To Engage or Not to Engage: The Case of an Emerging Innovation Ecosystem in Sweden

(1)

Master Thesis

HALMSTAD

UNIVERSITY

Master's Programme in Industrial Management and Innovation, 120 Credits

To Engage or Not to Engage: The Case of an Emerging Innovation Ecosystem in Sweden

Master's Thesis, 30 Credits

Halmstad 2020-06-22

Alireza Esmaeilzadeh, Harvey Blanco

(2)

i

ABSTRACT

The purpose of this study is to explore the engagement in an innovation ecosystem for knowledge co-creation. It aims at exploring the various aspects of ecosystems, innovation and knowledge which can drive or hinder actors to engage in collaboration in an innovation ecosystem. A single case study was selected as a research strategy (The OSMaaS project), as it provided us the opportunity to analyze an innovation ecosystem with specific characteristics that few has considered before. Semi-structured interview was used as data collection technique since this interview method offered us the required flexibility to explore in depth the individual experiences lived during the process of evaluating whether to engage or not to the OSMaaS project. Consequently, a hybrid approach of thematic analysis was selected as method for data analysis as it allowed us to interact with the interviewees or the empirical world, the concepts regarding innovation and ecosystems or theory, and the OSMaaS project or the case study. The findings show that aspects of ecosystems, innovation and Knowledge co-creation aspects such as co-opetition, ecosystem governance and structure, proximity, relative advantage, compatibility, complexity, trialability, observability, competitive advantage, and product development contain factors driving and hindering actors’ engagement in an innovation ecosystem. These factors are explained within this study and show what have drove and hindered actors to engage in the OSMaaS project.

Keywords: innovation ecosystem, co-opetition, knowledge co-creation, governance and structure, proximity, Innovation aspects, abductive research, thematic analysis, mobility as a service.

(3)

ii

ACKNOWLEDGEMENT

We wish to thank everyone that have helped us for writing this thesis. First and foremost, we wish to thank Maya Hoveskog and Fawzi Halila for their valuable support, feedback and guidance throughout our studies. Further, A special thanks to the OSMaaS project members that have participated in our thesis. Your valuable information and insights formed our empirical data. Thank you!

Halmstad, June 2020 Harvey Blanco and Alireza Esmaeilzadeh

(4)

iii

LIST OF FIGURES

Figure 1 A schematic view of the theoretical frame of reference of this study ... 12

Figure 2 Ecosystem Aspect ... 16

Figure 3 Innovation Aspect ... 17

Figure 4 Knowledge co-creation ... 19

Figure 5 Theoretical Frame of reference ... 20

Figure 6 Research onion, adapted from (Saunders et al., 2009) ... 21

Figure 7 Stages undertaken to code the data, a diagrammatic representation - adapted from Federay and Muir-Cochrane, 2006 ... 25

Figure 8 Graphic illustration of the OSMaaS innovation ecosystem. ... 38

(8)

vii

LIST OF TABLES

Table 1 A typology of innovation strategies (adapted from Wilhelm and Kohlbacher, 2011) 18

Table 2 Interviewees profile ... 23

Table 3 Definition of the terms used when analyzing the data ... 24

Table 4 Establishing trustworthiness during each phase of thematic analysis - adapted from Notwell et al., 2017 and Pearse, 2019 ... 26

Table 5 An example of the structural coding of the codebook – adapted from Boyatzis, 1998 and Guest et al., 2012 ... 28

Table 6 Example of the process used for reaching consensus when segmenting data ... 29

Table 7 An example of coding all data sources by applying the codes from the codebook .... 29

Table 8 An example of data-driven codes with segments of text from all participants ... 30

Table 9 An example of the process of connecting the codes and identifying sub-themes ... 31

Table 10 Example of corroborating and legitimating sub-themes to identify themes ... 33

Table 11 Description of the five principles of research ethics used for our study ... 35

Table 12 Profile of the organizations engaged to OSMaaS ... 40

Table 13 Drivers and hinders regarding ecosystem aspects ... 51

Table 14 Drivers and hinders regarding innovation aspects ... 53

Table 17 Drivers and hinders for actors to engage in the OSMaaS innovation ecosystem ... 60

Table 18 Drivers and hinders regarding the ecosystem aspects ... 66

Table 19 Drivers and hinders regarding the innovation aspects ... 69

Table 20 Drivers and hinders regarding the knowledge co-creation aspects ... 71

(9)

1

1 INTRODUCTION

Mobility as a service (MaaS) adds value by integrating different forms of transport services into a sole, on-demand mobility service (Hietanen, 2014). It aims to provide a flexible, personalized, service on demand, and user-centered mobility service by bundling different mobility services (Jittrapirom et al., 2017; Polydoropoulou et al., 2020). It is also a marketplace for mobility services in which customers can purchase from different suppliers (Willing et al., 2017). MaaS has the potential to make the transport system much more efficient and help the users to recognize and choose the daily mobility options (Strömberg et al., 2016). It emerges because of opportunities afforded by digital information platforms to plan and deliver multimodal mobility options in point-to-point trips and first/or last-mile travel to public transport journey (Wong et al., 2018). However, the way that MaaS will unfold in the future is still uncertain and depends on a series of technological, social and regulatory trends and developments (Polydoropoulou et al., 2020).

Mobility is no longer a physical asset to purchase, instead, it should incorporate all transport services including cars, buses, and rail as a single on-demand service (Ambrosino et al., 2016).

Therefore, enterprises within the automotive industry need to consider changing their business model from a product to a service-oriented model; combining products with services or refining and personalizing the services offered. For instance, nowadays, car manufacturers not only sell cars as a product but also provide “mobility-as-a-service” for their customers. This shifting focus from products to services leads to new value propositions to customers, which requires new value creation activities, new partnerships, and asks for new revenue models. in other words, novel service offering will require changes to existing business models (Athanasopoulou et al., 2019).

MaaS has been argued as being a part of the solution to prevent transportation problems. To date, however, progress from pilots to large-scale implementation has been slow due to different aspects affecting its development and implementation (Karlsson et al., 2020). This is mainly because MaaS does not comply with traditional business models but rather networked business models co-created in a network of actors where the development process is continuous and iterative by nature (Wong et al., 2018). The nature of collaboration between traditional transport and emerging mobility service is extremely different. Current business models are designed to be separated and competing; in contrast with the idea of collaboration. Therefore,

(10)

2

business models should be re-designed under consideration of mutual interdependence (Beutel et al., 2014). Business models need to be better integrated to MaaS to stimulate innovation among actors (Pangbourne et al., 2020). The success of MaaS significantly depends on understanding the unique particularities of each implementation area to design appropriate business models that can provide the necessary business availability to the involved actors.

These business models need to take into account the MaaS operation mission, objectives, and strategies (Polydoropoulou et al., 2020).

As Sochor et al. (2015) stated, the major obstacles towards the success of MaaS is that the environment in which urban mobility operates, is rather fragmented and it lacks a holistic approach. As a result, synergies could not be achieved between different modes of transport.

Decisions are often based on public actions and do not sufficiently address interfaces with the private sector and the contribution it could make to the achievement of urban mobility goals.

A second barrier, as explained above, is that business models of individual service providers do not necessarily fit within the scope of the mobility broker. The integration of services requires new and integrated business models. A third barrier is that, if public transport is partially subsidized through taxes, it will be hard for a mobility broker to purchase such as trips for less than what the individual traveler can with a monthly or yearly pass. Lastly, cooperation between the broker and public transport management becomes a question of policy.

For MaaS development and implementation to take off, a common vision and road map is needed, where public and private actors share the risk inherent in investing on a new unproven concept (Karlsson et al., 2020). MaaS ecosystems are formed by public and private actors who need to cooperate and compete in order to capture value. MaaS business ecosystem is the wider network of firms that influences how the MaaS provider, creates and captures value and comprises of a wide range of stakeholders including public authorities, public and private transport operators, data providers, IT companies, ticketing and payment service providers, telecommunications, financing companies, institutions, passenger associations, etc. The composition of each business ecosystem may be totally different due to the different configuration of its areas (Polydoropoulou et al., 2020). The advance of MaaS business models is vital for the prosperity of the emerging MaaS ecosystem (Karlsson et al., 2020).

Wong, et al. (2018), defines the network of actors within MaaS as: Transport operators (plus mobility service providers like parking operators), data providers, technology and platform providers, information and communication infrastructure, insurance companies, regulatory organizations, and university/research institutions. The government is an active player

(11)

3

assuming a broker role amongst these actors, even as an interface that magnifies the aggregation challenge. However, the broker role is a challenging proposition for governments since they may not only lack the incentive to innovate but also it may cause potential conflict of interest. However, Smith et al.(2018), argue that having public transport authorities in front seat roles would bring stability to MaaS and guarantee service existence and coverage. MaaS on the other hand, would create more flexibility for public transport. After all, mobility service providers, public transport authorities and regional authorities are commonly indicated as the key actors in the MaaS partnership (Polydoropoulou et al., 2020).

In summary, an innovation ecosystem with a variety of actors from different sectors is indispensable for the development of a comprehensive MaaS. However, each actor might need a different motive to join the ecosystem. In their research, Polydoropoulou et al. (2018), show that different actors might have different motivations to join the MaaS partnership. For example, private sector companies’ main motive is to gain a higher market share while small operators might join the partnership to create their network with bigger alliances. On the other hand, larger companies, like mobility service providers and public transport operators, might be more interested in quality demand data. Moreover, depending on the stage of development of an innovation ecosystem, the actors might engage the network with different reasons. Most of the studies on innovation ecosystem take place when the innovation ecosystem is already mature.

Regarding the necessity of collaboration for a successful development of MaaS, the concept of innovation ecosystems is chosen as the theoretical backbone of this study. Ecosystem as a concept stem from the analogy of biology and it proved to be a promising approach to innovation (Gómez-Uranga et al., 2014). Moore (1993) believes that, organizations are loosely interconnected and networked actors of an ecosystem in which they co-evolve their capabilities around an innovation by sharing knowledge, resources, technologies, and skills while competing and collaborating. However, the innovation ecosystem as a construct, is still too broad. Depending on the unit of analysis different scholars emphasized on different aspects of innovation ecosystems (Jacobides et al., 2018). To find our path through a jungle of definitions, we position our discussion around a more specific type of innovation ecosystem called “the Jam Central Model” described by Zahra and Nambisan (2012). The Jam Central Model refers to a group of independent actors that collaborate in order to envision and develop a radical innovation. Further, we provide a brief description of each group of aspects and its important aspects.

(12)

4

1.1 Ecosystem Aspects

Ecosystem aspects in this study, refers to the potential aspects related to the quality of network that might influence engagement to an innovation ecosystem. Nowadays, organizations require to continually use, create, and share knowledge in order to remain competitive. For this, organizations need to co-operate with other actors, among which some may also compete for similar technology or market. Co-opetition refers to a continually changing dynamic balance between competition and cooperation of independent actors. As a new business reality, co- opetition brings more complexity to the relationships among actors within an innovation ecosystem, compare to the linear models of innovation (Baldwin and von Hippel, 2011). The balance between harmony and rivalry, helps the innovation ecosystems to both exploit the existing knowledge within the network, and explore new knowledge for further developments.

As Wilhelm and Kohlbacher (2011) noted, although co-operation promotes trust for exploitation of knowledge, strong co-operative bonds may lead to structural lock-in.

Consequently, creative tensions inherited from competition, is a key to permitting radical knowledge creation. Briefly, in an inter-organizational context with a radical, multi-technology innovation goal, actors must maintain a balance between co-operation and competition in order to reach an effective knowledge co-creation.

The role of governance is critical when balancing the rivalry and harmony in a way that neither the innovation ecosystem become chaotic, nor it suffers from structural inertia. Gulati et al (2012) claim that an architect who designs the hierarchy, sets system-level goals, and establishes interfaces and standards, is necessary for the ecosystem to remain balanced, although, the relationship between the members and the hub is based on self-selection. On the other hand, other more recent studies suggest that more formal mechanisms like platform governance, management of standards, intellectual property rights, and other contractual forums are critical tools with which the hub firms can control the ecosystems, motivate the actors and discipline other ecosystem members (Jacobides et al., 2018).

Another aspect that might affect the actors’ engagement is proximity. Porter (1998) claims that, geography must not be an origin for competitive advantage any longer, however, in practice, location is still essential to competition. Boschma (2005, p. 61) discuss that, “geographical proximity per se is neither a necessary nor a sufficient condition for learning to take place.

Nevertheless, it facilitates interactive learning, most likely by strengthening the other dimensions of proximity. However, proximity may also have negative impacts on innovation

(13)

5

due to the problem of lock-in”. The dimensions of proximity that Boschma (2005) describes are cognitive, organizational, social, institutional and geographical proximity which are described in detail in the next chapter.

1.2 Innovation Aspects

Innovation aspects in this study, refers to the potential aspects related to the quality of an innovation, from the idea to implementation. The attributes of an innovation might also motivate or demotivate actors to join an innovation ecosystem. Rogers (1983) claims that the adopters’ perception of innovation attributes can predict the rate of adoption of an innovation.

Rate of adoption refers to the relative speed of an innovation to be adopted by the members of a social system. If actors perceive the diffusion of an innovation as successful, then it might motivate them to engage in the innovation ecosystem. In chapter two, we provide the attributes of innovation that may affect the decision of the actors when joining an innovation ecosystem.

1.3 Knowledge co-creation Aspects

Knowledge co-creation in this study, is simply the outcome of engagement in an innovation ecosystem. The quality of such outputs can potentially affect the willingness of the current actors to continue their collaboration, or for new actors to join an innovation ecosystem. As Drucker (1994) puts in, knowledge is the sole meaningful resource and a difficult asset to harness and locate. Nowadays, Organizations learned to collaborate in order to create new knowledge. This new knowledge results in a new competitive advantage in market or a new product that is developed within the ecosystem. Although, knowledge seems to be a main driver for engagement in an ecosystem, there are more fundamental aspects that shape the arrangement in background. For example, Wilhelm and Kohlbacher (2011), describes how co- opetition promotes knowledge co-creation by maintaining the balance between rivalry and harmony. In the next chapter, we also discuss the Knowledge co-creation aspects in more detail.

(14)

6

The purpose of the study is to explore the engagement in an innovation ecosystem for knowledge co-creation. This analysis is grounded in the early phase of an ongoing project;

more precisely, the OSMaaS¹ project and was guided by the following research question:

What are the drivers and hinders for actors to engage in an innovation ecosystem for knowledge co-creation?

In the next section, the layout of the thesis is presented.

1.4 Thesis Layout

This study consists of seven separate chapters. The first chapter is an introduction in which we provide background, problem discussion, the research purpose and question, and the thesis layout. Next, in the second chapter, we introduce our frame of reference by a review of the literature on innovation ecosystem, highlighting definitions, types of innovation ecosystems, typology of actors, and the aspects that might affect actors’ engagement.

In the third chapter, we describe our method with which we answer the research question.

Further, we explain and justify our research approach, research strategy, research context and sample, data collection method, and data analysis.

In the fourth chapter, the empirical case is introduced. For this, we describe the OSMaaS project which is our empirical context.

Fifth chapter introduces the findings of the empirical study. These findings are later used in the discussion.

In the sixth chapter, we discuss the findings of the research. The aim of this chapter is to compare the existing literature with the findings of our research.

Finally, the seventh chapter contains of a summary of the findings, managerial implications and a set of suggestions for further research.

1 Open, Self-organized, Mobility as a Service (OSMaaS project is the empirical case of this study and it is presented in chapter 4).

(15)

7

2 THEORETICAL FRAME OF REFERENCE

In this chapter, we provide the relevant theories to frame the references of the thesis. First, innovation ecosystem as a construct, its origins and relevant definitions of the concept are presented. Then, different types of innovation ecosystems and different types of potential actors within an innovation ecosystem is discussed. Further, ecosystems aspects, innovation aspects, and Knowledge co-creation aspects are described in more detail. This, in order to find potential drivers and hinders for actors to engage in an innovation ecosystem.

2.1 Innovation Ecosystem

In the literature of innovation, the ecosystem construct has emerged as a promising approach.

However, this construct has been developed under a variety of labels, each of which takes a specific perspective to define the phenomenon. In the past-half century, the literature on organizational routines has been affected by the analogy of biology (Gómez-Uranga et al., 2014). The concept of ecosystem had first been used by ecology researchers in 1930, before it was adopted by social sciences to conceptualize the global economy as an ecosystem in which organizations and customers interact with each other as living organisms (Valkokari, 2015).

Gomes et al. (2018) cites Moore (1993) as the pioneer author in the management literature introducing the business ecosystem concept. As Moore (1993) puts in, companies should be considered as part of an ecosystem with loosely interconnected network of actors, coevolving their competences around an innovation by competing, collaborating, and sharing skills, knowledge, resources and technologies. Doing also the analogy form the biological field, Iansiti and Levein (2004) claim that a loosely connected network of entities, including suppliers, distributors, makers of related products and services, outsourcing firms and a number of other organizations constitute business ecosystems, in which the delivery of a company’s own offering affect and are affected by each entity.

Hence, the ecosystem metaphor is subject of criticism by some authors, as for instance, Oh et al. (2016) which argues that the negative effect of using ecosystem as metaphor within a business context outweighs the benefits. The first reason is the lack of established rigorous correspondence rules between natural and business ecosystems. Secondly, unlike ecosystems in business context, natural ecosystems are not designed or engineered systems. Lastly, natural ecosystems do not have policies. Although the analogy from biology may not completely apply to economics, the linkage between the abstract fundamentals in both biology and social

(16)

8

sciences is proven to be beneficial (Gómez-Uranga et al., 2014). Despite the controversy built around the ecosystems construct, its literature serves as a theoretical lens to describe interorganizational networks and alliance portfolio (Overholm, 2015). Furthermore, ecosystems have also been used as framework for modeling value creation and value capture (Priem et al., 2013).

Within business context, there are a variety of conceptualizations of ecosystems (Gomes et al., 2018). Adner (2017, p. 42) proposed that “the ecosystem is defined by the alignment structure of the multilateral set of partners that need to interact in order for a focal value proposition to materialized”. Moore (2006) defines business ecosystem as “intentional communities of economic actors whose individual business activities share in some large measure the fate of the whole community”. Considering business ecosystem and innovation ecosystem as equal concepts, Adner (2006, p. 2) describes it as “the collaborative arrangements through which firms combine their individual offerings into a coherent, customer-facing solution”. Although some authors have used business ecosystem and innovation ecosystem as synonyms, Valkokari (2015) distinguishes these two constructs by showing the differences and logic of action.

Gomes et al. (2018) states that innovation ecosystem draws upon business ecosystem and argue that business ecosystem relates mainly to value capture while innovation ecosystem focuses on value creation. Zahra and Nambisan (2012, p. 220) provide another definition of the innovation ecosystem as, “a group of companies – and other entities including individuals, too, perhaps – that interact and share a set of dependencies as it produces the goods, technologies, and services customers need”. Gomes et al (2018, p. 45) also provides a definition for innovation ecosystem which is more aligned with this research; “an innovation ecosystem is set for the co- creation, or the jointly creation of value. It is composed of interconnected and interdependent networked actors which includes the focal firm, customers, suppliers, complementary innovators and other agents as regulators”. However, regarding the variety of definitions, we prefer to pick a more specific definition that fits to the current study. To do so, in the next section, different types of innovation ecosystem is presented.

2.2 Types of Innovation Ecosystem

Zahra and Nambisan (2012) distinguish four types of innovation ecosystem which differ in their nature of innovation space and governance. Although, the first three types of innovation ecosystems described by Zahra and Nambisan (2012) are not aligned with the focus of this

(17)

9

study, a short introduction as provided by those authors is helpful for thereafter positioning our discussion.

2.2.1 The Orchestra Model

The Orchestra model refers to a group of firms working jointly to exploit a market opportunity grounded on a clear innovation architecture/platform orchestrated by a dominating firm. For instance, in their ecosystems, Intel and Microsoft are the dominating firms while the other actors either create new products and services as a part of an integrated solution or operate on the main firms’ primary product/technology by adding value as a complementary offering.

2.2.2 The Creative Bazar Model

In the Creative Bazar model, the dominant firm searches for innovation in a global bazar of technologies, new ideas and products. Later, the dominant firm uses its own infrastructure to develop on the acquired ideas and commercialize them. For example, small actors within the biotechnology industry target specific areas of interest of large pharma companies to ensure the rapid commercialization for their R&D outputs.

2.2.3 The MOD Station Model

Furthermore, the MOD Station model allows customers to modify and distribute existing products or services. Here, the main firm exploits the new markets and technologies by using a community of innovators, experts, customers and scientists. In this case, the main firm plays the role of catalysts by providing the innovation architecture. In the PC-based video game industry, some companies follow this model to allow their customers to create ‘mods’

(modifications) of existing games and also to distribute them to other customers.

2.2.4 The Jam Central Model

The last model discussed by Zahra and Nambisan (2012), which is more aligned with this study, is the Jam Central Model. This is when a group of independent actors collaborate to imagine and develop an emergent or radical innovation. In this type of innovation ecosystem, the direction and goals arise organically from the collaboration. In other words, the nature of the innovation is rather improvisational. Moreover, the ecosystem is not governed by a centralized leader and the governance responsibility is distributed among actors. Linux open source software is a good example of such an ecosystem; IBM played a supportive role with

(18)

10

limited authority when setting goals and activities within the ecosystem. As a result, the company developed new businesses in collaboration with the open source community.

The Jam Central model focuses its efforts on knowledge development, with less emphasis on the commercialization of opportunities and market risks. However, any success in the commercialization of new technology/knowledge depends on broad collaboration among the different actors within the same or similar areas of specialization, combining their diverse and complex knowledge bases into the emergent field. A critical challenge here, is to protect assets and knowledge of the actors while keeping interactions and sharing open. The creation of new knowledge can generate new ecosystems and obsolete exiting ones, jeopardizing large companies when imagining the commercialization possibilities based on the new knowledge.

We chose this classification to position their discussion more specifically. Although the literature reveals more classifications of innovation ecosystems including Hub-based innovation ecosystem (Nambisan and Baron, 2013), Open innovation ecosystem (Chesbrough et al., 2014), Platform-based ecosystem (Gawer, 2014), and Digital innovation ecosystem (Rao and Jimenez, 2011), the characteristics and differences among them is out of the scope of this thesis.

Further, the typology of the actors in an innovation ecosystem has been discussed. This due to the fact that depending of the type of organizations, actors may join the ecosystem for different reasons. Therefore, it is important to take the organization type into consideration when exploring the hinders and drivers for engagement.

2.3 Typology of Actors

The typology of the actors involved in innovation ecosystems should also be taken into consideration when analyzing the drivers and hinders since the actors’ expectation for joining in such ecosystems may depend on their responsibilities and objectives. As an analytical model the triple-helix describes the variety of policy models and institutional arrangements, and their dynamics. The networks of relations create surplus value by continuously harmonizing the underlaying infrastructure towards the settled goals. However, reaching a common goal for the innovation ecosystem might be challenging since the driving force of the interactions between actors might differ on their expectation of benefits which might mean different things for different actors. Moreover, the continuous interactions between actors involve relational tensions. Nevertheless, these tensions need not to be resolved because the resolution of such tensions, might harm the dynamics of the system (Etzkowitz and Leydesdorff, 2000).

(19)

11

As concept, the triple-helix explains the development of simultaneous pair-wise collaboration of independent actors from different sectors including, public sector (agencies and government bodies), knowledge generating sector (research institutes, R&D centers and universities), and business sector. The diversification of actors and their interactive relationships creates a highly complex ecosystems which maximizes the exchange and knowledge co-creation and information (Etzkowitz and Leydesdorff, 2000). Furthermore, triple-helix cooperation patterns stretch mutual interdependencies within an innovation ecosystem which, in turn, leads to synergy effect of self-supportive growth and co-creation of value on a continual basis (Russell and Smorodinskaya, 2018; Smorodinskaya and Katukov, 2015). Although, the triple-helix model shows the fundamental actors of innovation ecosystems, it lacks some other critical players. MIT University, in a framework for building and accelerating innovation ecosystems, goes beyond the triple-helix model and suggests a more comprehensive classification with five types of stakeholders including Entrepreneur, Risk capital, University, Corporation, and Government (Budden and Murray, 2019).

Entrepreneurs refers to a specific type of start-up enterprises aiming to create competitive advantage through new innovations to scale and grow quickly. Although they are as entrepreneurial as the other actors, they might lack the dense networks of resources and key individuals. Therefore, networking is a good motive for them to join an innovation ecosystem in order to overcome the lack of founder resources. The role of these entrepreneurs is crucial for the success of an innovation ecosystem since they are in the frontline of innovation.

The presence of risk capital in an innovation ecosystem is necessary, but not sufficient. Their role must be more than simply funding the ecosystem. On the other hand, innovation ecosystems provide a geographically localized and efficient context for these actors to identify new ideas, teams and entrepreneurs. Moreover, the deep network acts as a source of endorsements and referrals for them to allocate the limited time and investments to the better ideas among all the investment choices.

The role of certain universities in the development of innovation ecosystems is iconic. For example, Stanford university played a crucial role in the emergence of Silicon Valley.

Universities’ role varies widely from technical and scientific training, novel science-based ideas, sophisticated facilities, entrepreneurial education, etc. However, their choice of focus between research and education determines their role in the ecosystem.

(20)

12

Large corporations, on the other hand, seem to be less attached to local innovation ecosystems since they have increasingly seen themselves as multi-national or global. However, like other actors in the ecosystem they have a role to play in and benefit from the dense network of connections. They can provide the ecosystem with facilities, infrastructure for innovation, on- the-job talent development, their convening power, and contribution to risk capital.

Although the presence of government is crucial for development of innovation ecosystems, their role is often controversial. Their contribution is mainly beneficial for shaping appropriate norms and rules within ecosystem networks, no matter as the leader or a key participant. It is worth mentioning that governments can participate in different levels including national level, regional level, or city level. Furthermore, they can also come from different governmental departments at each level, as well as, from political side or a more official side.

Some other types of actors may also participate in the ecosystem such as specialized service providers (like lawyers), Non-governmental organizations (NGOs), Some corporates (like financial institutions and banks), and accelerators but they are often funded by one of the main stakeholders of the ecosystem.

Further, we discuss different aspects that may promote or demote engagement in an innovation ecosystem for knowledge co-creation. a schematic view of the discussion is presented in figure 1.

2.4 Ecosystem Aspects

Depending on the unit of analysis different scholars emphasized on different aspects of innovation ecosystems (Jacobides et al., 2018). However, some authors believe that innovation

Figure 1 A schematic view of the theoretical frame of reference of this study

Ecosystem Aspects

Knowledge Co-creation

Aspects Innovation

Aspects

(21)

13

ecosystem is defined by three main characteristics. First, there is a stablished dependency amongst actors since survival and performance of the actors and the ecosystem itself are intently linked. Second, the actors and the ecosystem share common objectives and goals to serve a certain customer value preposition. Lastly, actors within the ecosystem share a set of complementary knowledge, skills, technologies and capabilities (Adner and Kapoor, 2010;

Iansiti and Levien, 2004; Nambisan and Baron, 2013).

Innovation ecosystem are open non-linear systems. They are characterized as altering multi- faceted motivations of interconnected actors, continuous structural transformation encouraged both internally and externally, and high receptivity to feedback. For innovation ecosystem to be self-adaptable towards rapid change, a non-hierarchic, collaborative model of governance, and agile network relationships, is needed (Russell and Smorodinskaya, 2018).

Further, we discuss the aspects on innovation ecosystems in order to find potential drivers and hinders for engagement of actors.

2.4.1 Co-opetition

As it appears in the literature of innovation ecosystem, the term collaboration has been used broadly to describe the relationship between networked actors, however, there is a lack of universal definition in the literature regarding this concept. According to Camarinha-Matos and Afsarmanesh (2008), collaboration is seen as a process in which mutually engaged participants share information, resources, responsibilities and risks to jointly plan, implement, and evaluate a program of activities aimed at achieving a common goal. The level of this collaboration is also controversial. As Russel and Smorodinskaya (2018, p. 116) describe, collaboration is “the most developed form of interactive communication”. According to Russel and Smorodinskaya (2018), for innovation ecosystem actors to co-create a new set of values, a sophisticated stage of interactions, common strategy, joint goals, joint identity and joint responsibilities, is crucial. On the other hand, authors defines the relationship between actors in an innovation ecosystem as loosely networked which to some extent is in contrast with the tight bonds that are described by Russel and Smorodinskaya (2018). The word collaboration is often used to describe co-operation. Therefore, a definition provided by Wilhelm and Kohlbacher (2011, p. 68) is taken in this study that states, “Co-operation is often described as a situation where compatible goals result in joint action between two firms with the goal of achieving mutual benefits.”

(22)

14

Despite the importance of collaboration when stretching relations among actors within innovation ecosystems, by no means it excludes competition between them. Wilhelm and Kohlbacher (2011, p. 68) define the competition as, “competition is usually said to occur when two parties strive for something only one can attain. Generally, competition can be understood as the rivalry caused when structural conditions underlying the relationship between two firms (for example, product similarity) are dependent on the same resources (such as potential customers for the product in question)”. Actors might compete in some business projects while co-operating in others. This phenomenon is known as Co-opetition which refers to a continually changing dynamic balance between competition and cooperation of independent actors. Co-opetition as a new business reality brings more complexity to the relationships among actors within an innovation ecosystem, compare to the linear models of innovation (Baldwin and von Hippel, 2011). However, relational contracts and coordinated activities under a common strategy can regulate the complexity of such co-opetition (Russell and Smorodinskaya, 2018).

The common goal of the actors when collaborating within an innovation ecosystem, is to jointly tackle the challenges towards the strong global competition. There is a strong relationship between the strength of actors’ linkages and the co-created value added within an innovation ecosystem. Therefore, collaboration implies a specific dynamic balance as well as several types of complex relationships. In this sense, when the level of cooperation between actors reaches a certain level - in terms of joint strategy, joint goals and joint identity - the innovation ecosystem emerges (Russell and Smorodinskaya, 2018).

2.4.2 Ecosystem Governance and Structure

Jacobides et al (2018) argue that what enables ecosystems to emerge, is modularity. Modularity lets a number of interdependent yet distinct organizations to coordinate with no formal hierarchical authorization. Ecosystems create added value through coordination of a set of roles that face similar regulations without customized contractual agreements with each partner.

However, most of the studies tend to highlight the role of a specific actor as the hub or “lead firm” in the emergence of an ecosystem (Williamson and De Meyer, 2012). Gulati et al (2012) argue that the role of an architect who defines the hierarchy, sets system-level goals, and establishes interfaces and standards, is indispensable. However, the relationship between the members and the hub is based on self-selection. On the other hand, some more recent studies suggest that more formal mechanisms like platform governance, intellectual property rights,

(23)

15

management of standards and other contractual forums are crucial tools in the hands of hub firms with which they motivate and discipline other ecosystem members (Jacobides et al., 2018).

The governance of an innovation ecosystem is related to the following actions: designing the role of the actors and coordinating the interactions between them, orchestration of resource flows among alliance members, and forging partnerships. Designing the role of the actors refers to (Dedehayir et al., 2018).

2.4.3 Proximity

Friedman (2005) in his book ‘The World Is Flat’ states that innovation can happens anytime and anywhere, however, in the modern economy the world of innovation may not seem so flat.

In reality, innovation is distributed unevenly throughout the world. As Porter (1998) puts in, while in theory geography should not be a source of competitive advantage anymore, in practice, location is still central to competition.

According to Boschma (2005) “geographical proximity per se is neither a necessary nor a sufficient condition for learning to take place. Nevertheless, it facilitates interactive learning, most likely by strengthening the other dimensions of proximity. However, proximity may also have negative impacts on innovation due to the problem of lock-in”. Boschma (2005) defines five different dimensions for proximity including cognitive, organizational, social, institutional and geographical. Cognitive proximity refers to the bounded rationality of the economic actors.

In other words, firms search in neighboring proximity to their existing knowledge base in order to reduce uncertainty. Organizational proximity means the limited capacity of coordination and exchange between different pieces of knowledge owned by different actors between and within organizations. Economic relations are also a function of social context. The more socially embedded are the firm’s relationships, the better its innovation performance and the more interactive learning. However, too much social proximity may reduce the learning capability of organizations. Institutional proximity implies to the norms, values and institutional arrangements at the micro-level which are embedded in specific exchange relations. The last form of proximity refers to the geographical distance between economical actors. Short distance has proven by a large body of literature to be helpful for bringing people together and exchanging tacit knowledge. However, without a restricted definition of geographical proximity, it seems to be combined with other forms of proximity. It is important to note that

(24)

16

regardless of the type of proximity, both too little and too much proximity harms the innovation and learning (Boschma, 2005).

Figure 2 summarizes the aspects involved in the ecosystem aspects.

2.5 Innovation Aspects

The idea behind an innovation might act as a driver or hinder for different actors to join an ecosystem. Rogers (1983, pp. 250–251) suggests a set of attributes to describe an innovation.

The individual receivers’ perception of these attributes can predict the rate of adoption of the innovation. Rate of adoption refers to the relative speed of an innovation to be adopted by the members of a social system. A brief definition of each attributes is shown as:

• “Relative advantage is the degree to which an innovation is perceived as better than the idea it supersedes”. Relative advantage and rate of adoption has a positive correlation.

• “Compatibility is the degree to which an innovation is perceived as consistent with the existing values, past experiences, and need of potential adopters”. Compatibility and rate of adoption has a positive correlation.

• “Complexity is the degree to which an innovation is perceived as relatively difficult to understand and to use”. Complexity and rate of adoption has a negative correlation.

• “Trialability is the degree to which an innovation may be experimented with on a limited basis”. Trialability and rate of adoption has a positive correlation.

• “Observability is the degree to which the results of an innovation are visible to others”.

Observability and rate of adoption has a positive correlation.

Figure 2 Ecosystem Aspect

Ecosystem Aspects

Aspects Innovation

Aspects

Co-opetition

Ecosystem Governance and Structure

Proximity

(25)

17

Figure 3 describes the attributes that defines the innovation aspects.

2.6 Knowledge co-creation Aspects

Knowledge is a primary economic resource and a crucial feature of post-industrial societies (Fong, 2003). Drucker (1994, p. 42) in his book “Post-capitalist society” goes beyond and states that knowledge is “the only meaningful resource” among all the other types of resources including land, labor, and machinery. As an intangible asset, knowledge is more difficult to harness and locate, but easier to loose (Fong, 2003). Organizations, more than ever, need to continuously use, create, and share knowledge in order to remain competitive. The creation of new knowledge is not limited to the borders of an organization. A team can bring the collective knowledge of its members together in order to serve the customers, which, in turn, creates competitive advantage for the team. Collaborative knowledge creation (CKC) refers to the condition when several partners (two or more) work together in order to create new knowledge and information. This new knowledge can be utilize by each of the partners for future developments and innovation (Hong et al., 2010). One of the common team arrangements is when actors come together to develop a new product. New product development as a knowledge intensive activity, involves cross-functional linkages in which a team of participants with different viewpoints join to co-create. Knowledge co-creation is a fundamental element

Figure 3 Innovation Aspect

Ecosystem Aspects

Aspects Innovation

Aspects

Relative Advantage Compatibility

Complexity Trilalability Observability

(26)

18

of product development. Actors need to constantly integrate new information into their understanding in order to tackle the technical challenges as well as other ever-changing requirements (Fong, 2003). Further, we dig into the nature of knowledge co-creation as a fundamental outcome of an innovation ecosystem.

As we mentioned before, the relationship between different actors in an innovation ecosystem is a balance between harmony and rivalry, or in other words, co-opetition. According to Wilhelm and Kohlbacher (2011), the close connection between co-operation and competition plays a crucial role in the process of knowledge co-creation. Before describing this impact, we might need to define two main concepts regarding the knowledge creation. First, knowledge exploitation which refers to improving the intellectual capital of an organization using the existing knowledge. Second, knowledge exploration which indicates the enhancement of the intellectual capital of an organization through creating unique new private knowledge. For organizations to build and increase their competitive advantage, both knowledge exploitation and exploration are indispensable. As Wilhelm and Kohlbacher (2011) concluded, although co-operation creates an environment of trust for knowledge sharing and exploitation of the knowledge among actors, strong co-operative ties may lead to structural inertia. In other words, the network becomes so inwardly focused that it ignores the external changes. The actors become increasingly alike and the network loses its ability to explore new knowledge.

Therefore, creative tension which roots in competition, is a key to permitting radical knowledge creation. Competition can decentralize the knowledge, which, in turn, leads to a better position for the actors and helps firms innovate. Wilhelm and Kohlbacher (2011) also links the structure, co-ordination, knowledge, strategy and innovation, which is shown in table 1.

Table 1 A typology of innovation strategies (adapted from Wilhelm and Kohlbacher, 2011)

In an inter-organizational context with a radical, multi-technology innovation goal, actors must maintain a balance between co-operation and competition in order to reach an effective knowledge co-creation. However, too much rivalry is still the main cause for alliance failure and the network requires a high level of governance in order to keep the balance between intense competition and harmonious inertia (Wilhelm and Kohlbacher, 2011).

Form of governance Mode of co-ordination Strategy Type of innovation

Hierarchy Co-operation Knowledge sharing Mainly incremental

Market Competition Knowledge Creation Radical (single-

technology innovation)

Network Co-opetition Knowledge co-creation Radical (Multi-

technology innovation)

(27)

19

Figure 4 shows the main wanted outcomes of Knowledge co-creation aspects in an innovation ecosystem.

To conclude, in this section, the trajectory of innovation ecosystem as a concept has been provided and the relevant definitions has been discussed. Further, different types of innovation ecosystems and the typology of potential actors involved in innovation ecosystems have been presented. Then, we dived into the aspects of innovation ecosystems in order to find the possible drivers and hinders for engagement of the actors. The first group of aspects investigates the quality of network and ecosystem. First, co-opetition as the continually changing dynamic balance between competition and cooperation of independent actors has been mentioned by many studies. Co-opetition by nature can potentially act as both driver and hinder depending on the ability of the ecosystem to keep the balance between harmony and rivalry. Second, the way an innovation ecosystem manages the structure and governs the whole network might motivate or demotivate the actors involved. third, different types of proximity from cognitive, organizational, social, institutional to geographical proximity can potentially act as a driver or hinder for engagement of the actors. The second group of aspects refers to the quality of innovation. The way the actors perceive an innovation might also affect their decision towards joining the ecosystem for development such an innovation. Finally, knowledge co-creation is the outcome of the ecosystem. Actors join an innovation ecosystem to jointly learn new knowledge. These Aspects have later been used for analyzing the empirical data of this study.

Figure 4 Knowledge co-creation

Ecosystem Aspects

Aspects Innovation

Aspects

Competitive advantage Product Development

(28)

20

Overall, our theoretical framework of reference provides a comprehensive overview of the potential aspects that might encourage actors to engage or not in collaboration with an innovation ecosystem for value-cocreation. The theoretical framework of reference was developed by combining both the relevant aspects related to innovation ecosystems and the attributes of innovation diffusion. Moreover, as can be seen in the subsequent chapters of this study, an empirical case was included to explore the extent to which these attributes influence engagement in practice, and to see whether additional attributes need to be incorporate in order to portray in a better fashion what drives and hinder actors to engage in an innovation ecosystem. The summary of the prepositions driving the theoretical framework of this study is following presented in figures 5.

Ecosystem Aspects

Aspects Innovation

Aspects

Competitive advantage Product Development Co-opetition

Ecosystem Governance and Structure

Proximity

Figure 5 Theoretical Frame of reference

Relative Advantage Compatibility

Complexity Trilalability Observability

(29)

21

3 METHOD

In this section, it is described the general plan used to answer the research question. This study used Sanders et al (2009) research onion stages to describe the development of the research method including: approach, research strategy, time horizon, and data collection. Figure 6 illustrates our research design.

Figure 6 Research onion, adapted from (Saunders et al., 2009)

3.1 An exploratory qualitative study

The intention of this study was to look for new insights and to explore in depth the interviewees’ knowledge and experiences to inform us about what have drove them to engage in an innovation ecosystem for knowledge co-creation, as well as, to generate hypothesis for further research. Therefore, to search for this information, we used qualitative research which allowed us to concentrate on individual voices and depth of experiences rather than over generalization which is more a characteristic of quantitative research (Sullivan and Sargeant, 2011).

3.2 Abductive approach

The research approach used for this study was abductive. An abductive approach was effective for this study since our objective was to discover new insights and variables for which actors

Approach: Abductive

Research strategy: Case study

Time horizon: Cross- sectional

Choice: Mono- method

Data collection:

Semi- structured

interview

Data analysis:

Thematic analysis

(30)

22

engaged in innovation ecosystems. Here, one major difference, as compared with both inductive and deductive approaches, is the role of the theoretical frame of reference (Dubois and Gadde, 2002). As this study relied on abduction, the original theoretical frame of reference was continuously modified, partially as a result of surprising empirical findings, but also of theoretical insights gained during the process of the study.

3.3 Research strategy

The research strategy used to answer the research question was case study. This involved an empirical investigation of an innovation ecosystem within its real life context using multiple sources of evidence (Saunders et al., 2009). A case study was chosen since we as researchers, endeavored to gain a rich understanding of the context of an innovation ecosystem and the process enacted by the actors for engagement, as well as to answer a ‘what?’ research question.

A single case study was selected (The OSMaaS project), as it provided us the opportunity to analyze an innovation ecosystem with specific characteristics that few has considered before.

The following characteristics describe why we decided to use the OSMaaS project as case for this study: (I) the project was set for the knowledge co-creation, (II) it is composed by interconnected and interdependent actors from different sectors which includes government, university and private organizations, (III) the nature of the innovation is rather improvisational, and (IV) although the project is managed by a leading organization, the governance responsibility is distributed among the actors. These characteristics of the OSMaaS project not only allowed us to study an innovation ecosystem, but also provided us a real scenario to explore why actors engaged in it.

3.4 Time horizon

To answer the research question of this study, we undertook a cross-sectional time horizon.

The study of a particular phenomenon (the OSMaaS project) was analyzed at a particular time (engagement). This fitted well with our perspective when reflecting what drove actors to engage in OSMaaS.

3.5 Mono-method as choice

To develop this study, we used a single data collection technique and a corresponding data analysis method. Semi-structured interview was used as a technique to collect empirical data and thematic analysis as the method for analyzing it.

(31)

23

3.6 Data collection

Semi-structured interview technique was used for collecting the data. The semi-structured interview process used a conversational and informal style in order to incite the full disclosure of information, as well as, to help the participants to feel free of pressure and open up to the exploration of unexpected or co-constructed themes (Sparkes and Smith, 2013). This interview method also offered us the required flexibility to explore in depth the individual experiences lived during the process of evaluating whether to engage or not to the OSMaaS project. The 37 questions used in the interview (see appendix 1) were built based on patterns of prepositions identified in the literature review. The structured part of the interview was used only with the objective of keeping the interviewees within the theoretical boundaries (Boyatzis, 1998) of innovation ecosystems.

The interviews were taken over a short period of time, from the 5^th of march to the 25^th of march of 2020. All the interviewees were representatives of the actors (organizations) engaged to the OSMaaS project. We targeted the individuals that were actively contributing in the workshops and meetings of the project. The initial list included twelve individuals; they received an invitation to an interview via email. However, a total of seven were able to participate in the interview. The five aimed individuals that did not join the interview, expressed their intention on contributing, nevertheless their agendas did not pair with the schedule proposed for the interviews. Some interviews were carried out via digital communication and others face to face; ranging between 35 to 90 minutes. All the interviews were recorded using the voice recorder application of Apple and later transcribed in full. Table 2 shows a description of the interviewees.

Table 2 Interviewees profile

Interviewee Profile

Interviewee Country Organization Position Int. Date

Interviewee 1 Sweden Polestar Business designer 18/3/2020

Interviewee 2 Sweden Halmstad University Product owner 18/3/2020

Interviewee 3 Sweden CEVT Innovation strategy director 19/3/2020

Interviewee 4 Sweden Halmstad University Professor of information technology 5/3/2020 Interviewee 5 Sweden Varberg municipality Infrastructure mobility/statistics and

analysis leader 25/3/2020

Interviewee 6 Sweden Wireless-Car Vice-President of strategy product

management and partnership 17/3/2020 Interviewee 7 Sweden Halmstad University Professor of computer science 11/3/2020

(32)

24

3.7 Data analysis

This section presents a comprehensive narrative of how we developed the process of analyzing the data. Before describing this process any further, in table 3, we define the terms used when carrying out the analysis. The definitions were taken from Guest et al. (2012) and (Boyatzis, 1998) since these follow the consensus of basic terms for qualitative analysis.

The method of analysis selected for the development of this study was a hybrid approach to qualitative methods of thematic analysis (Fereday and Muir-Cochrane, 2006). This method enabled the interaction between the interviewees or the empirical world and us; the concepts regarding innovation ecosystems or theory and us; and the OSMaaS project or the case study and us (Gama, 2019). The hybrid approach incorporated both deductive and inductive approaches (Fereday and Muir-Cochrane, 2006); in our case, the deductive approach was driven by prior-research and the inductive approach was driven by raw data generated through the semi-structured interviews.

The deductive version of thematic analysis took its departure point from theoretical prepositions generated from a review of the literature in innovation ecosystems and applied these to the data collection and its respective analysis (Pearse, 2019). The following deductive or prior-research-driven codes were generated form the literature review: collaboration, competition, knowledge-cocreation, ecosystem governance, ecosystem structure, proximity, relative advantage, compatibility, complexity, trialability, and observability. Using these prior- research codes, a code template was then developed. The code template served as a mean for

Table 3 Definition of the terms used when analyzing the data

Term Definition Source

Data “The textual representation of a conversation,

observation, or interaction”. (Guest et al., 2012, p. 50) Coding “the process by which a qualitative analysis links specific

codes to specific data segments”. (Guest et al., 2012, p.50) Code “A textual description of the sematic boundaries of a

theme”. (Guest et al., 2012, p. 50)

Prior-research code A code derived from the theoretical frame of reference (Boyatzis, 1998) Data-driven code A code derived from the raw data collected for the

interviewees. (Boyatzis, 1998)

Code template “Structured compendium of codes that includes a

description of how the codes relate to each other”. (Guest et al., 2012, p. 50) Sub-theme “Pre-unit of meaning that is observed in the data”. (Guest et al., 2012, p. 50) Theme “A unit of meaning that is observed in the data”. (Guest et al., 2012, p. 50)

(33)

25

organizing segments of raw or inductive data for its subsequent interpretation (Boyatzis, 1998;

Fereday and Muir-Cochrane, 2006; Pearse, 2019).

The inductive side of thematic analysis was generated from qualitative enquiry in the form of segments of data that were subsequently translated into data-driven codes (Boyatzis, 1998). In this study, the inductive coding process involved identifying important moments within the raw data and encoding them prior to the process of interpretation (Fereday and Muir-Cochrane, 2006). The good identification of an important moment in the data is defined as a ‘good code’

which embodies a segment of data that captures the qualitative richness of the phenomenon in study (Boyatzis, 1998). Information was encoded in order to organize the collected data; it facilitated the identification and development of themes.

The method encompassed six stages for data coding adapted by Federay and Muir-Cochrane (2006) and Pearse (2019) which contained: (I) literature review; (II) developing the codebook which involved labelling, defining and describing the prior-research code; (III) segmenting data and data driven code; (IV) applying the template of codes and additional coding; (V) connecting the codes and identifying subthemes, and (VI) corroborating and legitimating themes. In figure 7 an illustration of the stages used for data coding can be observed.

(I) Literature review

(II)Developing the codebook (prior- research-driven code)

(III)Segmenting data (data-driven code)

(IV)Applying the template of codes and additional coding

(V)Connecting the codes and identifying subthemes

(VI)Corroborating and legitimating themes

Figure 7 Stages undertaken to code the data, a diagrammatic representation - adapted from Federay and Muir-Cochrane, 2006

To Engage or Not to Engage: The Case of an Emerging Innovation Ecosystem in Sweden

Master Thesis

UNIVERSITY

Master's Programme in Industrial Management and Innovation, 120 Credits

To Engage or Not to Engage: The Case of an Emerging Innovation Ecosystem in Sweden

Master's Thesis, 30 Credits

Halmstad 2020-06-22

Alireza Esmaeilzadeh, Harvey Blanco

ABSTRACT

ACKNOWLEDGEMENT

TABLE OF CONTENTS

LIST OF FIGURES

LIST OF TABLES

1 INTRODUCTION

1.1 Ecosystem Aspects

1.2 Innovation Aspects

1.3 Knowledge co-creation Aspects

1.4 Thesis Layout

2 THEORETICAL FRAME OF REFERENCE

2.1 Innovation Ecosystem

2.2 Types of Innovation Ecosystem

2.3 Typology of Actors

2.4 Ecosystem Aspects

2.5 Innovation Aspects

2.6 Knowledge co-creation Aspects

3 METHOD

3.1 An exploratory qualitative study

3.2 Abductive approach

3.3 Research strategy

3.4 Time horizon

3.5 Mono-method as choice

3.6 Data collection

3.7 Data analysis