Review of the results of the study and answering the research questions

In order to address the first sub‐question, what the structural properties of ecosystem architecture are, the dissertation proposes a distinction between exogenous and endogenous ecosystem architecture design properties, based on the multiple‐lens examination of three different ecosystem architecture conceptions and the conceptual and empirical research in the complementary publications. The study reviewed three different, emerging streams of ecosystem architecture literature, namely boundary, design and orchestration, and conducted empirical, qualitative and quantitative research. Distinct structural properties of ecosystem architecture were identified and then further categorized into endogenous structural properties related more to a specific position in architecture and exogenous structural properties referring to the “building blocks” of the architecture. Exogenous structural properties, arising from the inherent characteristics of ecosystem architecture, are exogenous from the hub firm point of view and cannot easily be influenced by firms. They are, to a certain extent, taken for granted. To influence them requires more effort than influencing the endogenous structural properties, as they are more related to the architecture of the ecosystem than to the position which a company occupies in the given architecture. The endogenous structural properties are endogenous from the firm perspective and therefore, firms have the ability to influence them.

In order to address the second sub‐question, how ecosystem architectures are designed and managed, the study proposes a process view. These processes are divided into two: emergent processes and engineered processes. Emergent processes, such as advocating and renegotiating the architecture, are more closely linked to the exogenous structural properties of architecture and to the design of the architecture. In contrast, engineered processes, such as enhancing knowledge mobility, managing innovation appropriability, managing network stability, managing innovation coherence and managing innovation leverage are more related to the endogenous structural properties of ecosystem architecture and to the management of the architecture. Given that the endogenous structural properties are easier to influence than the exogenous structural properties, they can be subjected to a wider range of processes for designing and managing the ecosystem architecture. These processes include the emergent processes of advocating and renegotiating the architecture and the engineered processes. Out of these processes, managing network stability and managing innovation appropriability are suggested to be the most important and influential ones in terms of designing and managing the architecture, and the hub firm’s position and value capture potential in such architecture.

50

Finally, and in order to address the main research question of how the design of ecosystem architecture affects value creation and appropriation within a particular ecosystem, the dissertation proposes the framework depicted in Figure 4. The properties in Figure 4 are not presented in any specific order.

Figure 4: Exogenous and endogenous ecosystem architecture design properties

Based especially on the boundary and design conceptions of ecosystem architecture and the findings of the individual publications, the following exogenous structural properties of ecosystem architecture were identified: complementarity, asset mobility, and redundancy. In ecosystems, companies have the potential to, for example, pool complementary resources. In the most basic form this can be a quest to integrate and bundle partners’ products or even further to support the creation of complementary products. Complementarity can manifest itself in the forms of sharing information about the customers, marketing resources or technical support. As illustrated in publication II, the ecosystem architecture can be designed with different levels of complementarity, especially when the architecture is yet to emerge. Thus, ecosystem hubs can advocate or renegotiate a certain level of complementarity and influence the level of interdependence between the network members. Renegotiation can be based on rotating the hub position as illustrated in publication III, leading to different companies occupying complementary positions at different times. Further, complementarity needs to be managed with the process of managing innovation coherence in order to collectively generate value.

The more plentiful and replaceable the complementary assets are with each other and the less there are identical relevant resources and capabilities the more value the ecosystem can potentially generate. Asset mobility and redundancy can be influenced with the same emergent processes as complementarity to design the architecture. Publications I and II discussed the processes of advocating the architecture in more detail in terms of how to establish a structural template of division of labour when the architecture is yet to emerge. Processes can also include managing innovation leverage (i.e., by standardizing the technological interface between the hub and the complementor). Regarding exogenous ecosystem design properties, in general, it can be concluded that the more plentiful and replaceable the complementary assets are with each other and the less

51

there are identical relevant resources and capabilities, the more value the ecosystem can potentially generate.

In addition to the exogenous structural properties, the following endogenous structural properties of ecosystem architecture were identified: permeability of boundaries, stratification, bargaining power and bottleneck position. The endogenous structural properties are endogenous from a particular firm’s perspective and therefore, the firm has the ability to influence them. The proposed framework also posits that hub firms in particular are in a position to impact these properties within the ecosystem. Whereas exogenous ecosystem design properties influence the value creation potential of the ecosystem, the endogenous ecosystem design properties influence the value appropriation potential of a particular actor in the ecosystem more. Given that the endogenous structural properties are easier to influence than the exogenous structural properties, they can be subjected to a wider range of processes of designing and managing the ecosystem architecture (i.e., orchestration processes (Dhanaraj & Parkhe, 2006; Nambidan & Sawhney, 2011)). In general, it can be concluded that the endogenous ecosystem design properties of boundary permeability, stratification, bargaining power, and bottleneck position increase the value appropriation potential of a particular actor in the ecosystem.

In addition to the architecture design properties affecting value creation and appropriation potential, the study argues that there is a relationship between the exogenous and endogenous design properties which works in both directions. First, the model suggests that at a particular point of time, exogenous ecosystem design properties formulate the value creation conditions under which endogenous properties can be used to appropriate value. For example, an actor in a bottleneck position and with bargaining power can capitalize on high asset mobility and complementarity to effectively appropriate value from the ecosystem. On the other hand, high levels of asset redundancy may limit the possibilities to achieve a position with strong bargaining power, and also make it difficult for any actor to situate itself in a bottleneck position where it could act as a broker between specialized assets. In sum, the proposed model suggests that exogenous ecosystem design properties pose opportunities and limitations on the use of endogenous properties. The more fixed (i.e., built in a long‐term process) these properties are, the more strictly the opportunities and limitations are defined.

While exogenous design properties can be seen to define the opportunities and limitations for the use of endogenous properties at a particular point of time, the study argues that actors can affect exogenous properties over time. Most importantly, the model suggests that actors strong in endogenous properties have the possibility to modify and reconfigure the exogenous design properties of asset mobility, complementarity, and redundancy in order to advocate or renegotiate the architecture. For example, bargaining power or stratification (i.e., based on different decision making rights) can be used to increase asset complementarity or asset mobility or to lower asset redundancy. Recent research supports this type argumentation: Santos and Eisenhardt (2009) show how small entrepreneurial ventures can achieve a comfortable position in the industry architecture by influencing the structure of their sector in ways that would eventually fit their own capabilities (see also Gawer & Cusumano, 2002; Morris & Ferguson, 1993). By focusing on the structural properties, the study complements the process view of Ozcan and Eisenhardt (2009) on how companies can advocate a desired architecture. Further, whereas the boundary conception focuses only on the modification of the firm’s boundaries in order to increase the share of the value appropriated, the possibility to leverage the endogenous ecosystem design properties as an additional strategy and strategic action to renegotiate the established architecture and division of surplus is highlighted. In sum, endogenous ecosystem design properties can be leveraged to modify and reconfigure the exogenous properties (i.e., in order to advocate a specific architecture or