Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system

Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Technovation journal homepage: www.elsevier.com/locate/technovation

Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system Enrico Baraldi n, Malena Ingemansson Havenvid Department of Engineering Sciences, Uppsala University, Sweden Department of Industrial Economics and Technology Management, NTNU and the STS Centre, Uppsala University, Sweden

art ic l e i nf o

a b s t r a c t

Article history: Received 29 April 2014 Received in revised form 10 August 2015 Accepted 12 August 2015

Relying on an in-depth case study of the incubator related to the Swedish medical university Karolinska Institute's (KI), this paper identifies new analytical and strategic dimensions of incubation. Departing from the current literature's prevalent focus on incubators as organizations performing a predefined set of activities for incubatees (facility renting, coaching, training and connecting), we perform a multilevel analysis embracing, next to the organizational and the project-specific levels, also the broader institutional and inter-organizational level. Our analysis relies on seven key components of incubation, namely its time, place, sources, resources, control/governance, activities/services and outcomes. Further, we view incubators as strategic actors engaged in value creation on a broader arena than the strict incubation context, even an international arena, where incubators' choices and interactions can be analyzed with the help of concepts from various streams in the business strategy literature. The specific strategic drivers of business incubation that we identify in the KI incubator's case are six: positioning in the value chain, risk taking/time perspective, revenue model, governance/control, internationalization, and cooperation/ competition. The paper concludes with managerial implications urging incubators to take more of a strategic perspective rather than focussing only on the established components of their operations. & 2015 Elsevier Ltd. All rights reserved.

Keywords: Karolinska Institute Incubator Incubation system Value creation Strategy Network Pharmaceutical

1. Introduction The literature on business incubators focuses mostly on the organizational entity which performs a set of activities or services for incubated firms, such as facility renting, coaching, training and networking (Mian, 1996; 1997; Rice, 2002; Peters et al., 2004; Bergek and Norrman, 2008; Bøllingtoft, 2012; Vanderstraeten and Matthyssens, 2012, Lai and Lin, 2015). Further, the literature analyses certain dimensions characterizing the behavior of incubators, such as technological level and management support (Smith and Zhang, 2012: 228-9), internal resources (Somsuk and Laosirihongthong, 2014) or selection strategies, business support and mediation (Bergek and Norrman, 2008: 23-5), in order to categorize them into typologies (e.g., generalist Vs specialist, see Vanderstraeten and Matthyssens (2012)), incubating models (see e.g., Grimaldi and Grandi (2005): 113-5) and historical generations (Bruneel et al., 2012). While this research has the merit to describe the role and the functioning of incubators, it does so with an internal focus on incubators' operations as performed in relation to incubatees, rather n

Corresponding author. E-mail addresses: [email protected] (E. Baraldi), [email protected] (M. Ingemansson Havenvid).

than capturing the entire context embedding incubators. There are indeed studies that take a more holistic perspective on incubators (e.g., Etzkowitz, 2002) and consider external stakeholders (Alsos et al., 2011), although most often delimiting the analysis to universities (Rothaermel and Thursby, 2005a, 2005b). Other stakeholders, such as public agencies, financiers and established firms, may be acknowledged as part of the incubator′s context (see Mian (1997): 259), but their interactions with the incubators are not penetrated in detail in the extant literature. Moreover, when the context is considered, it appears to be restricted to universitycentred networks (see e.g., McAdam and Marlow (2008)) or the local community and region (see e.g., Ehret et al. (2012)), or it is expressed in abstract terms as an innovation or technological “system” (Bergek and Norrman, 2008: 24-5). Finally, within these views of the incubating context, incubators are assumed to be functionally bound to perform only activities strictly related with incubation, namely firm selection (see Aerts et al. (2007)), business support (see Vanderstraeten and Matthyssens (2012)) and mediation/networking (see Ahmad and Ingle (2011)). However, investigating the development, functioning and broader embedding context of a particular case of incubator, which does not fit the established analytical categories, offers the opportunity to expand our understanding of the phenomenon of business incubation. This is the case of the incubator related to the

http://dx.doi.org/10.1016/j.technovation.2015.08.003 0166-4972/& 2015 Elsevier Ltd. All rights reserved.

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

2

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Swedish medical university Karolinska Institute (KI): since the late 1990s, KI has built its own innovation-supporting system (Stone and Frank, 2001), encompassing a highly integrated set of organizations dealing with selecting medical discoveries across the Nordic countries and turning them into spin-offs, which KI's incubator directly controls and finances for up to 10–15 years with large funds, especially since one of these support organizations has been listed on the Stock Exchange (Baraldi and Waluszewski, 2011). KI's incubator includes the provisions not only of business support services but also of highly specialized drug discovery services to the incubated companies (Baraldi et al., 2014), which KI's incubator runs as time-bound drug development projects constantly monitored as if it were a venture capitalist (Baraldi and Ingemansson, 2013). KI's incubator can hardly be categorized according to the existing literature′s models and dimensions (e.g., Grimaldi and Grandi, 2005; Bergek and Norrman, 2008), including the three recognized generations of business incubators (Bruneel et al., 2012; Smith and Zhang, 2012: 230). In fact, as we shall discuss later, KI's incubator blends the features of several established models and is at best an outlier due to its extreme focus on science and a single industry (biomedical), maximal intervention in startups, strictest focus on “picking winning ideas” (Bergek and Norrman, 2008: 24), and extremely long incubation times. Moreover, KI's incubator includes a scope of activities which transcends those of a typical incubator, such as conducting actual product development and international operations. Against this background, this paper conducts an in-depth qualitative case study (Yin, 2003) of KI's incubator with the dual purpose of identifying (1) new analytical dimensions capturing the broader embeddedness and (2) strategic drivers addressing the long-term development of business incubators, thereby providing a more holistic perspective on this phenomenon. While most literature functionally restricts the role of incubators to just a few well-defined activities, with limited interfaces to the external context (typically to a local “entrepreneurial ecosystem”, Fetters et al., 2010), this paper views incubation as a multifaceted and more complex phenomenon, entailing or indeed requiring broader and deeper external interfaces. As for the first purpose, we perform a multilevel analysis embracing the broader institutional (Lundvall, 1999; Whitley 1994, 2000; Casper and Kettler, 2001; Casper and Matraves, 2003) and inter-organizational level (Anderson et al., 1994; Håkansson and Ford, 2002; Gadde et al., 2003), as well as the organizational and project-specific level of KI's incubator. In this way we can capture both the broader and the restricted context of incubation, including its connections to a local entrepreneurial ecosystem (Fetters et al., 2010; Abetti and Rancourt, 2008: 391). Research embracing the entire context of incubation is advocated by Hackett and Dilts (2004: 74) and Phan et al., (2005: 177), who stress the importance of investigating various levels of analysis ranging from national to regional innovation systems and from the incubator's environment to the single incubated firm or project (Ibid: 177). Thus, the first contribution of this paper is a three-level analytical scheme connecting incubation and its components especially to its broader embedding context, viewed not simply as a system to which incubators need to adapt, but as specific actors with whom incubators interact (Håkansson and Ford, 2002). As for the second purpose, Vanderstraeten and Matthyssens (2012: 656) suggest that “Changes in the incubation market have prompted scholars to devote more attention to how incubators can strategically position themselves”. Following these authors and Aaboen (2009) who view incubators as strategic actors, our analysis of such actors transcends the typical, operational incubation activities and embraces the strategic level. In this context, incubators' choices and interactions concern mainly the creation of

value in relation to external actors (Normann and Ramirez, 1993; Ramirez, 1999) and can be analyzed using concepts from various streams in the business strategy field: for instance, positions (Mintzberg, 1987) and positioning in a value system (Stabell and Fjeldstad, 1998), revenue models (Osterwalder and Pigneur, 2010), international operations (Andersson et al., 2002), and cooperation/ competition with other actors (Baraldi et al., 2007; Gnyawali and Park, 2011). Thus, our second contribution is shifting the focus from incubators′ daily operations (selection, support and mediation) to key strategic choices affecting their long-term development, expressed as a set of six strategic drivers extracted from the KI's incubator case. This study features an incubator which is highly specialized (Schwartz and Hornych, 2008) in bio-medicine, with the strong academic ties that operating in this industry entails (Schwartz and Hornych, 2010). Despite the diffusion of “bio-incubators” (over 20 in Germany and the UK alone, Ehret et al., 2012: 302), with a few notable exceptions, such as the work of Cooke et al. (2006) who compare 8 bio-incubators, research focussing explicitly on this particular type of incubators is still limited. Thus, the third contribution of this paper addresses this gap in the incubation literature by providing a detailed empirical account of a bio-incubator, stressing the unique challenges and industrial context which this type of incubators face. The paper is organized as follows: after developing our theoretical framework for a multi-level analysis of incubators, we present our methodology; then we delve into the case of KI's incubator, followed by a discussion section which identifies six strategic drivers of business incubation. We conclude the paper with suggestions for further research and practical implications.

2. Previous literature on incubators and theoretical framework This section first identifies a set of components characterizing the operations of business incubators, then it reviews the incubator literature in relation to the selected components (Section 2.1) and finally it combines them with three levels of analysis (Section 2.2). The literature provides a wealth of definitions of business incubators, which can be helpful to single out the central components in this concept. For instance, “organizations that supply joint location, services, business support and networks to early stage ventures” (Bergek and Norrman, 2008: 22; our italics); “a value-adding intervention system” that “…controls and links resources with the objective of facilitating the successful new venture development” (Hackett and Dilts, 2004: 57, our italics); or “a place where specific professional resources are organized to help the emergence and first development of new companies” (Albert and Gaynor, 2006: 134). Thus, place, resources, services, control and time issues appear as common themes in the incubator literature. While confirming the relevance of these five components, Grimaldi and Grandi (2005: 115) also add “origins of incubated ideas”, which correspond to the sources sustaining a prototypical “incubation process”, as the one proposed by Hackett and Dilts (2004): 57, Fig. 1) and leading to a set of incubation outcomes, measurable according to frameworks such as the one proposed by Mian (1997). In other words, a business incubator can be viewed as any organization (cf. Bergek and Norrman, 2008: 22) or even a company (Phan et al., 2005: 175) following a basic scheme of “input-processing-output” to create value (Ramirez, 1999) in a certain time and space, by controlling certain resources necessary for its operations/activities. Place, time, sources, resources, control (or governance), services (or activities) and outcomes are accordingly seven key components of incubation appearing as recurring themes in

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

3

the literature on this phenomenon. The next Section 2.1 systematizes and discusses the business incubation literature according to these seven components, while the following Section 2.2 combines these seven components with three levels of analysis inspired by Hackett and Dilts (2004: 63), namely an institutional/ inter-organizational, an organizational and a single project/incubatee level. These three levels define the layered context embedding incubators. Our theoretical framework encompasses therefore the seven key components as viewed from the three levels of context. We conclude this theoretical section by indicating that these key components can be combined to capture the choices that constitute the strategic drivers of incubation, which however will be extracted from our case study and elaborated in the discussion Section 5.

relationships to tenant firms. They moreover contribute to transferring and connecting the knowledge resources of universities to start-ups (Ibid; Rothaermel and Thursby, 2005a; Rubin et al., 2015). However, in order to achieve this, incubators need to hold a base of material and immaterial resources (Aaboen, 2009: 664; Mian, 1997: 257-8; Somsuk and Laosirihongthong, 2014), with the latter, such as business development expertize and reputation (Mian, 1996), playing an increasingly important role. While sharing material resources creates scale economies for incubators and incubatees (Bruneel et al., 2012: 111), a highly immaterial resource, such as the incubator′s network enables access to external resources, such as legitimacy and knowledge (Ibid: 112), especially via university linkages (Rothaermel and Thursby, 2005b). Resources are accordingly the fourth key component of incubation.

2.1. The key components defining business incubators and incubation

2.1.5. Control and governance These issues are our fifth key component of incubation, becoming salient in relation to how the goals of an incubator are set, including the choice between for-profit or non-profit, which is in turn related to the private, public or mixed ownership/sponsorship of the incubator (Peters et al., 2004; Grimaldi and Grandi, 2005). The presence of stakeholders with diverging goals, all trying to control the incubator, may require clear-cut governance rules so to avoid paralyzing conflicts (Phan et al., 2005: 178). On the other hand, the incubator itself forms a “controlled environment” (Ibid: 179; Hisrich, 1988) and exert a direct control over its incubatees via its monitoring function (Vanderstraeten and Matthyssens, 2012: 665).

2.1.1. Place Established definitions of incubators stress that incubators entail a shared environment and space (Hackett and Dilts, 2004: 57; Bruneel et al., 2012: 110), a place which is both physical and organizational (Smith and Zhang, 2012: 228) and provides logistical support to start-ups as well as the opportunity to collaborate with other organizations (Oakey, 2012: 68). Thus, a first component of incubators and incubation is place. 2.1.2. Time This is another important component of incubators for several reasons. Incubators contribute to the entrepreneurial spin-out process by ideally shortening its time (Clarysse et al., 2005); incubation is expected to be a temporary affair for a firm, as witnessed by graduation policies and exit rules after 3–5 years (Mian, 1997: 281; Bergek and Norrman, 2008: 23); each incubator also tends to provide increasingly better support to incubatees the longer it has been in operations thanks to learning effects (Allen & McCluskey, 1990), while the whole population of incubators has evolved with time across three generations entailing increasingly sophisticated and less physically-bound services to incubatees (Bruneel et al., 2012). 2.1.3. Sources Since incubation can be viewed as a process per se (see Hackett and Dilts (2004): 57) or as part of a larger process of entrepreneurial spin-out (Clarysse et al., 2005; Berbegal-Mirabent et al., 2015) or an even broader innovation journey (Van de Ven et al., 1999), a third important component of incubators are the sources which aliment this process (see Grimaldi and Grandi (2005): 115). To start with, incubators have screening practices (Aerts et al., 2007) applied to select the venues to be incubated. Moreover, incubators apply these selection rules to different potential sources of venues, ranging from strictly scientific ideas originating from a university to any entrepreneurial idea (Mian, 1997; Smith and Zhang, 2012: 229), and distinguishing between a focus on the quality of the entrepreneur or of the very idea (Bergek and Norrman, 2008: 23-4). Further, in defining their “tenant mix”, incubators can choose between specializing on specific technological domains and being generalists (Schwartz and Hornych, 2008; 2010), and between accepting many venues and picking only the expected winners (Bergek and Norrman, 2008: 23-4; Clarysse et al., 2005). Thus, sources are the third key component of incubation. 2.1.4. Resources According to Aaboen (2009: 657) incubators provide, mobilize and adjust such resources as office space, competence, finance and

2.1.6. Activities/services Looking at the operations of incubators, the sixth key element of incubation are the activities they perform, especially as services provided to incubatees (Aaboen, 2009: 666-8). For instance, Bergek and Norrman (2008: 23-5) summarize these activities into selection, business support and mediation, which can be performed in different ways thereby defining different typologies and generations of incubators (Bruneel et al., 2012). While space rental is a basic activity, incubatee selection is performed with different degrees of sophistication (Vanderstraeten and Matthyssens, 2012). Similarly, business support may include several other activities, ranging from counseling and coaching to education and access to financing (Mian, 1996; Aernoudt, 2004: 127). Finally, mediation refers to the incubator's intermediary role in connecting incubatees to other organizations in its external network (Bergek and Norrman, 2008: 24-5). 2.1.7. Outcomes Reflecting the incubation literature strong interest in assessing incubators and the many indicators suggested (Ibid: 22), the final component of our framework are the outcomes of the incubation process, corresponding to the values created internally by the incubator. Some outcomes concern the early end of this process, such as occupancy rates, others concern the performance of incubated firms (capitalization, patents etc.) all the way to graduation, and other outcomes concern the economic status of the incubator, such as revenues, funds and financial position (Mian, 1997: 281). 2.2. Three analytical levels and three types of contexts for incubation The seven components of incubation (place, time, sources, resources, control/governance, activities/services and outcomes) characterize the single incubator, which forms the primary context of incubation. However, the very same components are also dimensions that characterize the single incubatee (a project or a company), and the broader context that embeds incubators: in other words, those seven components are dimensions found at various levels of analysis, as visible also in the review of the business incubation literature by Hackett and Dilts (2004: 63). The

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

4

importance of investigating innovation at different levels of analysis is advocated among others by Drazin and Schoonhoven (1996) and, for the specific phenomenon of incubation, by Phan et al. (2005). However, while the latter authors propose four levels of analysis, namely entrepreneurial team, incubated firm, incubator and “systemic” (Ibid: 166, 177), we opt for three levels only, as we conflate teams and incubated firm within a single level. On the other hand, we specify as part of the “systemic” level both the broader institutional context (Whitley, 2000; Casper and Matraves, 2003) and the specific inter-organizational connections (Axelsson and Easton, 1992; Anderson et al., 1994) embedding the incubator. Another tripartite classification of the context embedding incubation is proposed by Hackett and Dilts (2004). Compared to their three analytical levels of incubatee, incubator and community (Ibid: 63, 66, 69), the first two levels fully correspond to our levels, whereas the third level we propose – the institutional and interorganizational context – is not geographically restricted to a locality but can embrace also transnational connections and institutions. Thus, while including the local “entrepreneurial ecosystem”, often centred on a university (Fetters et al., 2010; Abetti and Rancourt, 2008: 391), our third level of analysis also extends beyond such geographical boundaries. The institutional/inter-organizational level of analysis indicates how the incubator interacts with external actors and is accordingly highly relevant for its strategy, since focusing just on internal issues (such as incubatees or internal organization) misses its dependence on the external context (Pfeffer and Salancik, 1978). Moreover, analyzing the interface between incubator and the external context also reveals how incubators can expand their boundaries by internalizing resources and activities which are usually found in the surrounding ecosystem (as KI's incubator did with venture capital). Section 2.1 has reviewed the seven components of incubation departing from the organizational level (see the mid column of Table 1), namely the incubator's internal mechanism in terms of inputs, resources, processes (e.g., selection criteria and activities) and outputs. The place, the time, the sources, the resources, the control, the activities and the outcomes of business incubation assume different meanings if considered at the analytical level of the single incubatee/project or at the institutional and inter-organizational level. This section focuses accordingly on the relevant facets of each component as viewed in these two other contexts (see Table 1).

2.2.1. Single incubatee level (as a “project”) The place dimension for the single incubated firm includes both the co-location with other firms enabling networking and cooperation (Bøllingtoft, 2012) and economies of scale thanks to shared physical resources (Bruneel et al., 2012: 111), and the closeness to universities providing access to laboratories and reputational effects (Mian, 1996). As for time, incubatees are companies with a very short history and whose development, indeed survival, requires solving in the short term a set of critical junctures (Vohora et al., 2004: 168), often expressed also as milestones. The key sources that sustain the development of incubatees are the inventor (often a scientist) behind the core idea and the entrepreneur conducting the business (who may or may not be the same person, see Lundqvist (2014) on “surrogate entrepreneurs”), coupled with formal owners and financiers (Gompers and Lerner, 2001). Important resources for incubatees include ideas and knowledge (Rothaermel and Thursby, 2005a), IPRs such as exclusive licenses (Ibid: 308), a team with experience, equipment (Grimaldi and Grandi, 2005), relationships, such as those to universities (Mian, 1996; Rothaermel and Thursby, 2005b), and finance (Gompers and Lerner, 2001). As for the control and governance of the single incubated firm, there are several actors next to the incubator itself who exert influence, such as monitoring, setting project milestones and evaluating the new venture (Davila, 2005). These include the firm's founders, but also financiers, both public and private such as VCs (Ibid), and major customers or development partners (Tomkins, 2001; Dekker, 2004). Next to technology and business development, the main activities of incubatees include learning (Bruneel et al. 2012: 112), networking (Ibid; Bøllingtoft, 2012), and selling, if the venture approaches the threshold of sustainability (Vohora et al., 2004). Finally, the relevant outcomes at this level of analysis are the growth in terms of sales and the sustainability of the venture (Ibid; Roig-Tierno et al., 2015), but the opposite, namely failure or project termination, may well be the actual outcome. There are also a series of intermediate outcomes that can be even more valuable for the single incubatee, such as acquisition of funds (capital or grants), recruitment of key personnel and advisors, obtaining IPR protection and other technical achievements. 2.2.2. Institutional and inter-organizational level The broader context of incubation highlights how innovation depends on complex systems including key actors, such as knowledge providers, mediators and users (Edquist, 1997;

Table 1 Seven key components of incubation across three analytical levels. Single incubatees (the project)

Organizational level (the incubator)

Institutional and inter-organizational level

Place

Co-location

Facilities, campus/park

Global, national, regional

Time

Short-term, milestones

Exit/graduation rules, learning, 3 generations

Long-term, path-dependence

Sources

Entrepreneurs, scientists/inventors, owners, financiers

Scouting & selection. University Vs open, science, Several incubators and universities. Transnaideas Vs people. Specialist Vs generalist tional networks

Resources

Ideas, knowledge, IPRs, team, experience, Physical space, expertize, management, reputaequipment, relationships (univ. linkages), tion, legitimacy, relationships (to various stakeholders), resource transfer finance

Multiple technologies and supply chains, distributed competences, central network nodes

Control/governance

Incubator, founders, financiers (public, VC), potential customers, development partners

Private, public, mixed. Profit Vs no-profit. Influence/monitor incubatees. Powerful stakeholders

Distributed across several stakeholders. Several actors in the network try to control

Activities/services

Learning, business & technology development, networking, selling

Selection, space rental, business support, counseling and coaching, financing, mediation & networking

Development, production and use of technologies. In-licensing, company acquisitions, regulations, standardization

Outcomes (values created)

Obtaining funds and IPR protection, recruitments, growth, sustainability, termination

Occupancy, graduation (exit), revenues/funds, financial position

Innovations, economic development, job creation, profit/losses to established organizations

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Lundvall, 1999; Sharif, 2006), and on the institutions framing their interactions, at the level of entire countries (Whitley, 1994, 2000) or industries, such as the stringent regulations within biotech or pharmaceuticals (Casper and Kettler, 2001; Casper and Matraves, 2003). At a more fine-grained level, this broader context includes inter-organizational networks (Axelsson and Easton, 1992; Anderson et al., 1994; Håkansson, 1997) connecting the incubator with such other organizations as universities, science parks, established companies, public agencies, and many other stakeholders, including policymakers. Hansen et al. (2000) even urge incubators to build such networks to support incubatees and eventually survive themselves. The place of incubation is not constrained geographically but can embrace the global arena, depending on where those relevant external actors are located. This form of “internetworking” (Etzkowitz, 2002: 124) can in fact also embrace other incubators or foreign customers. As for time, what counts at this level of analysis are the long-term dynamics of the industries wherein the incubator and its incubatees operate and their pathdependent change processes (Rosenberg, 1994); while the sources sustaining incubation are other incubators, several universities, as well as the aforementioned actors, all connected across transnational networks (Dicken et al., 2001; Andersson et al., 2002; Perks and Jeffery, 2006). The resources impacting incubators within this broader context include the multiple technologies and supply chains (Gadde and Håkansson, 2001), which incubatees need to relate to depending on their specific industry (Mas-Verdú et al., 2015), and the competences distributed across several organizations (Lundvall et al., 2002; Håkansson and Waluszewski, 2007). In the aforementioned networks some actors assume a central role thanks to their numerous relations, unique competences or superior resources (Lorenzoni and Baden-Fuller, 1995), making it pivotal for incubators to relate with these actors. As for control and governance in the networks embedding incubators, several actors, ranging from public agencies to established companies, attempt simultaneously to influence and control both start-ups and their innovations (Strömsten and Waluszewski, 2012; Baraldi et al., 2014). Therefore, incubators are exposed to the pressures exerted by several stakeholders (Mian, 1997: 258-9) to achieve multiple, often conflicting goals, which typically requires compromises causing suboptimal solutions or the pursuit of the most easily measurable goal (Alsos et al., 2011). The activities at the inter-organizational level embrace both the industrial activities of technology development, production and utilization performed by established companies (Håkansson and Waluszewski, 2007) and their financially-oriented activities such as in-licensing of incubatees′ technologies and acquisition of companies. Other important institutional activities are the regulation (Casper and Matraves, 2003) and standardization of the industry playground (Van de Ven et al., 1999: 157-60), where also public organizations intervene. Finally, outcomes at this level of analysis refer to the broader impact of an incubator, locally but also internationally, in terms of innovations brought to the industrial sectors to which it relates (Hansen et al., 2000; Etzkowitz, 2002), economic development and job creation (Bergek and Norrman, 2008: 22) and also, more specifically, the profits (or losses) generated for certain established organizations. These outcomes also correspond to the values that the incubator creates in relation to external actors, as opposed to the internal values associated with the outcomes in the two other levels of analysis. Table 1 summarizes our framework, which combines the seven components of incubation with the three levels of analysis: this frame enables contextualizing business incubation by showing which facets of the seven components become salient at each analytical level. The seven components of this framework relate to the daily

5

operations of business incubation. However, moving towards the institutional and inter-organizational level, the basic components assume a more open nature, and can be changed and combined in different ways as the time perspective also broadens. Therefore, these combinations of the seven components assume a strategic relevance and become key choices that impact an incubator's longterm development and value creation in relation to external actors and stakeholders (Normann and Ramirez, 1993; Ramirez, 1999). Reflecting broader changes in the incubation contexts, such as increased competition between incubators, several authors have started taking a strategic perspective on business incubation (Vanderstraeten and Matthyssens, 2012; Aaboen, 2009; Schwartz and Hornych, 2008; Grimaldi and Grandi, 2005). While these works focus on incubators' strategic positioning, our discussion of the KI′s incubator case will expand the range of strategic choices and extract, next to “positioning in the value chain”, five additional strategic drivers, that is, key choices affecting the long-term development and value creation of an incubator in relation to the broader embedding context.

3. Methodology This section reviews first the research process behind the KI's incubator case (Section 3.1), then the data collected (Section 3.2) and finally how this data was analyzed (Section 3.3). 3.1. Research process KI's incubator presents several unique features making it an outlier compared to the incubator models identified in the literature: extreme focus on science and a single industry, maximal intervention in start-ups, strictest focus on picking expectedly winning ideas, and extremely long incubation times. As such, the extreme case of KI's incubator is useful as it reveals more information and basic mechanisms that are not evident in average or general cases (Flyvbjerg, 2006: 229). Therefore, investigating this incubator represents an opportunity to better understanding the dimensions of incubation, including the possibility of identifying new ones. Relying on an in-depth case study, we aimed to learn (Flyvbjerg, 2006) about the specific circumstances through which incubation happens in this particular setting. Since this method provided us with the opportunity of learning how particular events, decisions or actions took place (Yin, 2003), as well as why (Eisenhardt, 1989), we explored different facets of KI's incubator: the goals behind its creation, the function of each of its subunits, and the handling of incubated projects/companies. The KI's incubator case was constructed during 2010–2014 and included several part-studies, which entailed several streams of data collection and concurrent analyses in order to systematically cover the three analytical levels adopted in this paper – the institutional/inter-organizational, the organizational, and the project/incubatee level. The first study and analysis (2010–11) covered the innovation-supporting and incubation strategies of KI University as viewed at an organizational and especially institutional/ inter-organizational level. A second study (2011–13) targeted, mostly at an organizational level, the control and governance mechanisms applied by KI′s incubator also in relation to its interorganizational network. Thirdly, we conducted in parallel with the first two types of studies and until 2014 four studies and analyses of single projects/incubatees, which were however related to their organizational and institutional/inter-organizational contexts. These part-studies indicated that KI's incubator was a case that could offer a new perspective on incubation, due to its sophisticated intra-organizational network of tightly integrated organizational units constituting a single incubation system. Moreover, our

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

6

mapping of the main actors and resources of KI's incubator soon revealed also that it had expanded its boundaries by vertically integrating downstream into activities, such as full-scale financing and exits towards global partners, which are normally part of the “entrepreneurial ecosystem” (Fetters et al., 2010). However, despite its expanded boundaries, KI's incubator as a strategic actor still needs to relate and interact with this external ecosystem – a major difference from typical incubators being though that the entrepreneurial ecosystem of relevance for KI's incubator is not regional but highly international, due to its strict focus on pharmaceuticals, whose potential partners/customers are found only in a global arena. As much of KI's incubator's unicity resides in the broad international reach of its operations, we chose internationalization as one of our six strategic drivers (see Section 5). According to Dubois and Gadde (2002: 554) an in-depth case study is particularly suitable for understanding the interaction “between a phenomenon and its context”. A single-case strategy enabled us to identify and analyze both the internal context of KI's incubator (all the way to single incubated projects/companies) and the broader, external context in which the incubator operates both regionally and internationally. As an illustration of a unique, indeed extreme incubation approach, KI's incubator is suitably presented as a standalone case, whose details can provide new insights that can expand the general understanding of incubation. The research process which has led to the results of this study departed from two main theoretical bases: industrial networks (e.g., Håkansson and Ford, 2002) and commercialization of science (e.g., Vohora et al., 2004). However, following an abductive approach and systematic combining (Dubois and Gadde, 2002), during the overall research project we continuously combined the emerging empirical insights with both existing and additional theory (e.g., incubation-specific studies or management control literature), which were used to supplement and develop the original theory. In this process, we had to re-assess and further collect our data continuously. Even though the core strategic issue of value creation (e.g., Ramirez, 1999) was present from the start of our research process, in the later stages, the specific empirical findings from KI's incubation system were analyzed more explicitly through theories and concepts from the strategy literature, as a way to re-interpret and develop these insights in ways new to the incubation literature. This process is described in Section 3.3. 3.2. Data collection The three analytical levels discussed in the previous section have firmly guided our data collection in terms of choice of interviewees and respondents, as well as formulation of our interview guides (see Appendix for selected interview guides). The three analytical levels also provided us with a “mental agenda” (Yin, 2003) of how we described and analyzed KI's incubator. Below follows an outline of the case protocol of our data collection. In terms of primary sources of data, the aforementioned partstudies rely on 23 unique interviewees and respondents who provided data by means of 16 interviews, 2 informal meetings and a witness seminar (see Appendix for details on these data sources). On the single project/incubatee level, we interviewed corporate managers and scientists involved in starting, conducting and exiting projects: at least two persons for each of four such incubatees were interviewed (Aprea, Clanotech, Dilafor and Lipopeptide). On the organizational level, representatives of such key units as the holding company (KIHAB), the investment company (KD), the scouting company (KIAB) and the research service company (Actar) provided data about the emergence and current operations of KI's incubator. On the institutional/inter-organizational level, the same individuals with long-term experience of the pharmaceutical industry have answered specific questions about how this

incubator relates to external organizations and companies in the broader national and international context. The first respondents to be interviewed were selected due to their central roles in either founding KI's incubator or in its current operations. These early interviews helped us identify further people, organizations and commercialization projects to be included in our investigation. This “snowball sampling” (Biernacki and Waldorf, 1981) thus provided us with further investigation possibilities from which we could design an appropriate selection of sources to illustrate the dimensions of this particular incubator. All interviews were semi-structured and the interview guide was continuously adapted during the study to fit the respondent's particular experience and role in KI's incubation system, next to the specific level of analysis. Four examples of our interview guides are provided in the Appendix: for instance, managers holding executive positions in the central companies of KI's incubation system were asked to provide the historical background of its evolution, its main goals and challenges, business model, portfolio management and external connections (see interview guides A, B and C in the Appendix); whereas managers of incubated companies were asked to describe their development process, milestones, resources, challenges, control and governance mechanisms (see interview guide D in the Appendix). Interviews ranged in length between 40 min and 2 h. 15 out of 16 interviews were recorded and transcribed, whereas for the single interview lacking recording and the two informal meetings detailed notes were taken. We also arranged a “witness seminar” in 2011 with the goal of covering all the three levels of analysis in the KI's incubator case. This video-filmed seminar lasted three hour and was attended by the three key persons involved in establishing KI's incubation system and one scientist early involved in commercializing his research. The topics treated in this seminar were why and how this system was founded, how it operates nationally and internationally, as well as the challenges it faces. As the company KD is now listed on the Stock Exchange, we also analyzed public financial records of its business operations such as quarterly reports and official press releases over its investments in new projects/incubatees, mergers, joint-ventures, drug development progress, and exits (see Appendix). These secondary sources have been used to verify the statements of the interviewees and seminar attendees. Still, the case as featured in Section 4 was built predominantly on primary data from the interviews, informal meetings and the witness seminar. For reasons of space we opted not to include direct quotes from our respondents, but to indicate which specific sources apply to various passages in our text. 3.3. Analyzing the data The KI's incubator study is part of a larger research program aiming to understand the value creation processes of several Swedish universities' innovation-supporting structures, especially in relation to the broader external networks that embed innovations (Van de Ven et al., 1999; Håkansson and Ford, 2002; Gadde et al., 2003; Håkansson and Waluszewski, 2007). Shaping, developing and adapting value creation processes form the basis for strategic management (Ramirez, 1999), which induced us to view incubators as “strategic actors”, an approach which is gaining ground in the business incubator literature (cf. Vanderstraeten and Matthyssens, 2012: 656). Thus, the overarching questions underlying the investigation of KI's incubator have been: what types of values does this incubator create, how does it create values, and for whom? Given the level of abstractions of these questions, they were not explicitly included in our interview guides or used during our discussions with the respondents. Instead, answers to these questions were progressively arrived at by means of several steps

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

of analysis, conducted on the data collected at our three levels of analysis (single incubatee, incubator and institutional/inter-organizational). Much of our primary data has the form of informant narratives (Boje, 2001) about either the evolution of the entire KI's incubation system or specific organizations within it, including individual incubatees. A first step in our data analysis has been to compare the narratives provided by the various informants in terms of the key events, participating actors, motives and context (cf. Makkonen et al., 2012): this comparison led us to identify events, actors, motives and contextual elements recurring or complementing each other in the individual narratives, so to create the joint narrative featured in Section 4. Next to chronologies, our primary data also includes information concerning KI′s incubator′s structures (i.e., internal and external organizations) and routines (e.g., incubatee selection and evaluation models or portfolio management). Our analysis scheme extracted first the structures and routines from the individual respondents' accounts; then it compared structures and routines provided by respondents at the level of the incubator′s central steering organizations (KIHAB, KIAB and KD) with those provided at the level of single incubatees/projects. As for routines, our analysis compared the evidence about for instance governance, control and resource allocation to portfolio companies obtained from interviews with KD or KIHAB managers with evidence on the same issues from interviews with CEOs of portfolio companies in search of common patterns, both between the incubator and the incubatee levels and across all incubatees. Despite the different perspectives, the evidence from the four incubatees and from central managers largely converged on these issues and other routines, forming the basis of the unified rendering of KI's incubator's routines and processes in Section 4. As for the structures, our data analysis aimed at identifying both a common picture shared among the various respondents and complementary pictures that would enrich each other. Network maps were drawn in relation to most interviews. In particular, four incubatee-specific network maps were drawn, detailing the resources and actors embedding each incubatee. We searched for overlaps in the actors featured in the various network maps as well as for differences and additional actors. This analytical procedure allowed us to identify also as many external actors as possible, such as investors in KI's incubator and co-investors in its portfolio companies, which are now jointly featured in the network map of Fig. 1 in Section 4. Prior to defining the final version of the case story (as eventually featured in Section 4), the emergent joint narrative, common routines and structures describing KI's incubator were also analyzed in order to break down the case materials in terms of a number of components characterizing the operations of business incubators. This analysis was conducted as a continuous matching (Dubois and Gadde, 2002) between this processed data (narrative, routines and structures) and the literature on incubation, leading to identify seven such key components (Table 1). Thus, instead of first deducing these components from the literature and then applying them to the KI's incubator case or inductively extracting the seven components from the ready-made case, we concurrently developed both the theoretical components and the case story. This abductive approach followed a logic of “systematic combining” between theory and empirical material (Dubois and Gadde, 2002). Therefore, the theoretical frame of Section 2 was developed in parallel with and is also partly a result of the empirical material forming the case. When we eventually selected the seven components of incubation of Table 1 we could more explicitly “apply” them to the KI's incubator case, which enabled us to identify the specific characteristics of KI's incubator and subsequently search for any similarities and differences compared to typical incubators (as

7

defined in the literature). Then, following the theme of value creation (Ramirez, 1999), we further analyzed KI′s incubator as a “strategic actor” (cf. Aaboen, 2009; Vanderstraeten and Matthyssens, 2012) searching which values it aimed to create, how it was constructed to produce them, and the actual outcomes of value creation. This was the final step of our analysis, aimed at identifying the specific strategic challenges of this extreme type of incubator. It was at this point that the current version of the case story was written (see Section 4): even if KI's incubator's strategic challenges had already surfaced in our analysis, by matching the empirical materials with concepts from several streams in the strategy literature, we opted for compiling a case story relatively free from theoretical concepts. It is instead in the following discussion (Section 5) that we explicitly present those strategic challenges, namely six “strategic drivers”, that we consider as specific for KI's incubator, in the sense that they are not as pressing for common types of incubators. From an analytical point of view, we obtained these strategic drivers by combining with each other the seven components of incubation of Table 1 and by viewing them at the inter-organizational/institutional level and in relation to long-term value creation processes. More precisely, each of the six drivers identified in Section 5 was obtained by combining in a specific way the outcomes of incubation with some other components of Table 1. For instance, we obtained the strategic driver “positioning in the value chain” (stretching from entrepreneurial ideas to end markets) by combining the activities/services (what the incubator does concretely), the place (where in the value system it operates, Stabell and Fjeldstad, 1998), the sources (what inputs it starts operating from), and the outcomes (what value comes out from its activities); while we obtained the strategic driver “internationalization” (Forsgren et al., 2005; Vahlne and Johanson, 2013) by combining the geographic location of the incubator's sources, resources and outcomes (where the values are created) with the specific places where it operates. Afterwards, we explicitly connected each driver to the theoretical concepts from the strategy literature that inspired, in previous steps of analysis, its identification and refinement, as visible now in Section 5.

4. The case of Karolinska Institute′s incubation system This section presents first the broader institutional and interorganizational context embedding KI's incubator and subsequently delves into the specific organizational context, including how it affects the single incubatees and their projects. 4.1. The institutional and inter-organizational context The Karolinska Institute (KI) is an internationally renowned university and medical research institution accounting for several Nobel Prize winning scientists. Founded in 1810, it is now a university specializing in medical discovery and is regularly ranked as one of the world's best 100 universities, with 2000 researchers, 300 professors and 5500 undergraduate students. Since the mid1990s KI has constructed a highly sophisticated innovation-supporting system actively engaging in the commercialization of research discoveries. 4.1.1. Building an incubation system In 1995, the Swedish government founded Karolinska Institutet Holding Ltd. (KIHAB), now acting as an “umbrella organization” for a system of actors formed for the goal of commercializing research. KIHAB is wholly owned by the Swedish state and managed through KI as a governmental representative. However, the shy contribution of SEK5 million from the government to start up an

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

8

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Fig. 1. The inter-organizational network embedding KI′s incubation system.

incubation system was a far cry from the capital needed to develop and commercialize new pharmaceuticals (Witness seminar Meijer; Interview Bogentoft). This initiated a gruesome process of trying to acquire additional capital which eventually resulted in a sophisticated and unique innovation-supporting and financing system. This section describes the main events leading up to the current system as well as its basic commercialization mechanisms. Shortly after KIHAB, Karolinska Innovations AB (KIAB) was founded (1996) as a business development company with the original purpose of commercializing research by selling licenses. In parallel to this business, KIHAB started to form companies around specific commercialization projects in very early stages, as the idea was to acquire capital by finding investors to these companies and then creating full-scale R&D spin-offs (Witness seminar Meijer; Interview Fransson; Interview Bogentoft). However, due to the “teacher's exception”, granting in Sweden intellectual property to the inventor and not the university, selling licenses turned out to be very hard to do; there was no real interest in buying licenses compared to for instance in the USA, where it is the inventors that often buy the license from their university. Forming companies in very early stages did not attract any investors either, because these companies where very risky ventures due to the high failure rate of pharmaceutical development projects (less than one out of ten succeeds). The pressing issue of acquiring capital for this system thus called for a different solution (Witness seminar Thelin; Interview Bogentoft; Interview Sundberg). A modified model based on developing the commercialization projects even further before approaching potential investors was therefore implemented. Before approaching potential investors the projects would instead be developed into Phase II of clinical trials, testifying the efficacy of a drug, which meant a reduced economic risk for potential investors (Witness seminar Wigzell; Interview Kull). However, for KI's incubation system this meant that even more capital was needed as a greater part of the expensive development would need to be covered by KI's incubator before receiving any eventual revenue. Therefore, a number of time-limited funds were established through the newly founded Karolinska Investment Funds (KIF), which raised initially SEK 500 million and then an additional SEK 1.2 billion via three separate

funds. However, over the next few years it would become clear that these types of time-bound funds (7 years) did not suit the process of developing new pharmaceuticals, usually taking 10–15 years (Witness seminar Meijer; Interview Fransson; Interview Bogentoft). As a consequence, in 2007 the funds were turned into a long-term investment company called Karolinska Development Ltd. (KD). Three years later, in 2010, KD had as many as 730 shareholders, among which the largest was the Third Swedish National Pension Fund (Karolinska Development Interim Report, Jan–Sept 2010). 4.1.2. The inter-organizational system today As the portfolio of investments was expanding and the projects were entering more expensive development phases the yearly burn rate was approaching SEK250 million per year (Interview Fransson; Interview Meijer). More capital was needed. Therefore, in 2011 KD was floated on the Stockholm Exchange, which raised an additional SEK600 million bringing its equity to about SEK2 billion. This has now resulted in thousands of shareholders, of which the largest is still the Third National Pension Fund. In 2014 KIHAB held 8% of the shares in KD, granting KIHAB however voting shares enough to control KD. Another substantial investor in KI's incubator is the UK venture capital firm Rosetta Capital Ltd. specialized in life sciences. In 2013 KD transferred 13 of its portfolio companies to KDev Investment Ltd., a subsidiary of KD of which Rosetta Capital acquired 14% of shares, paid SEK220 million. Through this joint venture KD boosted its operating profit this year and Rosetta Capital got shares in a selection of KD's portfolio companies (Karolinska Development Annual Report 2013). Today (2014), KD's investment portfolio contains 33 projects corresponding to 23 companies. 25 of the projects (16 companies) concern pharmaceuticals, of which 16 are in clinical trials, and eight projects medical technologies (seven companies). This number and its total value make KD one of the largest investment companies focused on the life sciences (and particularly pharmaceuticals) in Scandinavia and in Europe. Under the ownership of KIHAB, KI's incubation system has grown to also include, next to KIAB and KD, KI Science Park Ltd. (KISP), which offers office and laboratory space to incubatees, and the spin-off company Actar

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

(Academic Targets), which offers small-scale research and clinical services mostly to the projects KD has invested in (Interview Wikström; Karolinska Development Annual Report 2013). During the last ten years, KIAB and KD have also evaluated and invested in ideas from other Swedish universities (e.g., Uppsala, Umeå, Gothenburg), as well as from foreign universities (e.g., NTNU, Norway, and Aarhus University, Denmark). This is part of a “scouting” strategy for bringing as many attractive investment options to the KD investors as possible (Interview Flink; Karolinska Annual Report 2009; Karolinska Annual Report 2013). However, the evaluation and investment model also stretches across national borders: to expand its own operations, KIHAB bought a 25% share in the Finnish company Biocelex Ltd, making KI's incubator directly enter into a foreign innovation-supporting system, namely the Turku Science Park. Biocelex has the same function in this Finnish context that KIAB has in KI's incubation system and has directly adopted the same evaluation and business development model based on evaluating potential investments for KD (Interview Meijer). Thus, through this international acquisition, KIHAB and KD exported their incubation model (evaluation and investments) to Finland. In 2013, KD also established agreements for cooperation with the American Mayo Clinic, the San Raffaele Hospital in Milan and the Medical University in Graz, Austria, aiming to mutually identifying new life-science investments (Karolinska Development Annual Report 2013). In addition to this branching out in terms of scouting for new ideas and making investments across national borders, KI's incubation system has several collaborators and co-investors both in the public and private sector. One of them is Almi Ltd., which is a publicly owned corporation with local units across Sweden with a mission to support new business ventures by offering counseling, seed capital and incubation services. Vinnova, the Swedish Agency for Innovation Systems, has also made investments in KI's incubation system and is collaborating with KIAB. Other co-investors are The Foundation for Baltic and East European Studies (Östersjöstiftelsen), controlled by the Swedish Government, and Praktikertjänst, the largest private health-care provider in Sweden, which both are co-owners of several of the portfolio companies where KD typically holds majority shares. In selecting and running its commercialization projects, KIAB and KD are also related to about 300 life science experts in the private sphere. These act as consultants for evaluating new ideas and have active board positions in the portfolio companies (Interview Meijer; Interview Caldirola). The next section specifically addresses how these incubatees are selected, developed and exited within KI's incubator. Fig. 1 displays this system's structure (inside the dotted oval) and inter-organizational connections. 4.2. KI's incubator's organizational context and the single incubated projects KI′s incubator is based on a specific “selection mechanism” for evaluating and selecting appropriate investments, and a specific “development and exit mechanism” for how to promote and evaluate a projects progress, and finally create an exit, or alternatively terminate the project. The first stage of the selection mechanism primarily involves two organizations belonging to KI's incubation system, KIAB and Actar (Interview Kull; Interview Wikström). KIAB is the first contact with KI's incubator for researchers and together with Actar this organization evaluates potential investments for the investment company KD. The ideas which KIAB selects as the most interesting for commercialization (about 10 out of a 100 ideas per year) must be patentable and correspond to a number of requirements concerning both economic and clinical potential (Interview Backman; Interview Wikström). In this evaluation, KIAB works closely with Actar, a spin-off company acting

9

as clinical test and development facility wholly owned by KD. Actar works mainly for KD and KIAB, but also as a contract research organization for companies external to KI's incubator. When evaluating and selecting potential investments, KIAB and Actar focus on three specific aspects of research results: (1) their relevance to a medical need, (2) their potential market, and (3) their economic risk (Interview Backman; Interview Kull). If the research result is evaluated as commercially interesting from the standpoint of all three aspects, it is presented as a potential investment for KD. Both KIAB and Actar are bound by a deal-flow agreement to present any potential investment to KD first. Only if KD chooses not to invest, is it possible for KIAB or Actar to pursue a different investor (Interview Kull; Interview Wikström). If KD decides to invest in a project (about two ideas per year), it becomes part of a “development and exit mechanism”. The decision needs to be approved by an Investment Advisory Board through a validation period of 18–24 months, and a detailed R&Dand commercial plan until clinical Phase II has to be in place. Subsequently, patenting and regulatory issues has to be dealt with, as well as the establishment of a formal structure for the investment in terms of a specific project team and/or through the forming of a company (Interview Kull; Interview Wikström). KD is a highly active owner of its investments: it normally has a majority ownership of its start-up companies and appoints the CEOs and the board members. The CEOs are usually senior managers from larger pharmaceutical corporations with extensive experience of large-scale pharmaceutical development. The boards consist of one or two representatives from KD, the inventor(s), and expert advisors from either academia or industry. KD develops and manages the incubatees by providing them with funds for approximately six to 18 months at a time, and after each period re-evaluates the project in terms of clinical progress (Interview Backman; Interview Mohell). As the goal is to reach a clinical Phase II, the projects are evaluated from the standpoint of certain milestones that need to be achieved in pursuit of that goal. If the milestones are not met the project or company runs the risk of being terminated (which happened in 30–40 cases over the years). KI's incubator is dependent on an eventual exit to a large pharmaceutical company acquiring the incubatee or a license for the potential product. Within this incubation system, KD is thus supplying the commercialization projects with financial and human resources in order to bring them to Phase II, which through an exit then will create revenue in terms of upfront payments, milestone payments and/or royalty income. Reflecting the impact of the broader institutional and inter-organizational context, this model is specifically adjusted to fit the needs of large pharmaceutical companies, which are putting less resources on R&D, but at the same time are unwilling to invest in premature pharmaceutical substances (before Phase II). For medical technology projects, the requirements are less strictly defined, but there is still the issue of timing in terms of identifying an appropriate type of exit at the right point in time, that is, when the product is ready to be launched on the market (Witness seminar Thelin; Interview Kull). In order to create revenue from the incubated projects and thus sustain the entire incubation system, every project is run in a very cost effective manner: it is run as a virtual organization manned with only a CEO and one-two other employees, while most other activities are outsourced to external consultants (Interview Caldirola; Interview Mohell). Also, as the progress of the projects are evaluated every six to 18 months, KD makes sure that resources are used efficiently and are not “wasted” on projects that do not show encouraging results (Witness seminar Wigzell; Interview Kull). As substantial revenues can only be created through exits, rather than while the incubatee is under the ownership of KD, the system is based on directing resources only to those projects which show

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

10

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

the potential of a future exit. The project's managers/CEOs in turn need to actively pursue industrial partners and a potential acquirer already from the start (Interview Mohell; Interview Nilsson). This is however hard to do in very early phases of clinical development due to high failure rates. While many projects struggle finding such partners, although they are approaching or completing a Phase II of a clinical trial, there are also examples of projects that succeed in doing so: for instance Dilafor (founded in 2003) managed in 2014 to license tafoxiparin, a newly developed drug for obstetrics and gynecology, to HongKong's Lee's Pharmaceuticals (Interview Ekman-Oderberg; www. karolinskadevelopment.com, March 2014). KI's incubation system has been successful in creating licensing deals for its technologies (about 30), but less so in getting its companies acquired, with an outcome of only one exit through sales (Witness seminar Meijer; Interview Bogentoft). The exited company, Cogmed (based on software for training short-term memory) was in 2010 sold to the multinational Pearson Publishing. An eight year deal granting royalties on the software's sales will hopefully at least cover KI incubator's development costs (Interview Flink; Interview Fransson).

5. Discussion: six strategic drivers for business incubation This section firstly provides an analysis of the KI's incubator case in Table 2 below, which applies systematically to the case the framework based on seven components and three analytical levels of analysis developed in Section 2 (see Table 1). Secondly, this section discusses how bringing together and combining certain components defines six strategic drivers representing key choices that, at the institutional/inter-organizational level, influence the incubator's long-term development and value creation in relation to this broader embedding context (Normann and Ramirez, 1993; Ramirez, 1999). KI's incubator consists of an entire system composed of highly integrated organizational units controlled by a single organization, KIHAB. These units act together as a single incubator, even though each unit performs highly specialized activities: idea scouting, selection, evaluation, validation, financing, control and governance, product development, deal preparation, and exits. This intraorganizational complexity is visible inside the dotted oval of Fig. 1 and is analyzed along the seven components of incubation in the central column of Table 2 below. Next to this internal complexity, KI's incubator is strongly engaged in operating and creating connections at the institutional and inter-organizational level, that is, towards the actors outside the dotted oval of Fig. 1 and in relation to the seven components in the right column of Table 2. These connections embed KI's incubator not only in the local entrepreneurial ecosystem (Fetters et al., 2010; Abetti and Rancourt, 2008: 391), but also beyond it at international level. Thus, KI's incubator fits well the concept of “networked incubator”, whereby “incubation transcends the incubator” (Hansen et al., 2000) and requires building external networks especially with established companies and experts so to create value for incubatees (Ibid: 76). However, in building or interacting with these external networks an equally important issue is how the incubator can create value for the external actors (Håkansson and Ford, 2002; Gadde et al., 2003; Baraldi, 2008). As value creation is at the core of business strategy (Ramirez, 1999) we discuss now the analysis in Table 2 with more focus on the institutional and inter-organizational level which is less explored in studies on incubation, even if this is the very level where values are created in relation to external actors. Similarly to other works in the business incubator literature such as Hansen et al. (2000), Aaboen (2009) and Vanderstraeten and Matthyssens

(2012), we consider now the incubator as a strategic actor who combines the first six components in the rows of Table 2 in order to create values internally and externally, corresponding to the seventh component of outcomes. Value creation in terms of place requires that KI′s incubator coordinates several organizations and facilities, either co-located in the Stockholm area or widespread geographically (e.g., Finland), and that it engages in international connections to commercialize science. Internationalization is accordingly a strategic issue for this incubator, which we develop further as a strategic driver below. As for time, value creation requires balancing the short timeframe of project milestones (6–18 months) with very long incubation times (10–15 years) and KI's incubator's dependence on the pharmas′ long-term development paths, with the high level of risks that this entails. Looking at the sources from which KI's incubator strives to create value, it strictly focuses on picking academic ideas within life sciences expected to become winners; but the chase for solutions that can eventually fit the needs of specific pharmas induces KI's incubator to broaden the search of sources to cooperating but also competing with several universities' incubators, which is paralleled by the need to access financial sources, also abroad. As for resources, KI's incubator holds a strong range of organizational resources (finances, skills, expertize, lab facilities) and relationships to experts and established companies: however, the major challenges to creating value for the external network is to match and combine these resources and incubatees′ technologies, on the one hand, with the complex established resources of pharmas, the science produced on the international scale, healthcare systems, KI's incubator's shareholders and investment partners, on the other hand. A further challenge is that some of these actors, namely pharmas and some VCs, are also competitors to KI's incubator in either developing drugs or selling start-ups to pharmas. The emergence of cooperation/competition in the two components of sources and resources suggests that this is a key strategic challenge for KI's incubator, which we take up as a specific strategic driver. The component of governance and control follows a logic of majority ownership within the organizational boundaries of KI's incubation system, which enables KIHAB/KD to exert strong control on how value is created via milestones and board appointments. However, at the inter-organizational level, the freedom of action of this incubator is delimited by more subtle forms of control coming from the network, such as quarterly report requirements, FDA's clinical test regulations and pharma customers with dominant purchase power in specific therapeutic areas. Due to the singularity of an incubator assuming majority control of its incubatees and the challenging impact of external control on KI's incubator's value creation and strategy, we consider “governance and control” as a strategic driver of its own. As for the activities and services, KI's incubator performs both routinely the “select-develop-exit” mechanism and more strategic activities such as steering projects (including their termination) and acquisitions of shares in other science parks. These internal activities create value also for external parties as they fit the new external division of labor in the pharmaceutical industry, whereby large firms outsourced early R&D to research institutes or smaller firms, waiting to invest only in later-stage, and lower risk, projects (cf. Ehret et al., 2012: 303). However, this incubator, and the actual value it can create, is highly dependent on the subsequent activities of refinement, production and use performed respectively by pharmas and healthcare organizations, which are all exposed to the strong regulation by the FDA. Hence, positioning in this value chain is a key strategic issue for KI's incubator, which we develop further as our first strategic driver below. Finally, the outcomes of this incubator point out that, while such intermediate outcomes as IPR and fund acquisition are not so

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

11

Table 2 KI's incubation system analyzed across seven dimensions and three contexts.

Place

Single incubatees (the project).

Organizational level (the incubator).

Institutional and inter-organizational level.

Most incubatees co-located within KI university's premises (KISP and nearby).

Core organizations and facilities inside or near KI Science Park (KISP).

Connections both to KI university and ideas also from other university and abroad. Owned Finnish subsidiary, Biocelex. Transferring academic research results to industry, irrespective of place.

Close coordination of idea screening, selection, patenting, financing, project management, choice of board members etc.

Checking milestones of portfolio companies every 6– Dependent on pharma firms' long-term paths of drug 18 months. development: chosen therapeutic areas, pipelines, existing patents and their life span. Exit/license after Phase II of clinical trial, hence some Lengthy approval processes for clinical trials (10–20 incubatees remain as long as 10–15 years. years).

Time

Requirement to achieve milestones of clinical progress every 6–18 months.

Sources

Ideas: Scientists from KI university but also Strict focus on “picking winning academic ideas” only Cooperating/competing with several other uniother Swedish and foreign universities and in pharmaceuticals (and med-tech). versities (e.g. agreement for cooperation with medical hospitals. providers and universities in the US and Europe -Italy and Austria). Funds: One major owner/financier (KD) and Selection criteria: medical relevance, market potenFinancing from various national and international a handful other public/private co-owners. tial, economic risk. organizations (public/private). Scientific ideas sourced also from another owned Adapting to the changed R&D needs of “big pharmas”: incubator (Biocelex, Turku SP). outsourced early research. A very restricted range of Pharmas eventually interested in any new solution from KI/KD.

Resources

CEOs with experience from pharmaceutical development in large corporations. Minimized permanent research staffs. Access to virtual organizations of consultants and life science experts.

KISP's office and laboratory space. KIAB's legal and patenting expertize, patent rights and commercialization skills. KD's finances and portfolio management skills.

Actar is the testing and clinical development facility for single incubatees. Financed via KD's equity capital. Indirect relationships with industry through CEOs and board members.

Actar as core R&D facility for the entire system. Relations to the academic and business expertize of 300 life science experts. KI university, KIHAB and KD's relationships with global life science and pharmaceutical companies (informal and licensing contracts).

Incubation system controlled by KIHAB despite floatation of KD. Public and private investors in KI's incubation system: 3rd National Pension Fund, other VCs. KD acts as venture capitalist with strong control (financial and strategic) on incubatees, starting from their “selection” and then the “develop and exit” mechanisms. Each incubatee/project must satisfy the big Co-ownership of selected incubatees (with Rosetta pharmas′ requirement of reaching Phase II Capital, Baltic Foundation and Praktikertjänst). of clinical trials. KD strongly connects incubatees: managed as part of a portfolio of projects, rather than informal networking.

Control/ Incubatees are majority-owned governance by KD. KD's tight control via milestones and clinical performance. Quarterly reports due to KD′s floatation on the Stock Exchange.

Activities/ services

Identifying substances for potential pharmaceuticals.

“Selection-development-exit” routine, with internal division of labor: KIAB/Actar -KD -company exit through sales Phase II.

Conducting early phases of clinical trials.

Very active ownership by KD: financing, business support, appointing staff, CEO and board members.

Scanning for industrial partners and future R&D support via Actar's laboratory services. owners of the company or license. Some interactions with health care and the Acquisition of shares in other science parks. Karolinska Hospital. Termination of projects. Outcomes (values created)

IPRs and fund acquisition, key recruitments for over 40 incubatees. 16 projects in Phase I/II of clinical trials (surviving key technical milestones). One project successfully concluded with product launch and company being acquired, but not a drug or med-tech (Cogmed, software). 30–40 terminated projects over the years.

Scientific research on an international scale. KI university: a world-known brand of scientific excellence. Multiple shareholders: 3rd Pension Fund, small shareholders on the Stock Exchange. Partnership with intl. investors (Rosetta Capital). VC firms as both partners and competitors to KD. Pharmas' established resource structures: technological platforms, patents, previous investments, R&D pipelines. May be competitors. Hospitals and health care for clinical trials. The Stock Exchange's “quarterly report-logic”. Need to conform to FDA's requirements on clinical trials. Dependence on finding buyers (big pharmas or VCs) of licenses or incubated projects: but some large firms will control the deal by being the only potential counterpart in a therapeutic area.

New external division of labor: big pharmas outsourcing early R&D to academia and contract research organizations (CROs). Academia and incubators perform early development, but depend on production and use being performed by big pharmas and healthcare system. FDA's regulation and standardization in selected fields and therapies.

Invested portfolio of 23 firms (33 projects), decreased by 10 firms/projects since 2011. SEK2.5 B raised capital (emissions, floatation). One full exit, Cogmed. 13 partial exits to Rosetta.

About 3 employees per company in the commercialized projects. SEK220 M paid by Rosetta Capital (now partner in 13 projects). 30 firms licensing KI's technologies.

Burn rate of portfolio of SEK250 M per year.

One innovation on the market: Cogmed's software via the new owner Pearson Publishing. New shareholders, but no dividends paid.

Equity of about SEK 2 billion.

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

12

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

problematic, project termination is a much more likely outcome (30–40 projects) than achieving clinical phases (16 projects): project conclusion with a complete exit has happened only once (Cogmed) during the 15 years KI's incubation system has been operational. After surfacing in our discussion on the time component, risks appears as a challenging issue also for actual outcomes, due to the high failure rate of drug development projects. Therefore, we consider “risk taking” and the associated “time perspective” as a main strategic driver for this incubator, a driver which also embraces financial resources and outcomes. As for financial values, despite large amounts of raised capital (SEK2.5 billion), no dividends have been paid to shareholders and the whole portfolio of companies/projects is expensive for KI's incubator to sustain, with a burn-rate of SEK250 million yearly: this can induce choices to reduce the number of companies, especially if the underlying projects do not reach satisfactory clinical results quickly. However, KD was able to obtain additional cash from a new partner (Rosetta) through a partial exit from 13 still-ongoing projects. Revenues and various models to obtain them are accordingly a pivotal concern for KI's incubator, which deserve to be developed into a specific strategic driver. As for non-monetary values, this incubator generated around 60 direct employments (3 per incubatee), one software innovation marketed now by a multinational publishing company, and 30 technologies being licensed to life science companies. In the above discussion of the seven components of KI's incubator, we identified six “strategic drivers”, that is, key choices impacting the incubator's long-term development and value creation (Ramirez, 1999) in relation to external actors. These six strategic choices are: “positioning in the value chain”, “risk taking and time perspective”, “revenue model”, “governance and control”, “internationalization”, and “cooperation/competition”. These drivers are not taken from a single strategy model or theory, but are inspired by several streams in the strategy literature, such as positioning (Mintzberg, 1987; Stabell and Fjeldstad, 1998), business venturing (Gompers and Lerner, 2001), business model canvass (Osterwalder and Pigneur, 2010), international business (Andersson et al., 2002; Vahlne, and Johanson, 2013), and business network perspective (Håkansson and Ford, 2002; Gadde et al., 2003). All six drivers have in common that they address value creation by combining the outcomes of incubation with some other of its key components (see Table 1), even if each driver does it in a specific way. 5.1. Positioning in the value chain This is a fundamental driver as it impacts and somehow frames the other five. This driver stresses that an incubator's value creation, both for incubatees and external actors, depends on the strategic position (Mintzberg, 1987) it assumes in terms of its scope of activities/services (what it does), place (where in the value system it operates, Stabell and Fjeldstad, 1998), sources (what inputs it starts operating from), and outcomes (what it creates in terms of values). The incubation literature proposes positioning choices in terms of sectorial focus Vs. diversified scope (Vanderstraeten and Matthyssens, 2012: 657; Schwartz and Hornych, 2008) and positioning along the “technological entrepreneurial value chain” (Phan et al., 2005: 170-1). However, the KI′s incubator case shows how a narrow sector focus (pharma/med tech) can be combined with a positioning along an even broader value chain, which does not stop shortly after new business formation, but also embraces the following steps of exit and revenue generation, both pivotal to KI′s incubator. Thus, KI's incubator is positioned differently from most incubators as it covers a very large spectrum in the “bioscience knowledge value chain” (Cooke et al., 2006: 1168): this incubator expanded its boundaries and internalizes now

several activities usually found in the surrounding “entrepreneurial ecosystem” (Fetters et al., 2010; Abetti and Rancourt, 2008: 391), such as majority ownership, venture funding and selling incubated projects. 5.2. Risk taking/time perspective As most incubators focus on creating value in the form of companies surviving the start-up phase (Bøllingtoft, 2012), they typically take low risks (contained investments) and have shortterm horizons (e.g., graduation rules, Mian, 1997: 281), especially if compared with VCs (Gompers and Lerner, 2001). Due to its broad positioning in the value chain and its focus on highly uncertain outcomes (Phase II drug projects and exits), KI's incubator takes much greater risks and needs a longer time perspective than a common incubator. The KI's incubator case and this strategic driver reveal therefore how, depending on the outcomes aimed at and the degree of governance/control (majority ownership by KI's incubator), both time and risk (i.e., resources invested with high uncertainty, Cooper, 2001: 125-7) can be considerably stretched for an incubator: 10–15 years (instead of the typical 3–5 years in most graduation rules) and massive financial resources are needed to perform drug R&D activities/services starting from sources in academic medical sciences. 5.3. Revenue model Departing from the more general concept of “incubation model” (i.e., how the incubation process is organized and managed, Bergek and Norrman, 2008: 22), this strategic choice concerns that part of the business model (Osterwalder and Pigneur, 2010) that sustains value creation by defining how the incubator obtains the cash flows for performing its activities/services and reach expected outcomes (cf. Mahadevan, 2000). In this regard, due to the massive financial resources and long time necessary to make valuable to pharmas its companies/projects based on pure science as sources of idea, KI's incubator cannot rely only on public subsidies but is a for-profit incubator (Grimaldi and Grandi, 2005; Bruneel et al., 2012). Being a profit-driven incubator is not unique among bio-incubators (see Cooke et al., 2006). However, what makes KI's incubator unique (and this strategic driver so important) is that KI's incubator needs not only license revenues, but also cash from its own equity emissions long before it can obtain cash from profitable exits: these equity emissions include even a floatation of the incubator on the Stock Exchange, so that this driver in turn also impacts the control/governance dimension below. 5.4. Governance and control This driver assumes a paramount importance for an incubator such as KI's incubator which is listed on the Stock Exchange, in terms of both internal accounting control over value creation activities and outcomes (Otley and Berry, 1980) and external control (Pfeffer and Salancik, 1978). Internally, KI's incubator exerts a tight vertical control/governance on its incubatees (Phan et al., 2005: 179; Vanderstraeten and Matthyssens, 2012: 665) by freely selecting levels and forms of control (e.g., milestones, board appointments, budgets, cf. Emmanuel et al., 1990), with a particular emphasis on the results, especially financial ones. KI's incubator's value creation activities are heavily dependent on financial resources (SEK 250 million yearly) over long time periods (10–15 years per project), which required floating the incubator on the Stock Exchange. Therefore, externally, KI's incubator is exposed to some degree of vertical control by its shareholders and of horizontal control (Johansson and Östman, 1995) by its partners (e.g., Rosetta), potential customers (dominant pharmas) and indirectly

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

connected actors (e.g., FDA). As identified also in the incubation literature (Mian, 1997: 258-9; Alsos et al., 2011), this external and indirect control (Pfeffer and Salancik, 1978) reduces the strategic freedom of KI's incubator, making it even more sensitive to external influences than other incubators. 5.5. Internationalization This strategic driver is important because of KI's incubator's geographically broad sources (scientific ideas from foreign universities), the need to find financial resources internationally, and the expected outcomes of selling drug projects to global pharmas. The value creation logic implied by this strategy even turned KI′s incubator into a multinational organization (Andersson et al., 2002) with subsidiaries in two places (Stockholm and Turku) and connections with resource structures found in transnational networks (Dicken et al., 2001; Perks and Jeffery, 2006). Even if international connections between bio-incubators are recognized as important for “co-incubation” of incubatees by Cooke et al., (2006), KI's incubator brings such connections one step further, embracing a broader group of foreign actors and establishing rapidly a direct international presence (Vahlne and Johanson, 2013). 5.6. Cooperation/competition The last driver assumes salience because, to create value for shareholders and pharma customers (two pivotal outcomes for survival), KI's incubator has to interact within an external network stretching from scientific sources to other actors controlling key resources such as experts, capitalists, healthcare and regulatory agencies (FDA). Strategizing in such a network corresponds to building and developing over 100 business relationships with the aforementioned actors (Håkansson and Ford, 2002; Gadde et al., 2003; Baraldi et al., 2007). However, KI's incubator faces the challenge that it can be simultaneously in competition with the very same cooperation partners, such as pharmas and venture capitalists (Gnyawali and Madhavan, 2001; Gnyawali and Park, 2011). Most incubators do interact with external stakeholders such as public agencies, universities, financiers and even potential customers of their incubatees (Hansen et al., 2000), often as part of their “mediation” activities (Bergek and Norrman, 2008; Ahmad and Ingle, 2011). However, the relationships between KI's incubator and external parties are deeper than those of other incubators, as they entail more mutual dependence (Håkansson and Snehota, 1995), both in technical and economic terms. Therefore, the strategic choices concerning cooperation with specific external actors are particularly important for KI's incubator, and make situations when also competition arises with the same actors particularly sensitive, due to complex interdependencies (as in the KD-Rosetta relationship).

13

(6) cooperation with a very large network of relationships, but also competition with a few of these partners. Even though developed from a single case, these six drivers (positioning in the value chain, risk taking/time perspective, revenue model, governance/control, internationalization, and cooperation/ competition), hold an abstract and general character. In fact, if viewed as dimensions they can assume different values: (1) positioning in the value chain can be broad or narrow (focussed on one specific activity); (2) risk taking can be high or low, combined with a short or long-term perspective; (3) revenue models vary in terms of the relative importance of different cash flows such as incubatee fees, subsidies, capital emissions, licenses and exits; (4) control can be tight or weak and viewed vertically and horizontally; (5) internationalization can concern few or many countries, with or without local presence; and (6) cooperation/ competition varies in terms of number and depth of cooperative and competitive relationships. Therefore, the six drivers can also be applied to describe, or even develop, the strategies of other incubators, especially in a moment when the “incubation industry” is under transition and several new models such as KI's incubator are likely to appear. A managerial implication for incubators is accordingly that, especially now when they are under scrutiny and often questioned in their contribution to innovation and economic development, focusing on traditional elements of their operations (selection/ graduation criteria, range of services or contacts provision to incubatees) may not be the best option for improving their performance and position: incubators need instead to take a more strategic perspective and address the six aforementioned drivers. Considering driver no. 6, as much as incubators need to be good at providing their incubatees with network contacts, they also need to develop their own networking strategies, based on interacting and building relationships in the broader network that embed them, much like any company needs to do (Baraldi, 2008). The limitations of this study relate to its exploratory nature and focus on a single occurrence, possibly a singularity, of incubation. Therefore, further research, ideally based on multiple case studies on selected incubators, may refine the six strategic drivers into dimensions enabling to identify new typologies of incubators, reflecting also their changed behavior. Such research should also enable understanding the co-occurrence of and the connections between the six strategic drivers: for instance between a certain revenue model and the type pf governance/control exerted by and on an incubator.

Acknowledgment This research has been funded by Svenska Handelbankens Research Foundation (Grant number P2009-0070:1).

6. Conclusions and implications Appendix A. Sources of data Our analysis of the KI's incubator case through seven components of incubation (place, time, sources, resources, governance/ control, activities/services and outcomes) and at three levels (single incubatee/project, organizational and institutional/interorganizational) identified six strategic drivers for incubators. According to these drivers, KI's incubator is characterized by: (1) a broad positioning covering almost the entire innovation value chain; (2) a high risk taking and long-term perspective; (3) a profit-based revenue model driven by a financial investment and venture capital logic; (4) a tight internal vertical control coupled with a certain degree of external vertical and horizontal control; (5) a substantial internationalization of its operations; and

Secondary sources: Karolinska Development Annual Report 2013. Karolinska Development Interim Report, Jan–Sept 2010. Karolinska Institutet's Annual Report, 2009. www.actar.se (Oct 2008, May 2009, Dec 2009). www.karolinskainnovations.ki.se (Dec 2008, March 2014). www.karolinskadevelopment.com (Dec 2009, March 2014). www.ki.se (continuously). Interviews, meetings and seminar (in chronological order): 1. Meeting with Lena Hansson, Hanna Jansson and Jessica Norrbom, Bio-entrepreneurship Unit, KI, 1-10-2009.

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

14

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2. Interview with Rune Fransson, Chairman KIHAB, 28-10-2010, tape recorded. 3. Meeting with Charlotta Dahlborg, Bio-entrepreneurship Unit, KI, 10-11-2010. 4. Interview with Lilian Wikström, CEO KIAB, 29-11-2010. 5. Interview with Conny Bogentoft, formerly CEO KIAB, formerly CEO KIF and formerly CEO and CSO Karolinska Development, 17-5-2011, tape recorded. 6. Interview with Ola Flink, formerly CEO KIAB, VP Portfolio Management Karolinska Development, 20-5-2011, tape recorded. 7. Interview with Folke Meijer, CEO KI Holding AB, 23-5-2011, tape recorded. 8. Interview with Carl Johan Sundberg, Director Bio-entrepreneurship Unit, KI, 17-6-2011, tape recorded. 9. Interview with Björn Kull, R&D Director, Actar AB, 06-09-2011, tape recorded. 10. Witness seminar with Hans Wigzell, former vice-chancellor KI, Hugo Thelin, former consultant KI Holding AB/KIAB, Folke Meijer, CEO KI Holding AB, and Urban Ungerstedt, Prof. Emeritus, KI, 10-10-2011, video filmed. 11. Interview with Ulrika Backman, former project leader of Clanotech at Actar AB, 24-10-2011, tape recorded. 12. Interview with Gunvor Ekman-Oderberg, Professor at KI and co-founder of Dilafor, 24-11-2011, tape recorded. 13. Interview with Jan Nilsson, CEO Lipopeptide, 11-25-2011, tape recorded. 14. Interview with Alvar Grönberg, Director of Preclinical R&D Lipopeptide, 25-11-2011, tape recorded. 15. Interview with Nina Mohell, CEO Aprea, 23-01-2012, tape recorded. 16. Interview with Patrizia Caldirola, CEO Clanotech, 30-01-2012, tape recorded. 17. Interview with Klas Wiman, Professor at KI and co-founder of Aprea, 07-03-2012, tape recorded. 18. Interview with Mona Ståhle, Professor at KI and co-founder of Lipopeptide, 26-03-2012, tape recorded. 19. Interview with Yihai Cao, Professor at KI and founder of Clanotech, 27-03-2012, tape recorded. Selected interview guides (A) Interview guide for Rune Fransson, Chairman KIHAB (2810-2010) (authors' translation of the original in Swedish) 1- What is your role in KI's incubation system? 2- What do you do in practice? 3- Can you please describe KI's overall strategy for commercialization of science? 4- Which organizational structure has KI built up? 5- Which organizational units does it include? What is the role of each unit? 6- Can you please draw a figure over this organizational structure? 7- Can you please explain the background behind the current strategy and structure for commercialization of science? Did you have role model in developing it? 8- How have this strategy and structure changed in the recent years? 9-How does KIHAB operate concretely, especially towards KI University, single researchers and external companies? 10- Can you please suggest us some specific spin-offs or technology originated from KI on which to conduct detailed case studies? (B) Interview guide for Conny Bogentoft, formerly CEO KIAB, formerly CEO KIF and formerly CEO and CSO Karolinska Development, KD (17-5-2011)

(authors' translation of the original in Swedish) 1- What is/has been your role in KI’s incubation system? 2- Why was this system built in the way it was? 3- Which are the strengths and the weaknesses of this incubation system? 4- What types of barriers has it encountered? 5- Which challenges is KI's incubator facing for the future? 6- Can you please explain Karolinska Development's (KD) business model? 7- How does KD manage its portfolio of companies? 8- How did the floatation of KD on the Stock Exchange proceed? Are you satisfied with its results? 9- Have KD-owned companies launched products to market? 10- Which are the 30 þ technologies which have been licensed by KIAB? To which external companies? For which revenues? 11- Can you please suggest examples of companies owned by KD, more or less successful, on which we may conduct in-depth case studies? (C) Interview guide for Lilian Wikström, CEO KIAB, (29-112010) (authors' translation of the original in Swedish) 1- What is your background and when did you become the CEO of KIAB? 2- What is the core business of KIAB? 3- What is KIAB's core competence? 4- How is your business financed? 5- Please, describe a normal working day for you – what are your main tasks of the day? 6- What is the procedure for selecting new incubatees? Please describe the entire process from selection to investment! 7- How do you decide which ideas are turned into companies and which are turned into licensing projects? 8- So far there is one “exit”, CogMed, can you describe the process of getting that idea into commercialization and exit? 9- Is KIAB collaborating with any large pharmaceutical companies that are potential buyers of the incubatees? 10- On the website of KIAB you describe the incubation system at KI as “unique”. In what way would you describe it as unique? 11- What is KIAB's relationship to the researchers providing the new ideas? 12- Which actors, both “inside” and “outside” the KI incubation system, are KIAB's main collaborators? 13- How does KIAB collaborate with KD? 14- What happens to the ideas that KD choose not to invest in? Are they passed on to other potential investors? 15- KD has about 40 incubator projects, how many of them have passed through the procedure applied by KIAB? 16- Have you identified any pros and cons with the procedures that KIAB applies to select new incubatees? 17- Could you please direct me to other people who are also knowledgeable of the overarching structure of KI's incubation system? (D) Interview guide for Ulrika Backman, former project leader of Clanotech at Actar AB, (24-10-2011) (authors' translation of the original in Swedish) 1. Please, describe your background and how you were employed at Actar. 2. You got involved in the research and commercialization of the incubatee “Clanotech”, please describe the medical invention of the project. 3. What stage was the project in as you got involved, and which was your role in the project? 4. What was the scientific, medical and economic risk assessment that was made at KIAB and Actar? 1. Which was the potential medical need? 2. Which was the potential market (the users)?

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

3. Wherein laid the economic risk? 5. How many were involved in the research and commercialization work? 6. Was Actar involved in any other projects at the time? If so, what were their content and challenges? 7. Describe the continuous evaluation and economic accounting process of the project: 1. To what degree have KD been involved in the project, and how have they evaluated it? 2. What were the milestones? 3. Did Clanotech manage to achieve them? How? 8. Which have been the main obstacles of the project? Speed, research, financially? 9. What have been the main drivers of the project? 10. How involved have the founding researchers been? Has this involvement changed over time? 11. What was the trigger for turning Clanotech into a company? Was it due to continuous success or pressure from the owners, or something else? 12. What have been the major changes of the project since it was turned into a company? For instance, how is it related to Actar now when it has become its own company compared to a project within Actar? 13. How would you define the nearest future for Clanotech? Are their critical milestones, is there a need for a quick exit, are there any potential investor? 14. Has anything changed for Clanotech now that KD has been floated on the Stock Exchange?

References Aaboen, L., 2009. Explaining incubators using firm analogy. Technovation 29, 657–670. Abetti, P.A., Rancourt, C., 2008. University incubators as agents for technology transfer and economic growth: case studies in USA, Ukraine and Finland. Int. J. Technol. Intell. Plan. 4 (4), 379–397. Aernoudt, R., 2004. Incubators: tool for Entrepreneurship? Small Bus. Econ. 23 (2), 127–135. Aerts, K., Matthyssens, P., Vandenbempt, K., 2007. Critical role and screening practices of European business incubators. Technovation 27, 254–267. Ahmad, A.J., Ingle, S., 2011. Relationships matter: case study of a university campus incubator. Int. J. Entrep. Behav. Res. 17 (6), 626–644. Albert, P., Gaynor, L., 2006. Technology business incubation management: lessons of experience. In: Bernasconi, M., Harris, S., Moensted, M. (Eds.), High-Tech Entrepreneurship: Managing Innovation, Variety and Uncertainty. Routledge, London, pp. 131–143. Allen, D.N., McCluskey, R., 1990. Structure, policy, services, and performance in the business incubator industry. Entrep. Theory Pract. 15 (2), 61–77. Alsos, G.A., Hytti, U., Ljunggren, E., 2011. Stakeholder theory approach to technology incubators. Int. J. Entrep. Behav. Res. 17 (6), 607–625. Anderson, J.C., Håkansson, H., Johanson, J., 1994. Dyadic business relationships within a business network context. J. Market. 58 (4), 1–15. Andersson, U., Forsgren, M., Holm, U., 2002. The strategic impact of external networks: subsidiary performance and competence development in the multinational corporation. Strateg. Manag. J. 23 (11), 979–996. Axelsson, B., Easton, G. (Eds.), 1992. Industrial Networks – A New View of Reality. Routledge, London. Baraldi, E., Brennan, R., Harrison, D., Tunisini, A., Zolkiewski, J., 2007. Strategic thinking and the imp approach: a comparative analysis. Indus. Market. Manag. 36 (7), 879–894. Baraldi, E., 2008. Strategy in industrial networks: experiences from IKEA. Calif. Manag. Rev. 50 (4), 99–126. Baraldi, E., Waluszewski, A., 2011. Betting on science or muddling through the network′. Two universities and one innovation commission. IMP J. 5 (3), 172–192. Baraldi, E., Ingemansson, M., 2013. Projects as an attempt to make science into business. In: Proceeding of the Embedding Commercialization Projects in Industrial Networks, Paper Presented at the 29th IMP Conference, Atlanta. Baraldi, E., Ingemansson, M., Launberg, A., 2014. Controlling the commercialization of science across organizational borders. Four cases from two major Swedish universities. Indus. Market. Manag. 43, 382–391. Berbegal-Mirabent, J., Rebeiro-Soriano, D., García, J.L.S., 2015. Can a magic recipe foster university spinn-off creation? J. Bus. Res. (in press)http://dx.doi.org/10. 1016/j.jbusres.2015.06.010

15

Bergek, A., Norrman, C., 2008. Incubator best practice: a framework. Technovation 28, 20–28. Biernacki, P., Waldorf, D., 1981. Snowball sampling: problem and techniques of chain referral sampling. Sociol. Methods Res. 10, 141–163. Boje, D.M., 2001. Narrative Methods for Organizational and Communication Research. Sage publications, London. Bøllingtoft, A., 2012. The bottom-up business incubator: Leverage to networking and cooperation practices in a self-generated, entrepreneurial-enabled environment. Technovation 32, 304–315. Bruneel, J., Ratinho, T., Clarysse, B., Groen, A., 2012. The evolution of business incubators: comparing demand and supply of business incubation services across different incubator generations. Technovation 32, 110–121. Casper, S., Kettler, H., 2001. National Institutional frameworks and the hybridization of entrepreneurial business models: The German and UK biotechnology sectors. Indus. Innov. 8 (1), 5–30. Clarysse, B., Wright, M., Lockett, A., Van de Velde, E., Vohora, A., 2005. Spinning out new ventures: a typology of incubation strategies from European research institutions. Journal of Business Venturing 20, 183–216. Casper, S., Matraves, C., 2003. Institutional frameworks and innovation in the German and UK pharmaceutical industry. Res. Policy 32, 1865–1879. Cooke, P., Levin, C., Wilson, R., Kaufmann, D., 2006. The biosciences knowledge value chain and comparative incubation models. J. Technol. Transf. 31 (1), 115–129. Cooper, R.G., 2001. The new product process: the stage-gate game plan. In: Cooper, R.G. (Ed.), Winning at New Products. Perseus Publishing, Reading, MA, pp. 114–153. Davila, T., 2005. An exploratory study on the emergence of management control systems: formalizing human resources in small growing firms. Account. Organ. Soc. 30, 223–248. Dekker, H.C., 2004. Control of inter-organizational relationships: evidence on appropriation concerns and coordination requirements. Account. Organ. Soc. 29, 27–49. Dicken, P., Kelly, P., Olds, K., Yeung, H., 2001. Chains and networks, territories and scales: Toward a relational framework for analysing the global economy. Global Netw. 1, 89–112. Drazin, R., Schoonhoven, C.B., 1996. Community, population, and organization effects on innovation: a multilevel perspective. Acad. Manag. J. 39 (5), 1065–1083. Dubois, A., Gadde, L.-E., 2002. Systematic combining: an abductive approach to case research. J. Bus. Res. 55, 553–560. Edquist, C., 1997. System of Innovation: Technologies, Institutions and Organizations. Pinter Publishers, London. Eisenhardt, K.M., 1989. Building theories from case study research. Acad. Manag. Rev. 14 (4), 532–550. Ehret, M., McDonald-Junor, D., Smith, D., 2012. High technology and economic development: the BioCity Nottingham technology incubator. Int. J. Entrep. Innov. 13 (4), 301–309. Emmanuel, C., Otley, D., Merchant, K., 1990. Accounting for management control. Chapman & Hall, London. Etzkowitz, H., 2002. Incubation of incubators: innovation as a triple helix of university-industry-government networks. Sci. Public Policy 29 (2), 115–128. Fetters, M.L., Green, P.G., Rice, M.P., Butler, J.S., 2010. The Development of University-Based Entrepreneurship Ecosystems: Global Practices. Edward Elgar Publishing, Cheltenham. Flyvbjerg, B., 2006. Five misunderstandings about case-study research. Qual. Inq. 12 (2), 219–245. Forsgren, M., Holm, U., Johanson, J. (Eds.), 2005. Managing the Embedded Multinational: A Business Network View. Edward Elgar, Cheltenham. Gadde, L.-E., Håkansson, H., 2001. Supply Network Strategies. Wiley, Chichester. Gadde, L.-E., Huemer, L., Håkansson, H., 2003. Strategizing in industrial networks. Indus. Market. Manag. 32 (5), 357–364. Grimaldi, R., Grandi, A., 2005. Business incubators and new venture creation: an assessment of incubating models. Technovation 25, 111–121. Gnyawali, D.R., Madhavan, R., 2001. Cooperative networks and competitive dynamics: a structural embeddedness perspective. Acad. Manag. Rev. 26 (3), 431–445. Gnyawali, D.R., Park, B.-J., 2011. Co-opetition between giants: collaboration with competitors for technological innovation. Res. Policy 40 (5), 650–663. Gompers, P., Lerner, J., 2001. The Money of Invention. Harvard Business School Press, Cambridge. Hackett, S.M., Dilts, D.M., 2004. A systematic review of business incubation research. J. Technol. Transf. 29, 55–82. Hansen, M.T., Chesbrough, H.W., Nohria, N., Sull, D.N., 2000. Networked Incubators. Harv. Bus. Rev. 78 (5), 74–84. Hisrich, R.D., 1988. New business formation through the enterprise development center: a model for new venture creation. IEEE Trans. Eng. Manag. 35 (4), 221–231. Håkansson, H., 1997. Organization networks. In: Sorge, A., Warner, M. (Eds.), The IEBM Handbook of Organizational Behaviour, pp. 232–239 1997. Håkansson, H., Snehota, I. (Eds.), 1995. Developing Relationships in Business Networks. Routledge, London. Håkansson, H., Ford, D., 2002. How should company interact in business networks? J. Bus. Res. 55 (2), 133–139, February 2002. Håkansson, H., Waluszewski, A. (Eds.), 2007. Knowledge and Innovation in Business and Industry. The importance of using others. Routledge, London, New York. Johansson, S.-E., Östman, L., 1995. Accounting Theory. Integrating Behaviour and Measurement. Pitman Publishing, London.

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i

16

E. Baraldi, M. Ingemansson Havenvid / Technovation ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Lai, W.-H., Lin, C.-C., 2015. Constructing business incubation service capabilities for tenants at post-entrepreneurial phase. Journal of Business Research (in press) http://dx.doi.org/10.1016/j.jbusres.2015.06.012. Lorenzoni, G., Baden-Fuller, C., 1995. Creating a strategic center to manage a web of partners. Calif. Manag. Rev. 37 (3), 146–163. Lundqvist, M.A., 2014. The importance of surrogate entrepreneurship for incubated Swedish technology ventures. Technovation 34 (2), 93–100. Lundvall, B.-Å., 1999. National business systems and national systems of innovation. Int. Stud. Manag. Organ. 29 (2), 60–77. Lundvall, B.-Å., Johnson, B., Andersen, E.S., Dalum, B., 2002. National systems of production, innovation and competence building. Res. Policy 31, 213–231. Mahadevan, B., 2000. Business models for internet based e-commerce: an anatomy. Calif. Manag. Rev. 42 (4), 55–69. Makkonen, H., Aarikka-Stenroos, L., Olkkonen, R., 2012. Narrative approach in business network process research—Implications for theory and methodology. Indus. Market. Manag. 41 (2), 287–299. Mas-Verdú, F., Ribeiro-Soriano, D., Roig-Tierno, N., 2015. Firm survival: the role of incubators and business characteristics. J. Bus. Res. 68 (4), 793–798. McAdam, M., Marlow, S., 2008. A preliminary investigation into networking activities within the university incubator. Int. J. Entrep. Behav. Res. 14 (4), 219–241. Mian, S.A., 1996. Assessing value-added contributions of technology business incubators to their tenants. Res. Policy 25, 325–335. Mian, S.A., 1997. Assessing and managing the university technology business incubator: an integrative framework. J. Bus. Ventur. 12 (4), 251–340. Mintzberg, H., 1987. The strategy concept I: five Ps for strategy. Calif. Manag. Rev. 30 (1), 11–24. Normann, R., Ramirez, R., 1993. From value chain to value constellation: designing interactive strategy. Harv. Bus. Rev. 71 (4), 65–77. Oakey, R., 2012. High-Technology Entrepreneurship. Routledge, London. Otley, D.T., Berry, A.J., 1980. Control, organisation and accounting. Account. Organ. Soc. 5, 231–244. Perks, H., Jeffery, R., 2006. Global network configuration for innovation: a study of international fibre innovation. R&D Manag. 36, 67–83. Peters, L., Rice, M., Sundararajan, M., 2004. The role of incubators in the entrepreneurial process. J. Technol. Transf. 29, 83–91. Pfeffer, J., Salancik, G.R., 1978. The External Control of Organizations: A Resource Dependence Perspective. Harper and Row, New York. Phan, P.H., Siegel, D.S., Wright, M., 2005. Science parks and incubators: observations, synthesis and future research. J. Bus. Ventur. 20, 165–182. Osterwalder, A., Pigneur, Y., 2010. Business Model Generation. Wiley & Sons, Hoboken. Ramirez, R., 1999. Value co-production: intellectual origins and implications for practice and research. Strateg. Manag. J. 20 (1), 49–65. Rice, M.P., 2002. Co-production of business assistance in business incubators. An exploratory study. J. Bus. Ventur. 17, 163–187. Roig-Tierno, N., Alcázar, J., Ribeiro-Navarrete, S., 2015. Use of infrastructures to support innovative entrepreneurship and business growth. Journal of Business Research (in press)http://dx.doi.org/10.1016/j.jbusres.2015.06.013.

Rosenberg, N., 1994. Path-dependent aspects of technological change. In: Rosenberg, N. (Ed.), Exploring the Black Box. Technology Economics, and History. Cambridge University Press, Cambridge, pp. 9–23 1994. Rothaermel, F.T., Thursby, M., 2005a. University–incubator firm knowledge flows: assessing their impact on incubator firm performance. Res. Policy 34, 305–320. Rothaermel, F.T., Thursby, M., 2005b. Incubator firm failure or graduation? The role of university linkages. Res. Policy 34, 1076–1090. Rubin, T.H., Aas, T.H., Stead, A., 2015. Knowledge flow in technological business incubators: evidence from Australia and Israel. Technovation 41–42, 11–24. Somsuk, N., Laosirihongthong, T., 2014. A fuzzy AHP to prioritize enabling factors for strategic management of university business incubators: resource-based view. Technol. Forecast. Soc. Change 85, 198–210. Schwartz, M., Hornych, C., 2008. Specialization as strategy for business incubators: an assessment of the Central German Multimedia Center. Technovation 28 (7), 436–449. Schwartz, M., Hornych, C., 2010. Cooperation patterns of incubator firms and the impact of incubator specialization: Empirical evidence from Germany. Technovation 30, 485–495. Sharif, N., 2006. Emergence and development of the national innovation systems approach. Res. Policy 35 (5), 745–766. Smith, D.J., Zhang, M., 2012. Introduction: the evolution of the incubator concept. Special Issue: technology incubators – promoting science businesses? Int. J. Entrep. Innov. 13 (4), 227–234. Stabell, C.B., Fjeldstad, O.D., 1998. Configuring value for competitive advantage: on chains, shops, and networks. Strateg. Manag. J. 19 (5), 413–437. Stone, R., Frank, L., 2001. Karolinska Inc. Science 293, 2374–2376. Strömsten, T., Waluszewski, A., 2012. Governance and resource interaction in networks. The role of venture capital in a biotech start-up. J. Bus. Res. 65, 232–244. Tomkins, C., 2001. Interdependencies, trust and information in relationships, alliances and networks. Account. Organ. Soc. 26 (2000), 161–191. Vahlne, J.-E., Johanson, J., 2013. The Uppsala model on evolution of the multinational business enterprise: from internalization to coordination of networks. Int. Market. Rev. 30 (3), 189–210. Vanderstraeten, J., Matthyssens, P., 2012. Service-based differentiation strategies for business incubators: exploring external and internal alignment. Technovation 32, 656–670. Van de Ven, A., Polley, D., Garud, R., Venkataraman, S., 1999. The Innovation Journey. Oxford University Press, New York, Oxford. Vohora, A., Wright, M., Lockett, A., 2004. Critical junctures in the development of university high-tech spinout companies. Res. Policy 33, 147–175. Whitley, R., 1994. Dominant forms of economic organization in market economies. Organ. Stud. 15 (2), 153–182. Whitley, R., 2000. The institutional structuring of innovation strategies: business systems, firm types and patterns of technical change in different market economies. Organ. Stud. 21 (5), 855–886. Yin, R.K., 2003. Case Study Research: Design and Methods, third edition. Sage Publications, Thousand Oaks, CA.

Please cite this article as: Baraldi, E., Ingemansson Havenvid, M., Identifying new dimensions of business incubation: A multi-level analysis of Karolinska Institute's incubation system. Technovation (2015), http://dx.doi.org/10.1016/j.technovation.2015.08.003i