/service (5) 3.10 4. Previous alliance experience (13) 3.07 5. Technical capabilities (11) 3.03 6. Complementarity of capabilities (1) 2.97 7. Managerial capabilities (9) 2.97
8. Intangible assets (7) 2.93
9. Special skills that you can learn (12) 2.90 10. Industry attractiveness (2) 2.87 11. Unique competencies (6) 2.87 12. Cost of Alternatives (14) 2.80 13. Partner's ability to acquire skills (10) 2.70 14. Willingness to share expertise (3) 2.53
Table 4.7 Partner selection criteria in order of importance, international research (source: Hitt et al. 2000)
(Behind the variables the position in current analysis)
4.2.2 Innovative Performance
As mentioned before, the innovative performance of the firms is analyzed using three different measures. Respondents were asked how many new products the firm developed in 2004, how many inventions were patented in 2004, and what percentage of the total turnover in 2004 was spend on R&D. Only two out of three measures proved usable. The number of inventions patented in 2004 was 0 in all questionnaires. This variable is therefore not used in further analysis. The percentage spend on R&D differs more; table 4.8 shows the distribution within the sample14.
R&D Expenditure Number Percentage
0 % - 5 % 11 21.6 % 5 % - 10 % 8 15.7 % 10 % - 20 % 19 37.3 % 20 % - 30 % 10 19.6 % 30 % - 40 % 1 2.0 % 40 % - 50 % 1 2.0 % > 50 % 1 2.0 %
Table 4.8 Percentage of the total budget spend on R&D
Most firms (48 = 94.1%) have an R&D budget between 0 and 30 %. The number of new products launched in 2004 also showed a reasonable variety. The answer to this question varied from 0 to 30 new products introduced in 200415 with and average o 3.45 (with a standard deviation of 5.98).
14 Four questionnaires returned blank at this item. Based on the characteristics of these respondents and the
overall average answer these respondents were presumed to have a R&D budget of 10-20% of the total budget.
15 One answer was 1,000 at this item. This respondent is entirely removed from the sample because this is an
4.3 Correlation Coefficients
Because of ordinal nature of selection variables, the Kendall tau-b correlation analysis is deployed. Appendix III contains the correlations of the selection characteristics and innovation and control variables. As shown in this correlation matrix, many selection variables are mutually correlated. However, between the selection variables and innovation variables less correlation is observed. Only the variables ‘cost of alternatives’ (p < 0.01) and ‘alliance experience’ (p < 0.10) are significantly negative correlated with innovate nature of the focal firm. Thus hypothesis 3 is confirmed by the results of the correlation analysis. The other four hypotheses are not supported by the correlation results.
The size of the selecting firms is significantly correlated with ‘intangible assets’ (p < 0.10), ‘financial assets’ (p < 0.01) and ‘technical capabilities’ (p < 0.10). The larger the selecting firm the more importance is being placed on the financial situation of the partner. Furthermore, the larger the selecting firm the less important the technical capabilities of partner firms become. It seems likely that larger firms have the technical capabilities they need themselves, so therefore experience less need to acquire external technology.
The alliance experience of the focal firm is significantly correlated with ‘industry attractiveness’ (p < 0.05), ‘market knowledge/access’ (p < 0.10), ‘intangible assets’ (p < 0.05) and ‘financial assets’ (p < 0.01). Worth noticing is the fact that the alliance and experience and firm size both have positive significant correlation with ‘intangible assets’ and ‘financial assets’. As the correlation matrix displays, the size of firms is positively correlated with the alliance experience (p < 0.05). Larger firms tend to have more alliances than smaller firms. Firms that are larger are likely to have more capacity to engage multiple alliances. Another significant feature is the phenomenon that the more firms gain alliance experience, the less marketing related issues tend to play a role. Both ‘industry attractiveness’ and ‘market access/knowledge’ are negatively correlated with alliance experience.
The scope of the alliance is significantly positive correlated with the ‘alliance experience’ (p < .10) and the special skills that can be learned from the partner (p < 0.05). The broader alliances are, the more important the special skills of the partner become. Alliance experience of the partner becomes also more important when the scope of the alliance is broader. It seems likely that the broader the alliance is, the more difficult it becomes to manage. Therefore, alliance experience might be more appreciated when the scope of the alliance is larger.
