EXHIBIT 6.1. Designing Interfaces for Improved Productivity. Source: authors’ research.
Improving Ecosystem Productivity
SO FAR, WE HAVE CONCENTRATED ON THE BENEFITS TO BE gained from pursuing ecosystem strategies. But in chapter 2 we also cautioned that ecosystems are generally less efficient than hierarchical organizations at delivering customer value, once the opportunity has been identified and the offering scoped out.
It may be necessary to accept some loss of efficiency in exchange for the greater knowledge creation, innovation, and flexibility that ecosystems offer. But if the inefficiencies become too great, the ecosystem may collapse under its own weight. Thus, the efficiency with which the ecosystem delivers value needs to be “good enough” to ensure that its inefficiencies do not swamp the gains elsewhere. The productivity of the system matters, and ecosystem leaders can play a key role in promoting it.
Overcoming Ecosystems’ Productivity Disadvantage
An ecosystem can never be as efficient as a vertically integrated company because bringing all activities under one roof allows a company to achieve the maximum alignment between activities; each one of them can be designed to dovetail seamlessly with the one that comes before and after it. Activities at different stages in the value chain can be synchronized, eliminating wasted capacity or bottlenecks that would undermine efficiency. One of the most important benefits of a vertically integrated value chain is that the capacity of each contributor can be matched to the others, and this so-called line-balancing is easier to perform because coordination is under the control of a single firm.
However, vertical integration and internal value chain optimization has a downside: it generally reduces flexibility. Some companies seek to address this problem by using internal marketplaces to coordinate their activities. Internal markets can achieve greater flexibility while maintaining efficiency because transactions inside a company seldom require the complex contracts that are otherwise necessary to guard against the failure to perform or price gouging by suppliers. The application of improvement processes such as Total Quality Management, lean management, and Six Sigma can ensure that all the units in an organization adhere to common standards. Transfer-pricing systems can be established to govern internal exchange. The effects of fluctuation in volumes and changes in the product mix can be handled by production planning and supply-chain management systems. Investments are coordinated by the company’s capital-budgeting process. Therefore, with the appropriate structures and processes in place, vertically integrated companies can enjoy the benefits of low transactions costs and high productivity while remaining fairly flexible.
However, by its very nature, vertical integration constrains a company to perform most of its key activities in-house. And these in-house units may be less capable, less innovative and less efficient than external suppliers who are forced to compete in the open market. To deal with these limitations, companies have learned to generate many of the advantages of vertical integration by establishing tightly coordinated chains of external suppliers. Integration between these suppliers is often strengthened by creating sourcing hubs that enable the sourcing of components to be coordinated centrally.1
Given that ecosystems are usually at an efficiency disadvantage compared to vertically integrated or centrally controlled value chains, a key question for the ecosystem leader is how to reduce this handicap and at the same time preserve the advantages of ecosystems, including more learning, more innovation and the flexibility to cope with volatile environments and rapid growth that vertically integrated companies or traditional supply chains cannot match.
For instance, Alibaba’s emerging ecosystem faced just this challenge when it decided to transition from being an information portal to an e-commerce website: it was both innovative and very flexible, but also hopelessly inefficient. It was difficult for Alibaba to ensure that payments would flow smoothly and securely from buyers to sellers. China’s credit card system was still in its infancy. Potential customers were worried about the security of online payments as well as the delays involved in making payments by alternative methods such as checks and postal orders.
These problems were aggravated by the fragmented nature of China’s logistics industry at the time, which comprised tens of thousands of providers. Their capabilities were weak, their services were unreliable, and their coverage was patchy. A delivery might pass through many hands, each representing a different level in a long transportation chain from the warehouse to a village, before it reached its final destination. In fact, Alibaba’s Jack Ma used to lament in those days that the “terrible” logistics were crippling the growth of internet retailing in China. The promise of the new ecosystem was enormous, but it may never have gotten out of its starting blocks if Alibaba had not been able to tackle the inefficiencies in its delivery and payment system, often by creating systems like Alipay that customers and partners could use. In other cases, it relied on building relationships with creative partners in logistics like Huawei and Haier.
Solving these kinds of problems to improve efficiency and minimize the productivity disadvantage of ecosystems while protecting the benefits is a fine line to tread. Pursuing too much efficiency risks killing the creativity needed to cope with uncertainty and damages the process of learning and innovating. So the goal is to improve the ecosystem’s productivity, rather than simply trying to maximize its efficiency at all costs. And there are some practical steps an ecosystem leader can take to address this challenge. First among these is to design a set of interfaces through which the partners can interact efficiently.
Designing Efficient Interfaces between Participants
Ecosystem leaders can play a huge part in addressing the productivity challenges their ecosystems face by designing interfaces between partners that reduce transaction costs.2
In a classic supply chain, most of the interfaces between parties are governed by contracts. That’s appropriate, and efficient, when performance can be measured objectively (think of a situation where a component or service must be delivered in the right quantity at the right time). However, in an ecosystem the interaction between partners is seldom that simple. For an ecosystem to succeed in innovating, partners need to interact in a way that helps each party to learn. Instead of designing the interfaces around transactions where one party delivers something to another, the ecosystem leader needs to design interfaces that will facilitate the flow of knowledge and enable innovation. As the participants in an ecosystem learn, their roles and contributions will change, so the interfaces need to be flexible enough to allow the ecosystem to evolve continually.
What the ecosystem leader needs to design, therefore, are interfaces that enable diverse participants to efficiently, but flexibly, exchange products, services and information that are constantly changing. We propose seven principles that can be used to guide the design of such interfaces (see exhibit 6.1).
Writing Appropriate Contracts
Traditional contracts usually won’t be up to the challenge of organizing the interfaces in an ecosystem. Recall that DS has more than eight hundred partners in its ecosystem, who continually update and develop their software solutions or development needs. Many of them do not even interact through DS but, instead, work directly with each other. Imagine trying to renegotiate contracts with hundreds and perhaps thousands of such partners in an ecosystem every time the products, services, information, or access provided to the ecosystem changes in unpredictable ways. Think about trying to re-price all the knowledge flowing between all the partners in an ecosystem and then, rewriting new contracts to govern the new exchanges. The ecosystem would drown in lawyers’ fees and be paralyzed by delays in getting the contracts renegotiated.
Of course, there’s nothing wrong with contracts per se. They are part and parcel of business interactions. But traditional performance contracts are insufficient to ensure effective collaboration between ecosystem partners.3 Aligning behavior is difficult enough in traditional supply chains—but in innovative ecosystems, typically littered with unknowns and high uncertainty, trying to specify contracts with concrete actions and deliverables, and fixed prices, is likely to become a hopeless task. And when unexpected changes occur, partners in the ecosystem may be affected unequally. This could invalidate the premise on which carefully tuned contractual agreements were based, sowing distrust. Ecosystem leaders therefore need to encourage partners to draw up contracts that focus on high-level outcomes, while leaving room for flexibility. Additionally, these contracts must be perceived as fair when things change, and also clearly clarify how dispute resolution will be handled.
However, the rapidly changing requirements, cost uncertainties, and difficult-to-measure performance that characterize most ecosystems prevent the partners from drawing up contracts that include every possible contingency. Rather than specifying detailed actions and detailed outcomes, contracts between ecosystem partners should be written to capture the open-ended nature of the collaboration required. A good analogy is an employment contract that defines a job description but does not specify detailed tasks and outcomes. Such open-ended contracts between the partners in an ecosystem provide an element of de facto oversight and incentives for the parties to make decisions as if they were part of the same organization.4 These kinds of contracts provide the ecosystem with the flexibility to respond to unforeseen situations.
Suitable contracts must also guarantee fairness. In an ecosystem, one partner rarely has complete information about all the others in the ecosystem. It is often difficult to judge whether changes in behavior by other partners are acceptable. The contract must ensure that decisions will be taken in such a way that they are demonstrably fair to all the partners in the ecosystem. In 2003, INSEAD’s Chan Kim and Renee Mauborgne proposed that fairness in a contract requires three qualities to be present. One, clarity about the rules of the games and the purpose that must be achieved. Two, provisions to ensure the engagement of all the affected partners in decisions that concern them. This is a way for the ecosystem leader to communicate respect for all the partners. Three, a transparent explanation of when and why the terms of agreement can be modified so as to avoid automatic suspicion and engender trust.5
However, to make these flexible contracts that focus more on the process of collaboration (rather than specifying precise tasks or outputs) work, partners must accept that they need to maintain a balance between the investments, risks, and rewards accruing to each party.6
Contracts also typically include clauses covering the mechanisms for dispute resolution—and this is essential in contracts between ecosystem partners too. But they need to take a different form than seen in most commercial contracts so that they can handle the high levels of innovation and flexibility characteristic of ecosystems. Rather than being a mere safety valve of last resort, they need to be cast as almost routine procedures for responding to a stream of contentious issues as they arise. Ecosystem partners need contract provisions that enable them to resolve the many inevitable differences that would arise between them simply, quickly, and collectively. The goal is not to break the deadlock of rare, and potentially terminal disputes, but instead to facilitate shared problem solving.7
An example of this kind of contract is the agreement that ARM and Vodaphone entered into in March 2019 governing their joint development of a product that integrates Vodafone’s global network and Internet of things (IoT) platform with Arm’s IoT software and services. It commits the parties to work together to develop a programmable, connected system on chip (SoC) designs that eliminate the need for traditional SIM cards like the ones used in mobile phones. But rather than detailing specific product features and deliverables, the contract captures the partners’ commitment to develop interconnectivity capabilities, standards, and security protocols that will allow other partners in the ecosystem to develop single IoT products that can be shipped and reliably connected anywhere. The aim is to enable the end user to securely deploy and manage massive numbers of IoT devices across global markets at lower cost and with less complexity.8
Enabling Data Exchange
Our second recommendation for improving ecosystem productivity is for the ecosystem leader to develop interfaces that can smooth the exchange of data among the partners. This starts with the simplest case: where data needs to flow between the ecosystem leader and other participants in the ecosystem.
The kind of interfaces that need to be established and the benefits to the ecosystem are well illustrated by the experience of The Guardian, a leading British newspaper. (We will discuss this case in more detail in the next chapter, but here it is useful to focus on the data exchange between The Guardian and its partners.) The Guardian’s open journalism ecosystem is an interesting example of how creating interfaces for the smooth exchange of data between an ecosystem leader and its partners can help an ecosystem develop and prosper. In the early days of the internet, The Guardian wanted to enable readers, free-lance journalists, and website developers to access its content, and also encourage them to integrate this into the applications and websites these independent users and developers would create. To enable this data exchange to happen efficiently, The Guardian decided to build an Application Program Interface (API) to facilitate the process. As we mentioned in chapter 4, APIs are sets of routines, protocols, and development tools that determine how software components should interact as data flows between them. Because the right APIs make interactions easy, they are the enablers of many successful ecosystems.
But this creates a potential problem for ecosystem leaders: leakage of valuable data to a mass of partners. The Guardian faced this problem. While it wanted to promote the goals of open journalism, it also wanted to limit the amount of information that its partners could access for free—a classic case of the trade-off between enabling partners and protecting its own interests that ecosystem leaders frequently face. To achieve the right balance, it needed a sophisticated API to manage its interactions with partners. The interface The Guardian came up with comprised two tiers. The For-Free tier was available to any website developer who registered, but it limited the number of calls they could make per second as well as the total number of calls every day from each application. It also restricted access to the amount of text in each piece, and limited access to a certain, albeit large (over 1.7 million in 2015) number of pieces of content. Additionally, the For-Free tier required developers to display the embedded advertising associated with the content they pulled from the newspaper’s website. The Paid-For tier, by contrast, gave subscribers access to the full range of content available from The Guardian, including articles, video, audio, and images, at a fee that varied with the number of calls made each day and the amount of data pulled though the interface. This tier also allowed partners to display third-party advertising without sharing revenues with The Guardian.
Just supplying data to partners, however, sharply limits the amount of value an ecosystem leader can create. The Guardian recognized this constraint. It could create and capture more value in its ecosystem if it could go beyond simply supplying data efficiently to its partners. If it developed an interface that facilitated efficient, two-way flows of information with its partners, it would be able to access valuable learning from its partners’ activities as well. So, it embedded in the interface an analytics code that allowed The Guardian to track the location and popularity of the content used by its partners, including any embedded advertising. Designing such an interface enabled the newspaper to draw data and accumulate learning from its extensive ecosystem, reaching beyond the activities in the ecosystem that it directly controlled.9
Because content and readership data are the lifeblood of the newspaper business, it made sense for The Guardian to design an interface that enabled it to keep strong control over the data that flowed between it and its partners in the ecosystem. But ecosystem leaders don’t always need to tightly manage the exchange of information, or even track who takes what and contributes how much. In the DS ecosystem of more than eight hundred business, education and software partners, it would be almost impossible to try to track information exchanges between the partners at a reasonable cost. Nor is this necessary in order for DS to create and capture value in the ecosystem. By creating the 3DSwYm interface, similar to a social media platform where partners can share information and exchange concepts and models, DS was able to help partners become more successful and promote its own products and services without controlling the data flows.
Is there a risk in not controlling such data exchange? Could there be leakage of important data outside the ecosystem? That risk exists. But it may be less important than one imagines. As we know from research on R&D collaborations, a lot of information-sharing between scientists and engineers takes place in an informal know-how trading market, similar to what one finds in an ecosystem.10 Engineers develop informal knowhow trading networks, and make judgments about who they consider to be an expert in a particular field. They build personal “rolodexes,” and when they have a problem, they seek out the relevant expert—even if the expert works for a rival—and seek his or her advice. If they know that the information is critical to the rival’s competitive advantage, the expert will not share much. However, in many cases, experts are willing to share considerable information because they expect to need, and receive, reciprocal help in the future. Engineers are actually quite good at estimating the value of the information they share, and the exchange of information proves to be effective. In fact, among experts in any field, the obligation to return a favor in a fair way is felt quite strongly.
The message for ecosystem leaders is clear: these informal exchange mechanisms can be harnessed to drive the ecosystem forward by putting in place interfaces and platforms that facilitate self-organization. Rather than creating a constraining set of structures and rules to manage knowledge and data flows in an ecosystem, the ecosystem leader might most effectively promote the success of its ecosystem by organizing conferences, creating private social networks, and setting up loosely organized joint development teams that help promote informal sharing and exchange of know-how. The value created through the vibrant exchange of knowledge among partners within the ecosystems of ARM and DS are testament to this approach.
Managing the Exchange of Complex Knowledge
Within an ecosystem, the flow of data between partners becomes particularly challenging when highly complex knowledge needs to be exchanged. More often than not, this complex knowledge is not written down, but resides in the heads of individuals. It also tends to be fragmented between different parts of an organization. As a result, it is especially difficult to assemble and transfer. This leads to our third recommendation to ecosystem leaders as they seek to shape interfaces that will make the ecosystem more efficient. To make sure complex knowledge flows smoothly across the ecosystem, the ecosystem leader needs to deploy an armory of approaches that include putting in place systems, tools and organizational structures that facilitate its own knowledge capture as well as the necessary flows between partners. The nature of complex knowledge, however, means that it is almost inevitable that the interfaces will require high levels of human interaction.
Accessing the complex knowledge embedded in the unfolding technological and product roadmaps that ARM needs to understand from its original equipment manufacturers and semiconductor fabrication partners is a case in point. This process involves more than collecting blueprints from each partner. ARM often has to piece together the future directions of a partner’s technologies and products from snippets of knowledge locked in the heads of employees in different departments. Worse still, that knowledge tends to be unclear, tentative and incomplete. As a result, the interfaces that an ecosystem leader such as ARM has to create to smooth the flow of complex knowledge across the ecosystem needs to be very different from those designed to share data.
As we saw earlier, ARM’s solution was to deploy specialized employees, in the form of partner managers, to act as an interface with those partners with whom it had to exchange large quantities of complex, tacit knowledge. Initially, each partner manager was assigned to a handful of partners, or in some cases, just one. The partner managers were responsible for piecing together an understanding of the product and technology road maps of its OEM partners (its “customers’ customers”) as well as the semiconductor partners to whom ARM licensed its designs. Their job description included: “developing sound, professional, long-term relationships with management, technical, and marketing personnel at the partner [firm] as a key role as well as creating, reporting, and executing on account strategies.”11 Another important part of the job was to “represent a partner to ARM, and provide feedback to ARM business units, engineering, marketing, and management regarding partner technical and business requirements,” and “identify and capture insights on the competitive landscape, trends, projections, and other relevant information.”
Relying on dedicated and experienced staff to access complex knowledge from partners makes a lot of sense when this knowledge is highly valuable to the ecosystem leader. Certainly, ARM’s use of partner managers as the primary interface with its collaborators in the ecosystem worked well at first. But as ARM’s ecosystem grew to include more than four hundred partners around the world, this approach to partnership management had to be modified. It wasn’t efficient or practical to dedicate expensive and scarce partner managers to every partner in the ecosystem. So ARM began to vary its interfaces according to the type of knowledge the partner could provide, the partner’s size, application segment in which it worked (such as mobile phones or automotive), type of technology involved, and the partner’s degree of influence on the future direction of its industry.
These actions reflect the classic trade-offs between value creation, flexibility, and efficiency that all ecosystem leaders have to make. As we saw in chapter 5, the first step ARM took to improve the efficiency of its interactions with its partners was to identify a small set of strategic partners who would be critical to the future success of its ecosystem. The top twenty strategic partners were assigned to one of ARM’s directors—the CEO and his direct reports—to manage the relationship. This ensured that even where the relationship was sales oriented, and a large ARM marketing and sales team was involved, top management could balance the short-term objective of ensuring revenues with the longer-term interests of sharing forthcoming technology roadmaps that would guide the future direction of the ecosystem.
The second step to improve effectiveness and efficiency was to split the traditional job of the partner manager into distinct roles performed by dedicated staff. A segment marketing organization was created for each end-use applications area. ARM grouped its activities into eight applications segments: Wireless, Storage, Imaging, Automotive, Consumer Entertainment, Networking, Security, and Industrial. These segment marketing teams had the task of working out, as ARM’s chief technology officer, Mike Muller, put it, “who are the players, who matters, and what do they want.”
ARM’s experience illustrates the need for ecosystem leaders to think carefully about the different types of interactions they need to have with partners within their ecosystem, and how the interface might need to vary between different divisions of the same partner. And a knowledge exchange designed to shape the future direction of technology or a product road map should use different channels and interfaces than those where the purpose is to secure sales or source products or services. Creating distinct channels to handle different types of interactions reduces the risk of damage to relationships as a result of mixed motives colliding or impeding the flow of knowledge by forcing it through an unsuitable or inefficient interface.
Codifying Knowledge
A fourth recommendation for ecosystem leaders to keep in mind when designing interfaces with partners is to recognize that some data and information is difficult to articulate and share. In such cases, the ecosystem leader may develop a system that codifies the knowledge. Codification can enable the knowledge to be universally shared or may render it accessible only to technical experts within the ecosystem who are familiar with the jargon used. In either case, it will facilitate the transfer of knowledge that would otherwise be imprisoned in one corner of the ecosystem.
The benefits of codification are well illustrated by the solution Thomson Reuters came up with when faced with the problem of providing information to support its customers trying to navigate the legal system. The information buried in legal documents is often difficult to share, partly because the implications of court proceedings are hard to extract and communicate. As a result, the productivity and efficiency of the legal system may be undermined. Thomson Reuters recognized that if it could better codify the patterns and implications of large amount of legal cases across the world, it could make it part of the information it provided to professionals in financial services, insurance, legal, taxation, and accounting. In chapter 4, we described the partnership that Thomson Reuters set up with CodeX, the Stanford Center for Legal Informatics. That is an example on how Thomson Reuters developed a codification system in collaboration with its ecosystem to improve productivity.
But codification of knowledge within the ecosystem can do more than improve productivity. It can also open up new opportunities to fuel innovation and create new value. The next step Thomson Reuters took shows how this potential can be realized. In 2015, Thomson Reuters sponsored the Legal Tech Open Innovation Challenge at CodeX, which was “an open call for legal professionals, programmers, entrepreneurs, data scientists and any other interested parties to develop new applications that improve the efficiency of the legal system by providing high-value analytics.”12 Participants were given access to federal court docket data, outside data, and Thomson Reuters’s open-source company identification system, PermID. The winner developed an application that used court data to create a model that predicted the likelihood that motions to dismiss would be successful before specific judges in US federal civil courts. As a result, Thomson Reuters was able, as Tim Baker, its director of innovation, put it: “To combine our proprietary content and domain expertise with a broad range of leading technologists to further fuel innovation and advance the practice of law.”13 This new model contributed to facilitating the exchange of complex legal data across the network of Thomson Reuters’s partners.
Hence, by making it easier to access, interpret, and exchange complex information, thereby improving the productivity and efficiency of practicing law, Thomson Reuters was able to strengthen its ecosystem and better serve its customers, who were “increasingly looking to the next generation of technology solutions to build and advance a better functioning legal system.” Enabling codification to help knowledge flow more easily throughout the ecosystem is thus a key contribution that the ecosystem leader can make to promote the network’s efficiency.
Our fifth recommendation for how ecosystem leaders can enhance productivity is to find ways to nurture trust—between the ecosystem leader and its partners, as well as among all the partners in the ecosystem. When the participants in an ecosystem trust each other, expensive safeguards become unnecessary and transaction costs fall. Trust thus makes the ecosystem more productive. Trust is indeed a key coordinating mechanism in any hybrid form of organization in between markets and hierarchies.14
The kinds of trust that can make the interactions in an ecosystem more efficient depend on the type of risks associated with what they are jointly trying to achieve. Two sources of risk are particularly important. One, the extent to which one party depends on the other, and two, whether or not the quality of what each party delivers can be objectively assessed.15
To understand how these two types of risk can be handled in an ecosystem, think about a very simple everyday analogy. Suppose you are engaging a painter to paint your home. People don’t choose a painter they have never used before, give her or him the keys to their house, and let him or her finish the painting when they are on vacation. Most will get the painter to start in a small, secluded part of the house, and then hurry home at the end of the day to assess the quality of the work. If the painter has done a poor job, they will fire him. And how do they ensure that the painter doesn’t take on another job and delay theirs? They withhold payment until the job is complete and may offer a bonus for finishing on time.
Some interactions in the ecosystem are much like those with a house painter. The relationship is asymmetric; one party depends on a partner’s performance, but the latter’s success does not depend on interaction with the former. In that case, contracts with performance measures and incentives are likely to provide the most efficient way of managing the interaction and ensuring trust, as long as the partner’s responsibilities can be precisely defined and performance is observable. Apple’s Apps Store and Amazon’s marketplace use this approach by requiring buyers and sellers to enter into contractual relationships, sometimes backed by incentives to encourage good performance.
However, as we have already flagged, it is rarely possible to detail all the tasks and responsibilities of each partner in an ecosystem and write these into contracts. In the process of creating new value, ecosystems typically involve partners in poorly defined and emergent processes of learning and innovation. Partners’ roles and interactions will change over time as the ecosystem evolves. Moreover, the performance of partners is often difficult to observe and measure. This is especially true when they contribute knowledge, capabilities and access to the process of building a complex interdependent ecosystem, where the causes of success or failure are hard to trace. It is in these situations that trust becomes extremely important to underpin smooth and efficient engagement between partners.
There are basically three ways to build trust: through one’s reputation based on competence, consistency, integrity, and so on; through repeated interaction gradually building familiarity; and through norms that create predictability and trustworthiness.
It is hard to build trust based solely on objective measures or simply through observation. Therefore, reputation almost always plays an important role. Take the example of an ecosystem leader wishing to attract a partner to stimulate the demand for the product or service that the ecosystem delivers by acting as a market maker. If the new partner fails to generate demand, was the company an untrustworthy shyster? It is impossible to tell. The ecosystem’s products might have been such a poor fit with the needs of potential customers that even the most competent and committed partner would not have been able to help it succeed. How do you solve this dilemma? The best way to develop trust is to focus on partners who enjoy an established reputation because you know that if they don’t do their best to stimulate demand in the market, their reputation will suffer.
For many of the interactions that are necessary for an ecosystem to deliver value, partners’ reputations can therefore play a key role. Attracting reputable partners will foster the trust required to reduce transaction costs in the ecosystem and make it productive. Ecosystem leaders need to make sure that when partners join the ecosystem, they put their reputations on the line. A good way to do this is to require that in order to join, key partners lend their brands to the products and services the ecosystem delivers, and prominently communicate their involvement in the ecosystem. This is particularly important in the outer reaches of an ecosystem, where performance is less observable and hard to measure. Here the ecosystem leader has to rely on attracting partners with strong reputations and helping others to build their reputations.
Recall that AWS, for example, fostered the growth of a trusted global partner ecosystem by setting criteria for entry into its AWS Partner Network (APN). That network also provided mechanisms by which partners could build their competence and reputation through a series of certifications.16 AWS maintained a partner directory that allowed potential customers to find trustworthy partners, where customers and approved software-and-service vendors could connect.
In other interactions between the partners in an ecosystem, performance may be more easily observable and objectively measurable. Consistent and repeated interactions will build familiarity and ultimately trust. Even then, success will depend on the quality of the interactions as well as the inputs. In these cases, dependence and trust in the relationship is symmetric. Symmetric relationships in an ecosystem, where success depends on understanding and coordination between partners, can only be developed by experimentation and learning by doing. The ecosystem leader therefore needs to find ways of promoting interactions between partners that provide opportunities to learn from each other and to experiment with joint activities. One way the ecosystem leader can do this is by sponsoring pilot projects where partners can explore and test ways of working together.
Alibaba, for example, when confronted with the need to improve logistics in China, had to identify seed partners who were willing to build trust by embarking on a journey of co-learning. It identified a small group of partners who believed that by working together, they could dramatically improve the ecosystem’s logistics capabilities. This required sustained effort with an uncertain result. And, as Alibaba’s Ming Zeng noted, the company then “worked almost day and night with these seed partners to get a prototype up and running as soon as possible. Once it was operating, learning and competence could be accumulated.”17
Contrary to early expectations, Alibaba found that the partners who joined weren’t industry leaders—these were often too wedded to their traditional models and even complacent—but up-and-coming companies that believed in the future of e-business, could see the upsides of the ecosystem, and were willing to build trust over time by working together.
Trust is often also a consequence of a calculated approach, via a sober assessment of the costs and benefits to the partners of exploiting each other’s vulnerability. The risks involved in committing to an ecosystem will only be acceptable to partners when they trust that the ecosystem as a whole is working to create mutual benefit as well as serving individual interests. It will fail if partners suspect that the ecosystem is riddled with hidden agendas working against them. The ecosystem leader, therefore, has an important role to play in building trust in the ecosystem as a positive force that treats participants fairly.
Just as with corporate leadership, communication has a key role here. Ecosystem leaders need to keep reminding participants of the benefits of being part of the ecosystem it is delivering. They need to celebrate partners who make investments or perform extra activities for the benefit of the ecosystem as a whole, to promote its prosperity in the longer term. They must promote forums, both online and face-to-face, where partners come together, listen to each other’s opinions, understand the implications of their decisions on specific partners and the wider ecosystem and potentially modify them, help develop a common view of events as they unfold, and look for opportunities for mutual benefit. Fostering this understanding of partners’ shared destiny and promoting these types of interactions helps reduce transactions costs, improves the ecosystem’s productivity, and creates a self-reinforcing cycle of commitment and adjustment that will help partners through the inevitable crises and hiccups that the ecosystem will experience.18
Agreeing on Governance Standards
Trust is also based on good norms that create predictability and trustworthiness. Therefore, our sixth recommendation is that the ecosystem leader should develop norms and standards that all of the parties in the ecosystem must subscribe to. An ecosystem will be more productive and efficient if customers, users, and partners don’t have to check each other’s credentials, because they know that the ecosystem leader ensures that they all abide by a set of minimum standards. Apple’s App Store, for example, insists that every offering meet certain minimum standards in five areas: Safety, Performance, Business, Design, and Legal. Safety, for example, covers standards for the exclusion of objectionable content, prevents outside links in apps designed to appeal to kids, and specifies requirements for the ways by which customers can get questions answered and access support. In the area of performance, meanwhile, Apple sets standards relating to completeness, hardware compatibility, and the rate at which the app can drain battery power or generate heat in a device.19
In creating its More Disruption Please (MDP) marketplace, athenahealth has taken standard-setting a step further to complement trust and facilitate interactions between partners in its ecosystem. Services sold on athenahealth’s platform must conform to the company’s privacy and security standards, and any solution accepted for the MDP marketplace must either lead to an increase in client revenues, reduce inefficiencies, or improve health outcomes. To check that these requirements are being met by partners, it has established a system for tracking performance metrics in real time.
As we saw in our discussion of APIs, standards covering the way product modules are designed, or the manner in which data is packaged and exchanged, help to smooth interactions between participants, and make the ecosystem work more efficiently. The ecosystem leader can also play a role in developing standardized interfaces to smooth the interactions between different partners. As we saw in in chapter 3, for example, DS developed a portal where all of its partners can access road maps for the different sectors to which it offers PLM solutions. Within its customized social network, DS then established standard protocols to enable the easy exchange of complex, three-dimensional designs. These standard interfaces and protocols helped reduce transactions costs, eliminate uncertainty, and increase flexibility. They helped partners avoid the need for developing complex agreements each time they altered their activities or roles. And when a situation was too complex to fit the standard protocols, and so required a nonstandard agreement, DS helped its ecosystem partners focus on crafting a contract that promoted collaborative activities and encouraged shared problem solving.
Penalizing Bad Behavior
Not every participant will live up to the standards that govern its ecosystem, so our final recommendation is that ecosystem leaders need to find ways to penalize those whose behavior threatens to undermine the ecosystem’s integrity, productivity and efficiency.
The problem is illustrated by Alibaba’s Taobao ecosystem. In its early days, it suffered from misrepresentation by sellers that deliberately misclassified their products in an attempt to get greater exposure. A seller of a computer mouse, for instance, would list it in a high traffic category such as men’s clothing to gain more visibility. To work efficiently and fairly, products and services must be properly categorized, so buyers can make purchase decisions from a comprehensive set of alternatives. Initially, Alibaba lacked a process for correcting categorization errors, and frustration and confusion built up among sellers. Alibaba needed a way of correcting errors and preventing unscrupulous sellers from gaming the system. Explains Taobao’s Zhang Yu: “The difficulty is that everyone in the ecosystem is trying to maximize their own interest, sometimes at the cost of others, which damages the fundamental principle of a healthy ecosystem, co-create and share. Therefore, there is a need for police, who can fight for the public interest.”20
Instead of dealing with the problem itself, Alibaba decided to leverage the ecosystem by recruiting new partners that could help. It introduced a categorization feedback-and-dispute resolution process, whereby a Market Judgement Committee would make all categorization decisions. Members of the committee were selected from qualified representatives of both buyers and sellers, and decisions were made by voting. By choosing this approach, Alibaba was able to enhance the effectiveness of each party and help the ecosystem become more robust.
Another type of destructive behavior that can plague ecosystems is free riding. Partners often need to make irreversible investments or other kinds of commitments based on the expectation that they will reap benefits in the future. The participants therefore face the moral hazard that some of their partners will try to ride on the backs of others, renegotiating the relationship after partners have already committed capital, effort, or technology.21 The ecosystem leader needs to control the risk of moral hazard by promoting transparency in interactions, and by imposing sanctions on, or even excluding, those who refuse to play fair. For example, one of the companies we studied had to exclude a significant development partner from the ecosystem after it found out that the latter had been consciously using software code from one of the other partners without having the permission to do so.
Balancing Innovation and Flexibility with Efficiency
In order to secure the competitiveness of your ecosystem against vertically integrated rivals and traditional supply chains, and to make sure that the benefits only an ecosystem can deliver aren’t outweighed by extra costs, you need to focus on creating efficient interfaces between the partners. This means answering the following questions:
1. Where it is appropriate to use contracts to govern relations in the ecosystem, are these sufficiently flexible, targeting high-level outcomes and performance standards rather than specific deliverables and actions?
2. Are these contracts perceived to be fair and do they facilitate shared problem solving, rather than simply acting as a safety-valve of last resort?
3. Have you constructed interfaces to enable efficient exchange of data between you and your partners as well as between partners? How much control do you need to exert over the data being exchanged?
4. When complex, tacit knowledge needs to be exchanged, have you put the tools, systems and structures in place to enable this, including mechanisms for high-touch human interaction where necessary?
5. Have you helped boost the efficiency of the ecosystem by putting in place mechanisms that help build trust between partners, including attracting reputable partners and encouraging them to put their brands and reputations on the line?
6. Have you established a set of minimum standards that all partners agree to abide by as a way of reducing risk and reducing the costs of transacting within the ecosystem?
7. What mechanisms have you put in place to penalize free-riders or other forms of bad behavior in the ecosystem?