Collaborative Innovation: How Does Organizational Structure Impact Firms’ Innovative Performance?

(1)

Collaborative Innovation: How Does Organizational Structure

Impact Firms’ Innovative Performance?

Author: Ismail Karlik August 18, 2016 Word count: 9402

(2)

Collaborative Innovation: How Does Organizational Structure Impact

Firms’ Innovative Performance?

University of Groningen

Faculty of Economics and Business

Master Thesis MSc Strategic Innovation Management Innovation Management & Organization

August 18, 2016 Ismail Karlik Aquamarijnstaat 62 9743RB Groningen i.karlik@hotmail.com Student number: 1911600

Supervisor: Prof. Dr. D.L.M. (Dries) Faems Second supervisor: Drs. H.J. (Holmer) Kok

Abstract: It is known that innovation is essential for growth and survival of an

organization in the long run. Although the importance of innovation is clear, managing innovation is not an effortless practice as it involves the process of organizational learning. This study empirically tests the impact of organizational structure in collaborative settings to identify the appropriate innovation strategy. Based on a sample of 356 SME’s, this study has found evidence that is partially in line with the hypotheses outline, suggesting that, (i) explorative collaboration has a positive significant effect on radical innovative performance, while exploitative collaboration has a positive significant effect on incremental innovative performance, and (ii) informal organizational structure strengthens the effect of innovation objectives aimed at creating new technologies and products through explorative collaboration.

(3)

1. Introduction

Innovation is known to be one of the most critical operators for gaining competitive advantage; likewise, innovating firms supply far more than just profits for their owners, as innovation implies economic growth, employment, and significant improvements to people’s lives (Acemoglu and Robinson, 2012; Baumol and Strom, 2007; Christensen and Raynor, 2003). This is also in line with the notion that innovation is essential for growth and survival of an organization in the long run (Baumol, 2002; Schumpeter, 1939). Although the importance of innovation is clear, managing innovation is not an effortless practice (Andriopoulos and Lewis 2010; Van de Ven et al., 1999).

The intricacy of managing innovation mainly stems from the need of firms to employ both exploitation and exploration (Andriopoulos and Lewis 2009; March’s, 1991), which can be achieved by ambidexterity; the capability to master both of the features (Birkinshaw and Gibson, 2004; Lin et al., 2012). Many scholars have emphasized the notion of ambidextrous organizations (Benner and Tushman, 2003; O’Reilly and Tushman, 2004) to deal with the contradictory demands of exploitation versus exploration (Andriopoulos and Lewis 2010; March, 1991). To satisfy the exploration-oriented objectives, Christensen and Overdorf (2000) illustrated the combination of exploitation activities, by new structures in the organization, spinouts and acquisitions, which is also perceived as a crucial part of an innovation strategy.

A highly relevant trend in dealing with these tensions is the interorganizational collaborative arrangements. Strategic alliance is one of the important ways to bridge the gap between short and long-term adapted developments (Brown and Eisenhardt, 1997; Todeva and Knoke 2005). Interorganizational collaboration is widely accepted as an important way of approaching innovation complementing the internal actions of organizations (Deeds and Rothaermel, 2003; Hagedoorn, 2002). Past literature reveals that collaboration within the organization with a variety of partners can improve their innovative capabilities (Santoro, 2000).

(5)

relationship between exploitative and explorative learning condition, it is widely accepted that a balance between the two is necessary to achieve continuing success (Gupta et al., 2006; Levinthal and March, 1993; March, 1991) and that they can be accomplished simultaneously (Auh and Menguc, 2005).

However, literature to date does not provide management with a proper understanding of how organizational structure in collaborative innovation can be beneficial in terms of innovative performance. This is highly relevant in terms of having the right type of alliance, as partners bring together complementary resources that can lead to synergistic value creation. Hence, twofold gaps in the literature are identified and analyzed in this research: (i) intra-firm innovation literature pays a lot of attention for the need of different management approaches for different innovation outcomes but ignores the option of external collaboration (ii) external collaboration literature emphasizes added value of collaboration but ignores firms’ internal design.

Based on the literature gap, this article aims to identify what effect different organizational structures have on interorganizational collaboration. This formulates the question, how firms committed to interorganizational collaborative innovation perform better in terms of innovative achievement. To be more specific, this study looks into organizational structure (formal or informal) as a moderator for interorganizational collaboration (explorative or exploitative). Empirical analysis is executed, using the data, collected from the firms located in the three provinces of the Northern Netherlands (Groningen, Friesland, and Drenthe) in collaboration with SNN (samenwerkingsverband Noord Nederland) and the Dutch Ministry of Economic Affairs. The survey is mainly based on the guidelines of the Community Innovation Survey (CIS), that is organized by Eurostat (affiliate of European Commission) and executed bi-anually. The aim of this survey is to obtain insight into the innovation practices and performance of companies, within the various EU member states (Eurostat, 2016). Methods section will elaborate more on the selected sample. Before descripting the methods in details, the literature section will look into empirical findings on the role of interorganizational collaboration and organizational structure on innovative performance, which will be followed by formulation of the hypothesis. An finally, after the methods section, results will be discussed and conclusion drawn, that will lead to the formulation of the implications.

(6)

need to clarify their objective in developing improved and/or new technologies and products. Second, firms that aim at improving radical innovative performance should seek for explorative collaboration, while those that aim at improving incremental innovative performance should seek exploitative collaboration. Third, firms aiming at radical innovation can benefit from explorative collaboration even more by combining it with informal organizational structure. To be more detailed, explorative collaboration (i.e. multidisciplinary innovation networks and living labs) in combination with; (i) loose informal control mechanism with support of informal relations and collaboration, (ii) strong emphasis on possibility to deviate from formal procedures when needed, and (iii) strong emphasis on adjusting according to the circumstances even if deviating from management principles, is even more effective in further increasing the radical innovative performance.

1.1 Research Questions

Based on the identified literature gap and the goals of my research, following research questions have been formulated:

RQ1: To what extend does formal organization structure affect the relationship between

exploitative interorganizational collaboration and incremental innovative performance?

RQ2: To what extend does informal organization structure affect the relationship

between explorative interorganizational collaboration and radical innovative performance?

2. Literature Review

2.1 The importance of ambidexterity

Making the right decision when investing in different types of activities, has been one of the main concerns of corporate strategy. There are two broad types of distinctive learning activites as provided by literature, exploration and exploitation, between which businesses share their attention and resources. The ability to involve in both is called ambidexterity, in which exploration includes activities such as searching, innovation, experimentation, discovery and risk taking, while exploitation is more characterized by efficiency, implementation, and refinement (Bailey et al., 2014; Cheng and Van de Ven 1996, March 1991).

(7)

economics (Ghemawat and Ricarti Costa 1993), strategic management (Winter and Szulanski 2001), and organization theory (Andriopoulos and Lewis 2010; Holmqvist 2004, Van den Bosch et al. 1999). It has become clear that these two activities demand substantially different processes, strategies, structures, capabilities, and cultures that is expected to impact firm performance in various ways.

To be more detailed, informal organizational structure, autonomy and chaos, loosely coupled systems, improvisation, and path breaking are more associated with exploration, while mechanistic structures, routinization, control and bureaucracy, tightly coupled systems and path dependence are associated with exploitation (Ancona et al. 2001). The financial returns, correlated with both activities also differ accordingly as exploration is perceived to have variable returns and distant in time, whereas exploitation is correlated with more certain returns and closer in time. This implies that businesses involving in explorative activities experience larger performance variation, while business with exploitation is more likely to experience more stable performance.

To conclude, the value of ambidexterity lies on the necessity of involving in conflicting tasks (e.g., differentiation versus low-cost production, investment in present versus prospective projects), in which there are always trade-offs to be made. Although it would be utopic to eliminate these trade-offs completely, most capable firms reconcile them to a great extent, and increase their abiding competitiveness (Bailey et al., 2014; Gibson and Birkinshaw, 2004). To maintain competitive advantage, firms are bound to manage both evolutionary and revolutionary changes (Tushman and O’Reilly, 1996).

2.2 Internal management approach & ambidextrous innovation

Literature on organizational theory has many arguments for building ambidextrous organization, including organizational structure (Raisch et al., 2009; Tushman and O’Reilly, 1996), trust in affairs with management and worker training (Adler and colleagues, 1999), shared vision, and recruitment and training (And Bartlett and Ghoshal, 1989). Although all elements are part of the story Gibson and Birkinshaw (2004) claim that ambidexterity in organizational arrangement is accomplished by “developing structural mechanisms to cope with the competing demands faced by the organization for alignment and adaptability” (p. 211).

(8)

innovation best. Brown and Eisenhardt (1997) introduces the semistructures that allow organizational units to change among the conflicting requirements of exploration and exploitation. The combination of organic and mechanistic structural elements into complex structures was illustrated by Adler and Borys (1996) and Sheremata (2000). While exploitative units are typically larger with tight processes and more centralization, explorative units, on the contrary, are usually relatively small with loose processes and decentralized (Andriopoulos & Lewis 2009; Benner & Tushman, 2003).

2.3 External collaboration & ambidextrous innovation

As mentioned in the prior discussion, past literature stressed the importance of internal resources for competitive advantages (Wernerfelt 1984). However, Conner (1991) suggested that firms employ rent-creating resources within both internal and external constraints. Resources superior or distinctive to those of rivals, may become the basis for competitive advantage when matched appropriately to environmental opportunities (Stettner and Lavie, 2013; Thompson and Strickland, 1990). In times of rapid change, knowledge is so broadly dispersed that a firm cannot have all internal capabilities needed for success. In such an environment internorganizational collaboration comes into play.

Interorganizational collaboration is widely accepted as an important way of approaching innovation (Baldwin and Hippel, 2011; Deeds and Rothaermel, 2003; Nieto and Santamaria, 2007). From an organizational leaning perspective, improvement in innovative capabilities can be achieved through interorganizational collaborations with a variety of partners, such as end users, universities, suppliers, and even competitors (Jiang et al., 2010; Santoro, 2000).

(9)

Organizational Structure Radical Innovative Performance H1 Exploratory Collaboration Exploitative Collaboration H2 Informal Formal Incremental Innovative Performance

External Collaboration Internal Management Performance

adaptation and flexibility (Doz, 1996), unintended knowledge spillovers (Teece, 2002), and different opinions for intended benefits (Lorange and Roos, 1992; Todeva and Knoke, 2005). Previous research has shown empirical evidence that there is a positive relation between interorganizational collaboration and innovative performance, despite these high failure rates (Im and Rai, 2008; Rogers, 2004).

3. Grounding Hypothesis

As the previous section has depicted already, research on internal management ignores the relevance of external collaboration for stimulating ambidextrous innovation while research on external collaboration ignores the importance of internal management. In order to achieve a desired innovative output, firms’ innovation strategy should look into both perspectives in which organizational learning remains central (Argote, 2012; Malhorta et al., 2005; Zahra and George, 2002; Santos-Vijande et al., 2012). In other words, an important part of the innovation strategy is to align interorganizational collaboration types (explorative/exploitative), organizational structure (formal/informal), and innovative performance (radical/incremental) (Klingebiel and Rammer, 2014). Figure 1 depicts the conceptual model in which the moderating role of organizational structure is illustrated.

(10)

Interorganizational collaboration is widely accepted as an important way of approaching innovation complementing the internal actions of organizations (Deeds and Rothaermel, 2003; Hagedoorn, 2002). However, it is important to know with which organizations to collaborate for different innovation purposes. The contrast of explorative and exploitative activities (March’s, 1991) has been distinguished by past research into explorative and exploitative collaboration (Rothaermel and Deeds, 2004). While explorative collaboration acquire discovery as their main driver, joint maximization of complementary assets has been the main aim of exploitative alliances (Koza and Lewin, 1998, p. 257). What is learned is profoundly linked to the conditions under which it is learned (Brown and Duguid, 1991). Supporting this claim, the form of interorganizational collaboration is alleged to be altered according to the specific knowledge to be exchanged, depending on the type of innovative performance (Argote and Miron-Spektor, 2011; Parkhe, 1993). Presented this way, the decision to cooperate is closely related to the make-or-buy decision. Here, it is understood that in order to acquire resources and skills that are not available internally, firms can consider turning to collaboration for both explorative and exploitative efforts.

Results indicate that explorative and exploitative collaborations have different effects on innovative performance, as the type of organization(s) to collaborate with differs as well (Nieto and Santamaria, 2007; Faems et. al, 2005; Von Hippel et al., 1999).

(11)

Geenhuizen, 2011) are appropriate methods to foster radical innovative output. As a consequence, the extent to which organizations tap into new knowledge is dependent on their participation in such activities (Levinthal and March, 1994), since exchanging know-how mainly requires the establishment of relationships in which exchange occurs within a learned and shared code (Kedia and Mooty 2013).

On the other hand, Bogers et al. (2010) suggests that ‘co-creation’ can be effectively used to improve products and better fit the needs of end users. In other words, producers focus on existing capabilities to create value by solving consumer needs (Prahalad and Ramaswamy, 2004). Hence, exploitative interorganizational collaboration, in the form of co-creation (Bogers et al., 2010; Prahalad and Ramaswamy, 2004) is an appropriate method to foster incremental innovative output.

This section will extend this reasoning, by looking at the moderating effect of organizational structure to consequently formulate the hypotheses about the innovation strategies synergizing external collaboration and internal management.

3.1 The moderating role of informal organizational structure

(12)

H1: Explorative interorganizational collaborations impact radical innovative performance

more positively in firms with informal organizational structure than in firms with formal organizational structure.

3.2 The moderating role of formal organizational structure

While explorative collaboration relies more on personal and informal modes of coordination and control to adopt new knowledge, Popadiuk and Choo (2006) suggest that exploitative collaboration aims at leveraging existing skills, clearly benefiting from measurable performance objectives monitored by formalized control mechanism. To be more specific, these are characterized by formal organizational structures with formalized work processes, centralized procedures, and clear job responsibilities. Hence, it is argued that firms with formal structures suit collaborative innovation initiatives that focus on efficiency rather than novelty (Christensen and Overdorf, 2000; Norman and Verganti, 2014; Oke, 2007). Drawing on this perspective, I argue that formal organizational structure strengthens the positive effect of exploitative interorganizational collaboration on innovation objectives aimed at improvement and further development of existing technologies and products. Therefore, the more formal a firms’ structure, the more incremental its innovative performance will be in a collaborative setting.

H2: Exploitative interorganizational collaborations impact incremental innovative

performance more positively in firms with formal organizational structure than in firms with informal organizational structure.

4. Methodology

(13)

interpretation of results with the comparison to hypotheses, reaching conclusions, providing implications.

4.1 Data collection and SNN

The Northern Netherlands Provinces alliance, known in Dutch as samenwerkingsverband Noord-Nederland (SNN), is an organization aiming at strengthening the Northern economy by working together with businesses, knowledge institutions, public organizations and other government agencies. The main objective is to create a valuable network fostering innovation. As SNN is a designated managing authority for EU regional development funding, the organization is also responsible for providing subsidies to spur regional innovation (SNN, 2016).

At the beginning of 2016, 2701 SMEs located in the North of the Netherlands (Groningen, Friesland, and Drenthe) were invited to participate in the ‘Benchmarking innovation in the North of the Netherlands project’. Specifically, these were the SME’s who previously applied for a subsidy at the SNN in the period 2010-2015. Furthermore, to minimalize selection problems, an additional 294 Northern-Netherlands SMEs have been approached who didn’t apply for a subsidy at the SNN. These contact details and financial information of these firms were retrieved from the ORBIS database. Substantially, a total of 432 Northern-Netherlands SMEs (RR: 14,4%) have filled in the online-survey, of which this research is based on.

4.2 Measures

(14)

4.2.1 Measurement of variables

This study aims to highlight whether organizational structure increases the positive effect of collaboration type on innovative performance. A distinction is made between radical and incremental innovative performance in order to draw a more complete picture of the phenomenon for strategic decision making. Furthermore, what this particular research is interested in, the moderating role of organizational structure, is divided between formal and informal organizational structure. An overview of the variables that are used to test this relationship are shown in table 1, and explained more in depth below. Furthermore, list of corresponding survey questions can be found under appendix 1.

Table 1. Variables used in this study

Measurement of variables

Description Notation

Dependent

Revenue (New products) Revenue (Improved products)

Radical innovative performance measured by the revenue obtained from products/services that are new to the market.

Incremental innovative performance measured by the revenue obtained from products/services that are new to the firm.

Percentage Percentage Main Informality Explorative collaboration Exploitative collaboration

Degree of formality/informality within the firm measured from three angles; control mechanisms, business procedures, and management principles.

Degree of collaboration the firm is engaged, in the form of multidisciplinary innovation-networks, and living lab.

Degree of collaboration the firm is engaged, in the form of co-creation with end users.

Categorical 1-5 Categorical 1-4 Categorical 1-4 Control Age Size R&D intensity Patent Secrecy Age of firm

Number of FTE’s (Full time equivalent)

Ratio of the total revenue spent on research and development Whether the firm has obtained patents

Whether the firm makes use of secrecy

Range Range 0-250 Percentage Binary Binary

(15)

one measuring the overall innovative performance (new + improved). The analyses uses the proportion of the revenue for income streams that belongs to the earlier mentioned business activities.

Indicators of interorganizational collaboration. Within interorganizational collaboration two collaboration types, exploitative and explorative, is distinguished. In order to measure these two collaboration types the survey is designed illustrating this distinction. Respondents that indicated to have engaged in co-creation with end users are labelled as ‘exploitative collaboration’ (Bogers et al. 2010), forming a knowledge leveraging alliance (March, 1991). On the other hand, firms that have engaged in interorganizational collaboration within multidisciplinary innovation networks (Ferrary and Granovetter 2009; Nieto and Santamaria, 2007), and living labs (Kusiak, 2007; Möller, 2010) are labelled as ‘explorative collaboration’, forming a knowledge-generating R&D alliance (March, 1991). Explorative collaboration is measured by combining the two questions, concerning multi-disciplinary innovation networks and living labs, rated on a 5-point Likert scale illustrating the degree of involvement in such collaboration type.

Indicators of organizational structure. To measure the effect of organizational structure on collaborative innovation additional questions have been developed into the CIS survey as it currently has not integrated such format yet. This makes a distinction of formal and informal firm structure engaging in collaborative innovation, and provide new insights in its effect on innovative performance. When measuring organization structure, three questions have been used to be rated on a 5-point Likert scale ranging from formal (1) to informal (5). The questions are formulated in a way to measure the organizational structure from three different angles; (i) control mechanisms (iii) business procedures, and (iii) management principles.

(16)

4.2.2 Interaction effect

In order to test the moderating effect of organizational structure variable on collaborative innovation, degree of formality and informality have been tested for potential interaction effects. Hereby, the moderator variable and predictors are multiplied to create the interaction term and to consequently measure the possible interaction effects. In order to identify correlating variables, a correlation matrix is looked at, which can be used to identify potential interaction effects. As our study is interested in moderation of organizational structure, we look into the relevant correlations only which are dealt with under the results section.

4.2.3 Validation of measures

A confirmatory factor analysis is conducted to establish convergent validity of the constructs. Streiner (2003) suggests that, “one of the central tenets of classical test theory is that scales should have a high degree of internal consistency, as evidenced by Cronbach’s alpha” (p. 217). This section depict the results of various methods aimed at testing the internal consistency of the variables measuring ‘organizational structure’ and ’explorative collaboration’. As results indicate in the internal consistency statistics (table 2), the variables are likely to factor well, as the Kaiser-Meyer-Olkin measure of sampling adequacy (.722) and the Bartlett’s Test of Sphericity (p < .01) confirm that factor analysis is appropriate for organizational structure. Same argument also holds for explorative collaboration with KMO statistic of ,654. Furthermore, it is also illustrated that Cronbach’s Alpha for organizational structure and explorative collaboration is estimated at ,843 and ,912, respectively. This means that 84,3% and 91,2% of the variance in the composite score for our variables is considered the internally consistent reliable variance (true score variance).

Table 2. Internal Consistency Reliability

Organizational Structure

Cronbach's Alpha ,843

Kaiser-Meyer-Olkin Measure of Sampling Adequacy ,722

Bartlett's Test of Sphericity Approx. Chi-Square 477,515

df 3

Sig. ,000

Explorative Collaboration

Cronbach's Alpha ,912

Kaiser-Meyer-Olkin Measure of Sampling Adequacy ,657

Bartlett's Test of Sphericity Approx. Chi-Square 426,983

df 2

(17)

Looking more closely at the output, it is observed that the inter-item correlation matrix illustrates correlation at the item level at around 0.6 range, implying that the data measures the same phenomenon. This suggests, we cannot speak of singularity (perfect correlation) or multicollinearity, as all correlations are below 0.8. This also explains the internal consistence reliability of 0.843.

4.2.4 Missing values

Although items are selected carefully, our sample size still reflects many missing values. In order to identify whether missing data points can be replaced with expected values, it needs to be verified that the missing values are missing at random. Hereby, Little’s MCAR (Missing Completely at Random) test is executed, to identify whether there is no systematic pattern that makes some data more likely to be missing than others. Anything less than 2% is fairly negligible, however, both dependent variables indicating innovation performance have 22,4%, 61 cases, missing values. This can be corrected by mean imputation through EM (Expectation Maximization) algorithm, which is a method to increase the cases while not affecting the mean. However, the test illustrates that the missing values are missing not at random (MNAR), p > 0.05. The attempt to improve the model performance by including a much larger number of cases has failed, which implies that the missing values could not be replaced with predicted values, maximum likelihood imputation or list wise deleting (Myers, 2011). Although not much could be done to increase the number of cases drastically, it is identified that some values were missing as a consequence of some sort of non-response. E.g. many respondents who have not been involved in R&D practices at all, did not indicate the proportion of their revenue allocated to R&D. By transforming those missing values, more cases got included into the analysis and were qualified for further interpretations.

4.2.5 Recoding variables

(18)

transformed in the analyses by taking the natural logarithm to account for a skewed distribution. Finally, in order to make the independent variables compatible, the moderator and predictor variables are standardize by means of mean-centering.

4.3 Research design & research method

This study aims to highlight the effect of the type of organizational structure on innovative performance. For this study we follow Van Aken, Berends and Van der Bij (2012), who suggest theory testing approach is suitable if the already existing literature are elaborated and not scattered, and yet lacks theoretical explanations which is the gap in the literature. As this is the case for collaborative innovation this paper follows a theory testing approach proposed by Van Aken, Berends and Van der Bij (2012).

With respect to the nature of the study multiple regression analysis is performed to test the hypotheses. Lewis (2007) suggests a hierarchical regression to determine the quality of predictors in multiple regression analysis, as the predictors of this research are based on past research. Additional, hierarchical regression is more appropriate for social sciences research due to high correlation between predictors.

5. Results

5.1 Descriptive Statistics

This section will discuss the statistics that are used to describe the basic features of the data of this study. General descriptive statistics is illustrated under descriptive statistics & correlations (table 3). The 432 responses in the database are not reflecting the final sample yet, as not all respondents filled in the necessary questions to perform this analysis and other faulty cases are detected. Several conditions have to be satisfied in order to obtain a clean sample size. Due to the nature of this research, a filter variable for organizational structure and innovative performance is necessary. Furthermore, I narrow down the responses by applying a filter for SMEs, in which I exclude cases that have that have more than 250 employees.

(19)

(20)

Table 3. Descriptive Statistics & Correlations Correlations Variables N Mean SD Turnover (New products) Turnover (Improved

products) Age Size

R&D

intensity Patent Secrecy Informality

Explorative Collaboration

Exploitative Collaboration Turnover (New products)

Turnover (Improved products) Age Size R&D intensity Patent Secrecy Informality Explorative Collaboration Exploitative Collaboration 421 19,4573 1,71661 1 421 12,4769 1,60241 ,318*** 1 423 26,1584 ,96110 -,126*** -,015 1 378 46,6905 1,44444 ,069* ,160*** ,511*** 1 385 14,5299 1,55567 ,452*** ,128*** -,272*** -,054 1 421 ,1178 ,32272 ,219*** ,031 -,054 ,112** ,233*** 1 421 ,3741 ,48446 ,412*** ,203*** -,073 ,153*** ,429*** ,369*** 1 367 3,6297 1,54342 -,143*** -,079* ,226*** ,229*** -,292*** ,041 -,141*** 1 380 3,1031 1,10250 -,246*** -,102*** ,190 -,014 -,609*** -,131** -,387*** ,167*** 1 380 _{3,1655 1,57242} _-,277*** _-,167** _,130*** _-,063 _{-,567*** -,096** -,440***} _,175*** _,846*** ₁ Valid N (listwise): 356

(21)

5.2 Main & Interaction effect: Estimation and testing for assumptions

As mentioned earlier, this analysis makes use of multiple regression analysis which has been considered appropriate looking at the nature of our research. However, before doing so, several tests have been run to decide whether all assumptions for a multiple regression analysis are met.

First of all, it can be said that the assumption for the independence of observations has been met, as the design of the study shows that each respondent is only measured once. Furthermore, the design of this study illustrates that the observations on each variable to be independent of every other.

Also, the assumption of multicollinearity is tested by examining Variance Inflation Factors (VIF's). It appeared that the predictor had a high VIF’s score, which violated one of the assumption for linear regression. In this study any score below 2 has been considered acceptable. In order to correct for the high VIF’s score, both moderators and predictors are standardized. After mean centering the interval variables, moderator and predictor is multiplied to create the interaction term and to consequently measure the main and interaction effects without violations.

Furthermore, in figure 2, histograms of the standard residuals are examined to determine whether there is any violation of assumption with respect to normality. Although the shape of the residual distribution are slightly skewed for both dependent variables, overall the histograms don’t show major skewness.

Figure 2. Distributions of standardized residuals of Models 1 and 2.

(22)

Shapiro-Wilk (S-W) tests (table 4). Results for both models rejected the null hypothesis that residuals are normally distributed (p <.01).

Table 4. Normality of residuals

Kolmogorov - Smirnov Shapiro - Wilk

Statistic df p-value Statistic df p-value

Model 1 0.181 356 0.000 0.817 356 0.000

Model 2 0.219 356 0.000 0.757 356 0.000

However, Field (2013) argue that the results of OLS regressions should not be affected due to our sample size being sufficiently large (356 cases). Furthermore, both Q-Q plot and P-P plot, show no major violation of the normality assumption.

As the assumptions are tested the OLS estimation test has been run. The models 1 (radical innovative performance) and 2 (incremental innovative performance) for regression analyses are tested following the sequence illustrated under tables 5 and 6, which will be interpreted simultaneously.

Looking at the F-test of the overall significance, all models under the tables 5 and 6 are found to be highly significant (P<.01). Furthermore, we look at the best fit of the model which is explained by R square and R square adjusted. The R square adjusted only increases if new variables added to the model improves the model, and decreases with poor quality predictors. Model 1.c provided the best fit compared to the other models since R squared adjusted presents the highest value (.268) at the significance level of 1 percent. In other words, our first model represents 26.8% of the variation in radical innovative performance. In the second model, model 2b provided the best fit compared to the other models since R squared adjusted presents the highest value (.062). This implies that our second model represents 6.2% of the variation in incremental innovative performance. Although 1.c and 2.c are the most important models including the interaction effect, which is the starting point of this research, other models will also be looked at for additional insights.

(23)

show that R&D has only a positive significant effect on radical innovative performance (,432; p<,01). Furthermore, secrecy has a positive significant effect on both radical innovative performance (,852; p<,01) and incremental innovative performance (,404; p<,01). Our last control variables, age and patent, are not significant in any of the models. Observing the main effects, informality as such has no significant effect on neither radical nor incremental innovation performance. Furthermore, explorative collaboration is observed to have a positive significant impact on radical innovative performance (,231; p<,05), while exploitative collaboration is observed to have a positive significant impact on incremental innovative performance (,194; p<,05). Looking at the interaction effects, we can see that only explore*informality has a positive significant impact on radical innovative performance (,074; p<,01), while exploit*informality has no significant effect on incremental innovative performance. Table 5. Regression analysis: Radical Innovative Performance as dependent

Radical Innovative Performance

Model 1.a Model 1.b Model 1.c

Variables B SE B SE B SE Intercept ,946*** ,288 ,893*** ,295 ,810*** ,294 Control variables Age -,100 ,100 -,112 ,100 -,120 ,099 Size ,091 ,065 ,098* ,066 ,113** ,066 R&D intensity ,372*** ,058 ,420*** ,070 ,432*** ,069 Patent ,116 ,252 ,119 ,257 ,158 ,255 Secrecy ,821*** ,187 ,826*** ,195 ,852*** ,193 Main effects Informality -,008 ,029 -,011 ,029 Explorative Collaboration ,231** ,141 ,307** ,142 Exploitative Collaboration -,078 ,098 -,090 ,097 Interaction effects Explore*Informality ,074*** ,027 Model parameters R2 ,264 ,271 ,287 Adjusted R2 ,254*** ,255*** ,268*** N 356 356 356

(24)

Table 6. Regression analysis: Incremental Innovative Performance as dependent

Incremental Innovative Performance

Model 2.a Model 2.b Model 2.c

Variables B SE B SE B SE Intercept 1,087*** ,303 1,140*** ,309 1,146*** ,311 Control variables Age -,149 ,105 -,151 ,105 -,149 ,105 Size ,214*** ,069 ,221*** ,070 ,219*** ,070 R&D intensity ,068 ,061 ,042 ,073 ,042 ,073 Patent -,370 ,265 -,262 ,270 -,262 ,270 Secrecy ,511*** ,196 ,405** ,204 ,404** ,204 Main effects Informality -,034 ,030 -,034 ,030 Exploitative Collaboration ,194** ,102 ,197** ,103 Explorative Collaboration ,222 ,148 ,221 ,148 Interaction effects Exploit*Informality -,005 ,020 Model parameters R2 ,069 ,083 ,083 Adjusted R2 ,056*** ,062*** ,059*** N 356 356 356

* Correlation is significant at the 0,10 level. ** Correlation is significant at the 0,05 level. *** Correlation is significant at the 0,01 level.

6. Discussion

While previous studies have often analyzed collaborative innovation and internal management separately in effecting innovative performance, this study has looked into it as a part of an innovation strategies of which the moderating role of organizational structure is tested. Hereby, a distinction is made for interorganizational collaboration types (explorative and exploitative), organizational structure (formal/informal, measured by (i) control mechanisms (ii) business procedures, and (iii) management principles), and finally innovative performance (radical/incremental). The ultimate purpose of this study was to investigate the impact of organizational structure in collaborative settings to identify the appropriate innovation strategy. The findings of this research show several interesting insights that might be useful for academia and practitioners.

(25)

outcomes, claiming that explorative interorganizational collaboration support innovation objectives to create new technologies and products, while exploitative interorganizational collaboration support the innovation objectives to improve and further develop existing technologies and products (Faems et al., 2005, p. 247). Furthermore, organizational structure as such has no impact on innovation performance.

Second, the study result for interaction effects confirm the first hypotheses and indicate that informal organizational structure strengthens the positive effect of explorative interorganizational collaboration on innovation objectives aimed at creating new technologies and products. To be more specific, informal organizational structure is appropriate for firms engaging in collaborations in the form of multidisciplinary innovation networks and living labs, when striving to achieve radical innovation.

Third, this research has found no support for the second hypotheses, which suggested that formal organizational structure strengthens the positive effect of exploitative interorganizational collaboration on innovation objectives aimed at improvement and further development of existing technologies and products.

6.1 Managerial implications

(26)

6.2 Limitations and future research

Looking at the adjusted R2 of the tested models, it can clearly be observed that the models only explain a small proportion of the variation in the data, which might be due to the absence of detrimental variables explaining radical and incremental innovative performance. Also, our second model measuring incremental innovative performance has a significance level of 10%, although a significance level of 5% or lower would be statistically preferred. Furthermore, this research has partially made use of the survey format of CIS (Community Innovation Survey), which is designed by Eurostat to gather innovation statistics (Eurostat, 2016). Although the content of the survey is well-designed and rich, it was not specifically developed for this research. Hence, several additional questions have been added to fit the purpose of the hypotheses outlined. This could also have been done for other variables to provide a more complete picture of the phenomena. For this particular research, the weakness of the CIS survey is that it doesn’t measure the purpose of several business practices, such as R&D and interorganizational collaboration. In other words, e.g., firms are asked to fill in the proportion of their revenue on R&D, without indicating their desired innovative output (radical/ incremental). Finally, the dependent variables is measuring innovative performance in percentages, which brings along the challenge of meeting assumptions of normality of the residuals. Although we argue that a large sample size can tolerate this violation, future research could look into ways to overcome this issue. For this research, a Tobit regression would be more suitable as this allows to account for the presence of censored values (McDonald and Muffit, 1980). However, our usage of SPSS did not permit to perform such analyses. Future research might consider using a statistics software (like, STATA) that support advanced regression models.

6.3 Conclusion

(27)

creating new technologies and products through explorative interorganizational collaboration. However, no support has been found for the second hypotheses, which suggested that formal organizational structure support the improvement and further development of existing technologies and products through exploitative interorganizational collaboration.

Acknowledgements

(28)

Appendices

Appendix 1. Survey questions

Age, measured by:

Q5 Vraag: Jaar van oprichting van uw bedrijf: ______

Size, measured by:

Q36 Vraag: Wat was het totaal aantal werknemers van uw bedrijf in 2015 (in FTE) ______ FTE (Full Time Equivalents) (1)

SkilledLabor, measured by:

Q37 Vraag: Hoe waren deze FTE's verdeeld naar opleidingsniveau? ______ % lager opleidingsniveau (geen MBO) (1)

______ % middelbaar opleidingsniveau (MBO) (2) ______ % hoger opleidingsniveau (HBO/WO) (3)

RD, measured by:

Q24 Vraag: Welk percentage van de omzet werd uitgegeven aan interne O&O activiteiten in 2015? ______ % (1)

Patent & Secrecy, measured by:

Q21 Vraag: Heeft uw bedrijf de volgende beschermingsmethodes gebruikt voor innovaties die in de periode 2013-2015 werden geïntroduceerd?

Octrooien (1)  Ja  Nee

Geheimhouding (2)  Ja  Nee

Informality, measure by:

Q53 Vraag: Geef voor elk van de volgende vragen aan wat het beste bij uw bedrijf past. Binnen mijn bedrijf is er...

Strakke formele controle met behulp van geavanceerde controle en informatie systemen

(1) (2) (3) (4) (5) Losse, informele controle met

behulp van informele relaties en samenwerking Sterke klemtoon op het altijd

naleven van formele procedures

(1) (2) (3) (4) (5) Sterke klemtoon op

mogelijkheid om af te wijken van formele procedures wanneer dit nodig is Sterke klemtoon op het

vasthouden aan solide management principes zelfs wanneer er zich veranderingen voordoen

(1) (2) (3) (4) (5) Sterke klemtoon op het

aanpassen aan veranderende omstandigheden zelfs wanneer dit vraagt om af te wijken van

(29)

Explorative Collaboration, measure by:

Q52 Vraag: In welke mate was uw bedrijf in de periode 2013-2015 betrokken bij de volgende vormen van samenwerking?

Multidisciplinaire innovatienetwerken of

samenwerkingsprojecten met publieke en private partners uit verschillende bedrijfstakken

(1)Niet (2)Weinig (3)Middelmatig (4)Veel

Living lab of projecten waarbij publieke en private partners samenwerken met burgers om samen nieuwe producten of diensten te

ontwikkelen

Exploitative Collaboration, measure by:

Q52 Vraag: In welke mate was uw bedrijf in de periode 2013-2015 betrokken bij de volgende vormen van samenwerking?

Co-creatie of het actief betrekken van

eindgebruikers bij het ontwikkelen van nieuwe producten en diensten

Rad./Incr. Inno. Performance measured by, measure by:

Q14 Vraag: Hoe is de gerealiseerde omzet van uw bedrijf in 2015 bij benadering verspreid over de volgende types goederen/diensten? (Totaal moet optellen tot 100%)

______ 1. In 2013-2015 geïntroduceerde goederen- en diensteninnovaties die nieuw voor uw markt waren.

(30)

References

Acemoglu, D., & Robinson, J. A. (2012). Why nations fail: The origins of power, prosperity, and poverty. New York: Crown.

Ahuja, G. (2000). The Duality of Collaboration: Inducements and Opportunities in the Formation of Inter-firm Linkages. Strategic Management Journal 21(3):317–343.

Andriopoulos, C., & Lewis, M. W. (2010). Managing Innovation Paradoxes: Ambidexterity Lessons from Leading Product Design Companies. Long Range Planning, 43(1), 104-122. Andriopoulos, C., & Lewis, M. W. (2009). Exploitation-Exploration Tensions and

Organizational Ambidexterity: Managing Paradoxes of Innovation. Organization Science, 20(4), 696-717.

Argote, L. (2012). Organizational learning: Creating, retaining and transferring knowledge. Springer Science & Business Media.

Argote, L., & Miron-Spektor, E. (2011). Organizational learning: From experience to knowledge. Organization Science, 22(5), 1123-1137.

Baldwin, C., & Hippel, E. V. (2011). Modeling a Paradigm Shift: From Producer Innovation to User and Open Collaborative Innovation. Organization Science, 22(6), 1399-1417.

Bailey, J., Hora, M., & Gaimon, C. (2014). "The Ambidexterity Paradox: Balancing Exploration, Exploitation and Learning from Failure." Academy of Management Proceedings, 2014(1), 15132-15132.

Baumol, W.J. (2002). The Free-Market Innovation Machine. Princeton: Princeton University Press.

Benner, M.J. and Tushman, M.L. (2003). Exploitation, Exploration, and Process Management: The Productivity Dilemma Revisited. Academy of Management Review 28(2):238–256.

Bercovitz, J., & Feldman, M. (2006). Entpreprenerial universities and technology transfer: A conceptual framework for understanding knowledge-based economic development. The Journal of Technology Transfer, 31(1), 175-188.

Bogers, M., Afuah, A., & Bastian, B. (2010). Users as innovators: a review, critique, and future research directions. Journal of management.

Brown, S.L. and Eisenhardt, K.M. (1997). The Art of Continuous Change: Linking Complexity Theory and Time-Paced Evolution in Relentlessly Shifting Organizations. Administrative Science Quarterly 42(1):1–34.

(31)

Deeds, D.L. and Hill, C.W. (1996). Strategic Alliances and the Rate of New Product Development: An Empirical Study of Entrepreneurial Biotechnology Firms. Journal of Business Venturing 11(1):41–55.

Deeds, D.L. and Rothaermel, F.T. (2003). Honeymoons and Liabilities: The Relationship between Age and Performance in Research and Development Alliances. Journal of Product Innovation Management 20(6):468–485.

Doz, Y. and Hamel, G. (1997). The Use of Alliances in Implementing Technology

Strategies. In: Managing Strategic Innovation and Change: A Collection of Readings. M.L. Tushman and P. Anderson (eds.). New York: Oxford University Press, 556–580.

Eurostat (2016). Community Innovation Survey (CIS). Retrieved from ec.europa.eu: http://ec.europa.eu/eurostat/web/microdata/community-innovation-survey

Ferrary, M. & Granovetter, M. (2009). The role of venture capital firms in Silicon Valley's complex innovation network. Economy and Society, 38(2), 326-359.

Field, A., (2013). Discovering Statistics Using IBM SPSS. Sage.

George, G., Zahra, S.A. and Wood, D.R. (2002). The Effects of Business– University Alliances on Innovative Output and Financial Performance: A Study of Publicly Traded Biotechnology Companies. Journal of Business Venturing 17(6):577–609.

Gustafsson, A., Kristensson, P., Witell, L. (2012). "Customer co-creation in service innovation: a matter of communication?". Journal of Service Management 23 (3): 311 - 327

Hagedoorn, J. (2002). Inter-firm R&D Partnerships: An Overview of Major Trends and Patterns since 1960. Research Policy 31(4): 477–492.

Im, G., & Rai, A. (2008). Knowledge Sharing Ambidexterity in Long-Term Interorganizational Relationships. Management Science, 54(7), 1281-1296.

Jansen, J., Van den Bosch, F., Volberda, H. (2006) Exploratory Innovation, Exploitative Innovation, and Performance: Effects of Organizational Antecedents and Environmental Moderators. Management science 52 (11), 1661-1674

Jiang, R. J., Tao, Q. T., & Santoro, M. D. (2010). Alliance portfolio diversity and firm performance. Strat. Mgmt. J. Strategic Management Journal, 31(10), 1136-1144.

Kedia, B. L., & Mooty, S. E. (2013). Learning and Innovation in Collaborative Innovation Networks. Restoring America’s Global Competitiveness through Innovation, 3-27.

(32)

Kusiak, A. (2007). Innovation: the living laboratory perspective. Computer-Aided Design and Applications, 4(6), 863-876.

Lin, H., Mcdonough, E. F., Lin, S., & Lin, C. Y. (2012). Managing the Exploitation/

Exploration Paradox: The Role of a Learning Capability and Innovation Ambidexterity. Journal of Product Innovation Management J Prod Innov Manag, 30(2), 262-278. Lorange, P. and Roos, J. (1992). Strategic Alliances: Formation, Implementation, and Evolution. Oxford: Blackwell.

March, J.G. (1991). Exploration and Exploitation in Organizational Learning. Organization Science 2(1):71–88.

McDonald, J.F. and Moffitt, R.A. (1980). The Uses of Tobit Analysis. Review of Economics and Statistics 62(2):318–321.

Möller, K. (2010). Sense-making and agenda construction in emerging business

networks—How to direct radical innovation. Industrial Marketing Management, 39(3), 361-371.

Nieto, M. J., & Santamaría, L. (2007). The importance of diverse collaborative networks for the novelty of product innovation. Technovation, 27(6-7), 367-377.

Niitamo, V., Kulkki, S., Eriksson, M., & Hribernik, K. A. (2006). State-of-the-art and good practice in the field of living labs. 2006 IEEE International Technology Management Conference (ICE).

Norman, D. A., & Verganti, R. (2014). Incremental and radical innovation: Design research vs. technology and meaning change. Design Issues, 30(1), 78-96.

O’Reilly, C.A., III and Tushman, M.L. (2004). The Ambidextrous Organization. Harvard Business Review 82(4):74–82 (April).

Oke, A. (2007). Innovation types and innovation management practices in service companies. International Journal of Operations & Production Management, 27(6), 564-587.

Phelps, C. C. (2010). A longitudinal study of the influence of alliance network structure and composition on firm exploratory innovation. Academy of Management

Journal, 53(4), 890-913.

Popadić, M., Černe, M., & Milohnić, I. (2015). Organizational Ambidexterity, Exploration, Exploitation and Firms Innovation Performance. Organizacija, 48(2).

(33)

Raisch, S., Birkinshaw, J., Probst, G., & Tushman, M. L. (2009). Organizational

Ambidexterity: Balancing Exploitation and Exploration for Sustained Performance. Organization Science, 20(4), 685-695.

Rampersad. G., Quester, P., & Troshani, I. (2010). Managing innovation networks:

Exploratory evidence from ICT, biotechnology and nanotechnology networks. Industrial Marketing Management, 39(5), 793-805.

Rogers, M. (2004). Networks, Firm Size, and Innovation. Small Business Economics 22(2):141–153.

Rothaermel, F.T. and Deeds, D.L. (2004). Exploration and Exploitation Alliances in Biotechnology: A System of New Product Development. Strategic Management Journal. Sammarra, A. and Biggiero, L. (2008) Heterogeneity and Specificity of Inter-Firm

Knowledge Flows in Innovation Networks. Journal of Management Studies, 45, 785–814. Santos-Vijande, M. L., López-Sánchez, J. Á., & Trespalacios, J. A. (2012). How

organizational learning affects a firm's flexibility, competitive strategy, and performance. Journal of Business Research, 65(8), 1079-1089.

Schumpeter, J.A. (1939). Business Cycles: A Theoretical, Historical, and Statistical Analysis of the Capitalist Process. New York: McGraw-Hill.

SNN (2016). SNN’s four focal tasks: Coordination, subsidies, lobbying, networking. Retrieved from snn.eu: http://www.snn.eu/en/about-us/four-focal-tasks

Soetanto, D. P., & Geenhuizen, M. V. (2011). Social networks, university spin-off growth and promises of ‘living labs’. Regional Science Policy & Practice, 3(3), 305-321.

Stettner, U., & Lavie, D. (2013). Ambidexterity under scrutiny: Exploration and

exploitation via internal organization, alliances, and acquisitions. Strategic Management Journal, 35(13), 1903-1929, http://dx.doi. org/10.1002/smj.2195

Teece, D.J. (1986). Profiting from Technological Innovation: Implications for Integration, Collaboration, Licensing, and Public Policy. Research Policy 15(6):285–305.

Teece, D.J. (2002). Managing Intellectual Capital. Oxford: Oxford University Press. Todeva, E., & Knoke, D. (2005). Strategic alliances and models of collaboration. Management Decision, 43(1), 123-148.

Van de Ven, A.H., Polley, D., Garud, R. and Venkataraman, S. (1999). The Innovation Journey. New York: Oxford University Press.