Business Incubators in the European Union:
Describing the field & exploring performance
measures relationships
M.J.A. van Damme (2504968)
Thesis - MSc. Entrepreneurship
Supervisor: Dhr. Prof. Dr. R.J. van der Voort
August 2016
Table of Contents
1. Introduction ...2 1.1 Background ...2 1.2 Problem Definition ...3 1.3 Scientific Relevance ...3 1.4 Societal Relevance ...4 2. Literature Review ...5 2.1 Definition ...5 2.2 Types ...6 2.3 Practices ...8 2.4 Performance measures ... 11 3. Methodology ... 15 3.1 Research Design ... 15 3.2 Questionnaires ... 15 3.3 Hypotheses... 15 3.2.1 Goals ... 16 3.2.1 Performance criteria ... 16 3.4 Sampling... 18 3.5 Data Analysis ... 18 4. Results ... 19 4.1 Descriptive statistics ... 19 4.2 Hypotheses testing ... 30 4.2.1 Goals ... 30 4.2.2 Performance criteria ... 39 5. Discussion ... 47 6. Conclusion ... 51 Bibliography ... 52 Appendices ... 551. Introduction
Background
During recent years there has been a growth in entrepreneurial activity in the European Union (EU) (Figure 1), judging on the data of the Global Entrepreneurship Monitor (2015). Member states of the EU have been included in the analysis, except for Cyprus and Malta due to lacking data. The average TEA is measured by the average percentage of 18-64 population who are either a nascent entrepreneur or owner-manager of a new business.
Figure 1 Average TEA in the EU (GEM, 2015)
The small and medium-sized enterprises (SME) form the backbone of Europe’s economy, as they represent 99% of all business in the EU. They account for 67% of total private sector employment (EC, 2012) and 58% of gross value added (ECA, 2014). This realization has led to increased attention to the support of these SMEs. An important tool in providing support to young businesses are business incubators (BIs), which are proven to significantly increase the 3-year survival rate; from 58% without (Eurostat, 2016), to 88% with BI assistance from BIs part of the European Business and Innovation Centre Network (EBN, 2013). Therefore, large amounts of financial and regulatory support are given to these BIs by the regional, national and EU government.
The heightened interest in the support of SMEs through business incubation can also be seen in the academic literature, especially regarding the broader return for society on public investments in BIs.
0 1 2 3 4 5 6 7 8 9 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 % Year
Average Total early-stage Entrepreneurial Activity (TEA) in
the EU
The general opinion is that there needs to be a more rigorous evaluation of BI performance, including not only hard outcomes (cf. the measurements in the former paragraph) but also soft outcomes to paint a better picture of success (Dewson et al, 2000; Schwartz & Göthner, 2006; Bergek & Normann, 2008). It appears that financial performance measures are hard to be tested, because the main share of incubators and incubatees are new small firms of which long term data is not available (Bearse, 1998). Next to that, the measurement of performance is different for every stakeholder of the BI. It is thus more relevant to look at specific stakeholder’s goals and put them into perspective of their performance measures (Cornelius & Bhabra-Remedios, 2003).
Problem Definition
However, the literature is lacking an overview of the relationship between BI type and its goals. A key factor in the goals of a BI appears to be the sponsoring institution. But, the influence of sponsoring institutions on a BIs goals is yet to be discovered (Cornelius & Bhabra-Remedios, 2003). This research addresses this problem by categorizing a sample of BIs in the EU based on their types and practices, followed by an analysis on the possible relationship between the categories, and the goals and performance measures of the BI. The goals are measured by a questionnaire asking BI managers to type their organization and to rate the importance of several goals and performance measures to their organization. Guided by these objectives the following research questions are proposed:
RQ1: “How can the business incubators in the European Union be categorized, considering their types and practices?”
RQ2: “Is the type of the business incubator related to the goals perceived by the management of the business incubator?”
The literature review will provide the different categories and goals, after which hypotheses will be presented.
Scientific Relevance
Cornelius & Bhabra (2003) state that there is a gap in our knowledge about the impact of differing institutional sponsors. They note that this is a key factor in developing a performance evaluation framework for BIs. Furthermore, Bergek & Normann (2008) note that to determine the performance of different incubators it needs to be related to their incubator models. They also see a general demand for more rigorous evaluations of BIs, which is also noted by the academics due to the increase of
focus and different aspects and put different weights on performance criteria. This research tries to further enrich the literature by testing the relationships between the types of BIs, and their goals and performance criteria. It may help to better understand the differences in perceived performance for BIs. This may also help to determine appropriate weights for performance models such as PROMETHEE (Schwartz & Göthner, 2009).
Societal Relevance
Performance frameworks are mostly used by institutions such as governments to determine the effect of their sponsorship. An improvement in the understanding of the relevance of different performance measures for different types of BIs will be useful for those institutions. Being able to put the types into relation with their goals provides a better foundation for developing new frameworks. Not only sponsors use performance measures, but also BI managers themselves to determine the effect of their operations. It is therefore also relevant for them to better understand the domain of BI performance.
Next to that, it is relevant for those institutions to get a new overview of the BI field. E.g. for the EU government as the European Commission (EC) stopped updating their database of incubators in 2006, when the system was terminated. Due to this lack of continuity the EC does not have an up-to-date overview. This is especially useful for EU member states that are developing their business incubation network (EC, 2014). The recommendation in this paper is for the EC to update their knowledge by engaging in knowledge exchange.
2. Literature Review
This section tries to give an overview of the literature and relate the different incubator definitions, types, and practices. The constructed frameworks will guide as a template for the questionnaires.
Definition
The definitions of BIs in the literature mainly differ in width, goals, and position in the business lifecycle. In the broad definition BIs are considered to be any type of organization that focusses on providing an environment that fosters and supports entrepreneurship, where the narrow definition places constraints such as assistance only for very young startups.
During the early days of incubator research there was a strong focus on the physical aspects of a BI, rather than its activities. Also, the level of technology is assumed low. This is reflected by the definition of the European Commission used in a 2002 BI benchmarking paper: “‘A place where newly created
firms are concentrated in a limited space. Its aim is to improve the chance of growth and rate of survival of these firms by providing them with a modular building with common facilities (telefax, computing facilities, etc.) as well as with managerial support and back-up services. The main emphasis is on local development and job creation. The technology orientation is often marginal” (European Commission,
2002).
More focus on the activities is proposed by the definition of Hackett & Dilts (2004) in which they acknowledge that a BI is “a shared office space facility that seeks to provide its incubatees with a
strategic, value-adding intervention system of monitoring and business assistance”, while stressing that
a BI is more than just a physical environment and infrastructure. Namely, it also encompasses the network of individuals and organizations that work together in the BI environment. Aernhoudt (2004) agrees on BI being a larger concept than just an office space, but focusses more on the “interactive development process” which is supportive to start-up growth and innovation. The European Commission apparently agreed on the larger concept as it defines an incubator in 2010 as a place to facilitate entrepreneurs in “addressing their needs and develop their business ideas, and transform them
into sustainable realities”. An interesting addition is the note that an incubator can still be an incubator
even if no psychical services are offered; i.e. virtual incubation which offers its services online. The shift of attention in the literature from facilities and administrative services to business support was also acknowledged by Bergek & Normann (2008), who believe business support is the most important factor while co-location is an important advantage. They furthermore consider incubation related to the early phase of a venture’s life, excluding fully developed business ideas (cf. Klofsten, 2005).
Aaboen (2009) & Scillitoe & Chakrabarti (2010) follow the logic of aforementioned definitions about both offering space and business support, but see incubatee firms as ‘knowledge- and
technology-intensive new technology-based firms (NTBFs)’. The reason for this is the evolution of BIs over the years
following the observations of Aerts (2007), who describes the first generation as focussed on job creation and real estate appreciation, and the second as more service and network oriented. NTBFs represent the third generation of BIs, which have a stronger focus on high-tech and ICT. Bruneel et al. (2011) research on BI evolution however did not include NTBFs in the generations, while agreeing on the aforementioned shift from focus on physical infrastructure (first generation 1950s to 1980s), to business support (second generation in the 1980s), to networks (third generation from 1990s onwards). A recent definition of a BI from the European Court of Auditors (ECA, 2014) seems to implement most of the aforementioned aspects:
“A business incubator is an organisation designed to support the successful establishment and further
development of enterprises. It often offers access to physical business infrastructure, individually tailored business support services and networking opportunities”.
This definition does however lead to some confusion when looking at the concept of a business accelerator. It is defined as “A fixed-term, cohort-based program, including mentorship and educational components, that culminates in a public pitch event or demo-day” (Cohen & Hochberg, 2014). It is argued that it distinguishes itself from a BI by the fixed length, the intensity, the provision of a stipend and services, and the cohort-based nature. However, when applying the business accelerator concept to the definition of the ECA it would still fit, as it can be considered as the further development of enterprises. This clearly reflects the difference between a narrow and a wide definition of BIs. For this research the broad definition of the ECA (2014) will be used to get a full overview of BI types and practices in the EU.
Types
Choosing a broad definition means there will be a large variety of organizations that classify as a BI. The literature has a large body different taxonomies based on characteristics such as objectives, practices, sponsors, and development stage. To distinguish them from each other the different typologies in the literature will be reviewed here.
The European Commission (2002) provided us with a typology matrix (Figure 1) based on the level of technology and management support. The grey corner on the bottom represents the categories in
which BIs mostly find themselves. This assumes that BIs usually provide a high level of management support to incubatees with technology-based enterprises, although not always the case.
Figure 2. Typology of Business Incubators (European Commission, 2002)
The paper is continued with a distinction between ‘traditional’ and ‘new economy’ BIs. ‘Traditional’ BIs facilitate economic development by promoting entrepreneurship, innovation, employment opportunities and growth, which are mostly operated by national or local authorities. On the contrary, the profit-driven ‘new economy’ BIs are often virtual (offering no physical workspace) and are mainly funded by venture capital companies or developed by large consultancies to capitalize their expertise. Their focus is mostly on high-tech and internet-related activities, rather than the general focus of the ‘traditional’ BI. The distinction between the two clearly incorporates different characteristics; i.e. sponsors, goals, and sector.
Another effort to categorize BIs is done by Bøllingtoft & Ulhøi (2005) using an adapted model (Figure 2) from Allen & McCluskey (1990). The first building block of this model has been developed by Brooks (1986), building on the aforementioned distinction between BIs as mainly real estate providers and BIs as business assistance servicers. It is a two-type incubator continuum in which the start-up will first enter an ‘economic growth incubator’ to gain access to incubators resources and network, after which it will move into ‘real estate incubator’. Allen & McCluskey (1990) disagreed with this assumed sequential action and modified the model to a value-adding continuum with a distinction in for-profit and non-profit and objectives (cf. Figure 2). This is derived from the ‘facility life cycle model’ in Allen’s
finding tenants. When the BI is financially healthy they enter the business development stage where the emphasis shifts to managing the tenant firms. The maturation stage assumes a sophisticated enterprise support network and demand for additional user space. These stages imply that the management style of BIs changes focus when targeting different groups of tenants or when tenants evolved within the BI. Bøllingtoft & Ulhøi (2005) further developed the model by introducing value added trough collaboration and the types of collaboration for each specific BI. This model follows a broader definition of a BI than the model of the European Commission (2002), as it defines all of the organizations mentioned in that model as a BI.
Figure 3. The Business Incubator continuum (Bøllingtoft & Ulhøi, 2005)
Practices
Bergek & Normann (2008) observe that practices among BIs with similar goals differ, which implies that there are different views on how to achieve the goal best. This also shows the difference between types of BIs and their practices, as a similar type of BI can have different practices. It is therefore useful to also identify BIs based on their practices.
The ECA (2014) also makes a distinction based on the three typical phases of incubation practices. During the pre-incubation stage the objective is to increase the chance that entrepreneurs are able to create a business effectively, and proceed successfully to the next stage. The focus is mainly on developing business ideas, models, and plans. The incubation stage starts when the venture is created and lasts until it becomes self-sustainable and ready to operate independently, mostly around 3 years. Here the main services are access to finance, training and coaching for entrepreneurs, networks, office space, and sometimes access to laboratories, workshops, and prototyping facilities. The post-incubation phase is entered when the businesses are able to continue operations with external support. However, several services may still be needed to further develop their business.
Figure 4 Stages of the incubation process and their activities (ECA, 2014)
In an effort to develop a framework to describe the differences between incubator models Bergek & Normann (2008) distinguished 5 components; i.e. selection, infrastructure, business support, mediation, and graduation. First of all, selection concerns the actions taken in accepting or rejecting entry to ventures. Hackett & Dilts (2004) described five commonly used selection criteria: prior employment & experience of the entrepreneur, technical expertise of the entrepreneur, properties of the market where the venture is aiming at, properties of the product/service, and the profit potential of the venture. Bergek & Normann (2008) conclude that on the one hand BIs select primarily based on either the entrepreneur or the idea, where on the other hand the strictness of criteria is either based on ‘survival of the fittest’ or on ‘picking the winners’. With the rigid ‘picking the winners’ approach BIs come close to private venture capital firms, which ex ante perform a deep analysis on a venture’s potential success.
work. Business support refers to coaching and training of incubatees. Distinction is made based on the role of the BI in the incubation process. At one extreme the BI uses strong intervention, providing strong guidance and sometimes full management teams. At the other extreme a laissez-faire approach is used, were business support is refrained from unless the incubatee itself takes the initiative. Mediation is concerned with making the connection between incubatees and the outside world with the purpose of leveraging entrepreneurial talent and/or resources (Bøllingtoft & Ulhøi, 2005; Grimaldi and Grandi, 2005). This can be done through network mediation, in which the BI matches incubatees with other actors (Brooks, 1986; von Zedwitz, 2003). Another style is to engage in institutional mediation where the BI mediates the impacts of institutions on the incubatees, amplifying the positive and mollifying the negative (Hackett & Dilts, 2004b). Lastly, graduation is associated with exit policies, stating when incubatees should leave the BI. In practice it seems to be mostly time driven, while others focus more on the achievement of benchmarks (Colbert et al, 2010). Bergek & Normann (2008) note that infrastructure and graduation do not differ enough in practice and are therefore irrelevant to include in the model. However, infrastructure does seem to differ nowadays when looking at the rise of ‘virtual business incubators’ (EC, 2012). Also, when looking at the frameworks regarding BI types it seems that different types of BIs operate in different life phases of the venture, with corresponding goals and graduation policies. As this research is focussed on a broad set of BIs all of the five components will be used (as in Figure 3).
BI Practices Model
Selection Infrastructure Business Support Mediation Graduation Entrepreneur vs. Idea Localities Strong intervention vs. Laissez-faire Network vs. Institutional Time vs. Benchmarks Survival of the fittest
vs. Picking the
winners Physical
Administrative
Figure 5 Adapted Incubator Model of Bergek & Normann (2008)
Bergek & Normann (2008) also point out that several BIs’ mediation activities are limited to certain regions, while others are more international and rather focussing on a specific technological field (cf. Carayannis and von Zedwitz, 2005; Clarysse et al, 2005). They therefore also distinguish them based on the innovation system they primarily connect to; i.e. regional/national or technological/sectoral innovation systems. When both categories apply they are seen as a cluster.
Performance measures
In the early days of research on the performance of BIS Allen & Weinberg (1988) described that the ultimate goal of any BI is to reduce the failure rate of small business. In turn this has a positive effect om job creation, economic diversification, product commercialization, and other development outcomes. In this view the success of a BI can be measured using a ratio: (Number of firms exiting the BI/Number of firms discontinuing operations while still a tenant), to then be compared to their peers. In 2003 Wiggins & Gibson added to the literature by proposing a broader set of tasks that BIs have to accomplish in order to succeed. First of all, clear metrics need to be established in order to measure success. There also need to be a provision of entrepreneurial leadership and they need to develop and deliver value-added services to member companies. Furthermore, a rational selection process needs to be developed. Lastly, the BI needs to ensure access for the incubatees to human and financial resources. Justification for this multi-model approach is given by Peters et al. (2004), as the model that was developed by them to measure BI success showed significant differences in the number of incubatees graduating from different types of BIs. Although the significance is low, it is noted that probably the success of incubators is related mostly to the presence and quality of coaching and access to networks. They assume that when the characteristics and quality of these services are taken into account there will be a significant effect.
Hackett & Dilts (2004) provide an interesting outcome framework for incubated businesses (Figure 4). When looking at this model it is clear that graduating or terminating from a BI has different impacts depending on their situation. It is possible for and business to be terminated, but still have some benefits when the losses are minimized. Then the business still has a chance to ‘reincarnate’ after entrepreneurial lessons have been learned.
Businesses applying for incubation Possible outcomes for incubated businesses
a. Those that do not need incubation 1. Incubatee surviving and growing profitably
b. Those that cannot be helped through incubation
2. Incubatee surviving and growing towards profitably
c. Those that should be incubated through some resource gap: the “weak-but-promising” having a good business case, but lack of resources
3. Incubatee has survived, no growth, no profit = (“zombie business”)
4. Incubatee operations terminated while still in incubator, but losses minimized
5. Incubatee terminated while still in the incubator and losses were large
Figure 6 Outcome for incubated business (composed from Hackett & Dilts, 2004)
This adds to the point that to measure success properly, a more holistic approach needs to be taken rather than a simple graduation/termination ratio. Not only hard outcomes, which are clearly definable and quantifiable, are needed but also soft outcomes that represent the intermediate stage on the way to achieving the hard outcome. The consideration of soft outcomes gives valuable context for clients’ needs and progress, depicting a more valid and full picture of successes (Dewson et al, 2000). This is supported by Voisey et al. (2006) based on their observations of the Graduate Teleworking Initiative (GTi) project to examine success of a BI. They note that some positive outcomes were not represented in the quantitative methods used, showing a need for a qualitative element. This is especially relevant for non-profit or on government support reliant BIs to be able to receive a continuing stream of support. The need for a multi-element approach towards BI performance was also acknowledged by Bergek & Normann (2008). They stated that most of the previous BI assessment literature was focussed on the outcomes, while they believe there needs to be a shift towards a holistic approach where the goals of the incubators are taken into account and their performance is put in relation to their incubator models. This is because practices differ among BIs with similar goals, indicating that there are different opinions on how to achieve a certain goal. By putting BIs with similar goals in relation to relevant outcome indicators, it is in their opinion possible to determine which practices are most effective in reaching the goals.
This view is supported by Cornelius & Bhabra-Remedios (2003), as they elaborate on the three different approaches in organizational theory. The goal-based approach evaluates an organization’s performance based on the achieved of the goals set (cf. Etzioni, 1964). However, because goals differ among
organizations cross-organizational comparison is problematic. The systems approach compensates for this problem by using multiple generic performance measures for the whole industry (cf. Georgopoulos and Tannenbaum, 1957). This approach does make comparison easier, but fail to address differences between stakeholder group perspectives on performance (Murphey et al., 1996). The multiple constituency approach (cf. Thompson, 1967) tries to mitigate these downsides by factoring in these differences. The different stakeholders’ goals and performance measures are depicted in Figure 2. Although each of the stakeholders has its own goals, the ultimate goal of a BI is venture growth. This shows a potential for conflicting interests in achieving venture growth, due to the differences.
Figure 7 Model depicting different BI stakeholders related to goals & performance (Cornelius & Bhabra-Remedios, 2003)
An effort to create a multidimensional evaluation tool on the effectiveness of BIs is done by Schwartz & Göthner (2009) by applying the PROMETHEE outranking method (Brans et al., 1986) to the performance of BIs (see Figure 5). This method is deals with the appraisal and selection of a set of variables based on several criteria, with the objective of identifying pros and cons of the alternatives and obtaining a ranking among them. Outranking therefore refers to the degree of dominance of one variables over the other (Vincke, 1992), and limits the degree to which a disadvantage on a particular viewpoint may be compensated by advantages on another viewpoint (Pirlot, 1997). Schwartz & Göthner (2009) note that it is important that incubatee-level indicators are considered along with incubator-incubatee-level
average incubation time of graduate firms, measuring the underlying function and the acceleration of the entrepreneurial process. Secondly, the share of start-ups is measured in a region as a major goal for less-favoured regions is to have a BI foster and support the formation of new ventures. The third element is the share of high-tech firms to evaluate the technological competence or innovativeness of the incubator firms, measured by the relative R&D costs and knowledge-based business-related services. Furthermore, the client satisfaction is measured through surveys. The criteria to evaluate here are rental space, collectively shared facilities and services, business assistance, and networking activities. The next element is the overall survival rate of incubated businesses, which is the core of most performance measures of BIs. Lastly, the effect on the environment is accounted for by calculating the employment growth after graduation. It is noted that the results are heavily influenced by the preference functions and threshold values and that an optimal configuration does not exist. However, the framework does show potential for evaluating BI performance and for public institutions to base their support programs on. Special attention needs to be paid to the input and the preference structures of the model. This differs for different stakeholders, but is not included in the model. Improvement of the quality of the results is possible when their preferences are included. This research also focusses on the performance criteria preferences of BI managers, adding to the improvement of PROMETHEE.
PROMETHEE Evaluation Criteria 1. Average incubation time 2. Share of start-ups 3. Share of high-tech firms 4. Client satisfaction 5. Overall survival
6. Employment growth after graduation
3. Methodology
Research Design
As this paper tries to paint a picture of the BI landscape and performance measures in the EU this research is using a quantitative method; i.e. questionnaires. The rationale for this is that to get a comprehensive view on the large body of EU BIs it is the method that fits the time given for the research to gather a rich dataset. It is a formal study because the frameworks from the literature are used as a template for the data collection and the data tries to provide a valid representation of the current state and answer the research questions. The method of collection is sending out emails to as much as possible EU BIs with a request to fill out the questionnaire online and this is therefore a communication study. The time dimension is cross-sectional, as the data collection is done at the same time representing a snapshot of one point in time. This all occurs under actual environmental conditions, seen there is no staging or manipulation of conditions (Blumberg et al, 2014).
Questionnaires
The questions in the questionnaire are based on and follow the same structure as the frameworks presented in the literature review (See Appendix III). The first part consists of demographics such as location, size, amount of incubatees, industry (following the Global Industry Classification Standard), and the position of the employee. The second part is dedicated to the typology frames regarding sponsorship (Bøllingtoft & Ulhøi, 2005), incubation process stage (ECA, 2014), virtual/physical (EC, 2010), BI evolution (Bruneel et al., 2011), management/technical support (EC, 2002), and the BI continuum (Bøllingtoft & Ulhøi, 2005). The third part considers the practices frames regarding selection, business support, mediation, and graduation (Bergek & Normann, 2008). The last part measures the importance of different goals (Allen & Weinberg, 1988; Wiggins & Gibson, 2003; Peters et al., 2004) and the relative importance of the evaluation criteria from PROMETHEE (Schwartz & Göthner, 2009). The software used to develop and distribute the questionnaires is the online Qualtrics Survey Software, of which access is provided by the business department of the UvA.
Hypotheses
To be able to answer RQ2 several hypotheses are proposed in this paragraph, which will be tested in the results section. They are divided between an assumed relationship between types and goals, and between types and performance criteria used. The H0 is formulated as that there is no effect to be able
3.2.1 Goals
The main assumption is that the type of sponsorship of a BI has a strong relationship with the goals of the BI. Thus:
(I) H0: The type of sponsorship a BI receives is unrelated to its goals.
H1: The type of sponsorship a BI receives is related to its goals.
It can further be argued that the stages of the incubation process for which the BI provides services is related to its goals.
(II) H0: The stage of the incubation process in which the BI operates is not related to its goals.
H1: The stage of the incubation process in which the BI operates is related its goals.
The BI continuum of (Bøllingtoft & Ulhøi, 2005) categorizes 5 types of incubators. These types may be related to a BI’s goals, as they are partly determined by sponsors. This will be tested through:
(III) H0: The type of BI, according to the BI continuum, is not related to its goals.
H1: The type of BI, according to the BI continuum, is related to its goals.
As BIs are also categorized by its practices, it is interesting to discover possible relationships between their practices and their goals. The approach of the BI for the practices mentioned by Bergek & Normann (2008) guides the categorization. Based on this, the following hypothesis is proposed:
(IV) H0: The approach of BIs to its practices is not related to its goals. H1: The approach of BIs to its practices is related to its goals.
3.2.1 Performance criteria
As it is assumed that the type of sponsorship is related to a BI’s goals, it can also be assumed that the type of sponsorship is related to the importance of specific performance measurement criteria. Therefore:
(V) H0: The type of sponsorship a BI receives is not related to the criteria used to measure its
H1: The type of sponsorship a BI receives is related to the criteria used to measure its
performance.
Another assumption is that the stages of the incubation process for which the BI provides services is related to its performance measures. E.g. when a BI focusses only on the pre-incubation phase, it would be logical that the ‘Employment growth caused by incubated ventures after graduation’ is of lower importance for the BI than the ‘Share of start-ups in the region’. This leads to:
(VI) H0: The stage of the incubation process in which the BI operates is not related to the
criteria used to measure its performance.
H1: The stage of the incubation process in which the BI operates is related to the criteria
used to measure its performance.
The BI continuum typology is assumed to be related to its performance criteria used, following the same rationale that it is partly based on sponsors.
(VII) H0: The type of BI, according to the BI continuum, is not related to the criteria used to
measure its performance.
H1: The type of BI, according to the BI continuum, is related to the criteria used to measure
its performance.
The possible relationship between the approach of BI practices guided by the framework of Bergek & Normann (2008) will also be tested through the following hypothesis:
(VIII) H0: The approach of BIs to its practices is not related to the criteria used to measure its performance.
H1: The approach of BIs to its practices is related to the criteria used to measure its performance.
Sampling
The unit of analysis is the BI in the EU, conforming the definition given in the literature review. Because the BI itself cannot fill in the questionnaire a manager of the BI is asked to do this and thereby represent the BI. For the sake of lower cost, greater accuracy of results, greater speed of data collection, and availability of population elements a sampling approach is taken. To improve accuracy, the goal is to receive as many as possible responses in order to have a sufficient amount of elements in the sample to reduce the systematic variance. Therefore, 697 emails have been sent to EU BIs with a request to fill in the questionnaire. To reduce the standard error of estimate and represent the population in all respects, different sources are used to collect contact details of EU BIs (See Appendix I). By choosing a variety of sources the precision of the sample is increased. Because the common factor of the sources is the internet it means that the sampling method is non-random/non-probability, because each member of the sample does not have a known non-zero chance of being included. However, because an important aspect of BIs in innovation one can assume that most of them make use of the internet by advertising themselves there. Despite the efforts of sending out numerous emails, reminders, and phone calls the positive response rate was just 5.6%, with 39 respondents, of which 37 are useful after inspection of the data. This means that the accuracy might be low due to the small sample of the entire population. However, the data still provides indicators of possible relationships. These are then to be tested on a larger sample in future research.
Data Analysis
To measure certain relationships and provide descriptive statistics of the sample, the SPSS 23 (IBM) and the Excel 2016 (Microsoft) software package are used. The data in from the questionnaire is checked by hand and edited if needed, e.g. for missing data, wrong decimal separators, incorrect values. The results section will start with the demographics of the sample and some overall descriptive statistics. Then the relationships are tested through a Pearson two-tailed bivariate correlation analysis, with the significance level at <0.05. When the p-value is less than the significance level, it can be concluded that the observed effect reflects characteristics of the population due to the lower sampling error. Especially when it is <0.01, as this indicates a high significance.
4. Results
Descriptive statistics
The sample consists of 37 valid responses, with the BIs in 13 different countries (Table 1). Note that the total amount of countries is different than the sample size is because one BI was based in 3 countries.
BI Location (n=37) n Cyprus 1 Denmark 1 France 2 Germany 5 Ireland 1 Italy 1 Lituania 1 Netherlands 6 Norway 1 Portugal 4 Spain 2 Sweden 6 United Kingdom 8 13 39
Table 1 Location BIs
The founding year range of the BIs is relatively large, with a width of 32 years (Table 2). The mean is around 2004 and the median 2006, indicating a negative skew with more BIs in the recent years of the range. Altogether the mean age is 11.65 years.
Year founded (n=37) Range 1983-2015 Mean 2003.97 Median 2006 Mean age 11.65
Table 2 Founding year BI
The range of the amount of employees is also relatively large, with a width of 69 employees (Table 3). When comparing the mean with the median it can be concluded that there is a positive skew. There are many BIs with up to 10 employees and a few with relatively larger numbers.
Amount of employees
(n=36)
Range 1-70
Mean 11.34
Median 6
Table 3 Amount of employees in BIs
The respondents were asked about the amount of currently and previously incubated ventures (Table 4). The range here is very large with a width of respectively 3994 and 3998, and heavily positively skewed. When examining the data, it can be seen that one BIs has serviced to 4000 incubated ventures and still does. This distorts the overall picture of the sample, regarding incubated ventures. To create a better picture of the rest of the sample, this outlier has been removed and the values have been recalculated (Table 5). This is still positively skewed, but to a lower amount. This is also reflected in the lowered standard deviation. It indicates that there are several BIs in the sample with very large amounts of incubated ventures. Incubated ventures (n=35) Current Past Range 6-4000 2-4000 Mean 182.37 345.16
Median 32 100
Stdev 672.57 775.28
Table 4 Incubated ventures BIs are servicing to currently and in the past
Incubated ventures (Adjusted) (n=31) Current Past Range 6-500 2-2000 Mean 70.09 223.33 Median 30 100 Stdev 106.91 381.87
Table 5 Incubated ventures BIs are servicing to currently and in the past (Adjusted for outlier)
Appendix II shows the industries in which the BI’s incubated ventures operate, including their relative share (low, medium, or high). The largest industries of the incubated ventures are Software & Services (85%), Health Care Equipment & Services (79%), Technology Hardware & Equipment (72%), Commercial & Professional Services (66%), Energy (58%), Media (61%) and Consumer Services (52%). As all those values are over 50%, it means that most BIs in the sample have to some extent incubated ventures operating in these industries. However, only for the Software & Services industry over 50% is reported as a high share. This leads to the conlcusion that most BIs in the sample have a high share of incubated ventures operating in the Software & Services industry. The most reported highest share of industry is continued by the Technology Hardware & Equipment (33%) and Commercial & Professional Services (33%), indicating that around a third of the incubated ventures in the BIs operate in these fields.
The type of sponsorship is dominated by government support, with local government sponsorship on top with 28.73% of total sponsorship (Table 6). The median here is close to the mean, although positively skewed. In total the amount of government sponsorship as a share of the total sponsorship is 52.09%, which shows that governments provide the largest share of sponsorship. University sponsorship comes second, closely followed by for-profit organizations. The sponsorship of non-profit organizations is relatively low with 8.06%. Other types of sponsorship mentioned by the respondents are: private investment, own resources of consortium members, and customers/entrepreneurs. The medians for
than those that do. The standard deviation is high for most of the types, which means that the observations relatively differ a lot from the mean shares of types of sponsorship. Except for non-profit organizations, which seems to be relatively consistent in the sample.
Types of sponsorship (%)
(n=33)
n Mean Median Stdev
Universities 16 18.03 0 27.59 Local government 22 28.73 21 30.72 National government 15 14.61 0 26.74 EU government 13 8.76 0 18.16 Non-profit organizations 14 8.06 0 13.05 For-profit organizations 16 17.67 0 29.27 Other 4 4.28 0 18.09
Table 6 Types of sponsorship provided to BIs
The main incubation process stage in which BIs operate is the incubation stage, as most respondents indicate operating in this stage and because it has the largest share on average (Table 7). However, most BIs operate in multiple stages, judging on the count of BIs for each stage. The data is evenly distributed around the, as the means and medians are similar.
Incubation process stages (%)
(n=37)
n Mean Median Stdev
Pre-incubation 29 23.90 24 19.57
Incubation 35 46.73 50 19.28
Post-incubation 33 29.37 29 23.67
Table 7 Incubation process stage in which BIs operate
Psysical services
(n=37)
n %
Yes 31 84
No 6 16
Table 8 Offering of physical services by BIs to incubatees
The respondents ranked business support out of the three offered services the highest, as a lower score indicates that it’s closer to the number 1 ranking (Table 9). Networks are less important in respect to the services offered in BI, followed by physical infrastructure as the least important.
Importance offered services (n=33)
Mean Stdev
Physical infrastructure 2.21 0.82
Business support 1.72 0.84
Networks 2.06 0.75
Table 9 Importance of BIs services offered to incubatees
The share of management support offered in relation to the total support in the BIs is mostly high or otherwise medium (Table 10). Low management support is observed in just 10.81% of the cases. Technological support however is mostly medium or low, with a low amount of BIs with high technological support.
Share of support (%)
(n=37)
Low Medium High
Management support 10.81 32.43 56.76 Technological support 35.14 51.35 13.51
Table 10 Relative share of management and technological support provided by BIs to incubatees
The mean of the ranking of objectives has to be interpreted as that 1 is the most important and 9 the least important for BIs (Table 11). The most important factor on average is creating jobs and enhancing
standard deviations. The least important objective is selling proprietary services to tenants, and second least important real estate appreciation. However, real estate appreciation has the highest standard deviation of the sample, indicating the BIs disagree the most on this objective of the exact place but still ranks low.
Ranking of objectives
(n=37)
Mean Stdev
Real estate appreciation 6.62 2.59
Job creation & enhancing of the entrepreneurial climate 2.54 1.89 Capitalize collaborative and symbolic potentials 4.57 1.86 Commercialization of university research 5.38 2.25
Capitalize investment opportunity 5.16 2.30
Sell proprietary services to tenants 7.19 1.93
Regional/area development 4.43 2.38
Network development & nurture 3.08 1.89
Secure availability to risk capital 6.03 2.28
Table 11 Ranking of BI objectives
The main relative share of average collaboration is the highest for for-profit organizations, although when combining the local, national, and EU government this is the largest share with 33.39% (Table 12). The lowest shares of government collaboration is found at the EU and the national level, both below the other types of support (excluding the ‘Other’ category). Collaboration with universities is also well represented, being close to the share of for-profit organizations. The lowest share is for non-profit organizations, if the government collaborations are combined. The standard deviation is high for collaborations with universities and for-profit organizations, showing a large variance from the mean in the sample. Other collaborations mentioned are research organizations, and other EU initiatives and projects. Collaboration with (n=36) n Mean Stdev Universities 32 25.82 24.22 Local government 29 17.92 15.00
National government 20 9.18 11.31
EU government 17 6.29 10.48
Non-profit organizations 30 12.75 11.94 For-profit organizations 27 26.49 25.17
Other 3 1.56 5.50
Table 12 Share of BI collaboration with other organizations
The results in Table 13 need to be interpreted as that a value of 0 means the selection criteria is of no importance, where 4 means the highest importance. The values are relatively close to each other and fall within the category between moderately important to very important. Most important are the properties of the product/service, with the lowest standard deviation. It seems that BIs agree that this factor is of most importance. It is closely followed by properties of the product or service, and further by the properties of the market where the venture is aiming at. Relatively the least important is considered the prior employment and experience of the entrepreneur, with second least important the technical expertise of the entrepreneur. Overall the focus is in the selection of incubatees is on the idea rather than the entrepreneur.
Importance selection criteria
(n=37)
Mean Stdev
Prior employment & experience of the
entrepreneur 2.14 1.23
Technical expertise of the
entrepreneur 2.32 1.11
Properties of the market where the venture is aiming
at 2.57 1.09
Properties of the
product/service 2.84 0.99
Profit potential of the
venture 2.70 1.13
Most of the BIs believe that the markets will provide selection processes over time by separating the winners from the losers, believing in the survival of the fittest principle (Table 14). However, still a large share (40.54%) believes selection of incubatees needs to be done based on strict criteria, which is in line with the picking the winners approach.
Selection based on
(n=37)
n %
Strict criteria 15 40.54
Market forces 22 59.46
Table 14 Selection of incubatees by BIs based on strictness vs. market forces
Although most BIs indicated that they select incubatees based on market forces, the relative amount of incubatees that get selected over the amount that applied is low (Table 15). A little more than third gets accepted, but note the high standard deviation. This reflects the large difference in acceptance rate between the approaches before mentioned.
Selected incubatees of total applying (%)
(n=37)
Mean 37.81
Stdev 30.13
Table 15 Percentage of incubatees that get selected after applying
In Table 16 a value of 0 means that there is no intervention with business support unless the incubatee takes the initiative, and a value of 4 means that the incubatees are heavily guided with business support. On average the BIs tend to guide the ventures to some extent, although not heavily (2.51).
Business support approach
(n=37)
Mean Stdev
Level of intervention 2.51 1.07
The extent to which BIs engage in network and institutional mediation (Table 17) ranges from 1 (none at all) to 4 (a great deal). For both categories BIs reported that it ranges from a moderate amount to a lot, with respectively values of 2.81 and 2.24. Network mediation therefore is the most used in the sample. Mediation (n=37) Mean Stdev Network 2.81 0.94 Institutional 2.24 0.98
Table 17 Mediating activities provided by BIs to incubatees
Most BIs connect to respectively regional and technological innovation systems, with just a fifth connection to the notional innovation system (Table 18). 32.43 % reported that they both connect to regional or national and technological innovation systems, indicating a cluster.
Connection to innovation systems
(n=37) n % Regional 25 67.57 National 8 21.62 Technological 21 56.76 Cluster 12 32.43
Table 18 Connection of BIs to innovation systems
The time aspect in BIs graduation policies are most prevalent with about half of the sample reporting this as the most important factor (Table 19). Just 10.81% of BIs have graduation policies mostly influenced by benchmarks. Furthermore, almost 3 out of 10 BIs reported that they do not have specified graduation policies. Other policies mentioned where based on commercial opportunities, a combination of both, and based on the specific needs of each incubatee to maximize its value.
Graduation policy (n=37) n % Time 19 51.35 Benchmarks 4 10.81 Not specified 11 29.73 Other 3 8.11
Table 19 Graduation policies in place at BIs
The values in Table 20 are based on a scale of importance ranging from 0 (not important at all) to 4 (extremely important). The most important goal is considered to be providing access to networks, which on average is reported as in between very important and extremely important. This is followed respectively by developing and delivering value-added services to incubatees, providing access to financial resources, providing high quality coaching, providing entrepreneurial leadership, reducing failure rate of small businesses, and providing access to human resources which all range from moderately important to very important. Goals considered slightly to moderately important are reducing the average incubation time, developing a rational selection process, and selecting incubatees based on intuition. The standard deviations show that the BIs disagreed most on the importance of developing a rational selection process, and reducing the average incubation time.
Importance goals
(n=37)
Mean Stdev
Reducing failure rate of small businesses
2.32 1.29
Providing entrepreneurial leadership 2.54 0.99 Developing & delivering value-added
services to incubatees
2.89 1.15
Providing access to networks 3.24 0.86
Providing high quality coaching 2.70 0.97 Accelerating the entrepreneurial
aspect by reducing the average incubation time
Developing a rational selection process
1.59 1.32
Selecting future incubatees based on intuition
1.19 1.17
Providing access to financial resources
2.70 1.08
Providing access to human resources 2.08 0.95
Table 20 Importance of BIs goals/objectives
On a scale of 0 (fully process focussed) to 4 (fully results focussed) BIs on average reported 2.49 (Table 21). This shows that the results are slightly more important to BIs than the process.
Process vs. Results
(n=37)
Mean 2.49
Stdev 0.96
Table 21 Focus of BIs on process vs. results
Table 22 shows the how important each of the performance criteria are to the BI (with 0 as not at all important and 4 as extremely important). The overall survival rate of incubated businesses is on average considered to be very important to extremely important. Second most important is considered client satisfaction with a value indicating that it is considered very important. Employment growth caused by the incubated ventures after graduation is considered moderately important to very important. The share of high-tech firms in incubated ventures is reported as moderately important, with the share of start-ups in the region as just a little less important. The least important is considered the average incubation time which is in between slightly important to moderately important. The most disagreement of BIs can be found in the share of high-tech firms, share of start-ups in the region, and employment growth after graduation.
Table 22 Importance of performance criteria to evaluate BIs effectiveness
Hypotheses testing
The different variables of the hypotheses are tested for correlation and provided here in the tables. Significant results are displayed and evaluated to be able to accept or reject the hypotheses.
4.2.1 Goals
(I) H0: The type of sponsorship a BI receives is unrelated to its goals.
The data shows us several possible relationships between the type of sponsorship a BI receives and the goals it has. There is a weak positive correlation between the relative share of for-profit sponsorship and the importance of the goal to provide entrepreneurial leadership. A little stronger positive correlation is found between the for-profit sponsorship and selecting future incubatees based on intuition (Table 23).
Providing entrepreneurial leadership
Selecting future incubatees based on intuition Relative share for-profit Pearson Correlation ,354* ,397*
Importance performance criteria
(n=37)
Mean Stdev
Average incubation time 1.49 1.10
Share of start-ups in the region 1.73 1.24 Share of high-tech firms in
incubated ventures
1.95 1.29
Client satisfaction 2.97 1.04
Overall survival rate of incubated businesses
3.08 1.04
Employment growth caused by incubated ventures after graduation
organizations sponsorship Sig. (2-tailed) .043 .022 N 33 33
Table 23 Correlation and significance analysis for hypothesis (I)
Furthermore, national government sponsorship seems weakly to moderately negatively related to the importance of proving access to networks, just as the importance of providing high quality coaching (Table 24).
Providing access to networks Providing high quality coaching Relative share national government sponsorship Pearson Correlation -,412* -,432* Sig. (2-tailed) .017 .012 N 33 33
Table 24 Correlation and significance analysis for hypothesis (I)
Also, there is a weak relationship between the share of non-profit sponsorship and the importance of providing access to financial resources (Table 25).
Providing access to financial resources Relative share non-profit sponsorship Pearson Correlation ,363* Sig. (2-tailed) .038 N 33
All of the mentioned relationships are significant at the 0.05 p-value level, indicating there is evidence to reject H0. There seems to be at least some relationship between the share of sponsorship and the
importance of goals.
(II) H0: The stage of the incubation process in which the BI operates is not related to its goals.
The results only indicate a possible weak to moderate negative relationship between the share of which the BI operates in the post-incubation phase and the importance of the goal to reduce the failure rate of small businesses (Table 26). This does prove evidence to some extent that the incubation process phase is related to at least one goal. Therefore, we can reject H0 as the p-value is significant.
(III) H0: The type of BI, according to the BI continuum, is not related to its goals.
The analysis shows there is a moderately positive relationship between the ranking of the objective to secure availability to risk capital and the goal to reduce the failure rate of small businesses (Table 27). Keep in mind that this means that the lower the respondents ranked the risk capital objective, the higher they rated the importance of reducing the failure rate of small businesses.
Reducing failure rate of small businesses Secure availability to risk Pearson Correlation ,469** Sig. (2-tailed) .003
Reducing failure rate of small businesses
Post-incubation phase Pearson Correlation -,405* Sig. (2-tailed) .013 N 37
capital objective
N 37
Table 27 Correlation and significance analysis for hypothesis (III)
For the ranking of the objective of network development & nurture in relation to the goal to provide access to network it means that when the objective is ranked higher, the goal is reported more important (Table 28). The share of collaboration with universities shows to be weak negatively related to the importance of the goal to provide access to networks. The share of collaboration in relation to the goal with for-profit organizations shows the opposite effect, i.e. a weak positive relationship.
Providing access to networks Network development & nurture objective Pearson Correlation -,421** Sig. (2-tailed) .009 N 37 Collaboration with universities Pearson Correlation -,375* Sig. (2-tailed) .024 N 36 Collaboration with for-profit organizations: Pearson Correlation ,376* Sig. (2-tailed) .024 N 36
Table 28 Correlation and significance analysis for hypothesis (III)
Another weak negative relationship is found between the share of collaboration with universities and the importance of providing access to financial resources (Table 29).
Providing access to financial resources Collaboration with universities Pearson Correlation -,346* Sig. (2-tailed) .038 N 36
Table 29 Correlation and significance analysis for hypothesis (III)
Based on the significant relationships H0 can be rejected, as there is some relationship between the type
of BI according to the BI continuum and the importance of BIs goals.
(IV) H0: The approach of BIs to its practices is not related to its goals.
Selection
The approach to selecting incubatees is in several respects related to the goals of a BI. The importance of prior employment and experience of the entrepreneur has a weak positive relationship with the importance of the goal to provide entrepreneurial leadership (Table 30). The relationship with the importance of the goal to provide entrepreneurial leadership is stronger positively related with the importance of the technical expertise of the entrepreneur in selecting incubatees, with a highly significant p-value. The relationship to the importance of this goal is even stronger with the importance of the selection criterion regarding the profit potential of the venture, with again a highly significant p-value.
Providing entrepreneurial leadership Prior employment & experience of the entrepreneur Pearson Correlation ,373* Sig. (2-tailed) .023 N 37
Technical expertise of the entrepreneur Pearson Correlation ,470** Sig. (2-tailed) .003 N 37 Profit potential of the venture Pearson Correlation ,422** Sig. (2-tailed) .009 N 37
Table 30 Correlation and significance analysis for hypothesis (IV)
The importance of the profit potential of the venture in selecting incubatees has a weak positive relationship with the importance of the goal to reduce the average incubation time (Table 31).
Furthermore, there are weak positive relationships between the profit potential criterion with the goal to reduce the average incubation time and the goal to develop a rational selection process (Table 32).
Developing a rational selection process Profit
potential
Pearson Correlation
,364*
Reducing the average incubation time Profit potential of the venture Pearson Correlation ,372* Sig. (2-tailed) .024 N 37
of the venture Sig. (2-tailed) .027 N 37
Table 32 Correlation and significance analysis for hypothesis (IV)
Also there is a weak to moderately positive relationship between the attitude towards selection based on market forces with the goal to provide access to financial resources (Table 33).
Providing access to financial resources Selection based on market forces Pearson Correlation ,391* Sig. (2-tailed) .017 N 37
Table 33 Correlation and significance analysis for hypothesis (IV)
Business support
The level of business support intervention has a weak relationship with the goal to reduce the average incubation time (Table 34). Furthermore, the relationship of this level with the goal to provide access to networks is weak to moderately positive. A moderate positive relationship is found between this level and the goal to provide high quality coaching, with a high significance level.
Providing access to networks
Providing high quality coaching
Reducing the average incubation time Business support intervention Pearson Correlation ,402* ,473** ,352* Sig. (2-tailed) .014 .003 .033 N 37 37 37
Mediation
The goal to provide entrepreneurial leadership is moderately positively related to the level of institutional mediation, and weak to moderately positively related to the connection with the national innovation system (Table 35).
Providing entrepreneurial leadership Institutional mediation Pearson Correlation ,518** Sig. (2-tailed) .001 N 37 National innovation system Pearson Correlation ,382* Sig. (2-tailed) .020 N 37
Table 35 Correlation and significance analysis for hypothesis (IV)
Next to that, the goal to provide access to networks has a weak positive relationship with the level of network mediation (Table 36).
Providing access to networks Network mediation Pearson Correlation ,333* Sig. (2-tailed) .044 N 37
Also, the goal to select future incubatees based on intuition is weakly positively related to the connection with the technological/sectoral innovation system (Table 37).
Selecting future incubatees based on intuition Technological/Sectoral innovation system Pearson Correlation ,331* Sig. (2-tailed) .045 N 37
Table 37 Correlation and significance analysis for hypothesis (IV)
The goal to provide access to financial resources has a weak positive relationship with the level of network mediation, and a weak to moderate positive relationship with the connection to the national innovation system (Table 38).
Providing access to financial resources Network mediation Pearson Correlation ,355* Sig. (2-tailed) .031 N 37 National innovation system Pearson Correlation ,394* Sig. (2-tailed) .016 N 37
Lastly, there is a weak to moderate positive relationship between the level of institutional mediation and the goal to provide access to human resources (Table 39). A stronger positive relationship with this goals is found with the connection to the national innovation system, with a high significance level.
Providing access to human resources Institutional mediation Pearson Correlation ,393* Sig. (2-tailed) .016 N 37 National innovation system Pearson Correlation ,443** Sig. (2-tailed) .006 N 37
Table 39 Correlation and significance analysis for hypothesis (IV)
Altogether, there are enough indications that there are several relationships between the approach of the BI to its practices and their goals. Thus, we can reject H0 here.
4.2.2 Performance criteria
(V) H0: The type of sponsorship a BI receives is not related to the criteria used to measure its
performance.
Based on the analysis there is a moderate positive relationship between local government sponsorship and the importance in assessing performance based on the share of start-ups in the region, with a high significance level (Table 40).
Share of start-ups in the region Local government sponsorship Pearson Correlation ,444** Sig. (2-tailed) .010 N 33
Table 40 Correlation and significance analysis for hypothesis (V)
The category ‘Other sponsorship’ is also significant and shows a weak to moderate positive relationship (Table 41), however just 3 respondents indicated that they receive other types of support (Table 19). Therefore, this result can be seen irrelevant.
Employment growth caused by incubated ventures after graduation Other sponsorship Pearson Correlation -,427* Sig. (2-tailed) .013 N 33
Table 41 Correlation and significance analysis for hypothesis (V)
H0 is still rejected as there is strong evidence for the relationship between local government support
and the importance of the criterion regarding the share of start-ups in the region.
(VI) H0: The stage of the incubation process in which the BI operates is not related to the
criteria used to measure its performance.
All the p-values of the correlation analysis between the stage of the incubation process in which the BIs operate and the importance of performance criteria were insignificant. Therefore, we can accept H0,
(VII) H0: The type of BI, according to the BI continuum, is not related to the criteria used to
measure its performance.
The importance of the average incubation time performance criterion is weak positively related to the ranking of the objective to secure availability to risk capital (Table 42).
Average incubation time Secure availability to risk capital Pearson Correlation ,339* Sig. (2-tailed) .040 N 37
Table 42 Correlation and significance analysis for hypothesis (VII)
There also is a positive weak relationship between the ranking of the objective to secure availability to risk capital and the importance of the criterion based on the share of start-ups in the region (Table 43). This criterion also has a relationship with the share of collaboration of the BI with the local government, which is a little stronger and more significant.
Share of start-ups in the region Secure availability to risk capital Pearson Correlation ,367* Sig. (2-tailed) .026 N 37 Local government collaboration Pearson Correlation ,415* Sig. (2-tailed) .012 N 36
Furthermore, the importance of the criterion client satisfaction has a weak to moderate positive relationship with the ranking of the objective to secure availability to risk capital and a moderate negative relationship with a BI collaborating with universities, which is highly significant (Table 44). This indicates that a higher share of university collaboration, means less importance given to the client satisfaction criterion and the other way round. This is the opposite for BIs collaborating with the EU government, as the analysis shows a weak to moderate positive relationship.
Client satisfaction Secure availability to risk capital Pearson Correlation ,399* Sig. (2-tailed) .015 N 37 Universities collaboration Pearson Correlation -,478** Sig. (2-tailed) .003 N 36 EU government collaboration Pearson Correlation ,393* Sig. (2-tailed) .018 N 36
Table 44 Correlation and significance analysis for hypothesis (VII)
Lastly, there is a weak to moderate positive relationship between the share of BI collaboration with non-profit organizations and the importance of the share of high-tech firms in incubated ventures criterion (Table 45).
Share of high-tech firms in incubated ventures Non-profit organizations collaboration Pearson Correlation -,391* Sig. (2-tailed) .018 N 36
Table 45 Correlation and significance analysis for hypothesis (VII)
There are several indications given that there indeed is a relationship between the type of BI, according to the BI continuum, and the criteria used to measure its performance. Based on this H0 is
rejected.
(VIII) H0: The approach of BIs to BI practices is not related to the criteria used to measure its performance.
Selection
The importance of the technical expertise of the entrepreneur in the selection process for incubatees is moderately positively related to the importance of the performance criterion regarding the share of high-tech firms among the incubated ventures, which is highly significant (Table 46).
Share of high-tech firms in incubated ventures Technical expertise of the entrepreneur Pearson Correlation ,499** Sig. (2-tailed) .002 N 37
Also there is a weak positive relationship between the importance of the properties of the product or service of the venture when selecting incubatees and the importance of the overall survival rate of incubated business criterion (Table 47).
Overall survival rate of incubated businesses Properties of the product/service Pearson Correlation ,366* Sig. (2-tailed) .026 N 37
Table 47 Correlation and significance analysis for hypothesis (VIII)
A last relationship regarding the approach to selecting incubatees is the importance of profit potential of the venture and the importance of the performance criterion employment growth caused by incubated venture after graduation, which is weak positive (Table 48).
Employment growth caused by incubated ventures after graduation Profit potential of the venture Pearson Correlation ,371* Sig. (2-tailed) .024 N 37
Table 48 Correlation and significance analysis for hypothesis (VIII)
Business support
The level of business support intervention shows weak positive relationships with the performance criteria of average incubation time, share of high-tech firms in incubated ventures, and client satisfaction (Table 49).
Average incubation time Share of high-tech firms in incubated ventures Client satisfaction Business support intervention Pearson Correlation ,373* ,383* ,362* Sig. (2-tailed) .023 .019 .028 N 37 37 37
Table 49 Correlation and significance analysis for hypothesis (VIII)
Mediation
The connection with regional innovation systems is weak positively related to the share of start-ups in the region performance criterion (Table 50). A moderate relationship is found for the connection with this innovation system and the criterion regarding employment growth caused by incubated ventures after graduation and is highly significant.
Share of start-ups in the region
Employment growth caused by incubated ventures after graduation Regional innovation systems Pearson Correlation ,366* ,444** Sig. (2-tailed) .026 .006 N 37 37
Table 50 Correlation and significance analysis for hypothesis (VIII)
There also is a relationship between the connection of a BI with technological or sectoral innovation systems and the share of high-tech firms in incubated ventures performance criterion, which is weak to moderately positive (Table 51).
