EXTERNAL KNOWLEDGE DOMAINS AND SEARCH STRATEGIES FOR INNOVATION
by
Karina Andrea Barrios
A thesis submitted to the
Faculty of Economics and Business,
University of Groningen
in partial fulfillment of the requirements for the degree of
Master of Science
in
Strategy & Innovation
2 FACULTY OF ECONOMICS & BUSINESS
1st supervisor Florian Noseleit
2nd supervisor Charlie Carroll
Website www.rug.nl/feb
Telephone number +31 50 363 3453
Q-MODUS B.V.
Supervisor Gea Vellinga
Website www.q-modus.nl
www.innovatiespotter.nl
3 PREFACE
This thesis is the final proof of competence for obtaining the Master of Science degree in Business Administration, with a specialization in Strategy & Innovation. I have undertaken the research in combination with an internship at Q-modus, a small firm specialized in knowledge sharing, innovation and business development, established in Groningen. I learned a lot throughout the process, especially on how to combine an academic view with practice. Thereby, I enjoyed working with different colleagues towards the Innovatiespotter congress that took place on the 18th of September, 2012.
I would like to take this opportunity to express my gratitude to my firm supervisor, Gea Vellinga for her knowledge input and the support and time she devoted to me. Furthermore, my acknowledgements go out to my first university supervisor Florian Noseleit. I thank you for your feedback and constructive criticism at times I really needed them. Also, my second supervisor deserves a warm recognition. A thanks and a lot of love goes out to my family and boyfriend, who supported me throughout the entire course of my thesis. Finally, I would like to this opportunity to dedicate this thesis to my late friend, Lina Liu.
Groningen, January 2013
4 EXECUTIVE SUMMARY
To facilitate the search for external knowledge, Q-modus B.V. developed the Innovatiespotter; www.innovatiespotter.nl. This website is the front-end of an information system i.e. online database which consists of thousands company profiles of innovative, knowledge-intensive and specialized companies in the Northern Netherlands. Extensive and current business data are to be found by using key words. The Innovatiespotter attempts to provide a solution for: direct marketing, networking, acquisition, data enrichment, lead generation, market research and economic policy development. For Q-modus it is important to get a study performed about external search strategies and the sources involved to obtain knowledge for the commercialization of the Innovatiespotter, and the further development and growth of the firm.
Extending studies on the search for external knowledge, I investigate the influence of external knowledge sources and search strategies on innovative performance by developing items for the types of knowledge sources, i.e. knowledge domains. Moreover, in order to gain further insight into the role of external search on innovative performance, I contribute by investigating the relative importance of these strategies within each knowledge domain. The significance of this study also lies in the extension to prior research on the search for external knowledge to a different region, the Northern Netherlands.
Out of the literature review and the factor analysis, I developed seven hypotheses. These propose that the use of the external knowledge domains; (1) market players, including customers and suppliers; (2) market intermediaries, like consultants and R&D enterprises (3) institutional sources, such as universities and governmental research institutions; (4) specialized sources, like technical and environmental standards, and regulations; (5) other sources, such as trade associations and professional meetings positively influences the firm’s innovative performance. Also, the use of a broad search trajectory (how widely the firm explores knowledge domains) and a deep search trajectory (how deep the firm explores knowledge domains). The quantitative data derived by a questionnaire collected mostly at the Innovatiespotter congress that took place on the 18th of September 2012 (www.innovatiespotter.nl/congres), is related to the hypotheses.
5
suppliers can be really hard as it is based on personal contacts. This type of interaction can lead to ‘unsatisfactory’ innovations because of inertia in these types of relationships. Besides, the direct release of information to competitors could negatively impact the benefits (i.e. the turnover related to innovation) gained from innovation activities.
Secondly, this study shows that acquiring knowledge through the use of consultants and commercial laboratories/R&D enterprises (market intermediaries) will positively affect the probability of gaining innovative performance. The usage of these knowledge sources entails cooperation with scientific agents or with agents outside the industry chain which can be typed as research experts.
Thirdly, universities or higher education institutes, government research organizations, other public sector organizations like business links, government offices and public research institutes (i.e. institutional knowledge domain) as information sources are associated with high innovative performance. Institutions may impact innovation output by providing scientific research inputs for innovating firms and helping in translating academic codified knowledge into practical and accessible know-how.
Fourthly, I found that using technical standards, health and safety standards and regulations, and environmental standards and regulations (i.e. the specialized domain) as a source of innovation is not significantly linked to higher levels of innovative turnover. A possible reason for the lack of significant effect is that firms may fail to utilize such knowledge in sufficient quantities, owing to difficulties presented by the technical aspects of the information.
Fifthly, concerning the other knowledge domain, no significant relationship with the firm’s innovative performance exists. A reason may be that that the absorptive capacity may play an important role. It is the capability to recognize the value of new, external information and impacts firm’s ability to transform this type of knowledge into competitive advantages. Firms should try to increase their absorptive capacity. I should note that, the source ‘fairs/exhibitions’, by literature designated as a component of the other knowledge domain, was left out of this study.
6
external search scope of knowledge sources may be subject to diminishing marginal returns and that oversearch might also hinder innovative performance. Firms may face a tradeoff between scope and depth in their innovation objectives.
In addition, it appears that the usage of internal knowledge and industry growth are important factors in explaining the propensity of firms to use knowledge sources and strategies in their innovative activities. No pattern of effects can be observed for the variable firm size. This was not expected because a particular group of firms, small and medium sized enterprises, seem to rarely possess all the resources they require to innovate effectively. However, also prior empirical evidence does not confirm the role of the size of the firm in relation to innovation and there exists support that it is plausible that also big firms have rigidities in introducing novelty.
Finally, I found that the specialized domain is associated with search scope the most. An explanation is that the combination of standards and regulations and immediate practical implementation will require firms to use also other external knowledge sources. The results for search depth suggest that innovators that make use of the other and market players domains adopt deep search strategies. When using these sources firms build and sustain virtuous exchanges and collaborations with external actors. Maintaining these deep links requires relatively a lot of resources and attention, and therefore firm seem to focus on these sources.
7 MANAGEMENTSAMENVATTING
Om het zoeken naar externe kennis voor bedrijven te vergemakkelijken, heeft Q-modus B.V. de Innovatiespotter ontwikkeld (www.innovatiespotter.nl). Deze website is het front-end van een informatiesysteem c.q. online database dat bestaat uit duizenden bedrijfsprofielen van innovatieve, kennisintensieve en gespecialiseerde bedrijven in Noord-Nederland. Uitgebreide en actuele bedrijfsgegevens zijn te vinden door het gebruik van sleutelwoorden. De Innovatiespotter tracht een oplossing te bieden voor: direct marketing, netwerken, acquisitie, data verrijking, lead generation, marktonderzoek en economische ontwikkeling van het beleid. Voor Q-modus is het van belang om een studie over externe zoekstrategieën en de bronnen die gerelateerd zijn uit te laten voeren om kennis voor de commercialisering van de Innovatiespotter, en om verdere ontwikkeling en groei van het bedrijf te verkrijgen.
In het verlengde van uitgevoerde onderzoeken naar de zoektocht van externe kennis, onderzoek ik in opdracht van Q-modus de invloed van externe kennisbronnen en zoekstrategieën op innovatieve prestaties van bedrijven. Dit doe ik door het ontwikkelen van items voor de verschillende typen kennisbronnen, oftewel kennisdomeinen. Om meer inzicht te krijgen in de rol van externe zoektocht op innovatieve prestaties, draagt deze scriptie bij door het relatieve belang van deze strategieën binnen elk kennisdomein te onderzoeken. Daarbij ligt de betekenis van dit onderzoek in de regio dat gekozen is, Noord-Nederland.
Zeven hypothesen werden ontwikkeld. Deze stellen dat het gebruik van de externe kennisdomeinen; (1) spelers op de markt, bijvoorbeeld klanten en leveranciers, (2) tussenpersonen in de markt, zoals consultants en R&D-bedrijven (3) institutionele bronnen, zoals universiteiten en onderzoeksinstellingen van de overheid, (4) gespecialiseerde bronnen, bijvoorbeeld technische en milieunormen en regelgeving, (5) overige bronnen, zoals brancheorganisaties en professionele vergaderingen een positieve invloed hebben op innovatieve prestaties (omzet gerelateerd aan innovatie). Ook het gebruik van een breed zoektraject (hoe breed het bedrijf de kennisdomeinen verkent) en een diep zoektraject (hoe intensief de firma kennisdomeinen verkent) werd eraan gerelateerd. De kwantitatieve data werd grotendeels verkregen door een enquête afgenomen op het Innovatiespottercongres dat plaatsvond op 18 september 2012 (www.innovatiespotter.nl/congres). Hieronder de resultaten.
8
het verwerven van kennis door klanten, concurrenten en leveranciers moeilijk blijkt te zijn, omdat het gebaseerd is op persoonlijke contacten. Deze interacties kunnen leiden tot 'onvoldoende' innovaties als gevolg van inertie. Bovendien kan het direct vrijgeven van informatie aan concurrenten een negatieve invloed hebben op de omzet dat behaald is door opgedane innovatieve activiteiten.
Ten tweede, dit onderzoek toont aan dat het verwerven van kennis door middel van consultants en commerciële laboratoria/R&D-bedrijven een positieve invloed heeft op de kans op het verkrijgen van innovatieve prestaties. Het gebruik van deze kennisbronnen behelst samenwerking met deskundigen buiten de industrieketen.
Ten derde, universiteiten of instellingen voor hoger onderwijs, overheid onderzoeksorganisaties en andere overheidsorganisaties zijn als informatiebronnen geassocieerd met hogere innovatieve prestaties. Instellingen kunnen het innovatieve vermogen beïnvloeden door ingang naar wetenschappelijk onderzoek voor innoverende bedrijven te verschaffen en door het vertalen van wetenschappelijke, gecodificeerde kennis naar praktische en toegankelijke know-how.
Ten vierde, het gebruik van technische normen, gezondheids- en veiligheidsnormen en regelgeving en milieunormen en voorschriften (gespecialiseerde domein) als een bron van innovatie niet significant gekoppeld is aan hogere niveaus van innovatieve omzet. Een mogelijke reden is dat bedrijven falen om deze kennis in voldoende hoeveelheden te gebruiken wegens de moeilijke, technische aspecten van de informatie.
Ten vijfde, met betrekking tot het overige kennisdomein, er is geen significante relatie gevonden. Een reden kan zijn dat het absorptievermogen van de bedrijven een belangrijke rol heeft gespeeld. Het absorptievermogen is het vermogen om de waarde van nieuwe, externe informatie en de effecten ervan te herkennen en om deze kennis als concurrentievoordeel te gebruiken. Bedrijven moeten proberen hun absorptiecapaciteit te vergroten. N.B. de kennisbron 'beurzen/ tentoonstellingen', die door de literatuur aangeduid wordt als een component van het overige kennisdomein, werd weggelaten in dit onderzoek.
9
‘oversearch’ de innovatieve prestaties kan belemmeren. Bedrijven kunnen worden geconfronteerd met een afweging tussen het zoeken in de breedte en de diepte in hun innovatiedoelstellingen.
Daarnaast blijkt dat de groei van de industrie en het gebruik van interne kennis belangrijke factoren zijn met een positieve invloed. Geen effect wordt waargenomen voor bedrijfsgrootte. Dit werd niet verwacht, omdat een bepaalde groep bedrijven, de middelgrote en kleine bedrijven (MKB), zelden lijken te beschikken over alle middelen die ze nodig hebben om effectief te kunnen innoveren. Echter, ook voorafgaand empirisch bewijs kan de rol van de grootte van het bedrijf in verband met innovatie niet bevestigen. Het is aannemelijk dat ook grote bedrijven starheid bij de invoering van innovaties beleven.
Tot slot, het gespecialiseerde domein wordt het meest geassocieerd met en breed zoektraject. Een verklaring hiervoor is dat de combinatie van normen en voorschriften en de onmiddellijke praktische toepassing die moet plaatsvinden, bedrijven zal verplichten om ook gebruik maken van andere externe kennisbronnen. Een ander belangrijk resultaat is dat bedrijven die gebruik maken van de ‘overige bronnen’ en ‘marktpartijen’-domeinen voornamelijk het diepe zoekstrategieën hanteren. Bij het gebruik van deze bronnen worden bedrijven gesteund in de samenwerking met externe actoren. Dit vereist relatief veel middelen en aandacht, en dus lijken de bedrijven zich te richten op deze bronnen.
10 ABSTRACT
This thesis focuses on knowledge sources outside the firm that may be used for innovation. It examines the role of external knowledge domains and search strategies on innovative performance and learns what the relative importance of these strategies is within each external knowledge domain. Data collection is undertaken by means of questionnaires. The use of market intermediaries and institutional sources seem to positively affect innovative performance. The usage of the market players, other and the specialized domain do not. There was no support for the expectation that a broad search trajectory positively impacts innovative performance, whereas the use of a deep search trajectory did. I found that the specialized domain is associated with search scope the most. Innovators that make use of the other and market players domains seem to adopt deep search strategies.
11 TABLE OF CONTENTS
1 INTRODUCTION 13
2 LITERATURE REVIEW 16
Innovation and External Knowledge Sources 16
Market sources 17
Institutional sources 18
Specialized sources 18
Other sources 19
External Search Strategies 19
Search scope 19
Search depth 20
3 METHODOLOGY 22
Procedures and Sample 22
Measures 23
Innovative performance 23
External knowledge sources as knowledge domains 23
External search strategies 24
Control variables 24
Analyses 25
Factor analysis 25
12
One-sample T-test 26
4 RESULTS 27
Factor analyses 27
Descriptive Statistics and Correlations 28
Regression Results 29
T-Test Results 30
5 DISCUSSION 39
Findings and Theoretical Implications 39
Contributions 43
Limitations and Future Research 43
6 REFERENCES 45
7 APPENDICES 56
APPENDIX A: Flyer 57
13
1 INTRODUCTION
Within the business setting, innovation is considered to be vital for a firm as it generates various positive outcomes including a sustained competitive advantage (Zahra & George, 2002; Salavou, 2004; Forsman, 2009), an increased productivity (Reichstein & Salter, 2006), the greater ability to survive in the industry (Banbury & Mitchell, 1995), and the leverage of existing resources and capabilities (Leonard, 1998). Given the importance of innovation as a major competitive weapon for many firms (Oliver & Liebeskind, 1998), research in a wide range of disciplines, including economics, marketing, and organization behavior, have been paying attention to the drivers of innovation (Chandrashekaran, Mehta, Chandrashekaran, & Grewal, 1999). Traditionally, firms organized research and development (R&D) internally and relied on external knowledge only for relatively simple functions (Johnson & Kuehn, 1987; Powell, Koput & Smith-Doerr, 1996).
Nowadays, many firms seem to favor the use of outside knowledge (Rosenkopf & Nerkar, 2001; Ahuja & Lampert, 2001) instead of finding solutions on their own or within a closely related environment for innovation (Chen, Chen & Vanhaverbeke, 2011). The main reason is that in today’s business environment, characterized by globalization, industry convergence, and rapid technology change (Korine, Asakawa & Gomez,. 2002), “no company is smart enough to know what to do with every new opportunity it finds, and no company has enough resources to pursue all the opportunities it might execute” (Wolpert, 2002: 80).
Evidence in innovation literature shows that firms often search for knowledge and new ideas by for example deepening contacts with customers, suppliers, universities and competitors (Cohen & Levinthal, 1990), attending trade fairs and conferences, and using online databases (Chen et al., 2011; Fontana, Geuna & Matt, 2006). Hence, an increasing number of firms have shifted to a so called ‘open innovation’ model (Gassmann & Enkel, 2004).
14
involves intra-industry and extra-industry sources (Stiglitz, 2002; Paananen, 2012). Types of external knowledge sources which may be used by firms are; (1) market sources, including customers and suppliers; (2) institutional sources, such as universities and governmental research institutions; (3) specialized sources, like technical and environmental standards, and regulations; (4) other sources, such as trade associations, fairs, and exhibitions (Laursen & Salter, 2006). In addition, literature on open innovation highlights the importance of the use of knowledge sources through the strategies ‘search scope’ and ‘search depth’ (Katila & Ahuja, 2002; Smith, Collins & Clark, 2005; Zhang & Li, 2010). Search scope describes how widely firms explores knowledge domains (Chen et al., 2011), whereas search depth resembles the extent of intensity to which firms draw from external sources of innovative ideas (Laursen & Salter, 2006). The strategic decision to be made by firms is whether or not they should make use of a diverse set of sources within the innovation process (Mohnen & Röller, 2005; Classen et al., 2012).
Firms often invest considerable amounts of time, money and other resources in the external search for new innovative opportunities (Arundel & Bordoy, 2006). These investments ought to increase the firm’s ability to create, use, and recombine new and existing knowledge (Laursen, 2004). External knowledge then will provide valuable information that the firm can use to develop innovations and as a consequence (Fleming & Sorenson, 2004), they should be more successful innovators than firms that do not. Greater success in innovation should, in turn, lead to a higher turnover related to new or improved products and services (Community Innovation Statistics, 2006). Innovation literature indicates that the character of a search strategy concerning external knowledge sources may positively influence the firm’s innovative performance (Katila, 2002; Katila & Ahuja, 2002; Miller, Fern & Cardinal, 2007) and therefore help them achieve and sustain innovation (Chesbrough, 2003). Choices firms make about whether to use knowledge from different sources and the strategic decision whether or not the firm should make use of a diverse set of sources (Classen, Van Gils, Bammens, & Carree, 2012) may directly affect their innovative performance (Mohnen & Röller, 2005; Paananen, 2012).
15
external knowledge sources and strategies should be expanded to other regions (Chen et al., 2011). Another research gap is that their approach suffers as it does not allow for an analysis of the importance of scope and depth of external search on the different knowledge sources to innovative performance. It is first necessary to develop several fine-grained items for the types of knowledge sources (Laursen & Salter, 2006).
In order to extend these studies and fill the aforementioned research gaps, I investigate the influence of external knowledge sources and search strategies on innovative performance by developing items for the types of knowledge sources, i.e. knowledge domains, identified by Laursen & Salter (2006). Moreover, in order to gain further insight into the role of external search on innovative performance, I contribute by investigating the relative importance of these strategies within each knowledge domain. The significance of this study also lies in the extension to prior research on the search for external knowledge to a different region, the Northern Netherlands.
16
2 LITERATURE REVIEW
Innovation and External Knowledge Sources
The innovation process has been divided into three sub processes: front-end process, definition process and implementation process (Forsman, 2009). During the front-end process, external information is detected and acquired. Then, the information is assimilated to enable the selection and decision-making regarding the potential innovation efforts or projects. Throughout the definition process external knowledge is combined with internal knowledge, and during the implementation process the new knowledge combinations are transformed into the competitive advantage of a firm (Zahra & George, 2002). It is why management literature over the past decade emphasized the need for firms to access external knowledge sources in order to successfully innovate (Arundel & Bordoy, 2001). Studies on open innovation provide helpful insights to pinpoint the contributions of these sources in determining innovative performance of firms (Chen et al., 2011). For example, Freel’s (2000) study showed that the most innovative and better performing firms are generally more likely to have links with external organizations. Similarly, Nooteboom (2006) addresses that innovations particularly arise from interactions between firms and other institutions. Furthermore, Chen et al. (2011) value the use and integration of external knowledge sources, such as attending trade fairs and making use of consultants and online databases, as an important determinant in a firm’s ability to acquire competitive advantage due to their innovativeness. There seems to be consensus on the notion that firms which are more open to external knowledge sources are more likely to be more innovative than firms which are not (Chang & Huang, 2007).
17
Following the division of knowledge sources of Laursen and Salter (2006), the relationship between external knowledge sources will be separately related to innovative performance below.
Market sources. The market provides five external knowledge sources: customers, suppliers, competitors, consultants, and commercial laboratories and R&D enterprises (Laursen & Salter, 2006). It can be expected that knowledge from these sources will be correlated with firm-level innovation (Reichstein & Salter, 2006). Firms collaborating with customers primarily search for new ideas or ways to reduce uncertainty associated with market introduction of innovations (von Hippel, 1988) and may beget innovations among their customers (Reichstein & Salter, 2006). Knowledge of clients’ demands reduces the risk of failure for the innovating firm (Pittaway, Roberston, Munir, Denyer & Neely, 2004). Cooperation with suppliers of equipment, materials, components, or software generally aims at input quality improvements or cost reductions from innovations (Hagedoorn, 1993) and innovators often need to work closely with suppliers in order to develop new technologies (Von Hippel, 1988). Regardless of whether a firm sells its products directly to the retail market, or to other businesses, it is often necessary for firms to work closely with an upstream source, such as suppliers, to understand and utilize the full potential of the new technology (Rouvinen, 2002; Cabagnols & Le Bas, 2002). Collaboration with another source of innovation, competitors, is typically motivated by potential synergy effects (Das & Teng, 2000) or shared R&D costs (Miotti & Sachwald, 2003) influencing positively innovative performance. Additionally, consultants may play a key role in diffusing new practices across industry and provide critical inputs to help firms develop new products or processes (Reichstein & Salter, 2006). Organizational skills learned from consultants (Svetina & Prodan, 2008) and the partnering with commercial laboratories or R&D institutes (Machikita, Miyahara, Tsuji & Ueki, 2008) seem to have a positive effect on the probability of technology-oriented innovations (Grimpe & Sofka, 2009). Thus, the use of these knowledge sources seem to positively influence innovation. The following hypothesis can be stated:
18
Institutional sources. Outside institutions which can be used for knowledge are universities or higher education institutes, government research organizations, other public sector organizations like business links, government offices and public research institutes (PRIs) (Laursen & Salter, 2006). Cooperation with universities and research institutes pursues radical breakthrough product innovations that may open up entire new markets or market segments (Monjon & Waelbroeck, 2003). These knowledge institutions are actively involved in a set of relationships occurring in the business environment (Gunasekara, 2006) and are particularly seen as lead players in the innovative activity of firms providing scientific research inputs for innovating firms (Keeble & Wilkinson, 2000). In addition to channeling information and knowledge, these institutions can also help translate academic codified knowledge into practical and accessible know-how (Gambarotto & Solari, 2004). Their contribution is important in enabling firms to develop thinking that steps outside their particular business system (Liyanage 1995; Pittaway et al., 2004). Institutional sources are often an important source of innovations, especially in emerging technology (Zucker, Darby & Brewer, 1998). Thus, I propose:
Hypothesis 2: the use of the external knowledge domain ‘institutional’ positively influences the firm’s overall innovative performance.
19 Hypothesis 3: the use of the external knowledge domain ‘specialized’ positively
influences the firm’s overall innovative performance.
Other sources. Finally, other sources are: professional conferences, meetings, trade associations, technical or trade press, computer databases, and fairs/exhibitions (Laursen & Salter, 2006). According to Simmie (2004) face-to-face meetings such as professional conferences, meetings, and fairs/exhibitions are important for the knowledge and information they employ in innovation, especially when they are intermittent. Also trade associations have been argued to be an important factor affecting innovation performance (Pittaway et al., 2004) because they are specifically aimed at promoting innovation. They ideally act as neutral knowledge broker but also act as an important conduit for the development of informal relationships (personal relations between individuals), which are the basis for the development of network relationships for innovation (Hanna & Walsh, 2002). Technical or trade press and computer databases are important sources of information about what is happening across firm boundaries, without the competitive consequences or costs of working with market-based sources such as customers and consultants (Rosenkopf & Almeida, 2003; Mol & Birkinshaw, 2009). Therefore, I propose:
Hypothesis 4: the use of the external knowledge domain ‘other’ positively influences the firm’s overall innovative performance.
External Search Strategies
To gather knowledge from external sources, firms implement an external search strategy (Laursen & Salter, 2004); search scope or search depth. These concepts are seen as two components of the openness of the openness of a firm’s external search (Laursen & Salter, 2006). The results of the research by Chesbrough (2003) strongly suggest that searching widely (search scope) and deeply (search depth) across a variety of search channels can provide ideas and resources that help firms gain and exploit innovative opportunities.
20
innovative activities (Laursen & Salter, 2006). The higher the number of different knowledge areas firms search across, the broader the span of their search is (Capaldo & Petruzzelli, 2011). Katila and Ahuja (2002) were the first to provide empirical support for a positive relationship between search scope (how widely a firm explores external knowledge) and innovation. Others reveal that broader search horizons are of importance for innovation processes, in particular in small and medium-sized enterprises (Street & Cameron, 2007). The scope of search across geographic, scientific, organizational and technological boundaries has been associated with a variety of innovation outcomes including invention frequency (Katila & Ahuja, 2002), invention impact (Rosenkopf & Nerkar, 2001) and breakthrough inventions (Ahuja & Lampert, 2001). Also Chiang and Hung (2010) advocate that search scope is positively related to innovation performance.
There are three ways in which search scope can contribute to innovation. First of all, innovation is an information intensive activity (Salman & Saives, 2005). Information regarding other firms’ product offerings and innovation activities can make innovation opportunities more visible to firms (Ahuja, 2000). Second, a broadened search scope can enrich a firm’s knowledge pool and provide more choices to solve problems (Katila & Ahuja, 2002). There is a limit to the number of new products that can be created by using the same set of knowledge elements (Zhang & Li, 2010). Search with a broad scope can increase a new product introduction by adding new elements in its knowledge pool and thus improves the possibility for it to find new useful combinations of these elements (Katila & Ahuja, 2002). Third, a broadened search scope can also help firms locate external complementary resources and capabilities that are critical for their innovative performance (Porter, 1998; Wolpert, 2002). Firms are then able to integrate and recombine various knowledge elements in external knowledge space to create their new or improved products and services. Thus, broad trajectories seem to stimulate innovation by introducing novelty (Davis & Eisenhardt, 2011) and therefore the following hypothesis can be stated:
Hypothesis 5: the use of a broad search trajectory positively influences the firm’s innovative performance.
21
a firm’s contacts with different external sources provides a mechanism for understanding the way firms search deeply within the innovation system and how these external sources are integrated into internal innovative efforts (Laursen & Salter, 2004). The construct ‘search depth’ indicates from how many channels the focal firm intensively sources ideas for innovations. For example, it may reflect the intensity of co-operation with external partners (Chen et al., 2011). The study by Chiang and Hung (2010) on search depth concludes that this strategy is positively related to the innovating firm’s innovation performance, especially when it comes to incremental innovation. For instance, by intensively relying on a partner the firm can gain trust (Chang & Huang, 2007) and therefore access that can result in outcomes for the firm such as enhanced competitive advantage and competitive awareness (Human & Provan, 1996), increased sales growth, higher market value (Stuart, Hoang & Hybels, 1999) and eventually innovation (Stuart, 2000). Stuart et al. (1999) point out that the greater the (technological) capabilities of partners the higher the rate of innovation of the business, and the more prominent the external partners the better access the business has to financial resources (Stuart, 2000). Davis and Eisenhardt (2011) state that deep trajectories stimulate innovation, at least until the limit of useful combinations is reached. Therefore, I propose the following hypothesis:
Hypothesis 6: the use of a deep search trajectory positively influences the firm’s innovative performance.
For a visual representation of the literature review, see Figure 1.
FIGURE 1 Conceptual model
22
3 METHODOLOGY
Procedures and Sample
Data was collected by means of questionnaires. Respondents were contacted in person on the Innovatiespottercongres (http://www.innovatiespotter.nl/congres), a match making event organized by Q-modus and partners, that took place on the 18th of September, 2012. The program consisted of speakers engaged in innovation (e.g. prof. dr. Aard Groen), workshops by several partners (e.g. PNO Consultants) were held and a networking reception took place. Firms from different industries, interested in matchmaking and innovation, and located in the Northern Netherlands attended the event. Those who consent received a paper questionnaire in Dutch. I tried to encourage their participation by offering a demo account on a service Q-modus offers, the Innovatiespotter (www.innovatiespotter.nl). This website is the front-end of an information system i.e. online database which consists of thousands company profiles of innovative, knowledge-intensive and specialized companies in the Northern Netherlands. Extensive and current business data are to be found by using key words. In addition, an online questionnaire was set up through Thesis Tools (www.thesistools.com). A flyer, containing the survey’s link was given away in a goody bag after the event and was posted on the company’s website. See appendix I and II for the flyer and the questionnaire respectively. All respondents with missing values for key variables were excluded. In the online questionnaire, questions related to key variables, i.e. dependent and independent variables, were set as required in order to keep the missing key values to a minimum. Firm information was complemented by using the database of Q-modus.
23
agriculture, forestry and fishing (Dutch: agricultuur, bosbouw en visserij), building (Dutch: bouwnijverheid), and the health and welfare (Dutch: gezonheids- en welzijnszorg) industry.
Measures
I followed suggestions for developing valid measures. Therefore, the measurement scales were specifically generated for this study and were based on descriptions and measures of related constructs in the literature (Bagozzi, Yi & Phillips, 1991). Laursen and Salter (2006) used the same variable structure in their study based on the Community Innovation Survey (2006). Further, similar types of measure structures have been used in many other studies (e.g. Leiponen & Helfat, 2010; Paananen, 2012).
Innovative performance. In this study, the dependent variable is innovative performance. This variable is measured by the turnover related to new or improved products/services (Arundel & Dorboy, 2001; Community Innovation Statistics, 2006). As it is a time-dependent variable, the length of the observation period becomes an important issue (Weinzimmer, Nystrom & Freeman, 1998). This study collects data from 2009 to 2011. The three-year period was selected in line with the Community Innovation Statistics (2006) survey. Innovative performance is measured by the survey item: ‘Which percentage of your total turnover in 2011 is related to new or improved products/services introduced during 2009-2011?’ with answering categories 0%, 1-5%, 6-10%, 11-20%, >20% (Liu, 2010).
24
able to create new items, factor analysis is performed. Factor analyses will show which types of external knowledge sources and thus, how many knowledge domains there are.
External search strategies. For search scope, all knowledge domains (factors) were coded as a binary variable. The scores of the different knowledge domains which differed from ‘0’ (no use) were coded as ‘1’. This means that the responding firm did use this external knowledge domain. Search scope was measured by summing the outcomes of these binary variables and one new variable was created. The score on this variable, search scope, represents the number of external knowledge domains the innovating company draws on for knowledge for innovation and thus how widely the firm explores new knowledge (Laursen & Salter, 2006). For search depth also each of the knowledge domains (factors) were coded into a dummy. Now, firms with a score of 3 (high use) on the scale of external knowledge domain were coded as ‘1’, and all others were coded as ‘0’. Search depth was measured by summing the outcomes per firm and here also one variable was created. The construct search depth indicates from how many knowledge domains the focal firm intensively or deeply sources ideas for innovations. Larger numbers on this variable indicate that the company relies on intensively sourcing information from external knowledge domains to come up with innovations (Laursen & Salter, 2004). Important to note, the variables search scope and search depth can take any integer value between 0 and the number of factors i.e. knowledge domains.
25
investigating the usage of external knowledge sources and strategies of firms (Laursen & Salter, 2006; Leiponen & Helfat, 2010; Classen et al., 2012).
Analyses
Factor analysis. There are criteria which need to show that factor analysis on the knowledge items is allowed. These are; the Cronbach’s coefficient alpha (α) with the corrected item-total (Hair, Black, Babin, Anderson & Tatham, 2006), the Bartlett’s test, and the KMO statistic (Huizingh, 2006). The alpha should be at least .60 (Huizingh, 2006; Hair et al., 2006) and an item is discarded when the correlation with the total is lower than .200 (Everitt, 2002). The reliability of the scale for the market sources (α = .60), institutional sources (α = .83), specialized sources (α = .77) is confirmed. Even though the alpha for the specialized sources would be higher when deleting the item technical standards, I choose not to because the item has a more than acceptable level of corrected item–total correlation; .441. Otherwise information would be lost for which the increase in reliability would not be able to compensate for (Nunnally & Bernstein, 1994). The reliability of the other knowledge scale was initially not sufficient (α = .54) but became sufficient after deleting the item fairs, exhibitions (α = .60). No results in the four inter-item correlation matrices were negative; no variables need to be recoded. Furthermore, the Bartlett’s test shows for all the scales a significance level of .000 and the KMO statistic is large enough (> .5) for all the sources (market; .565, institutional; .757, specialized; .552, other; .602). The results show that factor analysis is appropriate.
Selecting the appropriate number of factors was done by retaining the items with Eigenvalues greater than 1. The rotated component method Varimax is used to identify which items belong to which factor. A new variable is made by computing the mean of the relevant items and by rounding it. This is how the variables keep the ordinal measurement scale.
26
categories of the dependent variable represent an order. For the analyses I use the SPSS (version 16.0) ordinal regression procedure, which can estimate a variety of ordinal regression models, including the ordered logit. In the analyses, control variables are entered before the independent variables.
27
4 RESULTS
Factor Analyses
Regarding the institutional, specialized and other items, the factor analyses indicate one factor (66.72%, 69.30%, 55.64% variance, respectively). The analysis for the market knowledge domain indicates two factors are to be created. The items suppliers (.710), clients or customers (.766) and competitors (.562) are computed to become the factor ‘market players’ (variance 21.64%) and the items consultants (.872) and commercial lab/R&D companies (.888) are named ‘market intermediaries’ (variance 38.77%). Accordingly, there can be said that the market knowledge domain identified by Laursen and Salter (2006) consists of two constructs. Therefore Hypothesis 1 is divided:
Hypothesis 1a: the usage of the market knowledge domain – players positively influences the firm’s innovative performance.
Hypothesis 1b: the usage of the market knowledge domain - intermediaries positively influences the firm’s innovative performance.
Accordingly, the conceptual model of this thesis is reviewed, see Figure 2.
FIGURE 2
Reviewed conceptual model
28
Concluding, the constructed factors which represent the knowledge domains and thus the independent variables resulting from the factor analysis are market players (H1a), market intermediaries (H1b), institutional (H2), specialized (H3) and other (H4). For the output, see Tables 1 to 4.
Descriptive Statistics and Correlations
Table 5 presents the usage of external knowledge domains for innovation activities. Overall, the most important knowledge domain is market players, followed closely by the other knowledge domain. Market intermediaries, institutional and specialized sources seem to be equally important. As might be expected, the results indicate that the firms’ innovation activities are strongly determined by relations between themselves and their competitors, customers and suppliers (von Hippel, 1988).
Moreover, the descriptive statistics show that 62.2% of the firms can be typed as micro (0-10), 26.9% small and medium-sized (10-250), and 11.0% large (> 250). Furthermore, nearly 40% of the firms had a turnover of over 20% in 2011 related to new or improved products/services introduced in 2009-2011. 9.5% of the firms had a turnover of 0%, 20.3% a turnover between 1 and 5%, 17.6% between 6 and 10%, and 13.5% between 11-20%. Finally, none of the firms indicated they did not use knowledge within the firm for innovation. Nor were there any firms which used this source to a small extent; 27.3% of the firms uses internal knowledge at a medium level and 72.7% at a high level.
The interrelatedness among the variables is tested by performing a correlation analysis. Table 6 presents the means, standard deviations, and correlations among the variables examined in the study. When looking at the external search strategies, the average search scope of a company within the sample is 4.29. In other words, on average each firm used somewhat more than four different external knowledge domains for innovation-related activities. Besdies, the average search depth is 0.68. Hence, on average each firm sources intensively ideas for innovations from approximately one knowledge domain.
29
significantly (p < .05) at a correlation level (ρ) greater than .8 (Katz, 2006). Between the control variable internal knowledge and search depth a correlation relationship exists (ρ = .284; p <.05). Industry growth has a significant correlation relationship with market intermediaries (ρ = .309; p < .01), institutional (ρ = .449; p < .01), specialized (ρ = .350; p < .01) and search scope (ρ = .301; p < .05). Finally, firm size correlates significantly with market intermediaries (ρ = .331; p < .01), institutional (ρ = .328; p < .01), specialized (ρ = .377; p < .01), other (ρ = .228; p < .05), search scope (ρ = .330; p < .01) and search depth (ρ = .389; p < .01).
Regression Results
In total seven different regression models were tested. The models include the baseline model, i.e. the control variables, and the different independent variables in order to test the hypotheses. Table 7 presents the results. All models except for model 2 present a good fit (p < .05): the Pearson chi-square goodness-of-fit measure always had a significance level over 5%. The Nagelkerke R-squared indicates per model that it can account for the relevant percentage of the variance in innovative performance with the according number of observations (N). This measure ranges 19.1% to 41.7%. These summary measures suggest satisfactory ordinal logistic regression models (O’Connell, 2006). The reference groups presenting the largest sample size are chosen. Selecting this criterion means that the standard errors will be smaller and the confidence intervals will be narrower because the model has a larger comparison group and can therefore make more precise estimates (Katz, 2006).
Of the controls, a pattern of effects can be observed for internal knowledge and industry growth. Industry growth (Model 2; β = .398, p < .10, SE = .230, Model 5; β = .386, p < .10, SE = .211, Model 7; β = .703, p < .05, SE = .305) has a positive impact on the ability of the firm to produce innovations, as expected. The other control, internal knowledge, seems to be significant for the models 3 (β = 1.229, p < .05, SE = .556), 4 (β = 1.033, p < .10, SE = .542), and 1 (β = -1.029, p < .05, SE = .522) and presents a positive relationship with innovative performance. When internal knowledge is used at a medium level instead of a high level, the odds to gain innovative performance will be lower.
30
which affirms that the usage of market intermediaries’ sources is positively related to innovative performance of the firm, I do find evidence. The parameter in comparing high to medium usage for this particular knowledge domain is significant and positive for turnover related to innovation (Model 2; β = 1.753, p < .10, SE = .904). Furthermore, when these sources are not used compared to a medium use, it has a negative effect on innovative performance (Model 7; β = -19.545, p < .01, SE = 1.649). This indicates that the market intermediaries’ knowledge domain is an important factor in explaining innovative performance.
In the case of not emphasizing institutions seems to have less effect on innovative performance as compared to emphasizing institutions to a medium degree, thus I find support for Hypothesis 2 which suggests that the usage of the institutional knowledge domain positively influences the firm’s innovative performance, when controlling for the external search strategies used (Model 7; β = -18.098, p < .01, SE = 1.277).
On the other hand, no evidence is found for Hypothesis 3 that suggests that the usage of the specialized knowledge domain positively influences the firm’s innovative performance. Also, the results for Hypothesis 4, concerning the other knowledge domain, do not present a significant relationship with the firm’s innovative performance.
Hypothesis 5 suggests that the usage of the search scope strategy positively influences the firm’s innovative performance. From the results, it can be seen that the effect of this strategy is statistically significant at the 10 percent level (Model 6; β = -.524, p < .10, SE = .280) and at the 1 percent level (Model 7; β = -19.871, p < .01, SE = .864), but it has a negative sign, indicating a reversed effect between the usage of this strategy and innovative performance.
Finally, Hypothesis 6 predicts that the usage of the strategy search depth will positively influence the firm’s innovative performance. This hypothesis finds support; search depth seems to be a great predictor of innovative performance (Model 6; β = .699, p < .10, SE = .358).
T-Test Results
31
Factor
Item I II
1. Suppliers of equipment, materials, components, or software .067 .710
2. Clients or customers .137 .766
3. Competitors .088 .562
4. Consultants .872 .170
5. Commercial laboratories/R&D enterprises .888 .082
Note. Varimax rotations were performed. Each item's highest loading is presented in boldface.
Eigenvalues and percentage of variance accounted for by Factors 1 and 2 were 1.938 (38.77%) and 1.082 (21.64%), respectively.
TABLE 2
Factor Analysis of Institutional Sources
Factor
Item I
1. Universities or higher education institutes .789
2. Government research organizations .883
3. Government research organizations .854
4. Other public sector organizations, like business links, government offices and public research institutes .733
TABLE 3
Factor Analysis of Specialized Sources
Factor
Item I
1. Technical standards .694
2. Health and safety standards and regulations .932
3. Environmental standards and regulations .854
Note. Eigenvalues and percentage of variance accounted for by Factor 1 is 2.079 (69.30%).
TABLE 4
Factor Analysis of Other Sources
Factor
Item I
1. Other professional conferences .766
2. Trade associations .810
3. Technical/trade press, computer databases .648
34 TABLE 5
External Sources of Knowledge for Innovation
Percentages
External Knowledge Domain Not used Low Medium High
Market Players 1.3 25.3 56.0 17.3
Market Intermediairies 24.3 29.7 32.4 13.5
Institutional 20.3 32.4 40.5 6.8
Specialized 21.8 32.1 35.9 10.3
35
TABLE 6
Descriptive Statistics and Spearman Correlation Matrix
36
TABLE 7
Results Ordinal Logistic Regression
Independent Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7
Controls
Internal Knowledge (ref. high) Not used Low Medium . . . . -1.029** (.522) . . . . -.761 (.561) . . . . -1.229** (.556) . . . . -1.033* (.542) . . . . -.835 (.548) . . . . -.508 (.573) . . . . -.382 (.671) Industry Growth .160 (.193) .398* (.230) .268 (.231) .269 (.212) .386* (.211) .362 (.236) .703** (.305) Firm Size (ref. 11-50 employees)
0-10 = 1 51-250 = 3 >250 = 2 -.146 (.600) 1.522 (1.218) -1.358 (.896) .574 (.685) 1.974 (1.327) -.572 (.979) .009 (.616) 1.648 (1.309) -.889 (.948) -.404 (.636) 1.471 (1.222) -1.706 (1.087) .078 (.626) 1.586 (1.244) -1.297 (.914) .190 (.647) 1.565 (1.251) -1.333 (.934) .421 (.823) 1.893 (1.547) -.705 (1.272) Predictors
37
High = 2 .990
(.742)
-.046 (1.875) Market Intermediaries (ref.
medium) Not used = 0 Low = 1 High = 2 .870 (.702) 1.042 (.665) 1.753* (.904) -19.545*** (1.649) .764 (.857) .743 (1.444) Institutional (ref. medium)
Not used = 0 Low = 1 High = 2 1.098 (.708) .802 (.610) .965 (1.226) -18.098*** (1.277) .785 (.777) -.541 (2.183) Specialized (ref. medium)
Not used = 0 Low = 1 High = 2 .772 (.674) .791 (.602) .976 (1.034) -19.686 (.000) .954 (.771) .340 (1.531) Other (ref. medium):
38 Low = 1 High = 2 . .685 (.551) 1.234 (.759) . . . .759 (.723) Search Scope -.524* (.280) -19.871*** (.864) Search Depth .699* (.358) .937 (1.138) Number of cases (N) 82 66 66 69 68 63 63
Model fit sig. .016 .013 .029 .025 .008 .010 .048
Goodness-of-fit Pearson sig. .686 .037 .379 .817 .405 .101 .611
Pseudo R² Nagelkerke .191 .322 .241 .237 .277 .270 .417
Note. Warning for cells with zero frequencies for all models *p < .10, ** p < .05, *** p < .01
Table 8
Results One-Sample T-test
Variable Mean values
Findings and Theoretical Implications
This study was designed to gain a better understanding of the influence of the usage of external knowledge domains and related strategies on the innovative performance of firms. Several findings and theoretical implications have been found.
Firstly, my findings suggest that the usage of knowledge sources of the market players’ domain does not find support in positively influencing the firm’s innovative performance. A possible reason may be that gaining knowledge for innovation from customers, competitors and suppliers can be really hard as it is based on personal contacts (Arundel & Bordoy, 2001). For example, the relation between the firm and its suppliers and customers is a user-producer relationship where there can be negative sides of user–producer interaction in the context of innovation (Laursen, 2011). Lundvall (1988) denotes that this type of interaction can lead to ‘unsatisfactory’ innovations because of inertia in these types of relationships. Similarly, Christensen (1997) argues that when incumbent firms fail as innovators, it is because existing customers keep them captive and make them follow established technological trajectories, even when new and better opportunities emerge. The use of customer knowledge has an important downside as customers may often be conservative, forcing producer firms to search for new solutions along established paths, while shying away from truly new and promising opportunities (Laursen, 2011). In addition, the direct release of information to competitors could directly impact the benefits, i.e. the turnover related to innovation gained from innovation activities negatively (Arundel & Bordoy, 2001).
40
& Duysters, 2010) The idea is that intermediaries whose commercial goal is to bring heterogeneous parties together and co-develop innovations, and not that they just exploit the knowledge (Obstfeld, 2005).
Thirdly, this research demonstrates that universities or higher education institutes, government research organizations, other public sector organizations like business links, government offices and public research institutes (institutional knowledge domain) as information sources are associated with high innovative performance. It confirms the hypothesis stated in this thesis, as institutions may impact innovation output by providing scientific research inputs for innovating firms (Keeble & Wilkinson, 2000) and helping to translate academic knowledge into practical and accessible know-how (Gambarotto & Solari, 2004).
Fourthly, I found that using technical standards, health and safety standards and regulations, and environmental standards and regulations (specialized knowledge domain) as a source of innovation is not significantly linked to higher levels of innovative turnover. A possible reason for the lack of significant effect for specialized knowledge, which in literature also has been termed as regulatory information, is that firms may fail to utilize such knowledge in sufficient quantities, owing to difficulties presented by the technical aspects of the information (Criscuolo, Haskel & Slaugther, 2010). And although the role of regulation in the innovation process is the subject of ongoing debate (Jaffe & Palmer, 1997), still there is little research on the relationship between regulation and innovation (Reichsten & Salter, 2006). More research should be conducted.
41
(Rothaermel & Thursby, 2005). When firms lack the knowledge base to be able to absorb the required knowledge, the salience of such a souce of innovation for innovative firms can be limited (Laursen & Salter, 2004; Kirkels & Duysters, 2010). So firms should try to increase their absorptive capacity. I should note that, according to the results of the factor analyses the source ‘fairs/exhibitions’, by literature designated as a component of the other knowledge domain (e.g. Laursen & Salter, 2006), was left out of this study.
Furthermore, there was no support for the expectation that the use of a broad search trajectory positively impacts the firm’s innovative performance, whereas the use of a deep search trajectory did seem to yield higher innovation performance. Previous research indicated that the external search scope of knowledge sources may be subject to diminishing marginal returns and that oversearch might also hinder innovative performance (Classen, Van Gils, Bammens, & Carree, 2012). In other words, the higher costs and complexity of simultaneously managing multiple partnerships can lead to diseconomies for firms with constrained managerial resources (Belderbos et al., 2006; Lampert & Semadeni, 2010). Thus, firms may reach a tipping point (Laursen & Salter, 2006) where the use of an additional source actually decreases innovation performance. Therewithal, it could be the case that firms compensate for a lack of search scope by a higher search depth (the repeated use of a single knowledge domain). As other scholars pointed out, firms may face a tradeoff between scope and depth in their innovation objectives (Nelson & Winter, 1982; Cohen & Levinthal, 1990; Helfat, 1994; Classen et al., 2012).
42
resources they require to innovate effectively (Jacobs, 2007). Compared to large firms, these firms seem to have greater financial constraints, more manpower bottlenecks in terms of too few or inadequately qualified personnel, and often do not have other products to compensate for a lack of sales and profits (Kaufmann & Tödtling, 2002; Street & Cameron, 2007). Financial problems are encountered when the firm’s innovative activities are constrained by high costs, the risk of innovation and/or a lack of sources of finance, whereas resource problems are encountered when the firm is constrained by a lack of qualified personnel, information on the markets and technological knowledge (Paananen, 2012). This rationale is in accordance with the Schumpeterian view, where it is assumed that big firms have the resources and possess a monopolistic power that enables them to face the inherent risk of innovation (Van de Vrande, De Jong, Vanhaverbeke, & De Rochemont, 2009). However, also prior empirical evidence does not confirm the role of the size of the firm in relation to innovation and there is support that it is plausible that also big firms have rigidities in introducing novelty (Caloghirou, Kastelli & Tsakanikas, 2004). Problems with statistics, sectorial specificities or technological characteristics of innovation interfere and make the relation between size and innovation much more complex (Freeman & Soete, 1997).
43
collaborations with external actors (Zhang & Li, 2010). Maintaining these deep links requires relatively a lot of resources and attention (Laursen & Salter, 2006), and therefore firm seem to focus on these sources. The knowledge domains specialized, institutional and market intermediaries follow the other and market players domains in the relative importance for search depth, respectively.
Contributions
A major contribution of this study is that it distinguishes among sources of knowledge by assigning them to knowledge domains The measures created can reflect the picture of the innovating firm’s interacting with a network of multiple different actors – such as clients, suppliers, and a wide range of institutions and at the same time reduce the data scholars need to acquire. Furthermore, this thesis specifies the relationship between the knowledge domains and related search strategies with innovative performance. Additionally, the importance per knowledge domain is researched. The final contribution is the increased generalizability it creates for some recent papers on the same subject (e.g. Laursen & Salter, 2006; Chen et al., 2011).
Limitations and Future Research
44
45
6 REFERENCES
Ahuja G. 2000. Collaboration networks, structural holes, and innovation: a longitudinal study. Administrative Science Quarterly, 45: 425–455.
Ahuja, G., & Lampert, C. M. 2001. Entrepreneurship in the large corporation: a longitudinal study of how established firms create breakthrough inventions, Strategic Management Journal, 22, 521–543.
Arundel, A., & Bordoy, C. 2001. The relative value of internal and external information sources to innovation. Paper presented at the final KNOW workshop in Athens, May 26-27.
Arundel, A., & Bordoy, C. 2006. Collaboration & innovation outputs. In . Caloghirou, A. Constantelou & N. Vonortas N. (Ed.), Knowledge flows in European industry: 158 – 182. Oxon: Routledge, Taylor and Francis Group.
Autio, E., Hameri, A. P., & Vuola, O. 2004. A framework of industrial knowledge spillovers inbig-science centers, Research Policy, 33, 107–126.
Bagozzi, R. P., Yi, Y., & Phillips, L. W. 1991. Assessing construct validity in organizational research. Administrative Science Quarterly, 36: 421–458.
Banbury, C. M., & Mitchell, W. 1995. The effect of introducing important incremental innovations on market share and business survival. Strategic Management Journal, 16: 161-182.
Belderbos, R., Carree, M., Diederen, B., Lokshin, B., & Veugelers, R. 2004. Heterogeneity in R&D cooperation strategies. International Journal of Industrial Organization, 22: 1237– 1263.
46
sources: Complements or substitutes for innovative performance? Technovation, 24: 29-39.
Capaldo, A., & Petruzzelli, A. M. 2011. In search of alliance-level relational capabilities:
Balancing innovation value creation and appropriability in R&D alliances. Scandinavian Journal of Management, 27: 273—286.
Chandrashekaran, M., Mehta, R., Chandrashekaran, R., & Grewal, R. 1999. Market motives, distinctive capabilities, and domestic inertia: A hybrid model of innovation generation. Journal of Marketing Research, 36 (1): 95-113.
Chang, C., & Huang, H. 2007. Determinants of incremental innovation: A process view
incorporating theories of transaction cost, social exchange and embeddedness. NTU Management Review, 18 (1): 81-116.
Chen, J., Chen, Y., & Vanhaverbeke, W. 2011. The influence of scope, depth, and orientation of external technology sources on the innovative performance of Chinese firms. Technovation, 31: 362–373.
Chesbrough, H. 2003. Open Innovation. Cambridge, MA: Harvard University Press.
Chesbrough, H., Vanhaverbeke, W., & West, J. 2006. Open Innovation: Researching a new paradigm. Oxford: University Press.
Chiang, Y., & Hung, K. 2010. Exploring open search strategies and perceived innovation performance from the perspective of inter-organizational knowledge flows. R&D Management, 40 (3): 292–299.
47
Classen, N., Van Gils, A., Bammens, Y., & Carree, M. 2012. Accessing resources from innovation partners: The search breadth of family SMEs. Journal of Small Business Management, 50 (2): 191–215.
Cohen, W. M., & Levinthal, D. A. 1990. Absorptive capacity: a new perspective on learning and Innovation. Administrative Science Quarterly, 35: 128–152.
Community Innovation Statistics. 2006. From today’s Community Innovation Surveys to better surveys tomorrow, 1-22.
Criscuolo, C., Haskel, J. E., & Slaughter, M. J. 2010. Global engagement and the innovation activities of firms. International Journal of Industrial Organization, 28 (2): 191- 202.
Das, T., & Teng, B. 2000. A resource-based theory of strategic alliances. Journal of Management, 26, 31–61.
Davis, J. P., & Eisenhardt, K. M. 2011. Rotating leadership and collaborative innovation: Recombination processes in symbiotic relationships. Administrative Science Quarterly, 56 (2): 159-201.
De Campos, A. T. 2009. Patterns of networks and innovation in Spanish service firms. Paper presented at the summer conference on Copenhagen Business School, Frederiksberg, Denmark.
Dundon, E. 2002. The seeds of innovation: cultivating the synergy that fosters new ideas. Toronto: Amacom.
48
Fernhaber, S. A., McDougall-Covin, P. P., & Shepherd, D. A. 2009. International entrepreneurship: Leveraging internal and external knowledge sources. Strategic Entrepreneurship Journal, 3: 297–320.
Fleming, L., & Sorenson, O. 2004. Science as a map in technological search. Strategic Management Journal, 25, 909-928.
Fontana, R., Geuna, A., & Matt, M. 2006. Factors affecting university–industry R&D projects: The importance of searching, screening and signaling. Research Policy, 35: 309–323.
Forsman, H. 2009. Improving innovation capabilities of small enterprises: Cluster strategy as a tool. International Journal of Innovation Management, 13(2): 221–243.
Freel, M. 2000. External linkages and product innovation in small manufacturing firms. Entrepreneurship & Regional Development, 12: 245-266.
Freeman, C., & Soete, L. 1997. The Economics of Industrial Innovation (3rd ed.). Cambridge, MA: The MIT Press.
Gambarotto, F., & Solari, S. 2004. The role of reconnection of competencies and institutions in the collective learning process. Presented at Department of Economics, University of Padova, Italy.
Gassmann, O., & Enkel, E. 2004. Towards a theory of open innovation: Three core process archetypes. Proceedings of the R&D Management Conference, Lisbon, Portugal.
Granado, M. M., Lopez, R. P., & Sanz, Y. R. 2011. Institutional spillovers, public financial support and R&D cooperation strategies in Spanish firms. Paper presented at the Simposio de la Asociación Española de Economía, Universidad de Málaga, Spain.
