August 31st , 2015
Stages of innovation and their role in the relationship between ties with service
intermediaries and innovation performance.
Master’s Thesis
Student: Matey Nohchev, Student’s № 10826475 MSc Business Administration – Strategy Track
University of Amsterdam, Faculty of Economics and Business
Supervisor: Dr. Alexander S. Alexiev VU University Amsterdam
1 Statement of Originality
This document is written by Matey Nohchev who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
2 ABSTRACT
Professional service firms have been recognized to positively influence innovation performance of young companies. Nonetheless, innovation has long been seen as a multi-staged process, in which different activities and actors are at play at each phase. In order to empirically examine the effect of innovation stages on the relationship between ties with service intermediaries and innovation performance, a study of 85 young tech ventures in the Netherlands and the UK has been conducted. The research revealed controversial results when it comes to the direct relationship between the use of service intermediaries and innovation performance - only talent search firms positively affect innovation performance. Moreover, results show a significant moderating effect of the stages of the innovation process on this relationship. Innovation performance scores higher when accounting and financial service firms and law service firms are used in the adoption stage, whereas technology service firms are used in the implementation stage. No trend was found for talent search firms at either stages and for the use of distinct ties with service intermediaries at specific stages by this sample of companies.
3 Table of Contents
Statement of Originality ... 1
ABSTRACT ... 2
I. Introduction ... 5
II. Literature review ... 8
1. Innovation and Stages of the Innovation Process ... 8
2. New ventures ... 9
3. Role of Service Intermediaries ... 10
4. Service intermediaries and product innovation in new ventures... 11
5. Identifying literature gap and research question. ... 12
III. Theory development and hypothesis: ... 14
1. The moderating role of the innovation process stages on ties with service intermediaries. ... 14
2. Accounting and financial service firms, innovation stages and innovation performance ... 16
3. Law service firms, innovation stages and innovation performance ... 18
4. Talent search firms, innovation stages and innovation performance ... 19
5. Technology service firms, innovation stages and innovation performance ... 21
IV. Methodology ... 23 1. Measures ... 26 2. Dependent Variable ... 26 3. Independent Variable ... 27 4. Moderating Variable ... 27 5. Controls ... 28 V. Results ... 29 1. Descriptive statistics. ... 29 2. Reliability ... 30 3. Correlation analysis ... 31 4. Regression analysis ... 32 VI. Discussion... 44
1. Effect of ties with service intermediaries on innovation performance ... 45
1.2. Ties with accounting and financial service firms ... 45
1.3. Ties with law service firms ... 46
4
1.5. Ties with technology service firms ... 47
2. The effect of innovation stages on the use of ties with service intermediaries. ... 48
3. Theoretical contributions ... 48
4. Managerial implications ... 50
5. Limitations and propositions for future research ... 51
VII. Conclusion ... 52
References:... 54
Appendix 1. Cover Letter ... 59
Appendix 2. Survey ... 61
Appendix 3. Position profiles ... 67
5 I. Introduction
The importance of innovation has long been discussed in the academic literature. Regarding organizations, innovation is crucial for adaptation and renewal of organizations (Nohria and Gulati 1996; Banbury and Mitchell, 1995; Christensen, 1997; Cefis and Marsili 2006). For instance, firms engage in innovation in order to increase their short-term performance and to enhance their innovation capabilities for improving long term output (Caldwell, Herold and Fedor, 2004). Furthermore, Cefis and Marsili (2006), in their study on manufacturing firms, confirmed the positive role of innovation on survivability.
However, innovation is regarded as a multi-stage process (Baregheh, Rowley and Sambrook, 2009; Desouza et al., 2009; Hansen and Birkinshaw, 2007) and studying this multidimensional process is essential, as it can provide insights about the temporal strategic choices of managers, regarding innovation and the success of their outcomes. Wolpert (2002) argues that organizations tend to seek external partnerships at a late stage during the innovation process, which hinders such ventures on making an impact in the field, because the business opportunity is already clear to the competition. In addition, successful innovation is reliant on involving partners at early stages in the opportunity exploration process (Wolpert 2002). Moreover, Sung and Cho (2011) found that different agents play distinct roles at different stages of the innovation process.
It is important to note nevertheless, that intermediaries can play a key role within the innovation process of organizations (Wolpert, 2002; Howells 2006; Zaheer and McEvily, 1999; Bessant and Rush, 1995). Moreover, the effects of innovation are of greater importance for small and young firms, due to their risk exposure and these effects reduce the risk of organizational mortality (Cefis and Marsili, 2006). Therefore, the role of service intermediaries can be essential
6 for the innovation performance of new ventures. This is in line with the recent research of Zhang and Li (2010) and Wu, Li and Wang (2014) that show a positive causality between product innovation and ties with service intermediaries.
Given the benefits of innovation and professional service firms on organizations, it is suggested that a more detailed study in the context of young organizations and their relationships in the innovation process stages is needed; new ventures are at disadvantage, due to their short history related liability (Stinchcombe 1965; Freeman and Carroll, 1983; O’Farrell and Hitchens, 1988; Zhang and Li 2010; Zhou and Li, 2012).
It has been already established that ties with service intermediaries have a positive influence on the product innovation of new ventures (Zhang and Li, 2010; Wu et. al, 2014) and also the importance of viewing the innovation process in stages has been noted (Wolpert, 2002; Sung and Cho, 2011). So, this paper will build on the findings of Zhang and Li (2010) and its aim is to study the relationship between the different stages of the innovation process and the effect of ties with service intermediaries on the product innovation in the context of new ventures. As noted above, it has been stated that partnerships at different stages and innovation outcomes of established organizations are connected (Wolpert, 2002). Hence, this paper will focus on the moderating effect of the innovation stages on the relationship between new ventures’ ties with service firms and their effect on product innovation performance, along the advancement of the stages.
The purpose of this study is to bring further insight, in a temporal perspective, into the strategic choices of the new ventures, regarding the different stages of innovation and the specific ties they need at each stage. In regard to practical implications, such a research can provide clarifications of the possible mistakes, in a temporal perspective, that managers of new
7 ventures make when seeking external consulting with their product innovation. In addition, findings can impact managerial choices, pertaining to their chronological perspective on strategy. The evidence can give them indications about the stages of product development in which it is optimal to engage in a relationship with a certain category of a service intermediary and types of service intermediaries that are most important at discrete stages. Thus, this will aid young organizations in managing their ties with service firms more effectively throughout the product innovation stages. Consequently, one can argue that investigating this matter will enrich the existing theoretical understanding of ties and their impact on product innovation. The findings of this research can put the emphasis on the importance of certain type of connections, which are needed at distinct stages of the new product development, in order to enhance the innovation process. Furthermore, this study can provide insights, regarding potential obstacles to a rapid new product-to-market process, related to a misuse or overuse of connections.
Noting the compelling nature of the potential evidence, and in order to address the significance of the omitted aspect of the literature, a research question is proposed:
“What is the role of the stages of the product innovation process in the relationship between ties with service intermediaries and innovation performance in the context of new ventures?”
In addition, to investigate this question more specifically, the research will be done in the realm of answering two sub-questions:
• What is the effect of the innovation stages on the relationship between innovation performance and the use of service intermediaries by new ventures?
• What is the direct effect of the innovation stages on the use of ties by new ventures with specific service intermediaries?
8 II. Literature review
1. Innovation and Stages of the Innovation Process First of all, what exactly does the term innovation mean?
The literature on innovation has proposed various definitions of the concept, which is exemplified by the paper by Baregheh et al. (2009), where the authors captured and analyzed extant meanings of the term. However, for the use of the term to be consistent in this study, the definition proposed by Baregheh et al. (2009) will be used. The authors tried to encompass the main aspects of the term from the studied pool of innovation explanations and suggested a definition: “Innovation is the multi-stage process whereby organizations transform ideas into new/improved products, service or processes, in order to advance, compete and differentiate themselves successfully in their marketplace”. Moreover, due to the specificity of this study the term innovation will be related explicitly to new product development.
As mentioned above, innovation is recognized as a multi-staged process (Baregheh et. al, 2009; Damanpour and Schneider, 2006; Meyer and Goes, 1988; Desouza et al. 2009; Hansen and Birkinshaw, 2007; Sung and Cho, 2011); however there have been different views on the stages present in the innovation process in organizations. Desouza et al. (2009), in their study on 30 US and European companies with innovation procedures, indicated as “robust”, distinguished five stages of processes: idea generation and mobilization, screening and advocacy, experimentation, commercialization, and diffusion and implementation. Moreover, Baregheh et al. (2009), in their review on the literature on definitions of innovation, again recognize five stages of this process: creation, generation, implementation, development and adoption. Hansen and Birkinshaw (2007) proposed a more generalized model, in comparison to the aforementioned models, referred to as “The Innovation Value Chain”. Even though the model employs three stages, there are
9 similarities with the previous models, as each stage includes a number of sub-processes, which reasonably correspond to the five stages. The stages are idea generation, relating to the internal or external concept creation, idea conversion, relating to the screening and development of the idea, and idea diffusion, relating to spreading the idea across the organization. Sung and Cho (2011) go even further and simplify the process to two stages - adoption and implementation.
It can be seen that in spite of the variety of innovation stage count and categorization, there is a resemblance between the extents of proposed models. For the purpose of this study, however, the innovation process model that will be used is the one proposed by Sung and Cho (2011).
2. New ventures
New ventures tend to be at a disadvantage because even a modest search for external opportunities for innovation will be at a too high cost to them, due to limited resources (Zhang and Li, 2010; Eisenhardt and Schoonhoven 1990).
Thus, venturing into too many new ideas will be inefficient for them (Zhou and Li, 2012). They suffer, from what Stinchcombe (1965) pointed out as “liability of newness”. Young organizations lack comprehensive exploration capabilities, due to their limited external contacts and resources. Furthermore, such firms have a higher risk of failure, compared to established organizations because they have a low level of legitimacy, are dependent on external cooperation and have low competitive capabilities (Stinchcombe, 1965). This argument is further empirically tested and supported by Freeman and Carroll (1983) in their study of age dependence in organizational death rates in three disparate populations of organizations. They find that young ventures have a higher mortality rate than older ones. Finally, new ventures lack a track record,
10 which could affect their ability to raise capital, constraining both their creation and expansion (Hitchens and O’Farrell, 1988).
3. Role of Service Intermediaries
What are service intermediaries? Howells (2006) summarized in detail the literature on the roles of intermediaries and the process of intermediation in innovation. The author concluded that the main functions of service intermediaries, which come up in the literature, are as information repositories and disseminators, and technology transfer agents. Howells (2006: 720), in his research, defined a service intermediary as “an organization or body that acts an agent or broker in any aspect of the innovation process between two or more parties”. Similarly, Wolpert (2002) argued that service intermediaries can aid in the information exchange across companies while maintaining confidentiality. Zhang and Li (2010: 89) refer to service intermediaries as “professional service organizations, which provide organizations with supportive services in areas, such as accounting and finance, talent search, law, and technology”. Moreover, the literature on service intermediaries has covered the aspect of how these “middlemen” play an important role in the performance of organizations (Wolpert, 2002; Howells 2006; Zaheer and McEvily, 1999; Bessant and Rush, 1995).
Bessant and Rush (1995) present certain aspects of the innovation process in which intermediaries, or “consultants”, as called by the authors, can have a positive effect on organizations. Consultants can directly provide organizations with the needed knowledge; share their experience from previous cases; act as middlemen, linking organizations with specialist service; aid their clients by defining their innovation related issues. Many organizations are unable to prioritize and comprehend these issues, which prevents them from recognizing external
11 opportunities and employing external resources. However, intermediaries can help companies with these matters, by developing a framework for change or suggesting where such issues can be solved (Bessant and Rush, 1995).
Likewise, Zaheer and McEvily (1999) state that firms may be uncertain about which capabilities are worth pursuing, therefore a network of ties with organizations, which can provide such information, can be valuable. In addition, the authors argue that in order for organizations to be competitive, they must search for opportunities for improving their capabilities and acquiring new ones. Furthermore Wolpert (2002) argues that companies that are not looking outward for knowledge may not be able to discover and exploit new opportunities and capabilities. Zaheer and McEvily’s (1999) research on job shop manufacturers in a geographical cluster revealed that relationships with service intermediaries, namely technical support organizations, universities, training centers and research centers will have a positive effect on the firms, in terms of acquiring competitive capabilities. The function of these institutions is to disseminate knowledge and information (e.g. Howells, 2006; Wolpert 2002). Furthermore connections with such service intermediaries can reduce the search costs of organizations, which can include locating necessary knowledge and expertise, in order to obtain competitive capabilities (Zaheer and McEvily, 1999). Likewise, Wolpert (2002) stated that such firms can also aid in coupling companies with mutually beneficial capabilities.
4. Service intermediaries and product innovation in new ventures
Recent research has presented the positive role of service intermediaries on new venture innovation. Zhang and Li (2010) studied how ties of new ventures with service intermediaries affect their product innovation in a technology cluster in China. The intermediaries that are
12 brought to attention are technology service firms, accounting and financial service firms, law firms, and talent search firms. The collected data of 202 firms proved their hypothesis that ties of young organizations with professional service firms are positively connected with new venture’s product innovation. They studied three moderating variables that affected the positive relationships between the aforementioned ties – the perceived industry growth of managers in studied ventures, the relationships of the ventures with foreign firms in terms of distributing foreign products, and product export of new ventures.
Furthermore, Wu, Li and Wang (2014) found in their study that ties of knowledge-intensive business services intermediary with new ventures has a positive impact on their product innovation, which is in line with the findings of Zhang and Li (2010).
As noted beforehand, this study will build upon the research of Zhang and Li (2010) and its purpose is to further add to the investigation of new ventures and the role of service intermediaries in their innovation process. Consequently, the organizations regarded to as service intermediaries in this study, will be the ones examined by Zhang and Li (2010): accounting and financial service firms, law firms, talent search firms and technology service firms.
5. Identifying literature gap and research question.
So far the literature has presented the liable state in which new ventures reside, the importance of innovation for organizational survival, especially in small and young firms (Cefis and Marsili, 2006), and the important role of intermediaries in innovation both in established organizations (Wolpert, 2002; Howells 2006; Zaheer and McEvily, 1999; Bessant and Rush, 1995) and new ventures (Zhang and Li, 2010; Wu, Li and Wang 2014).
However, the literature has been silent about when do new ventures actually seek to establish ties with service intermediaries, in order to support their innovation on new product
13 development. Wolpert (2002) argued that organizations usually establish ties at later stages, which hinders their impact on the field and that successful innovation requires partners at early stages of the innovation process. Yet, the influence of this connection on new ventures and an explanation of the stages were absent. Furthermore, the author does not specify the types of service intermediaries, which have a positive effect on product innovation and at which stages they were established and utilized. Likewise Zhang and Li (2010) recognized the positive effect of ties with service intermediaries of new ventures on their product innovation, though the authors did not take into account the stages of the innovation process. Also, the authors omitted the possible moderating effect of the different types of ties at each stage of the innovation process on its positive outcome. One can argue that the use of ties varies across the new product development. For example, Sung and Cho (2011) studied the dynamics of the roles of top management, external environment, innovation and employees during two stages of the innovation process - adoption and implementation. They argue that these four actors play discrete roles at each stage. However, the authors did not focus on the role of service intermediaries as key players in the innovation process stages and studied a company well established into its field.
Taking into account the above-mentioned omitted aspects of the literature, this paper will address the role of ties with service intermediaries during the stages of the product innovation process on the innovation performance in the context of new ventures. In addition, both the adoption and implementation stages will be analyzed in order to assess if there is a pattern related to using distinct service firms at each stage of the innovation process.
14 III. Theory development and hypothesis:
In order to address the sub-questions and consequently the research question, a set of hypotheses, related to the stages of the innovation process and new ventures’ ties with service intermediaries will be proposed and a method for testing these hypotheses will be suggested.
Regarding the stages of innovation that are going to be used in this research, one can argue that a detailed model for these stages, such as the one Desouza et al. (2009) or Baregheh et al. (2009) propose, will be difficult to apply to a new organization. The reason for this is, as Desouza et al. (2009) mention, their model is based on established companies using a “robust” innovation process. As mentioned beforehand, new ventures tend to suffer from the “liability of newness” (Stinchcombe, 1965), due to their limited resources (Zhang and Li, 2010; Eisenhardt and Schoonhoven 1990) and lack of track record (Hitchens and O’Farrell, 1988). Thus it can be assumed that such organizations may fail to follow such a comprehensive innovation process, as the one Desouza et al. (2009) describe, due to the fact that young firms do not have a history of numerous developed products and further are unable to develop multiple goods at once. This may lead to omitting important stages, being in multiple stages at once or repeating stages. Therefore, in order to be more accurate at which stage the new ventures reside in at the current moment of their product development, the more simplified two stage model, employed by Sung and Cho (2011) will be used, which includes the phases adoption and implementation.
1. The moderating role of the innovation process stages on ties with service intermediaries.
When it comes to new product development, the process starts with the adoption stage (e.g. Sung and Cho, 2011). It is related to the decision of the company to pursue an innovation, as it expects to reap benefits from it (Klein & Sorra, 1996; West & Anderson, 1996). This stage
15 includes the process of idea generation, screening and the adoption of a project for further development. The implementation stage (e.g. Sung and Cho, 2011) follows after the adoption stage and relates to a move from the decision to pursue a certain innovation to a regular use of the innovation or its standardization (Holahan, Aronson, Jurkat, & Schoorman, 2004; Klein and Sorra, 1996). Thus this stage comprises of processes related to prototype manufacturing, testing, and bringing the idea to the market.
Regarding the question about when do new ventures actually seek to use a tie with a service intermediary, in order to support their innovation on new product development, it can be argued that the stage, in which the company currently resides, plays a key role on the preference of the young organizations for a specific tie. Sung and Cho (2011) found that different actors play distinct roles during specific stages. Furthermore, Hansen and Birkinshaw (2007) discovered that managers turn to different sources for collaboration during the advancement of the stages of the Innovation Value Chain. Finally, it is important to consider the fact that young organizations will have little experience, related to working with other companies (Desouza et. al, 2009). Thus one can argue that new ventures will engage in seeking partnerships, only when a certain need comes forth, rather than early in their innovation process, in order to anticipate potential issues and delay. This follows the notion of Wolpert (2002), who mentioned that companies tend to seek partnerships during later stages of their innovation processes. Taking into consideration all of the above, it is proposed:
Hypothesis 1: The adoption and implementation processes will moderate the positive relationship between the use of accounting and financial firms, law firms, talent search firms and technology service firms by new ventures, and innovation performance, in terms of enhancing or lessening it.
16 2. Accounting and financial service firms, innovation stages and innovation
performance
When it comes to the adoption stage of the innovation process, accounting and financial service firms can engage in various roles. Desouza et al. (2009) noted that in young organizations, screening through ideas and analyzing the cost-benefits from them may be inadequate, or even lacking, which could potentially lead to the death of the business. However, accounting and financial service firms can assist in strategic business consulting (Zhang and Li, 2010), which can aid organizations in screening out the ideas with less financial merit, thus sparing resources. The decision to venture into too many ideas in the adoption stage can be counterproductive for new ventures, so one can argue that in order to reduce the risk of delving into inferior ideas new ventures may seek the help of such companies. In addition, Desouza et al. (2009) state that in order for organizations to be successful at their screening and adoption process, there must be a strategic fit between the idea and both the company’s mission and values, and capabilities and resources.
These service entities can also aid in connecting new ventures to financial resource suppliers, in order to benefit the innovation process (Schoonhoven et al., 1990). Finally, when new ideas are generated and screened, in order to select those that will continue to the implementation stage, initial funding is required (Hansen and Birkinshaw, 2007). Therefore young organizations can seek help from such service companies in order to deal with such matters.
The implementation stage requires creating the means for utilization and production (Desouza et al., 2009). In this phase, organizations take part in building and testing prototypes, and further participating in processes related to delivering the product to the market and
17 commercializing it. This is in line with the argument by Desouza et al. (2009), who state that testing a prototype can be resource intensive, especially if the idea lacks strategic fit or is too expensive.
Nonetheless, Suchman (2000) argues that small companies frequently use their ties with law service firms when it comes to business consulting, due to the fact such organizations are unable to meet the expense of specialized consulting firms. Therefore young organizations may refrain from the use of accounting and financial service firms, not only to reduce costs, but also to benefit from the business acumen of lawyers, which according to Suchman (2000) are similar to conventional business advisors. Moreover, the author mentions that lawyers play a key role as “dealmakers” – connecting venture capitalists with young organizations. Furthermore, Zhang and Li (2010) state that new ventures engaged in high technology industries can use also technology service firms as a funding source and such companies can aid in the commercialization of a young organization’s product through their own infrastructure. Such service companies can also make available young organizations both with capital for future innovations (Zhang and Li, 2010) and also aid them in locating sources of complementary technologies and specialized expertise, crucial for the innovation implementation (Howells, 2006). Thus, one can argue that for new organizations it would be much more beneficial to work closely with technology service and law service firms, due to the benefit of both capital and expertize.
Considering the aforesaid, it is suggested:
Hypothesis 2a: The use of ties with accounting and financial firms by new ventures in the adoption stage will yield a higher innovation performance than in the implementation stage.
Hypothesis 2b: The use of ties witch accounting and financial firms by new ventures will be more positively related with the adoption stage than with the implementation stage.
18 3. Law service firms, innovation stages and innovation performance
Regarding law service firms, as mentioned above, it has been documented that such companies provide general business consultation and legal assistance along with support in constituting contracts (Suchman, 2000). In addition, Wolpert (2007) stated that such service companies often pick up ideas for new products and novel ways of doing business from existing competitors and non-competitors. However, as Hansen and Birkinshaw (2007) mentioned, organizations will first rely on their own capabilities for idea generation. Therefore in the adoption stage, even though such service entities could be beneficial in terms of providing insights about possible projects with merit, it can be argued that new ventures would stick to their own agenda. Moreover, new technology ventures have a pattern of forming around preconceived ideas, which in the first stage of the innovation process, are not tangible, thus not yet eligible for intellectual property protection.
As noted above, the implementation stage is typically related with the creation and testing of prototypes. At this stage, it is typical for organizations to protect their product by contacting legal entities, which, as Suchman (2000) mentioned, are proficient with intellectual property policies. The author also states that new organizations, as clients of law service firms, tend to lack the business acumen to drive their company forward. In this notion Castle (2001) and Kimberly & Evanisko (1981) state that a key condition for organizational effectiveness involves meeting market and technological demands, coming from external sources, such as customers, competitors and suppliers, in an adequate way. Thus, steering the project in the most effective way will be crucial for the innovation performance. Furthermore, as mentioned above, such service intermediaries can assist young organizations in structuring agreements (Suchman, 2000) and often become aware of new business models (Wolpert, 2007), which can be crucial for
19 licensing contracts. Therefore, working closely with such service entities during this stage of the innovation process will be of high merit for the innovation success and the new ventures’ competitive advantage.
Bearing in mind all of the above, regarding law service firms, it is proposed:
Hypothesis 3a: The use of ties with law service firms in the implementation stage will be related with a higher innovation performance, than in the adoption stage.
Hypothesis 3b: The use of ties witch law service firms by new ventures will be more positively related with the implementation stage than with the adoption stage.
4. Talent search firms, innovation stages and innovation performance
Bahrami and Evans (1995) state that talent search firms are used widely by startups and also established organizations for the purpose of employing new talents. These service companies are engaged in facilitating the cross-company migration of people, which aids in both developing social networks and strengthening the operational effectiveness of firms (Zhang and Li, 2010). Furthermore, Boeker (1997) states that such personnel movement enables young organizations to gain access to the knowledge generated by competitors and non-competitors and to familiarize themselves with other companies’ strategy. However, as stated earlier, new ventures tend to be cautious when it comes to jeopardizing intellectual property, which would be especially true for the adoption stage, when no tangible product is available, whereas the migration of people increases the risk of sensitive information disclosure. Furthermore, managers in organizations tend to first look inward, to their current employee concept creation capabilities, when it comes to the idea generation process (Hansen and Birkinshaw, 2007).
20 On the other hand, Sung and Cho (2011) argue that at the implementation stage, after the idea has been adopted, the active involvement of employees could be central for a successful implementation. They state that employee-driven implementation could improve both the company’s innovative capability and its organizational performance. Moreover, when the personnel are the main driver of the innovation implementation, they could develop a sense of ownership, become motivated to support the benefits for the company from the innovation (Armenakis, Harris, & Mossholder, 1993; Holt, 2002; Jones et al., 2005; Miller, Johnson, & Grau, 1994). Therefore at this stage, talent search firms could aid new ventures in locating the adequate and determined personnel to contribute to the implementation of the new product. Thus, during this phase, one can assume that the idea has already been legally protected, thus the migration of personnel can be facilitated, without the fear of information outflow. New ventures could only benefit from the involvement of people, possessing expert knowledge, which would support product development, and who may provide information, related to competitors’ strategy.
Taking into consideration the above mentioned information, it is proposed:
Hypothesis 4a: The use of ties with talent search firms by new ventures in the implementation stage will be related with a higher innovation performance, than in the adoption stage.
Hypothesis 4b: The use of ties with talent search firms by new ventures will be negatively related with the adoption stage.
21 5. Technology service firms, innovation stages and innovation performance
When it comes to technology service firms, such can act as brokers of technology, in terms of supplying new ventures with information about technologies and innovations other firms are engaged in, which could potentially aid new product development (Bessant and Rush, 1995; Hargadon and Sutton, 1997). However, Desouza et al. (2009) argue that young organizations with a “brittle” innovation process are unlikely to involve external sources at an early stage of the innovation process, due to the fact that they fear the loss of intellectual property. This would be especially true for technology service firms, which are aware of the current technological state and demands, and have a large network of contacts, thus a higher chance of information leakage. Moreover, in the adoption stage, as noted above, new ventures have not yet legally protected their ideas, as these are still concepts and no tangible product is available.
Regarding the implementation stage, however, Howells (2006) argues that ties with such service intermediaries can aid organizations in locating external sources of technologies and specific know-how, which can be crucial for the innovation implementation by adding value to it. Furthermore, Desouza et al. (2009) argue that organizations will engage in relationships with external organizations when undertaking uncertain and complex trials, which, as mentioned above, can require heavy resource expenditures. Facilitating a new product line will create additional costs, which can deplete the already limited resources of young companies. Furthermore, as mentioned above, the lack of track record can hinder raising capital, which constrains both creation and expansion (O’Farrell and Hitchens, 1988). Nevertheless, as mentioned previously, through technology service companies, young companies can
22 commercialize their new product and acquire capital to support forthcoming innovations (Zhang and Li, 2010).
Taking into attention the drawbacks and benefits of the use of technology service firms along the two stages, one can debate that:
Hypothesis 5a: The use of ties with technology service firms in the implementation stage will be related with a higher innovation performance, than in the adoption stage.
Hypothesis 5b: The use of ties with technology service firms will be negatively related with the adoption stage.
A conceptual model, comprising all the hypotheses, outlined in the theoretical framework and containing the employed control variables can be seen in Figure 1. Conceptual Model.
Figure 1. Conceptual Model
Ties with service intermediaries:
• Accounting and financial service firms
• Law service firms • Talent search firms • Technology service firms
Stages of the innovation process: • Adoption • Implementation Innovation performance Control variables: • Venture age • Founding team size • Founding team education • Venture size
• Employee education • Environmental Uncertainty
H1, H2a, H3a, H4a, H5a H2b, H3b, H4b, H5b
23 IV. Methodology
A sample size of 1151 new ventures, 631 of which were located in the Netherlands and 520 in the UK, and operating in high technology industries were selected for the intended research. Due to the limited timeframe in which the data were collected – from the 18th of May 2015 until the 10th of June 2015, and the large size of the sample, two graduate students were engaged in the research. The survey used was comprised of questions aimed at the same target sample, but related to two similar, yet different research topics. A response rate of 7.4 % was recorded or 85 companies, 69 of which were Dutch and 16 English.
A multi-step approach was applied, in order to locate the companies. First, Capital IQ and Orbis were used as reference databases for locating the companies. Due to the limited timeframe used for conducting this research, the criteria for technology ventures proposed by Li and Atuahene-Gima (2001) (e.g. Zhang and Li, 2010), namely (1) its founding team must be composed of engineers or scientists, (2) thirty percent or more of its employees must be technical employees, and (3) it must spend three percent or more of total sales on research and development could not be applied. However certain filters were applied in order to arrive at the appropriate sample of companies - only companies engaged in product development and which operated in the well-known realm of high-technology industries were used. The criteria for high technology industries that were applied was adopted from the one by the U.S. Bureau of Labor Statistics, and based on the high-technology employment research of Hecker (1999). A list of the sectors, in which the companies from the sample reside, can be found in Table 1. Second, in order to follow the notion of new ventures, only companies which are eight years or younger were used in this research (e.g. Zhang and Li, 2010). Finally, to ensure the legitimacy of the ventures’ age, and to exclude the risk of using data from re-registered ventures, the dates from the databases were cross-referenced with the respondents’ answers regarding the age of the
24 company.
Table 1. Industries, comprising the high technology sector 1. Aerospace and Defense
2. Automobiles and Components 3. Healthcare Equipment and Supplies 4. Biotechnology
5. Pharmaceuticals 6. Internet Software
7. Technology Hardware and Equipment
8. Semiconductors and Semiconductor Equipment 9. Health Care Technology
The original questionnaire was prepared in English and then was translated into Dutch by a native Dutch graduate student, fluent in both languages and experienced with business terminology. To ensure the validity of the Dutch translation, the Survey was translated back into English by a second resident of the Netherlands, again fluent in both languages and with a business background. This was done in order to locate misunderstandings, which could possibly arise in the initial translation. Before submitting the survey, a pre-test was conducted with several native Dutch speaking managers of new companies, with the purpose of assessing both the translation and relevance of questions. The final questionnaire included the option to switch between languages, which further added to eliminating any misinterpretations given the fact that English is the second most spoken language in the Netherlands after Dutch. The original survey can be found in Appendix 2, Survey.
Data were collected through surveys sent via electronic e-mails. The surveys were aimed at top managers in new ventures, who were directly involved in a product innovation process, however to confirm the adequacy of the responses, respondents were asked to state their current position and their level of involvement in strategic decision making. The cover letter, used to address the ventures is presented in Appendix 1, Cover Letter. The majority of positions that
25 were listed were founder/owner, CEO/CFO/COO, Operations/Marketing/Commercial director. Three respondents listed office manager as a position, however, judging from their involvement in strategic decision making – 3 and above on a 5 point Likert scale, their average experience in the industry (6.3 years) and the number of employees in the ventures (14 and below), one can argue that these responses are valid for analysis and are not a matter of concern. The profiles of the respondents can be seen in the Appendix 3, Table 12. Respondents’ position profiles. In addition, as noted above and similar to Zhang and Li (2010), to ensure the quality of the data, managers were asked about the extent of their involvement in strategic decision making, using a five-point Likert scale (1 = Not at all; 5 = To a large extent). All of the respondents scored a 3 (To a moderate extent) or above on this scale. Moreover, they were asked to input the years of work experience in their current industry, in order to, as Zhang and Li (2010) stated, safeguard the appropriateness of their answers related to the industry environment questions. The average work experience of the managers in the sample was 9.1 years. Also, the average venture age was 3.61 years - the set criteria for a new venture was 8 years (e.g. Zhang and Li, 2010) (Table 2.). The average founding team size was 2.26 people, and the average venture size, estimated in full time employees (FTE), was 7.05 employees. All these data suggest that the featured respondents are experienced professionals in their field and are well-informed in the realm of both their current products and their industry as a whole, thus suggesting for good quality of the data. Further information, regarding sample characteristics can be seen in Table 2.
26 Table 2. Descriptive Statistics – Respondent and Venture profile
N Minimum Maximum Mean Std.
Deviation Years of experience in the industry 85 1 34 9.02 7.797 Involvement in strategic decision making 85 3 5 4.73 .543
Venture age 85 0 8 3.61 1.914
Founding team size 85 1 8 2.26 1.216
Founding team education 85 1 5 3.65 1.624
Venture size (in FTE) 85 0 57 7.05 8.000
Employee education 85 1 5 3.72 1.419
Foreign invested venture 85 1 2 1.95 .213
Venture origin 85 1 2 1.07 .258
1. Measures
Measures, which required the respondents to rate their perception, namely the dependent variable innovation performance, the independent variable ties with service intermediaries, and the control variables environmental uncertainty, founder education and employee education, were rated using a five point Likert scale. (1 = Not at all and 5 = To a large extent). A category question was used for the moderating variable stage and the controls foreign invested venture and venture origin; whereas the constructs venture age, founder size and employee size were measured using quantity questions. A breakdown of the variables, containing more than one item can be found in Appendix 4, Table 13. Multi-construct variables.
2. Dependent Variable
In order to measure product innovation performance the already proven construct, provided by Zhang and Li (2010), namely; asking the respondents to rate the “extent to which their ventures were successful relative to their major competitors” (p.96) was used. The authors used 5 aspects to measure innovation performance, which were also applied in the survey, i.e. (1) frequently introducing new products, (2) being first in new product introductions in the market,
27 (3) quickly launching new products into the market, (4) developing new products with superior quality, and (5) using new products to penetrate markets. As the authors noted, this measure encompasses the key elements, used for evaluating new product performance (Brown and Eisenhardt, 1995).
3. Independent Variable
With the intention of measuring ties with service intermediaries, the approach employed by Zhang and Li (2010) was used. Respondents were enquired to rate the extent to which their companies are working closely with (1) accounting and financial firms, (2) law service firms, (3) talent search firms, and (4) technology service firms. However, unlike the authors, which created a composite measure of ties with multiple service intermediaries, the extent of a relationship with a specific service intermediary in each stage was taken into account.
The authors used general questions instead of the alternative method, which includes listing the names of the service intermediaries, due to the fact that in China some respondents consider such information a corporate secret (Peng and Luo, 2000). A similar approach was used; however the reason for this is related to the time boundaries, in which the research had to be made. Furthermore, each listed service intermediary was accompanied with an explanation and an example.
4. Moderating Variable
To specify in which of the innovation stages are new ventures currently residing, respondents were enquired to choose between the two options provided: (1) adoption and (2) implementation. So that any misinterpretations of the terms are excluded, each of the options was complemented with a short explanation of the activities, associated with the stage. As noted above, new ventures lack the track history, thus are likely to have a “brittle” innovation process
28 (Desouza et al., 2009) and may reside in multiple stages or fail to follow all the stages. Therefore to ensure the quality of the data and its reliability, the possibility to select both of the options was excluded, and respondents were compelled to indicate a distinct stage in which their most recent innovative initiative resides.
5. Controls
In order for this research to build on the findings of Zhang and Li (2010), venture age, founding team size, founding team education, venture size, employee education, foreign invested venture, venture origin and environmental uncertainty were taken as control variables as recommended by the authors. Regarding venture age, the respondents were asked to indicate the number of years, which have passed since the founding of the company. For founding team size, respondents were asked to indicate the number of actual founders and for venture size - they were enquired about the number of full time employees up until the date of survey submission. When it comes to founding team education and employee education the respondents were asked to rate the extent to which each group is comprised of engineers or scientists (e.g. Zhang and Li, 2010). Regarding measuring foreign invested venture and venture origin dichotomous questions were used with the corresponding possible answers: 1 = independent venture, 0 = corporate venture and 1 = yes, 0 = no (whether the venture is independent or not). Environmental uncertainty was assessed using the three measures, applied by Zhang and Li (2010) and adapted form Miller (1978), i.e. asking the respondents “to rate the degree to which they agreed with the following statements regarding their principal industry over the past three years” (p. 97): (1) it has been difficult to forecast how technologies will change in this industry, (2) competitor’s actions have been highly unpredictable, and (3) product market conditions have been changing very fast. As noted by the authors, these statements encompass the pace of change regarding the
29 technology and product market conditions and their predictability in the industries, in which the ventures reside.
V. Results
1. Descriptive statistics.
The first step in the analysis that was taken was related to locating missing data in the dataset. A frequency test was conducted for the following variables: company age, founder count, founder education, employee count, employee education, foreign invested venture, independent venture, environmental uncertainty, stage, ties with service intermediaries and innovation performance. No missing data was found for the given variables. The next step was to analyze the distribution of the data.
The collected data, used to measure the variable Stage - having only 2 categories - Adoption and Implementation - showed that 19 companies were in the first stage of the innovation process (22.4% of the sample), whereas 66 were in the second stage (77.7% of the sample) (Table 3.)
No normal distribution was found except for one of the five constructs comprising the dependent variable Innovation performance: Inno1 (skewness -.390 and kurtosis -.414) (Appendix 4, Table 13. and Graph 5). Inno2 and Inno5 both had a moderate negative skewness and kurtosis range of -0.5 to -1, presenting for a somewhat flat clustering, leaning to the right (Appendix 4; Table 13. and Graphs 6, 9). Inno3 was fairly symmetrical (skewness -.246) and flat (kurtosis -.640), while Inno4 showed a substantial negative skewness (-1.214), thus leaning to the right with a quite peaked distribution (kurtosis 1.224) (Appendix 4, Table 13. and Graphs 7, 8).
30 Ties with accounting and financial firms and law firms had a normal skewness in the range of 0 and 0.5, with a kurtosis between 0.5 and 1, showing for a fairly symmetrical and flat distribution (Appendix 4, Table 13. and Graphs 1, 2). Ties with talent search firms had a substantial positive skewness of 1.249 and a kurtosis of .739, showing for a dense, rather peaked, distribution to the left of the graph (Appendix 4, Table 13. and Graph 3). Ties with technology service firms was symmetrically distributed and flat (kurtosis = -1.238) (Appendix 4, Table 13. and Graph 4). The results from the frequency test can be found on Table 3.
Table 3. Descriptive Statistics of the variables stage, ties with service intermediaries and innovation performance
N Min. Max. Mean Std.
Deviation Stage:
Adoption 85 0 1 .22 .419
Implementation 85 0 1 .78 .419
Ties with service intermediaries: Ties with accounting. and financial firms
Ties with law service firms 85 1 5 2.68 1.147
Ties with talent search firms 85 1 5 1.88 1.149
Ties with technology service firms 85 1 5 3.01 1.332
Innovation Performance:
We are frequently introducing new products (Inno1) 85 1 5 3.45 1.075 We are being first in new product introductions in the
market (Inno2)
85 1 5 3.60 1.197
We are quickly launching new products into the market (Inno3)
85 1 5 3.24 1.172
We are developing new products with superior quality (Inno4)
85 1 5 4.04 1.029
We are using new products to penetrate markets (Inno5) 85 1 5 3.67 1.199
2. Reliability
In order to check the reliability of the studied constructs, a reliability test was performed for the following variables: Innovation Performance and Environmental uncertainty. All the
31 items from the dependent variable innovation performance scored a Cronbach’s alpha between 0.7 and 0.8, indicating for a high level of internal consistency (Pallant, 2007) and making them reliable for use in the analysis. Environmental uncertainty scored a .490, showing a poor reliability scale (Table 4.). However, one could argue, that such an item can be used in the analysis, taking into account the argument, that such a factor has a high “face validity” and “strong factor loadings” (e.g. Zhang and Li, 2010; c.f. Li and Atuahene-Gima, 2002). Thus for the sake of the analysis, a decision was made to include this variable in the following analysis. Finally it should be noted that no negative items were presented in the statistics of both variables, indicating for the absence of counter-indicative items.
Table 4. Statistics of the items comprising innovation performance and environmental uncertainty
Item Cronbach's Alpha if Item Deleted Mean Std. Deviation N Innovation performance: α =.774 Inno1 .710 3.45 1.075 85 Inno2 .760 3.60 1.197 85 Inno3 .712 3.24 1.172 85 Inno4 .756 4.04 1.029 85 Inno5 .721 3.67 1.199 85 Environmental uncertainty: α = .490
It has been difficult to forecast how technologies will change in this industry (Ind1)
.419 3.14 1.093 85
Competitors’ actions have been highly unpredictable (Ind2) .245 2.72 1.031 85 Product market conditions have been changing very fast
(Ind3)
.498 3.54 1.129 85
3. Correlation analysis
Before running the correlation analysis, the means of several variables needed to be estimated and dummy variables for all the dichotomous variables were created. All the items of innovation performance and environmental uncertainty were recoded correspondingly into the
32 new variables InnoTOT and IndTOT. Two dummy variables were computed for the dichotomous moderator stage – Adopt (0 – not in adoption stage, 1 – in adoption stage) and Implem (0 – not in implementation stage, 1 – in implementation stage), thus eliminating the need to include both of the items in the table – only adoption was entered. Regarding the dichotomous control variables - only 4 respondents pointed their company as a foreign invested venture, making for a % 4.7 of the total sample and 6 presented their venture as a corporate one, comprising 7.1 % of the total sample. Thus the variables Foreign Invested Venture and Independent Venture were excluded from the analysis due to low variation within the studied sample.
The next paragraphs will discuss the results from the correlation analysis. First, the independent and dependent variable will be examined, followed by the moderating variable. Second, correlations between some of the items of the independent variable will be discussed. Finally, correlations between the dependent variable and control variables will be presented.
When it comes to the innovation performance, no correlation involving statistical significance between the independent variable ties with service intermediaries and the aforementioned dependent variable was detected. This was the case for all four items of the independent variable, observed separately. As each of the items measures a different relationship construct, one should avoid seeking a correlation between innovation performance and the mean of all the items, comprising ties with service intermediaries. Regarding the moderating variable, Stage, depicted by the dummy variables Adopt and Implement, similar results were presented – there was no correlation, significant at either the p < 0.05 or p < 0.01 levels.
With relation to the four items of ties with service intermediaries, a positive correlation was spotted between law service firms and accounting and financial service firms (r = 0.59; N = 85; p < 0.001) and between technology firms and accounting and financial service firms(r = 0.22;
33 N = 85; p < 0.05), suggesting for multicollinearity. Graham (2003) argued that even low leves of multicullinearity can cause bias in the analysis. He further demonstrated that it can cause (1) inaccurate model parameterization, (2) decreased statistical power, and (3) exclusion of significant predictor variables during model creation. Thus in order to test for multicollinearity, a multiple regression analysis was performend to compute the tolerance statistic for each item of the independent variable. The regression involved only the four items comprising ties with service intermediaries. Each of the items was used as a dependent variable, while the other 3 were input as predictors, making for a total of 4 regression analysis. All four items scored a Tolerance Statistic between 0.6 and 1, thus showing an insufficient amount of variance shared among the constructs of the independet variable, in order to cause a concern for multicollinearity.
Finally, looking at the control variables in the correlations table, no correlations that were statistically significant between the dependent variable innovation performance and the control variables venture age, founder size, founder education, employee size, employee education and environmental uncertainty were found. However, a positive statistically significant correlation between the controls employee education and founder education (r = 0.59; N = 85; p < 0.01) was located. The full set of correlations between all variables is presented in Table 5.
34 Table 5. Correlations
Mean SD 1 2 3 4 5 6 7 8 9 10 11 12
1 Venture age 3.61 1.914
2 Founding team size 2.26 1.216 .018
3 Founding team education 3.65 1.624 .079 .005
4 Venture size (FTE) 7.05 8.000 .128 .185 -.116
5 Employee education 3.72 1.419 -.027 -.026 .587** .057 6 Environmental Uncertainty
(Mean)
3.13 .763 -.035 -.050 -.022 .141 .039 (.49)
7 Ties with accounting and financial firms
3.16 1.163 .043 -.098 .019 .222* .209 .154
8 Ties with law service firms 2.68 1.147 -.060 .085 -.035 .213 .120 .090 .593** 9 Ties with talent search firms 1.88 1.149 -.009 .056 -.137 .258* .060 .154 .139 .089 10 Ties with technology service
firms
3.01 1.332 .135 -.215* -.147 .165 -.162 .080 .222* .143 .273*
11 Adoption .22 .419 .096 -.138 -.110 .036 .027 .141 .046 -.024 .055 .102
12 Innovation Performance (Mean) 3.60 .824 .010 .134 .090 -.013 .128 .113 .053 .022 .211 .026 -.123 (.77) **. Correlation is significant at the 0.01 level (2-tailed).
32 4. Regression analysis
In order to test the hypothesis suggested in the theoretical framework, a hierarchical multiple regression was performed. First the positive relationship between ties with service intermediaries and innovation performance, suggested by Zhang and Li (2010) was tested, as well as the relationship between the control variables and the dependent variable. Second, seeing as the first set of hypothesis (hypothesis marked “a”) test for moderation, interaction variables, between each of the items of the independent variable ties with service intermediaries and the moderating variable stage were created. The final step was to test the second set of hypothesis (hypothesis marked “b”), involving the direct effect between ties with service intermediaries and stage of the innovation process.
To investigate the positive relationship between ties with service intermediaries and innovation performance, a hierarchical regression was performed after first introducing the control variables. In the first step of the regression analysis the following predictors were input: founder size, founder education, employee size, employee education, venture age and environmental uncertainty; thus representing Model 1. In the second step (Model 2) all of the items, comprising the independent variable were entered: ties with accounting and financial service firms, law service firms, talent search firms and technology service firms. Statistical significance was found only for ties with talent search firms at the p < .1 level (b = .142, p = .091 (Table 6.).
33 Table 6. Ties with multiple service intermediaries
Variable Model 1 Model 2
B Sig. B Sig. (Constant) 2.658 .000 2.402 .000 Founder education .004 .953 .027 .708 Employee education .073 .365 .053 .534 Employee size -.007 .557 -.012 .330 Founder size .105 .171 .112 .173
Environmental uncertainty (Mean) .136 .262 .106 .390
Venture age .010 .839 .008 .878
Ties with Accounting and financial service firms .030 .776
Ties with Law service firms -.022 .826
Ties with Talent search firms .149 .091
Ties with Technology service firms .020 .799
R square .053 .098
R square change .053 .045
F .729 .801
Following Zhang and Li (2010) after testing for the effects with multiple service intermediaries, a regression analysis was made for ties with each service firm separately. This meant again including all the controls in the first step of the hierarchical regression and only one item of the independent variable in the second step. However, similar to the results above, none of the analysis showed a statistical significance for any of the service intermediaries with the exception of talent search firms. The regressions showed that accounting and financial service firms, law service firms and technology service firms have no significant relationship to the dependent variable. Only talent service firms showed a statistically significant positive relationship with innovation performance at the 10% level (b = .157, p =.061) and explained a 4.3% variance in innovation performance (Table 7.)
34 Table 7. Ties with Talent Search Firms
Variable Model 1 Model 2
B Sig. B Sig. (Constant) 2.658 .000 2.481 .000 Founder education .004 .953 .027 .703 Employee education .073 .365 .052 .515 Employee size -.007 .557 -.012 .336 Founder size .105 .171 .101 .183
Environmental uncertainty (Mean) .136 .262 .109 .364
Venture age .01 .839 .011 .82
Talent search firms .157 .061
R square .053 .096
R square change .053 .043
When it comes to the control variables, no statistically significant relationship was found between any of the aforementioned variables, and the dependent variable innovation performance.
The next step that was taken was to test the first set of hypothesis, marked “a”, involving the moderator stage. However, as mentioned above, first interaction variables were created between each tie with a specific service intermediary and the stages of the innovation process. First, the interaction variables between the ties with service intermediaries and the adoption stage were created. This resulted in the creation of acc_x_adopt (accounting and financial firms * adoption stage), law_x_adopt (law firms * adoption stage), tal_x_adopt (talent search firms * adoption stage), and tec_x_adopt (technology service firms * adoption stage). Consequently, the interaction variables between the items of the independent variable and the implementation stage were created. Thus, acc_x_implem (accounting and financial firms * implementation stage), law_x_implem (law firms * implementation stage), tal_x_implem (talent search firms * implementation stage), and tec_x_implem (technology service firms * implementation stage) were computed.
A multiple hierarchical regression was computed for each item of the moderating variable stage and its interaction with each item from the independent variable ties with service intermediaries, after controlling for founder size, founder education, employee size,
35 employee education, venture age and environmental uncertainty, making for a total of eight scenarios. The step-wise hierarchal regression for each of the eight scenarios was performed in the following manner – in the first step all of the controls were introduced (corresponding to Model 1); in the second step, one of the two items of the moderating variable stage was introduced, along with one of the items form the independent variable ties with service intermediaries. This step corresponds to Model 2 in the regression. In the third step, corresponding to Model 3 the interaction variable between the two items from Model 2 was entered.
In order to test the second set of hypothesis, marked “b”, again, multiple hierarchical regression was used. Each of the four items of the independent variable ties with service intermediaries was used as an outcome variable in the regression. In the first step, corresponding to Model 1 all the controls were entered, whereas in the second step, corresponding to Model 2 an item of the moderating variable was input. The fact that dummy variables were used for the items of innovation stage, and the fact that the two items are self-exclusive, omitted the need for a second regression with the opposite item, thus resulting in 4 scenarios.
The first hypothesis, involving moderation, which was tested, was: The adoption and implementation processes will moderate the positive relationship between the use of accounting and financial firms, law firms, talent search firms and technology service firms by new ventures, and innovation performance, in terms of enhancing or lessening it. When it comes to the adoption stage, only accounting and financial service firms and law service firms showed significant positive results. Regarding accounting and financial service firms, in Table 8a it can be seen that Model 1 and Model 2 were not statistically significant: F (6,78) = .729, p = .627 (Model 1) and F (8, 76) = .71, p = .499 (Model 2). After introducing the interaction variable acc_x_adopt in step three, the model explained 11.1% of the variance as
36 a whole and was statistically significant at the 10% level F(9, 75) = 1.04, p = .066. The model explained an additional 5.8% variance in innovation performance. Concerning law firms, again, Model 1 (F (6, 78) = .729, p = .627) and Model 2 (F (8, 76) = .705, p = .525) were not statistically significant, whereas Model 3, with the introduction of law_x_adopt variable, was significant and positive (F (9, 75) = 1.372, p = .014 < .05) and explained a 14.8% variance in the dependent variable – an additional 8.8% (Table 9a). The hierarchical regression analysis for the implementation stage, yielded significant negative results for accounting and financial service firms and law service firms, whereas technology service firms (Table 10a) showed significant positive results. Talent search firms (Table 11a) did not provide any significant results for any of the stages. Taking all of the abovementioned into account, Hypothesis 1 received moderate support.
With relation to Hypothesis 2a, which suggested that using accounting and financial firms in the adoption stage will relate to a higher innovation performance, than in the implementation stage, one can see the results in Table 8a. As mentioned above, after introducing the interaction between adoption and ties with accounting and financial service firms, Model 3 shows significant results at the 10% level, explains an additional 5,8% in variance and also shows a positive and significant moderating effect on innovation performance (β = .372, p = .066 < .1). When the results from the interaction between accounting and financial service firms and implementation in Model 3 are examined, one can see a negative significant moderating effect on innovation performance at the 10% level (β = -.372, p = .066 < .1) (Table 8a.). Therefore the adoption stage is associated with a positive relationship between accounting and financial service firms and innovation performance, whereas the implementation stage is related with a negative one, due to the fact, as mentioned above, dummy variables were created for the stages. Thus Hypothesis 2a is supported.
