Setting the Rules of the Road

The rapid rise of a few powerful digital ecosystems disguises a harsh reality about this business model: Less than 15% of business ecosystems are sustainable in the long run.1 When we examined 110 failed ecosystems in a variety of industries, we found that more than a third of the failures stemmed from their governance models — that is, the explicit and/or implicit structures, rules, and practices that frame and direct the behavior and interplay of ecosystem participants.2

Business ecosystems are prone to different types of governance failures. One reason why the BlackBerry OS lost its competition with Apple’s iOS and Google’s Android was because Research In Motion failed to open its app ecosystem widely to developers until it was too late.3 Conversely, the video game industry fell into recession during the so-called Atari Shock in the 1980s in part because of overly open access to its ecosystem, which resulted in a flood of inferior games. Badly behaved platform participants, conflicts among ecosystem partners, and backlash from consumers or regulators are other indicators of governance flaws that can bring down an ecosystem.4

Many orchestrators struggle to find an effective governance model because managing an ecosystem is very different from managing an integrated company or a linear supply chain. Ecosystems rely on voluntary collaboration among independent partners rather than clearly defined customer-supplier relationships and transactional contracts. The orchestrator cannot exert hierarchical control but must convince partners to join and collaborate in the ecosystem. These challenges are exacerbated by the dynamic nature of many ecosystems, which develop and evolve quickly and continually add new products, services, and members.

Ecosystem leaders who understand the components of a comprehensive governance model and glean insights from ecosystem successes and failures can make more informed and explicit governance decisions. In doing so, they can improve the odds that their ecosystems will be among the lucky few that survive and prosper over the long term.

The Key Components of Ecosystem Governance

Good governance supports an ecosystem’s ability to create value, manage risk, and optimize value distribution among its partners. To wield governance in support of these ends and capture a competitive advantage, ecosystem orchestrators cannot treat it as an afterthought. Instead, they must systematically think through and actively design the following five elements of their ecosystem’s governance model. (See “Ecosystem Governance Framework.”)

Mission. A strong sense of shared mission is a potent instrument for attracting and retaining ecosystem partners and encouraging desired behaviors. Orchestrators who identify a clear and distinctive ecosystem purpose early in development and anchor it in a well-articulated set of values can motivate and align partners, particularly when this involves solving a significant problem or making an important contribution to society. An ecosystem’s purpose and values also can help orchestrators attract and retain the right partners and encourage desirable behaviors without undue reliance on complex rules and written standards.

Access. Controlling access allows an orchestrator to select only those partners and participants that meet specific requirements and agree to certain standards and behaviors, and to exclude all others. The rules governing access also can help determine partner commitment by requiring an investment or offering an incentive for joining the ecosystem and/or defining the level of exclusivity that partners must provide to the ecosystem.

Participation. Participation is controlled by the distribution of decision rights and the degree to which partners are invited to contribute to the formulation of ecosystem governance and strategy over time. It also includes the rules for conflict resolution among ecosystem stakeholders. Transparency is critical: Partners need a clear view into the rules and strategy of an ecosystem to actively participate in it and determine their own strategies. Offering partners a high degree of participation in discussions of governance and strategy can bolster their commitment and willingness to invest resources in an ecosystem.

Conduct. Orchestrators can directly influence the behavior of participants in their ecosystem using input control, process control, and output control. Input control, which is often automated using application programming interfaces (APIs) or integrated development environments, specifies the requirements for the partners’ contributions to the ecosystem, including standards and instruments of quality control and the approval of new contributions. Process control uses prescribed procedures and technical standards to regulate the behavior of partners as they interact with one another and the platform. Output control uses mechanisms such as customer feedback, curation, or algorithmic control to regulate the quality of products and services delivered through the ecosystem.

Sharing. The final building block of ecosystem governance defines the data and property rights of partners and the distribution of value among them. Data and property rights regulate ownership and use of the data and intellectual property that are contributed to — or created within — the ecosystem. Value distribution takes various forms, such as a market-based approach that allows partners to set their own prices; coordinated pricing and negotiated value distribution; and centralized pricing and value distribution that is determined solely by the orchestrator.

Using these five elements as a lens to analyze the governance models of more than 80 business ecosystems in a variety of industries yielded four foundational recommendations that can guide ecosystem leaders.

1. Align your ecosystem’s governance model with its strategic priorities. The strategic priorities of business ecosystems vary by their competitive situation and developmental maturity. No matter what the strategic priorities of an orchestrator, the dimensions of governance can be manipulated to support their attainment.

Ecosystem growth, for example, can be fostered by lowering entry barriers, easing the controls on conduct, and/or offering a more generous distribution of value. The Android ecosystem used all of these governance levers to gain scale in the early days of its competition with iOS. Google maximized access by opening it to all developers. It used open-source coding that offered partners the freedom to introduce variations and to integrate their own applications. To attract and support the developer community and spark innovation, Google initially awarded cash prizes to developers for superior applications.5

The governance model can help orchestrators maintain the quality of an ecosystem’s offerings. Belay does this by controlling access: It allows only qualified suppliers of virtual support services, such as administrative assistants and bookkeepers, on its gig economy platform. Kiva, the nonprofit crowdfunding platform that created a field partner network spanning 76 countries to make loans to low-income entrepreneurs, uses the conduct element to ensure loan quality. It has a strict process control system that defines every step on its platform, from applying for loans to underwriting, approving, and posting them; raising and disbursing funds; and receiving loan payments. The online publishing platform Medium uses the dimension of value distribution to incentivize the creation of high-quality content: It pays the writers in its partner program based on the level of reader engagement their articles generate.

If the strategic focus is on improving alignment among the partners of an ecosystem, again the different dimensions of governance can help. The smart-farming platform FieldView uses the dimension of entry to create alignment. It hand-picks its partners to ensure that each offers unique services that complement its overall offering, and it uses individual partner contracts to specify the products and services that partners are allowed to offer.6 The Linux open-source operating system creates alignment among its developer community by leveraging several governance dimensions: a common mission, strict technical guidelines and processes for conduct, and administrative decision rights that are assigned to specific users.

Generally speaking, very open governance models support a rain forest paradigm, in which a diverse set of autonomous players bolster systemic creativity and adaptability. Such models are well suited for ecosystems that must react quickly to changing technologies and customer preferences and whose strategic focus is rapid growth, exploration, and decentralized innovation. In contrast, very closed governance models support a walled garden paradigm by enabling consistency, alignment, and control. They are preferred when an ecosystem’s strategic priorities are focused on ensuring quality, improving efficiency, and coordinating innovation that requires committed partners.

But the strategic priorities of ecosystem orchestrators are rarely this clear-cut. Often, they have competing priorities, such as the need to promote a diverse set of participants while also ensuring the quality of their offerings. Finding the right balance between the two paradigms can mean the difference between success and failure.

Nuanced choices regarding the dimensions of governance can help orchestrators simultaneously achieve conflicting objectives. Apple’s iPhone ecosystem, for instance, achieved rapid growth by offering low-barrier access to app developers while at the same time ensuring a high level of quality and consistency by centralizing decision rights and using extensive quality checks before approving newly developed apps for the platform.

2. Use your governance model to stand apart from competitors. Just as countries and companies compete and prosper depending on the quality of their governance, so too can ecosystem governance serve as a source of competitive differentiation.7 This is particularly noticeable when the orchestrators of competing ecosystems come from different industries. In smart mining, equipment manufacturers Caterpillar and Komatsu have built closed ecosystems focused on their own products, whereas technology companies like Dassault Systèmes and Cisco are building universal internet-of-things platforms that are more open to third parties.

When orchestrators are in the same sector, they can develop different governance profiles to differentiate their competing ecosystems. In the smart-home arena, Apple established a rather closed governance model for its iHome ecosystem compared with competitors Amazon, Google, and Samsung. The key differences in the governance of Apple’s ecosystem include stricter access rules, extensive quality control for new applications, and more restrictive data-sharing policies.8 The company is trying to differentiate itself with a more coherent user experience, even though this could limit its growth rate.

The orchestrators of new ecosystems can adopt an open governance model to counter the network effects enjoyed by incumbents.9 By 2004, Microsoft’s Internet Explorer had won the browser war after capturing nearly 95% of the market. With no serious competitors, however, Microsoft underinvested in the browser’s development and took a relatively closed approach to third-party innovation. Google countered with an open governance model for the Chrome browser and opened a web store for third-party Chrome applications in 2011. Its browser became the market leader soon after.10

Of course, no competitive strategy is risk-free. Orkut, Google’s first attempt at a social network, was launched in 2004, the same year as Facebook. Facebook initially followed a closed governance model, limiting access to users with university email accounts and allowing them to interact only with users at their schools.11 It later opened up access but kept relatively strict privacy controls in place. In contrast, Orkut decided to compete on the basis of an open governance model. The platform applied fewer restrictions on user behavior and less-strict privacy controls than Facebook. This resulted in fast initial growth, but also in many fake profiles and lower-quality interactions, which contributed to Orkut’s demise in 2014.12

Moreover, while competing ecosystems initially experiment with diverse governance models and use them for competitive differentiation, over time the more successful models eradicate the weaker ones. As orchestrators learn what works and as competition becomes more oligopolistic, governance models tend to converge. In December 2019, a group of large players that included Amazon, Apple, and Google launched the Connected Home over IP alliance to increase the compatibility of smart-home products for consumers, a move that will lead to greater harmonization of their governance models.

If one ecosystem gains a competitive advantage by adapting its governance model, others may be forced to do the same to keep up. For example, most of today’s social media networks have similar governance models and tend to move in lockstep on emerging issues, such as establishing independent oversight boards for content curation.

3. Use governance to ensure social acceptance. Societal and regulatory scrutiny of ecosystems is on the rise. Sharing-economy and gig-economy platforms are coming under fire for avoiding costly requirements related to safety, insurance, hygiene, and workers’ rights. Social networks are criticized for lax data privacy policies and the dissemination of false and misleading information. E-commerce marketplaces are accused of undue price pressure and unfairly giving their own products an advantage. There are mounting concerns about the alleged concentration and misuse of power by dominant digital platforms.

Despite and perhaps due to their great success, an increasing number of orchestrators see their ecosystems being challenged by regulators, and they run the risk of being broken up or losing their licenses to operate. Beyond regulation, the mutual trust that is foundational to the success of an ecosystem is threatened: In previous research, we found that a lack of trust critically contributed to over half of the 110 ecosystem failures we studied.13

Thus, good governance is rapidly becoming a prerequisite for building social capital and securing the social legitimacy required by business ecosystems. To meet this dictate, orchestrators must understand that it is not sufficient to optimize the value proposition and experience of customers; they must also enhance the value and experience of other ecosystem contributors and external stakeholders. Moreover, the governance model must be designed to engender and maintain social acceptance, as well as legal compliance, over the long term and in the face of changing demands. Superior governance, understood in this way, must be consistent and fair.

Consistency means that the mechanisms of governance are transparent and easy to understand, comprehensive, internally consistent, and stable over time. HopSkipDrive uses consistent governance to establish trust for its particularly sensitive offering: ride-sharing services for children. The platform accepts only drivers who adhere to multiple ecosystem guidelines and have at least five years of caregiving experience; a clean multiagency, fingerprint-based background check; and a good driving record for at least the past three years. It employs strict process control, tracking each ride in real time to detect unsafe driving behavior and proactively address any issues. Moreover, the platform regularly reports on the frequency of safety-related issues such as traffic incidents, collisions, and distracted driving.14

Fairness means that governance complies with local laws and norms, avoids biases (for example, in data algorithms and access), and creates trust among participants (for example, by forbidding the misuse of orchestrator power). The importance of fair governance is illustrated by the recent rise of platform cooperatives. These platforms, which include hospitality platform Fairbnb.coop, ride-hailing service The Drivers Cooperative, and stock photography platform Stocksy United, are owned and governed by their contributors.15 Many of them have emerged in reaction to exploitation by ecosystem orchestrators.

4. Adapt your governance model over time. Adaptability is a key strength of a successful ecosystem. Typically, this adaptability stems from a modular setup that features a stable core (or platform) and interfaces, with highly variable components that can be easily added or subtracted. This enables ecosystems to evolve along with changes in the competitive environment, the needs of orchestrators and participants, social mores, and technology. This same kind of adaptability must also be reflected in the governance model of an ecosystem.

Consider Steam, the video game distribution platform. Originally launched in 2003 as an online platform for the distribution and patching of Valve games, its initial governance model was rather closed. In 2004, some 50 titles were available on the platform. Then Steam started to open up by lowering the barriers to entry, reducing input controls, and negotiating its first partnerships with game publishers. The number of available games quickly doubled.

Between 2007 and 2010, Steam further opened to external developers with the launch of the Steamworks software development kit and a number of APIs that facilitated data sharing with partners. Major publishers joined the platform, driving the total games available to more than 1,000.

Over time, Steam increasingly relied on user-based mechanisms for governance of the ecosystem. For example, in 2012 it introduced Greenlight, which allowed users to vote on which new games would be added to the platform. User feedback and reviews became increasingly important for the output control needed to secure the quality of the platform’s games.

Another major shift in governance took place in 2017, when Valve recognized that, with the growing size of the Steam ecosystem, Greenlight had become a major bottleneck for onboarding new developers and games. Accordingly, it replaced Greenlight with the Steam Direct submissions portal for developers, to further reduce entry restrictions and input control while relying on more effective user feedback mechanisms to ensure quality. By 2020, Steam had evolved into a very liberal and open ecosystem, with more than 10,000 games and 120 million monthly active players.

Steam illustrates a phenomenon that we see in many successful ecosystems: They tend to start with rather closed governance and become more open over time. One reason for this: The development of an ecosystem is strongly path-dependent, with early decisions having a significant impact on the trajectory and future scope of the ecosystem. Hence, many ecosystem orchestrators prefer a closed governance model at the beginning to control quality and behavior and to get the ecosystem on the path to success.

There is one caveat: While most business ecosystems tend to become more open as they grow, some ecosystems begin to tighten governance once they reach a certain size and market position. This can be a reaction to misuse of the platform, such as ride-hailing services reinforcing background checks for drivers after passengers are attacked, or to public or regulatory pressure, such as social networks strengthening input control to prevent the spread of misinformation during the pandemic.

The orchestrators of successful ecosystems also tighten governance to increase their own value capture, particularly if they have achieved a leading market position and partners are increasingly dependent on the platform. Reinforcing the rules for participation, conduct, and sharing are the most effective levers for achieving this. Google tightened its grip on Android after handset manufacturers and other partners began developing variations of the operating system that were incompatible with some applications and made it difficult for Google to profit from the platform by selling advertising and software.16 By moving the APIs for app development from the operating system layer to the Google Play app store, the company regained control of large parts of the Android ecosystem and could adapt its governance largely independently of the underlying operating system.17

As ecosystems become more widespread and established, the quality of their governance is an increasingly important success factor. But there is no single best way to design your governance model: It will be contingent on the strategic priorities, competitive dynamics, societal demands, and life-cycle stage of the ecosystem.

If you are — or want to become — an orchestrator, you should not treat governance as an afterthought but should instead think through and actively design the governance model. You need to understand the benefits and risks of being open or closed, align governance and strategy, and resolve strategic trade-offs by balancing the different dimensions of governance. You ought to put yourself into the shoes of ecosystem partners and users to understand the impact of your governance decisions on their incentives to participate and contribute. You should also analyze the governance models of competing ecosystems and use your governance choices to gain competitive advantage. And you will have to adapt your governance model over time to react to changes in user preferences, technology, competition, and strategy.

If you are joining an ecosystem as a partner, you need to understand how its governance model works and evolves over time. An ecosystem’s governance will strongly influence how attractive it is for you. You should consider governance as an important criterion for deciding in which ecosystem to participate by asking yourself several questions: How well does the purpose and culture of the ecosystem resonate with your values and preferences? What kinds of commitments and ecosystem-specific investments are required that may limit your future flexibility? Are the transparency and decision rights that you need to understand and influence the development of the ecosystem in place? To what extent do regulations for input, process, and output limit your customer access and freedom to operate? And are regulations in place that ensure that you benefit in a fair way from the data, intellectual property, and value that you contribute to the ecosystem?

Good governance is an essential key to the success of both ecosystem orchestrators and their partners.