How family firms use governance mechanisms to mitigate the risks of ecosystems: a case study from healthcare

This study aims to increase our understanding of how family firms, acting as ecosystem orchestrators, mitigate perceived relational and performance risk in ecosystems via the use of governance mechanisms. We use the ecosystem governance literature to understand whether family firms’ characteristics lead to a unique governance approach in an ecosystem setting. Our findings and theoretical implications are threefold; (1) formal and informal governance mechanisms act as complements to mitigate relational and performance risks; (2) the choice of governance mechanisms differs for relational and performance risk; (3) changes in perceived risk over time result in shifts in the relative dominance of formal/informal governance mechanisms. This study analyzes how family firms, acting as ecosystem orchestrators, mitigate perceived relational and performance risk via the use of governance mechanisms. Using data from a Belgian family firm that initiated an ecosystem in healthcare, the study shows that in a family firm-led ecosystem, formal and informal governance mechanisms act as complements to deal with relational and performance risk. We show that to deal with relational/performance risk, different combinations of governance mechanisms should be used. When ecosystem members perceive changes over time in perceived risk, the ecosystem orchestrator should change the balance between formal/informal governance mechanisms. Our study suggests that family firm orchestrators benefit from stimulating the active participation of ecosystem members. Family firm orchestrators might want to consider changing needs of a growing ecosystem, as, in larger ecosystems, frequent communication between the orchestrator and members can be more difficult. We argue that family firm orchestrators should remain mindful of the balance between formal and informal governance mechanisms.


Introduction
Family firms offer stimulating innovation environments based on unique characteristics, such as long-term investment horizons (Lambrechts et al., 2017). Due to insufficient diversity of resources to support internal innovation processes, family firms engage in open innovation (OI) initiatives, e.g., ecosystems, to complement internal knowledge and enhance innovation performance and competitive advantage (Gjergij et al., 2019). Family firms are risk-averse due to fear of losing control over resources as a result of OI partners' potential opportunistic behavior (relational risk) and/or unstable, unpredictable environmental conditions (performance risk) (Das & Teng, 2001;Xavier Molina-Morales et al., 2011). Research suggests that family firms minimize risks by taking the lead in ecosystem creation/governance. Thus, family firms remain in control over resources and influence OI processes (Lambrechts et al., 2017).
Governing ecosystems is challenging for family firms because they prefer collaboration with few previous collaboration partners to lower risks and stay in control (Gjergij et al., 2019). Family firms are not always experienced in managing collaborations with many diverse organizations within ecosystems. Research suggests that family firms use formal (control-based) and informal (trust-based) governance mechanisms to manage the internal organization and govern collaborations. These mechanisms are purported to allow them to manage the risks associated with internal/external innovation processes (Boucken et al., 2020;Lambrechts et al., 2017) and provide them with the opportunity to influence strategic decision-making of the ecosystem and achieve goal congruence and shared vision (Lambrechts et al., 2017;Pellegrini & Lazzarotti, 2019). Informal governance mechanisms are important in this respect (Bouncken et al., 2020).
Research in the family firm literature has focused on gaining an understanding of why governance mechanisms are important, how they govern their organizations internally and how they govern small, short-term OI structures. A similar pattern holds for family firms' risk approach; research has described the relationship between risk and (informal) governance, but from an internal organizational perspective (Casprini et al., 2017) and has provided perspectives on the openness of family firms (Bouncken et al., 2020;Lambrechts et al., 2017). Research has not focused on how family firms choose to govern ecosystems and how they mitigate perceived risks, despite recent calls (Chrisman et al., 2018;De Massis & Foss, 2018). The ecosystem literature provides guidelines for understanding the types of governance mechanisms that exist and how they relate to risk. This literature has shown that the use of formal (control-based) and informal (trust-based) governance mechanisms mitigates relational/performance risks. It has shown that different combinations of relational/ performance risk results in varying usages of formal/ informal governance mechanisms (De Man & Roijakkers, 2009). The choice of governance mechanisms can change over time, depending on ecosystems' developmental stage (Ritala et al., 2013).
Despite the benefits that ecosystems can offer for family firms (Lambrechts et al., 2017), we have a limited understanding of how they use governance mechanisms to manage ecosystems. An understanding of how family firms govern ecosystems and mitigate perceived risks could support them in becoming more innovative, coping with long-term future market changes, and help them contribute to societal wealth and employment (Casado, 2008). Given their characteristics, e.g., long-term horizons and fear of losing control (Casprini et al., 2017;Lazzarotti et al., 2017), we expect that they govern ecosystems differently than traditional firms. We pursue the research question: 1.1 How do family firms use formal and informal governance mechanisms to mitigate relational and performance risk when governing ecosystems?
We aim to increase understanding of how family firms, as ecosystem orchestrators, mitigate perceived relational/performance risk via governance mechanisms. We use the ecosystem governance literature to understand whether family firms' characteristics lead to a unique governance approach within ecosystems. We make three empirical contributions to the literature: (1) formal/informal governance mechanisms act as complements to mitigate relational and performance risks; (2) the choice of governance mechanisms differs for relational/performance risk; (3) changes in perceived risk over time result in shifts in the relative dominance of formal/informal mechanisms.

Why family firms engage in OI
OI refers to companies opening up, by using internal and external sources, paths to market to advance shared innovation processes (Gassmann & Enkel, 2004). OI offers structures such as alliances, platforms, and ecosystems. Scholars argue that family firms have insufficient diversity of resources for internal innovation processes and that they can use OI to complement their internal knowledge base (Pellegrini & Lazzarotti, 2019). Researchers suggest that OI enhances innovation performance and creates a competitive advantage for family firms (Gjergji et al., 2019). Two frameworks have influenced the literature on why they engage in OI: the resource-based view and behavioral theory. The family firm literature influenced by the resourcebased view builds upon family firms' characteristics: (1) long-term investment horizons and (2) importance of reputation. Family firms take longer investment horizons compared to traditional firms because they want to safeguard wealth for future generations (Lambrechts et al., 2017). Their goals do not focus on short-term financial gains, but take a long-term perspective to create (financial) growth (Chrisman, 2019). They are often concerned with reputation and have stronger incentives to protect it (Williams et al., 2018). They have more long-term relationships with partners compared to traditional firms (Gjergi et al., 2019). The resource-based view suggests that, because of these characteristics, family firms are better able than traditional firms to effectively engage in OI (De Massis et al., 2015). Scholars building upon the resource-based view argue that OI supports family firms to continue innovation processes and that family firms' internal characteristics enhance their ability to capture value (Gjergij et al., 2019).
Scholars drawing on behavioral theory refer to other characteristics: (1) fear of losing control and (2) concentrated ownership. Family firms fear to lose control and are more risk averse compared to traditional firms (Pellegrini & Lazzarotti, 2019). This is because ownership is concentrated within one family and coupled with this family's involvement in management (Chrisman et al., 2018). To protect the family's wealth, family firms take less risk (Naldi et al., 2007). This theory assumes that family firms do not open up because they fear to lose control, are more conservative, and fear to harm wealth (Pellegrini & Lazzarotti, 2019). Scholars argue that family firms are more risk averse compared to traditional firms and are more attached to traditions (Gjergij et al., 2019). They prefer collaboration with close partners and have low search breadth (searching for partners within a limited, known group) (Alberti et al., 2014).
Scholars increasingly combine the resource-based view and behavioral theory to provide an explanation for the openness of family firms (Casprini et al., 2017). They suggest family firms adopt a low search breadth and high search depth to reduce the relational risk associated with OI. By intensively working with a few partners, they reduce relational risk (Casprini et al., 2017). Their risk-averse attitude may not have a negative influence on openness, but results in a unique OI approach.

From why to how -family firm-led ecosystem governance
Governance is described as a combination of selfinterest, coercion, and trust to manage repeated social transactions and potentially opportunistic behavior of employees (Steier, 2001). Family firm internal governance combines formal/informal governance mechanisms to influence the behavior of employees; as explained by Daily et al. (2003), it is geared at "the determination of the broad uses to which organizational resources will be deployed and the resolution of conflicts among the myriad participants in organizations." (p. 371). It aims at setting the strategic direction and creating a dynamic equilibrium between the firm's and stakeholders' interests (Chrisman et al., 2018). Formal and informal governance mechanisms are used by family firms to manage the equilibrium between the firm's and employees' interests (Chrisman et al., 2018). Formal governance mechanisms aim at demotivating and restricting potentially opportunistic behavior of employees by means of controlling behavior (Ireland et al., 2002). Formal governance mechanisms monitor and incentivize employees (Madison et al., 2016) to behave in line with the firm's goals. Examples include a board of directors, monitoring of activities, incentive compensation plans, and contractual obligations (Madison et al., 2016). Informal governance is based on trust between stakeholders that no one will behave opportunistically, thereby showing respect (Xavier Molina-Morales et al., 2011). Informal governance mechanisms aim at encouraging cooperation and involvement to facilitate partner alignment (Madison et al., 2016). Examples are a board of advisors, relational control, organizational culture, training, participation, alignment of interests, common values, and shared attitudes (Chrisman, 2019).
Scholars differ in their explanation as to whether family firms internally prefer formal or informal governance mechanisms. Some argue that they prefer informal governance mechanisms as these enable them to focus on their relational network and reduce search, screening, evaluation, and enforcement costs (Cabrera-Suarez et al., 2015). Steier (2001) argues that younger family firms prefer the use of informal governance mechanisms, whereas older firms prefer formal governance mechanisms because complex organizational structures make it difficult to create alignments between stakeholders. Chrisman et al. (2018) argue that informal/formal governance mechanisms reinforce or counteract each other, though they conclude that informal governance mechanisms are more important for family firms than for others. Madison et al. (2016) add that formal and informal governance mechanisms coexist, as informal mechanisms create collective cultures and formal mechanisms monitor the firm.
Recent research explores how family firms take an external governance approach for different OI structures. This literature does not provide clear definitions of governance and focuses on explaining what type of mechanisms are used (Bouncken et al., 2020). To capture and create value from OI, family firms need to use governance mechanisms. Bouncken et al. (2020) show that they rely on informal governance mechanisms, such as trust, when managing alliances. Formal governance mechanisms can be a limiting factor for family firms, as it limits the flexibility they need for innovation (Bouncken et al., 2020). Scholars have explored how family firms manage networks, arguing that they prefer to use informal governance mechanisms when collaborating with prior partners and formal governance mechanisms when collaborating with unknown partners (Carrasco-Hernandez and Jimenez-Jimenez, 2013). This shows that they prefer to collaborate with previous partners, thereby building upon repeated ties.
Ecosystems consist of various interdependent actors that interact and exploit and/or explore knowledge collaboratively with the aim of creating/capturing long-term value at the ecosystem level (Kapoor, 2018). Each member provides access to their specialized, complementary resources for developing innovations (Kapoor, 2018). A focal firm orchestrates the collaboration process and takes responsibility for governance and goal achievement (Kapoor, 2018). Ecosystem governance is defined as the combination of formal/informal governance mechanisms ( Table 4 in Appendix 1) which coordinate, align, and safeguard partners' contribution and participation and define structures, processes, and culture to do so (Ritala et al., 2013).
Formal and informal governance mechanisms are combined in governance models (Ritala et al., 2013). Research shows that different combinations of perceived relational and performance risk result in varying combinations of formal/informal governance mechanisms (De Man & Roijakkers, 2009; Table 5 in Appendix 1). The focal firm's perception of different levels of these risks results in choosing different combinations of formal and informal ecosystem governance mechanisms. Performance risk is about factors that influence ecosystem success, whereas relational risk is about the probability that members might behave opportunistically and the consequences thereof (Das & Teng, 2001). In terms of performance risk, it could be that a national government changes laws and regulations, and as a result, certain innovations might no longer be allowed to be sold. When performance risk is perceived as high, environmental factors have a large influence on the behavior of the ecosystem, whereas low-performance risk implies that environmental factors do not directly influence ecosystem behavior. Examples include market volatility and geopolitical factors. When the relational risk is perceived as high, members behave opportunistically, posing risks for others, whereas low relational risk results in members behaving in an altruistic manner (De Man & Roijakkers, 2009). Both types are distinctive: relational risk focuses on partners' behavior, whereas performance risk focuses on external factors, which might negatively influence the ecosystem. Even if partners are intrinsically motivated (low relational risk), performance risk may negatively influence the ecosystem's success (Das & Teng, 2001).
When performance/relational risk is low, firms use either formal or informal governance mechanisms because no need exists to exert control over the ecosystem and the use of either formal or informal governance mechanisms suffices. When performance and relational risk are both high, firms tend to use formal and informal governance mechanisms because informal mechanisms enhance trust between partners, whereas formal mechanisms act as a safeguard for the orchestrator's personal value maximization (De Man & Roijakkers, 2009). When performance risk is high and relational risk is low, focal firms exclusively use informal governance mechanisms because partners pose no threat to the value maximization of the orchestrator, allowing them to remain flexible in responding to external threats (De Man & Roijakkers, 2009). When performance risk is low and relational risk is high, firms rely on formal governance mechanisms. As a result, the orchestrator is focused on protecting his firm from potentially opportunistic partner behavior and on maximizing his own results (De Man & Roijakkers, 2009). Different ecosystem development phases (Ooms et al., 2020) require different governance approaches, leading to changes over time in governance mechanism combinations. In initiation, the focus is on attracting relevant stakeholders and developing social relationships, whereas, in growth, the focus shifts to realizing goals. In both phases, a combination of formal and informal governance mechanisms is used (Ooms et al., 2020), yet in initiation, informal governance is more dominant, whereas, in growth, formal and informal mechanisms act as complements (Olander et al., 2010). The ecosystem governance literature poses the following: (1) Ecosystem orchestrators use various combinations of informal/formal governance mechanisms to mitigate specific combinations of perceived relational/performance risk; (2) different ecosystem development phases require different governance approaches, leading to changes in mechanism combinations over time.

Method
An explanatory single case study design is employed to increase our understanding of how family firms orchestrate ecosystems by using governance mechanisms to mitigate perceived relational/performance risk. Explanatory studies provide information on questions not yet understood in-depth and help to create an understanding of how phenomena arise (Yin, 2018). Our study is deductive: We observe how family firms govern ecosystems by using the ecosystem governance literature to underpin our data analysis by stipulating data categories (Yin, 2018). Explanatory research allows us to use predefined data categories to understand the phenomena in our case (Yin, 2018).
We study an in-depth single case study of the PRoF (Patient Room of the Future) ecosystem: an international interdisciplinary collaboration among professionals aiming to create new visions for the healthcare sector by designing healthcare concepts (see Appendix 2). The PRoF ecosystem case is unique because (Yin, 2018) (1) it is one of few European healthcare ecosystems in existence for several years, aiming to simultaneously influence the sector and develop innovative concepts; (2) access to key ecosystem actors, research articles on the case, and previous experience with the ecosystem supported the research process; (3) the size of the case is manageable, such that we can identify several groups within the ecosystem and collect rich data from these groups (avoiding too much complexity) (Yin, 2018); (4) the ecosystem is initiated and led by a family firm.
Our study takes a retrospective view and involves multiple sources, including semi-structured interviews, websites, earlier research on the case (Lambrechts et al., 2017;Vanhaverbeke, 2017), and an educational case description (Vanhaverbeke & Verhoeve, 2016). Thus, the interviewees were asked to structurally recall events from the past on how they perceived relational/performance risk and what kind of governance mechanisms were used. This allows for a greater understanding of how/why events occur and how people experience them. By asking interviewees to share their stories, they were able to both describe events and explain their significance (Murphy et al., 2019).
We conducted seven interviews, each lasting 45 to 60 min during a six-month period in 2017 and an additional interview in 2020 with the family firm orchestrator, to check whether any changes have occurred since 2017. A deductive approach requires seven to eight interviews to reach data saturation and to identify research themes (Francis et al., 2009). Each interview was recorded and transcribed. Secondary data were used to get a first idea of the ecosystem structure, performed activities, and the family firm's background. Secondary data revealed that the ecosystem consisted of three groups (commercial, expert, and field expert) and interviewees were sampled from these groups. Table 1 provides an overview of the interviewees based on group membership. This table also includes details of the interviewees, such as gender, their role, the type of organization they work for, and the point in time that they entered the ecosystem. The commercial group consists of for-profit companies, the expert group consists of academics and practitioners knowledgeable of healthcare, and the field expert group consists of patients, caretakers, hospitals, and nursing homes. These groups were further structured into a large ((field) experts) and small group (industry) and a board of directors.
Groups were identified via an interview with the orchestrator and corroborated by secondary data. In choosing interview members, we ensured that representatives of different groups were fairly represented. Interviewees were identified via a chain-referral sampling approach; the orchestrator provided contact details of potential interviewees. The orchestrator was aware of our research goal and provided contact details of potential interviewees that varied in terms of type and length of membership to help gain an understanding of how different member types with different membership lengths perceive the use of governance mechanisms/risks. Secondary data were used to check the appropriateness of potential interviewees and to ensure that the interviewee group was diverse. In the interviews, we aimed to understand the role played by formal and informal governance mechanisms in the ecosystem, how they were used, and whether there were changes in the usage of governance mechanisms over time. We also traced risks associated with ecosystems and how the choice of governance mechanisms was related to risks. These questions fit our goals, given that our objective is to investigate the extent to which insights from the ecosystem governance literature can be applied to ecosystems facilitated by family firms. The interview guide used for interviewing the orchestrator differed from the interview guide used for interviewing ecosystem members as the orchestrator fulfills a different role. The interview guides were based on the four concepts under study.
Analyses were carried out by using a deductive coding strategy. We started with a set of manually predefined codes based on the literature that were developed to provide a guideline for organizing interview data and secondary data in related segments (Linneberg & Korsgaard, 2019). We defined four coding categories: formal governance mechanisms, informal governance mechanisms, relational risk, and performance risk. These categories were chosen because they demonstrate how family firms use informal and formal governance mechanisms to mitigate performance and relational risk. We operationalized governance mechanisms by asking questions to understand what kind of mechanisms were used and how the ecosystem members perceived their use. The mechanisms were operationalized by asking questions about examples of governance mechanisms based on the literature. Regarding the operationalization of risks, we did not ask directly how interviewees perceived relational/performance risks; we looked at the literature to provide us with an idea of the characteristics of both risk types and asked questions to learn how ecosystem members perceived relational and performance risk based on these characteristics. We found in the literature that the unpredictability of the market influences the perception of performance risk. If the market is perceived as highly unpredictable, the perceived performance risk is high (De Man & Roijakkers, 2009). Based on the input of all interviewees on elements pertaining to both relational and performance risk, we categorized risks as low or high. The categorization was based on definitions from the ecosystem governance literature and was arrived at more indirectly compared to the categorization of governance mechanisms. We developed a codebook to organize the data for further interpretation ( Table 6 in Appendix 1). We developed a template (see Table 2 below; Crabtree & Miller, 1999) that was used to manually code the interview and secondary data to find out how the family firm orchestrates the ecosystem. One author developed the codebook, and others checked, discussed, and refined it, thus increasing inter-rater reliability (Linneberg & Korsgaard, 2019).
We used data reduction techniques in line with Miles and Huberman (1994). Initial analysis was followed by group discussions in our interdisciplinary research team. To validate interpretations of data, we discussed preliminary findings with the orchestrator. The combination of interview and secondary data (data triangulation) allowed for result comparisons and helped increase internal validity, as secondary data were used to crosscheck interview findings. The semi-structured interview guide supported the interviewees to recall experiences in a structured yet flexible manner, thereby increasing the validity of the findings (Murphy et al., 2019).

Governance during initiation
The orchestrator had experience with alliances and the family firm integrated OI into its business objectives, which included becoming a leading manufacturing company. Prior to starting the ecosystem, the firm had experience with OI for diversification and customer co-creation (Lambrechts et al., 2017). The orchestrator initiated the ecosystem, given that he felt the need to collaborate with a wider variety of partners. The orchestrator knew how important trust was between partners and started to select members based on shared norms and values, a common passion for improving healthcare, and commitment. He thought that these criteria would stimulate motivated members to join and would inspire them to contribute voluntarily to the goals. The orchestrator kept control over the organizations entering the ecosystem, staying in control of his company's knowledge and resources. The orchestrator also employed other governance mechanisms to decrease the chance that partners would behave opportunistically (decrease the perceived relational risk). From initiation onwards, the orchestrator employed informal governance mechanisms such as clear and open communication structures, regular (informal) meetings, and collaboration. One way in which collaboration and communication were stimulated was through brainwave sessions. These sessions aimed at creating a safe environment to openly share ideas and provide feedback. The orchestrator hoped that the presence of a safe environment would contribute to the partners' willingness and motivation to contribute. Table 7 in Appendix 1 provides an overview of quotes from both the initiation and development phases. Table 3 below provides information on the expected contributions and perceived risks of each partner. The orchestrator also employed formal governance mechanisms to protect himself and his partners from potential opportunistic behavior, aiming at decreasing relational risk via formal mechanisms. He created an organizational structure consisting of a large and small group, a core group, and a board of directors (council of the wise). The core group, consisting of the orchestrator and several partners, was in charge of the daily operations. The council of the wise, consisting of the orchestrator and several CEOs of members, made fundamental decisions regarding the future and decided upon rules. Some rules pertain to participative decision-making regarding new partner selection; the orchestrator and the small group collectively decide whether a new commercial member would be allowed to join. The orchestrator decided that commercial members had to pay a membership fee based on their company size, and he designed rules on innovation standards and financial agreements.
Some non-commercial members distrusted commercial members, feeling afraid they would pursue their own interests at the expense of others (highperceived relational risk). The orchestrator noticed the issue and decided to facilitate a dialog (informal governance mechanism), and the caregivers and patient members realized that the commercial members were intrinsically motivated to contribute (decreased perceived relational risk). According to partners, the presence of shared norms/values contributed to mutual trust building and partners feared that discussions about formal mechanisms would undermine trust. Interviewee 2: "I feel that most ecosystem members were chosen by [the orchestrator] based on trust, and earlier personal experiences, because you do not share your ideas with people you do not know, competitors or people with Feedback on concepts, testbed for parts of the concept Low High 7 OI expertise Low High potential conflicting interests." The way in which the orchestrator managed relationships attenuated the feeling among some that some partners might pursue their own interests at the expense of others. The brainwave sessions were contributing to trust building because members felt that open discussions motivated them to participate and learn from each other. Interviewee 4: "There are a number of companies that innovate more quickly or more actively to stay in PRoF." Partners felt that deceiving each other was no option because the potential removal from the ecosystem could result in loss of reputation within and beyond the ecosystem. Given that members trusted each other, they felt that the orchestrator was managing the ecosystem effectively and did not see a need for more formal governance to reduce the chance that others would behave opportunistically (reduce perceived relational risk). Most ecosystem members did not even realize that several formal mechanisms were in place. After implementing both formal/informal mechanisms, members did not fear that partners would behave opportunistically and felt that most members were intrinsically motivated, resulting in low perceived relational risk (Das & Teng, 2001). Whereas relational risks were managed consciously and successfully managed, managing performance risks turned out to be more difficult. Both the orchestrator and members initially perceived high levels of performance risk; they felt that external factors jointly had a negative effect on the ecosystem's potential success (Das & Teng, 2001). They felt that the healthcare sector was unstable and turbulent: "The evolution of the healthcare sector is large and undetermined (interviewee 6)." Members also felt that healthcare was complex and presented them with uncertain future developments. Interviewee 7: "I think the healthcare sector is evolving so quickly that we have to constantly reinvent ourselves." These perceptions were the result of governmental changes in regulations and laws, changes in funding, as well as changing societal trends. Given the difficulty of influencing these factors, members initially perceived high-performance risk. The orchestrator's main aim to create the ecosystem was to stimulate changes that would enable alignment with the external environment. The orchestrator hoped that the total ecosystem would be able to influence external factors potentially harming the ecosystem's success, thereby lowering performance risk. The orchestrator believed that change could be stimulated contingent on a combination of perceived low relational risk and a number of additional informal mechanisms. He believed that perceived low relational risk was important; because affecting change in the external environment was only possible if the ecosystem was healthy (members trust each other, participate, and contribute). In a healthy ecosystem (perceived low relational risk), the focus is on aligning the ecosystem with external factors rather than on managing internal relationships. The orchestrator attempted to influence the perceived performance directly by awarding several prizes at special PRoF conferences, at which the orchestrator handed out prizes to industry and academic leaders that had made significant contributions to healthcare.
The orchestrator predominantly used informal governance mechanisms combined with some formal governance mechanisms to manage relational/ performance risk. When the ecosystem was small and the orchestrator knew most partners, partners trusted that others would contribute the resources they promised to and would not steal each other's knowledge. This is also how he managed his firm, where he was CEO, member of the board of advisors, and member of the board of directors, thus maintaining control. The way in which he managed the ecosystem shows similarities to the way in which he governed the family firm.

Changes over time
The relational risk was present both at the start of the ecosystem and over time. Over time, the ecosystem grew significantly, and members joined with previous collaboration experiences from outside the ecosystem. Early members felt that the risk of opportunistic behavior increased and feared that new members would not be as motivated as earlier members to contribute and participate. Growth made several original governance mechanisms difficult to implement. The expanding ecosystem also increased the risk that members were no longer complementary, increasing the fear that others would use/steal each other's knowledge. This was mostly the case for commercial partners, probably because they feared that competitors would steal their knowledge. The fear of opportunistic behavior was less salient among large group partners, probably because they did not share personal products/services/knowledge. Some partners' intrinsic motivation to contribute to shared goals decreased. It became more difficult for the orchestrator to manage the ecosystem, he started to communicate less frequently, and leadership became more centralized. "The ecosystem is like an orchestra that keeps growing and growing with [the orchestrator] as conductor, but at a certain moment the ecosystem needs something different than the informal trust-based relations to continue in the long run" (interviewee 2). Members that were used to the openness of the original smaller ecosystem did not appreciate this centralized governance style and their intrinsic motivation decreased. Interviewee 3: "I feel that the power is a bit too much centralized, and, as a result, I feel that the decisions are made by only a few members." One member emphasized that he started to use company-specific intellectual property rights to protect his technology. Growth resulted in an increased perceived relational risk among commercial members. It was difficult for them to leave, given their fear of reputation damage potentially triggered by discontinuing their membership, leading to lock-in effects.
Members working for healthcare organizations that joined the ecosystem later in time did now know about the earlier, more decentralized governance system. They had less to lose in terms of their organization's knowledge and perceived lower levels of relational risk. They felt the enthusiasm that the original members had felt at the start: "It is mostly the passion, the willingness and the way in which we approach things, that connects us and shows us the need to collaborate" (interviewee 5). Several members felt that the orchestrator was preparing a successor, in the form of a non-family employee: "I feel that [the orchestrator] wants to stay or wants the leading position to stay within his organization, and it makes a difference that [name of employee] is already part of the family firm and the ecosystem" (interviewee 5). One member felt hesitant about the possibility of finding a suitable successor because, as in so many family firms, the success of the ecosystem was strongly connected to the identity and vision of the orchestrator.
To manage growth and increased relational risk for commercial partners, the orchestrator started to shift his focus from predominantly using informal governance mechanisms to using more formal governance mechanisms. Several original informal and formal mechanisms continued to remain in use. The orchestrator believed that in the ever-growing ecosystem in which it was no longer possible to know every partner, an increase in formal governance mechanisms might help to create a foundation of trust and reignite commitment. He added more rules to manage the increasing member numbers. Rules were set up for existing partners to be able to maintain their membership (e.g., by contributing a number of hours to projects). The ecosystem structure was changed to make it succession-proof. The new structure aimed to render the ecosystem more independent from the family firm by making it more financially sustainable via participation in European subsidies. Up to this point, the ecosystem had been financed mostly by the family firm. The council of the wise changed into jewel, a structure similar to a board of directors. Jewel's aim was to guard the ecosystem's norms and values and consisted of field experts. Jewel was "… an organ that could be consulted when [the ecosystem] faces a new problem, for which no rules exist yet" (interviewee 7). The orchestrator realized that he also had to continue to work with informal mechanisms to reignite member commitment. He did this by providing occasions for face-to-face contact to give members the feeling that they knew each other and provide them with a sense of what to expect from each other, despite increasing memberships.
Perceived performance risk remained high because of changes in laws and regulations that dramatically altered the healthcare landscape. The orchestrator increased the emphasis on the ecosystem's aims to trigger changes in the market. The goal was for the ecosystem as a whole to attempt to shape those external factors that might jeopardize long-term success (influence performance risk). He believed that low relational risk was important to be able to trigger these types of changes and he wanted to ensure members' commitment by asking them to sign a letter of intent declaring their official intention to contribute to goals. Twelve members refused to sign the letter and left. The orchestrator and jewel started to search for complementary initiatives in Europe to collaboratively influence the sector. In the past, the focus had been on finding comparable ecosystems. Those initiatives failed because no comparable ecosystems were found. In this stage, the orchestrator attempted to use mostly formal governance mechanisms to attenuate perceived performance risk. Despite attempts to influence the course of the industry, the ecosystem did not succeed in creating alignment with the external environment and performance risk remained high.
As the ecosystem developed, the orchestrator increasingly made use of formal mechanisms. As it grew larger, it became more difficult for the orchestrator to know all partners and to make sure that the interests of all partners are aligned (breadth increased at the cost of depth). By using formal mechanisms, the orchestrator was better able to mitigate potential relational risk for his own company and for most of the members. Informal/formal governance mechanisms acted as complements, and the family firm reduced the potential relational risk by smartly managing the increasing breadth of the ecosystem. Whereas in the firm, the orchestrator was aiming at a family member to succeed him; in the ecosystem, he was exploring options for funding besides exploring whether one of his employees could take over his role.

Concluding discussion
First, our case revealed that the use of governance mechanisms was related to the nature of perceived risks and formal and informal governance mechanisms were acting as complements. At initiation, perceived relational risk was low and performance risk was high. As the ecosystem grew, perceived relational risk increased while performance risk stayed high. Our study suggests that formal governance mechanisms mitigate both relational and performance risks, whereas informal mechanisms mostly mitigate relational risks. At initiation, informal governance served to manage relational risk by creating inter-firm trust and positive reputational effects among collaborators, whereas formal governance provided a structure that enabled the management of performance risk. As the ecosystem grew, formal governance was used to manage relational risk. Formal and informal mechanisms mostly aimed to reduce relational risk. This is because relational risk is internal (to the ecosystem) in nature and can be influenced, whereas performance risk is external in nature and cannot be as easily influenced.
Our finding is in line with the ecosystem literature (Ritala et al., 2013) and the family firm literature regarding internal governance (Nordstrom & Jennings, 2018). The ecosystem literature has shown that formal and informal governance mechanisms act as complements to create value for the orchestrator and members (Ritala et al., 2013). The family firm literature has argued that family firms use formal and informal governance mechanisms in a complementary manner to govern their firms. This is because family firms' characteristics emphasize long-term investment horizons, the need for monitoring the realization of objectives, and for minimizing opportunistic behavior, as well as building sustainable relationships with partners (Madison et al., 2016;Nordstrom & Jennings, 2018). This finding is not in line with insights from the family firm alliance governance literature, suggesting that for mutual knowledge creation in alliances, formal and informal governance mechanisms are neither complementary nor substitutes (Bouncken et al., 2020). Scholars show that mutual knowledge creation in alliances benefits most from informal governance mechanisms, suggesting that only informal governance mechanisms are necessary for alliance management (Bouncken et al., 2020). This difference might be explained by the lower and less diverse number of partners, and alliances' short-term goal orientation, resulting in a lower need to use formal mechanisms to stay in control.
The congruence between a family firm's internal governance and family firm-led ecosystems in terms of the complementary use of formal/informal governance mechanisms may be explained by a fit between family firm preferences and the ecosystem's nature. Family firms prefer long-term time horizons to establish socio-economic wealth, aligning well with the longevity of ecosystems. Ecosystems require the creation of long-term relationships, which is in line with the family firm's preference to engage in collaboration with repeated social ties to reduce the likelihood of opportunism (Das & Teng, 2001). Ecosystems might require different governance mechanisms than alliances.
Second, our findings suggest that the choice of governance mechanisms differs for relational and performance risk. When collaborating with well-known partners in an alliance structure, relational risks for family firms might be lower and informal mechanisms might suffice (Bouncken et al., 2020). In an ecosystem with more (unknown) partners, relational risks for orchestrators are potentially higher. This increases the need for informal and formal governance mechanisms to manage risks. Our results suggest that even in ecosystems, informal governance mechanisms remain important to create a sense of trust and cohesion and enable the reaching of long-term goals. This is in line with the ecosystem literature, suggesting that collaboration with other organizations imposes risks (Adner, 2006). Interestingly, family firms might reduce relational risk by taking the orchestrator role in an ecosystem instead of becoming members of a traditional firm-led ecosystem. An ecosystem led by a traditional firm might result in a mismatch between time horizons, objectives, and behavior of the traditional firm orchestrator and the family firm ecosystem member, which would result in relational risks for the family firm (Rau et al., 2019). By taking the lead, family firms are better able to maintain control over their business activities and demotivate members from behaving opportunistically (Bigliardi and Galati, 2017).
Our research suggests that performance risk can be influenced by a combination of low relational risk and formal governance mechanisms. The orchestrator felt that it was difficult to lower performance risk directly by the use of informal mechanisms and believed that low relational risk was important because affecting change in the external environment was only possible if the ecosystem was healthy (low perceived relational risk). Then, the focus was on aligning the ecosystem with external factors rather than managing internal relationships. Formal mechanisms were used in an attempt to influence external factors that might potentially harm success. This finding is not in line with the ecosystem literature, which suggests that there is no direct relation between relational and performance risk (De Man & Roijakkers, 2009). The difference between the family firm and the ecosystem literature might be explained by the high context specificity of family firm research, as suggested by the scholars (Rau et al., 2019). The high context specificity of family firm research might result in the observed different relations between performance and relational risk. Our findings suggest that relational risk can be influenced by a combination of formal/ informal mechanisms, whereas performance risk is mainly influenced by formal mechanisms.
Third, changes in perceived risks over time result in shifts in the relative dominance of formal and informal governance mechanisms. Initially, informal governance mechanisms were dominant, and the orchestrator employed limited numbers of formal governance mechanisms, probably because at initiation, the orchestrator preferred to collaborate with persons from organizations he personally knew or with persons from a network of organizations that were recommended by advisors. When collaborating with people that are trusted or have a good reputation, the relational risk is low (Das & Teng, 2001). There is a low perceived need to protect the family firm from opportunistic behavior and to predominantly use formal governance mechanisms to monitor and evaluate members' behavior. At initiation, the orchestrator used a wide range of informal governance mechanisms and only a few formal mechanisms. Performance risk remained high, despite attempts to influence it via low relational risk and formal governance mechanisms. This is not in line with ecosystem research, suggesting that highperformance and low relational risk should result in the sole use of informal governance mechanisms (De Man & Roijakkers, 2009), but it is in line with findings in this literature, suggesting that formal and informal governance act as complements when ecosystems develop (Ooms et al., 2020). The difference in findings regarding the complementary nature of formal/informal mechanisms might be explained by the relative lack of research in the ecosystem literature on the relationship between risks and evolution.
As the ecosystem grew, formal mechanisms gained prominence, but informal governance remained dominant. Ecosystem growth could influence the perceived risks for both members and the orchestrator, resulting in changes in governance to mitigate changes in perceived risks. As the ecosystem grew larger and organizations joined which were not known personally by the orchestrator, perceived risk for opportunistic behavior increased. The use of formal mechanisms seems to have become more important to protect members and the orchestrator from opportunistic behavior (Das & Teng, 2001). The performance risk remained high, despite attempts to influence it. This is in line with the ecosystem literature, suggesting that high-performance and high relational risk should result in the use of formal and informal governance mechanisms acting as complements. In the context of family firms orchestrating an ecosystem, this might be explained by the preference of firms to build long-term social relationships with partners (requiring informal governance mechanisms) and to stay in control of their resources and contributions over time (requiring formal governance mechanisms). Our research contributes to the family firm literature by showing that the unique characteristics of the family firm context require that formal/informal governance mechanisms be used complementarily across all stages of development, although the relative dominance of both types can change over time. Our findings are in line with the ecosystem literature that has investigated the relationship between governance and development phases, which shows the importance of informal governance in strengthening internal relationships between members, and the importance of formal governance in handling the increasing complexity of the ecosystem as it develops (Ooms et al., 2020).

Managerial implications
First, orchestrators might benefit from stimulating the active participation of members because this increases their intrinsic motivation, provides an effective channel for solving management issues, and creates community ownership (Chamala, 1995;Kelly and van Vlaenderen, 1995). Second, orchestrators might consider changing needs of a growing ecosystem. It can be difficult in larger ecosystems to enable frequent communication between the orchestrator and members and to design a decentralized governance structure, which can lead to less active participation. Orchestrators might benefit from considering how to effectively communicate with members as the ecosystem grows to ensure that members remain motivated. Third, orchestrators should remain mindful of the balance between formal and informal mechanisms. When starting an ecosystem, informal governance mechanisms are important to generate trust between members and to reduce perceived levels of relational risk. Relational can increase in larger ecosystems, as the partners possibly no longer know each other.

Future directions and limitations
Besides the ecosystem literature, other theories exist, such as psychological ownership (Rau et al., 2019), that provide fruitful avenues. Future research might focus on how the orchestrators' psychological ownership affects their choices of governance mechanisms to deal with risks. Different levels of psychological ownership might relate to different orchestrator's fears and motivations, which might result in different choices. Our research shows that the family firm opted for a family firm successor internally and a non-family firm successor in the ecosystem. It is interesting to understand whether and how different levels of psychological ownership might explain different approaches toward ecosystem and internal firm succession.
This study is based on one empirical case study, and though its allowance for in-depth research, it also limits the generalizability of findings (Yin, 2018). Future research should examine how family firms govern ecosystems across ecosystems and test the generalizability of our findings. Furthermore, we did not take a longitudinal approach, but retrospective approach relying on interviewees' memory, which might affect the results' validity. By using secondary data to corroborate interview findings, the confidence in our findings is increased. Future studies should employ a longitudinal design, thereby collecting data at different points in time in the ecosystem lifecycle. This approach could provide a better understanding of causal effects and valuable insights regarding the influence of next generations on ecosystem governance evolution (Rau et al., 2019) and the influence of the family firm's owner-manager on governance over time (Chrisman & Patel, 2012). Our coding strategy entails the risk that we might have focused too much on specific concepts under study, which could have biased our findings (Linneberg & Korsgaard, 2019). Future research might benefit from an inductive coding strategy and looking beyond the ecosystem governance literature for defining codes, e.g., the family firm alliance governance literature (Bouncken et al., 2020).

Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. Appendix 1.

Appendix 2. Case description
Boone International, started by the Boone family, was founded in 1950 as an ordinary furniture producer and developed into a world producer of furniture over the years. In 1985, Jan van Hecke bought the company from the Boone family and became the CEO. Van Hecke owns 66% of the shares of the company, whereas a silent, nonfamily member owns 34%. Whereas Van Hecke has led the company by himself for quite some years and continues to be the CEO of Boone International, two of Van Hecke's sons have been part of the board of advisors since 2011. As a result, the company is fully managed by the Van Hecke family, and Jan van Hecke will be succeeded either by one of his sons or by an external party. The original Boone family is no longer involved in the company Boone originally focused on the production of traditional furniture, but a worldwide crisis around 1989 resulted in a shift toward space-saving furniture, as Van Hecke recognized trends in the industry such as houses getting smaller, people living longer at home, and the pressure on the elderly care institutions. In addition to the shift in the type of furniture, Van Hecke also made a choice to focus on the healthcare sector, as he expected that a need would arise here for a space-saving furniture. Van Hecke hoped that these changes would enable Boone to distinguish itself more from other furniture manufacturers and enable it to play a frontrunner role in the world market. The shift toward a manufacturer specialized in the field of space-saving furniture for the healthcare sector was more complex than originally thought, mostly because of the extremely regulated nature of the healthcare sector Because of the highly regulated nature of the healthcare sector, Van Hecke soon realized that it was not that easy to quickly develop and sell new innovative space-saving furniture in this market. He decided to get around the normal system of wide-range testing to be in line with healthcare sector regulations by starting a showroom. By providing a showroom for customers at Boone's own business site, the company could receive early customer feedback and ensure that the innovations can be incorporated into the plans for new healthcare institutions that are planned to be built, even before they are on the market. This would allow innovations to be improved and further developed before being implemented in practice, thereby speeding up the innovation process After reading an academic study on the future of the furniture and wood industry, combined with his first experiences with the showroom, Van Hecke realized that OI could have a big impact on Boone International. With the support of an external OI expert, Boone International started to participate in several alliances, but after several years, Van Hecke realized that the complexity of the healthcare sector required multilateral collaborations. Van Hecke therefore initiated a meeting in 2009, for which he invited a wide variety of actors in the healthcare sector aimed at inspiring them to co-innovate together with Boone. The attendance rate was low, but the participants had diverse backgrounds and together came up with the idea to develop the Patient Room of the Future. The PRoF ecosystem was established in 2009 with the goal of developing and introducing new concepts such as the Patient Room of the Future to inspire care and well-being Based on his earlier experiences, Van Hecke developed an ecosystem structure and several ecosystem processes to govern the ecosystem. He set up the ecosystem's innovation strategy, managed conflicts, created member selection procedures, decided upon the underlying structures and processes, and maintained relationships among members. The PRoF ecosystem was split into a large "ecosystem" (consisting of non-commercial partners such as healthcare experts, care institutions, hospitals, patients) and a small "ecosystem" (commercial partners, such as companies). The first provided feedback and generated ideas; the latter paid a membership fee to take part in the ecosystem and received feedback. The ecosystem used a 9:1 rule; for every small group member, nine large group members were required to guarantee its non-commercial character and create a flat structure. The ecosystem was    • When a new member asks to join who is possibly a border case (i.e., it is not clear whether the new member will be complementary to the present members), the request is presented to the other members who then decide whether he/she may join.
[4] • Then [next year] we will use our new patented technology to fill in a new piece of technology.
[4] • I have the impression that power is too much centralized. [3] • We still work with the university chair, and the chair was recently given to another university.
[1] • There are no waiting lists anymore in the elderly houses, but new challenges arise. [1] Informal governance mechanisms (first row initially, second row over time • The more we talked about a juridical structure, the more discussion tended to arise that undermined the trust and informal connections.
[2] • The orchestrator manages the trust and steps in when the trust tends to break. [7] • I think the healthcare sector is evolving that quickly that we have to reinvent ourselves constantly.
[4] • … the healthcare sector is extremely regulated. [1] • There is now often no communication, you are informed afterwards, updated less frequently and we participated less in the general discussion meetings.
[3] • It took a while before these people [patients and caregivers] started to trust the industry and realized that PRoF goes beyond commercial interests. [1] • … we should prevent that people become ill and that change is slowly happening in health care.
[6] • The evolution of the healthcare sector is large and undetermined.
[3] Fig. 1 Timeline PRoF ecosystem designed around brainwave sessions, in which the large group generated ideas and screened them based on eight keywords. The ideas were then clustered into broader concepts and assessed by the small group. The small group members selected the best concepts out of the broader concepts developed by the large group, and each small group member developed a specific part toward the final concept. The development process aimed at creating new concepts; the large group provided feedback and assessed whether their ideas were properly integrated. The prototype had to be finished within six months and ready for the market within 12 months Fig. 1 provides an overview of the evolution of the ecosystem. In 2013, PRoF started to expand its activities beyond only concept rooms, introducing the PRoF Awards and Theme days. These activities aimed at rewarding remarkable people and organizations for their contribution to social well-being (Awards) and sharing and creating knowledge around one particular theme within and beyond the ecosystem boundaries (Theme days). In 2014, seven members went to the Altenpflege Messe, winning the innovation award with concept 4.0. In addition, a PRoF Chair was introduced in 2014, strengthening the cooperation between academia and the ecosystem. Despite its success, at the end of 2014, the ecosystem was put on hold for a year because the strategy of selling full concepts was not working out as expected. A second reason to put the ecosystem on hold was that some members would have liked to expand internationally, whereas others wanted to pursue a regional scope. The issue with selling full concepts was mostly the result of the highly regulated healthcare market. While the ecosystem was on hold, PRoF decided to change its approach from selling full concepts into selling separate elements and pursuing a national scope. The ecosystem expects to deliver concept 6.0 that will be related to the sharing economy in 2021. In addition to the delivery of the full concept rooms, the ecosystem has also started a working group consisting of caregivers and field experts that has focused on social issues and has aimed to raise awareness and create solutions for social problems Remaining true to its early days, the PRoF ecosystem's goal is still focused on developing and introducing innovation models to inspire care and wellbeing. The ways in which this goal is realized have changed over the years, for example, by including a social element in addition to the technical one. The PRoF ecosystem nowadays consists of 576 individuals, representing 324 organizations. The ecosystem's showroom attracts approximately 4000 visitors per year.