Platform Revolution

In the first quarter of 2015, Brian P. Kelley, the CEO of Keurig Green Mountain coffee company, had some explaining to do. The company had just launched the Keurig 2.0, a next-generation coffee brewer touted as the future king of brewing. The reigning king, the Keurig 1.0, had become ubiquitous in homes, offices, and hotels, and Green Mountain’s expensive coffee cartridges had helped fuel the growth of what had been a regional brewer into a firm worth more than $18 billion.

Yet with the launch of Keurig 2.0, sales did not rise further. Instead, they fell by 12 percent.

The problem had its origins back in 2012, when key patents on Keurig’s coffee pod design expired. Taking advantage of this change, competing coffee makers began selling pods that were compatible with Keurig’s machines and significantly cheaper. These rivals were like extension developers offering new sources of value to Keurig users. Of course, their existence, and the competition they presented to the official Keurig pods, caused an erosion of Keurig’s market share.

To fight back, Keurig included in its brewer version 2.0 a scanning device that prevented the use of any pods not marked with a special proprietary emblem. Consumers were furious. Many denounced Keurig on shopping websites; thousands watched YouTube videos purporting to demonstrate how to hack the system, tricking it into accepting coffee pods unauthorized by Keurig. Buyers lamented the firm’s “ridiculous corporate greed” and bemoaned the fact that the Amazon rating system made it impossible for them to give the new Keurig a zero star rating.¹

By attempting to grab an even bigger share of the profits from its coffee-making platform, Green Mountain had angered its community and forfeited profits. The king of coffee had violated three fundamental rules of good governance:

Governance is the set of rules concerning who gets to participate in an ecosystem, how to divide the value, and how to resolve conflicts.² To understand good community governance is to understand the set of rules for orchestrating an ecosystem.³

Green Mountain failed at the task of ecosystem governance. The Keurig in all its incarnations is simply a product platform, a single one-sided market serving a community of coffee drinkers. It could be much more successful as a beverage ecosystem with other value-adding options, a range of vetted suppliers, and an array of other high-quality services that customers would appreciate. Instead, Green Mountain chose to exclude suppliers that its customers found valuable, eliminating variety and freedom of choice in order to retain control. It overreached for a share of the value produced by its system. And it unilaterally placed its interest ahead of all others. Keurig users were the losers—and soon enough, so was Green Mountain.

The goal of good governance is to create wealth, fairly distributed among all those who add value. As we saw in chapter 2, the new technology-driven communities known as platform businesses are creating a vast amount of new wealth outside the firm, and these external benefits must be designed and managed fairly. Because these value-creating networks grow faster outside the firm than inside, ruling the ecosystem wisely puts a premium on not ruling it selfishly.

If navigating the rules for governance is hard for one-sided platforms like the Keurig brewing system, it’s exponentially harder when platforms are multisided. After all, multisided platforms involve numerous interests that don’t always align. This makes it difficult for platform managers to ensure that various participants create value for one another, and it makes it likely that conflicts will emerge that governance rules must resolve as fairly and efficiently as possible.

This is a juggling act that even giants and geniuses often get wrong. Facebook, for example, has alienated users with its privacy policies.⁴ LinkedIn has angered its developers by turning off their access to APIs.⁵ And Twitter has expropriated technologies developed by other members of its ecosystem while permitting Twitter users to harass one another. As Twitter’s CEO Dick Costolo said, “We suck at dealing with abuse.”⁶

In the complexity of the governance issues they face, today’s biggest platform businesses resemble nation-states. With more than 1.5 billion users, Facebook oversees a “population” larger than China’s. Google handles 64 percent of the online searches in the U.S. and 90 percent of those in Europe, while Alibaba handles more than 1 trillion yuan ($162 billion) worth of transactions a year and accounts for 70 percent of all commercial shipments in China.⁷ Platform businesses at this scale control economic systems that are bigger than all but the biggest national economies. No wonder Brad Burnham, one of the lead investors at Union Square Ventures, responded to the introduction of Facebook Credits—a short-lived system of virtual currency for use in playing online games—by wondering what the move said about Facebook’s monetary policy.⁸ In a similar vein, we might ask: In choosing to apply unilateral software standards as opposed to multilateral standards (as we saw in chapter 7), what kind of foreign policy is Apple pursuing? Is Twitter following an industrial policy based on investment in “state-owned” services or one relying on decentralized development by others? What does Google’s approach to censorship in China tell us about the company’s human rights policy?

Like it or not, firms like these already serve as the unofficial and unelected regulators of millions of lives. For this reason, platforms have much to learn from cities and states, which have had thousands of years to evolve principles of good governance. Like the platform businesses of today, cities and states have long had to wrestle with the question of how best to create wealth and distribute it fairly. Increasing evidence suggests that just governance is a crucially important factor in the ability of a state to create wealth—even more important than such obviously valuable assets as natural resources, navigable waterways, and favorable agricultural conditions.

Consider the modern city-state of Singapore. In 1959, when Lee Kuan Yew became prime minister, it had almost no natural resources. For defense and clean water, it relied on the Federation of Malaya, the predecessor of the nation of Malaysia (established in 1963). Corruption was rampant. Per capita GDP hovered below $430.⁹ Ethnic strife between Malays and Chinese, religious strife between Muslims and Buddhists, and political strife between capitalists and communists hamstrung progress.

Lee Kuan Yew brought economic vitality to Singapore by changing its system of governance. Having studied at the London School of Economics and with a law degree from Fitzwilliam College, Cambridge, Yew introduced the British system of justice and rule by law. Then he attacked corruption. To make graft less attractive, he raised the salaries of civil servants to match those of similar workers in the private sector. New public servants had to wear white as a sign of purity when they assumed office. The anti-corruption rules were enforced strictly: the minister of state for the environment, a staunch supporter of Yew, committed suicide rather than face prosecution on bribery charges.¹⁰ Fairer, more open government was encouraged by the creation of multicultural councils empowered to give voice to religious and ethnic groups willing to work within the system. Singapore now boasts a government that ranks with those of New Zealand and the Scandinavian countries as one of the least corrupt in the world. This is significant, in part, because each 1 percent drop in corruption and misuse of public power for private gain is associated with a 1.7 percent rise in GDP.¹¹

Although Yew has been criticized in the West for throttling political dissent, the economic results produced by his good-governance campaign are impressive. By 2015, Singapore’s per capita GDP was $55,182, higher than that of the United States. During the fifty-five years from 1960 to 2015, Singapore’s annual growth rate was 6.69 percent, almost 2 percent higher than that of Malaysia, from which it separated in 1965.¹²

Similar evidence of the importance of good governance to wealth creation can be seen by comparing the GDP growth and innovation rates in communist East Germany and North Korea with those in their near-twin siblings, West Germany and South Korea.¹³ Good governance matters.

Good governance is important in both nation-states and platform businesses because absolutely free markets, in which people and organizations interact with no rules, restrictions, or safeguards, can’t always be relied upon to produce results that are fair and satisfactory to those involved.

An example can be seen on eBay, where some participants inevitably have greater knowledge, market savvy, and bargaining skill than others. In most cases, the interactions that result are basically fair, even when a particular interaction produces a “winner” and a “loser.” But sometimes, outcomes that appear manipulative or even deceptive occur. For example, a group of eBay members, noticing that some inexperienced sellers were prone to misidentifying goods—for example, by misspelling Louis Vuitton with one “T” or Abercrombie and Fitch as “Abercrombee” or “Fich”—began taking advantage of these errors as middlemen. They would actively seek out mislabeled items, which were usually languishing unnoticed on the auction site, snatch them up at bargain prices, and then resell them for a huge markup under their proper names.

In one celebrated example, the owner of an antique beer bottle that had been kept in his family for fifty years decided to list it on eBay. Unfortunately, the seller had no idea of the true value of his heirloom. The beer had been produced in the 1850s as the result of a brewing contest intended to provide “life-sustaining ale” to a crew navigating the Arctic in hopes of discovering the fabled Northwest Passage from the Atlantic to the Pacific. (At the time, it was believed—erroneously—that ale could prevent scurvy.)¹⁴ The expedition failed, but a little of the original brew survived, and by the time of the eBay sale, two bottles were known to exist, both eagerly sought after by beer collectors and history buffs.

Unaware of all this—and careless to boot—the seller listed the precious bottle on eBay under the heading “Allsop’s Arctic Ale—full and corked with wax seal,” and proposed an opening bid of $299. The brand name should have been spelled “Allsopp’s,” with an extra P—a tiny error, but one big enough to confuse the serious collectors who would have been interested in the item. A shrewd scavenger for mislabeled bargains spotted the listing and jumped in as the sole bidder. He bought the bottle for $304 and relisted it on eBay three days after receiving it. When collectors heard about it, the brew fetched bids exceeding $78,100.¹⁵

The Case of the Ambiguous Ale is an example of market failure—a situation in which “good” interactions (fair and mutually satisfactory) fail to occur, or “bad” interactions do. If you can’t find an item you want on eBay, then a good interaction has failed to occur. If you did find an item you wanted but got cheated, abused, or deceived, then a bad interaction has occurred. In general, there are four main causes of market failures: information asymmetry, externalities, monopoly power, and risk.

Information asymmetry arises whenever one party to an interaction knows facts that other parties don’t and uses that knowledge for personal advantage. Consider the problem of counterfeit goods, when the seller knows that the goods are fake but does not inform the buyer. Fakes account for Skullcandy headphones with lousy sound quality, Gucci handbags with broken stitching, Duracell batteries that don’t hold a charge, OtterBox mobile phone cases that are not drop-proof, and Viagra that provides no lift. Estimated at more than $350 billion worldwide, the size of the counterfeit goods market exceeds the trade in illegal drugs (estimated at $321 billion).¹⁶

Externalities occur when spillover costs or benefits accrue to anyone not involved in a given interaction. Imagine that one of your friends provides your private contact information to a gaming company in exchange for a few digital points. This would be a bad interaction because it violates your privacy rights, and so it is an example of a negative externality.

The concept of a positive externality is a bit more ambiguous. Consider what happens when Netflix analyzes the movie-watching behavior of someone whose tastes match yours and uses this data to give you a more accurate movie recommendation. This would be a positive externality, because it provides you with a benefit based on an interaction in which you’re not directly involved. Individuals who benefit from positive externalities aren’t likely to complain about them—but they are considered problems from a business design standpoint, since they reflect value that is not being fully captured by the platform. In an ideal world—at least, according to economic theory—every value created would somehow be accounted for and credited accurately to the entity responsible for it.

A concept closely related to the positive externality is the public good, whose value is not fully captured by the party that created it. Individuals generally create too few public goods unless some governing mechanism is designed to recognize and reward them.

Monopoly power arises when one supplier in an ecosystem becomes too powerful because of its control of the supply of a widely sought good, and uses this power to demand higher prices or special favors. At the height of its popularity (2009–10), game-maker Zynga became excessively powerful on Facebook, leading to conflicts over issues such as the sharing of user information, the split of gaming revenues, and the cost to Zynga of ads on the social network. eBay has experienced similar problems dealing with so-called power sellers.

Risk is the possibility that something unexpected and essentially unpredictable may go wrong, turning a good interaction into a bad one. Risk is a perennial problem in all markets, not just on platforms. A well-designed market generally develops tools and systems that serve to mitigate the effects of risk, thereby encouraging participants to engage in more interactions.

The literature on corporate governance is vast, especially in the field of finance. However, platform governance involves design principles that traditional finance theory overlooks. The single most heavily cited article on corporate governance is a literature survey that considers only “the ways in which suppliers of finance to corporations assure themselves of getting a return on their investment.”¹⁷ The focus here is on the information asymmetry arising from the separation of ownership and control—a critical element of governance design, but far from sufficient.¹⁸ Information asymmetry between the community of users and the firm also matters, and their interests too must be aligned.

Additionally, platform governance rules must pay special heed to externalities. These are endemic in network markets, since, as we’ve seen when examining network effects, the spillover benefits users generate are a source of platform value. Understanding this forces a shift in corporate governance from a narrow focus on shareholder value to a broader view of stakeholder value.

Market designer and Nobel Prize-winning economist Alvin Roth described a model of governance that uses four broad levers to address market failures.¹⁹ According to Roth, a well-designed market increases the safety of the market via transparency, quality, or insurance, thereby enabling good interactions to occur. It provides thickness, which enables participants from different sides of a multisided market to find one another more easily. It minimizes congestion, which hampers successful searches when too many people participate or low quality drives out high. And it minimizes repugnant activity—which explains why platform designers forbid porn on iTunes, human organ sales on Alibaba, and child labor on Upwork. According to Roth, good governance occurs when market managers use these levers to address market failures.

A broader view of platform governance uses insights borrowed from the practices of nation-states as modeled by constitutional law scholar Lawrence Lessig. In Lessig’s formulation, systems of control involve four main sets of tools: laws, norms, architecture, and markets.²⁰

A familiar example can be used to clarify these four kinds of tools. Suppose leaders of a particular ecosystem want to reduce the harmful effects of smoking. Laws could be passed to ban cigarette sales to minors or forbid smoking in public spaces. Norms—informal codes of behavior shaped by culture—could be applied by using social pressure or advertising to stigmatize smoking and make it appear “uncool.” Architecture could be used to develop physical designs that reduce the impact of smoking—for example, air filters that clean the air, or smokeless devices that substitute for cigarettes. And market mechanisms could be used by taxing tobacco products or subsidizing “quit smoking” programs. Historically, those who want to control social behavior—including platform managers—have employed all four of these tools.

Let’s consider some of the ways in which these four kinds of tools can be used by platform managers as part of a governance system.

Laws. Of course, many laws created and enforced by nation-states—laws, in the traditional sense of the term—apply to platform businesses and their participants. Sometimes, the application of such laws can be tricky. For example, legal sanctions that punish bad actors offer one traditional way of addressing the problem of risk. However, applying these sanctions requires determining who is responsible for problems that arise and who should bear the blame when they do—something that is not always simple or straightforward.

When it comes to platform businesses, this is far from a purely theoretical issue. We’ve mentioned some of the serious legal problems faced by platforms earlier: individuals who listed properties on Airbnb had their homes used for brothels and raves; people offering personal services on Craigslist have been murdered.²¹ Case law does not generally hold platforms accountable for misdeeds of platform users, even though the owners of the platform may be reasonably well positioned to regulate and control the behavior of the users. Thus, individual platform participants usually have to bear the downside risk, at least as far as national and local laws are concerned. (We’ll return to this topic in chapter 11, on regulation.)

Applying Lessig’s concept of “law” to governance within the platform business takes us into a different arena. The “laws” of a platform are its explicit rules—for example, the terms of service drafted by lawyers or the rules of stakeholder behavior drafted by the platform’s designers. These laws moderate behavior at both the user and the ecosystem level. For example, at the user level, the Apple rule that allows a user to share digital content among up to six devices or family members prevents unlimited sharing while nonetheless providing economic incentives for the purchase of Apple services and making a reasonable amount of sharing convenient.²² At the ecosystem level, the Apple rule compelling app developers to submit all code for review, combined with the rule that releases Apple from any duty of confidentiality, allows Apple to proliferate best practices.²³

Platform laws should be, and usually are, transparent. Stack Overflow, the most successful online community for answering programming questions, offers an explicit list of rules for earning points as well as the rights and privileges those points confer. One point confers the right to ask and answer questions. Fifteen points confers the right to vote up someone else’s content. At 125 points, you gain the right to vote down content—which also costs one point. And at 200 points, you’ve added so much value that you earn the privilege to see fewer ads. This system of explicit, transparent laws solves a public goods problem by encouraging members to share their best insights with everyone else on the platform.²⁴

An exception to the principle of transparency applies to laws that might facilitate bad behavior. Dating sites discovered this the hard way. When the sites applied laws that gave stalkers a quick “hand slap” when they misbehaved, the stalkers soon learned to avoid the specific trigger that flagged them. If, instead, the platform delayed the negative feedback, the stalker had a harder time learning how he’d been caught, which led to a more powerful and lasting disincentive.

Similarly, when a troll on a user-generated content site has his account deleted, he often returns under a new identity. Smart platform managers started making nuisance posts invisible to everyone but the troll. Unable to inflame community sentiment, the troll retreated.

The underlying principle: Give fast, open feedback when applying laws that define good behavior, but give slow, opaque feedback when applying laws that punish bad behavior.

Norms. One of the greatest assets any platform—indeed, any business—can have is a dedicated community. This doesn’t happen by accident. Vibrant communities are nurtured by skilled platform managers in order to develop norms, cultures, and expectations that generate lasting sources of value.

iStockphoto, today one of the world’s largest markets for crowdsourced photographs, was originally founded by Bruce Livingstone to sell CD-ROM collections of images by direct mail. As that business tanked, Bruce and his partners hated the thought that their work might go to waste. So they began giving away their images online.²⁵ Within months, they were discovered by thousands of people who not only downloaded images but asked to share their own images as well. To maintain the quality Bruce prided himself on and to eliminate spam, pornography, and copyright infringement, he made sure that every image was scrutinized by an iStockphoto inspector. This was a painstaking and costly process. Bruce found himself working sixteen-hour days.²⁶

Realizing that individual human inspection could not scale, Bruce turned to crowd curation. He devised a system under which people who uploaded quality content could earn their way to becoming inspectors and community organizers. Groups emerged to handle specific categories of images—for example, images linked to locations like “New York” or categories like “Food.” Bruce himself worked relentlessly to praise, give feedback, and build his community. Under the online name Bitter, he regularly posted comments on the platform’s homepage promoting members and their work, such as when he noted “great new stuff from Delirium, also tasty food series from Izusek.”²⁷

These efforts established a powerful set of norms that came to govern the iStockphoto community. These norms included feedback, high-quality content, open engagement, and a natural role progression to greater levels of authority. Applying these norms, the community produced a remarkable body of stock photos, a classic and valuable public good.

As the story of iStockphoto suggests, norms do not arise in a vacuum. They reflect behaviors, which means that they can be constructed through the intelligent application of the discipline of behavior design.

Nir Eyal, who has worked in both advertising and game development, describes behavior design as a recurring sequence of trigger, action, reward, and investment.²⁸

The trigger is a platform-based signal, message, or alert such as an email, a web link, a news item, or an app notification. This prompts the platform member to take some action in response. The action, in turn, produces a reward for the member, usually one with some variable or unanticipated value, since variable reward mechanisms like slot machines and lotteries are habit-forming. Finally, the platform asks the member to make an investment of time, data, social capital, or money. The investment deepens the participant’s commitment and reinforces the behavior pattern platform managers want to see.

Here’s an example of how the process works. Barbara is a member of Facebook. One day, an interesting photo shows up in Barbara’s news feed—perhaps a picture of a gorgeous sunlit beach in Maui, Barbara’s favorite vacation spot. This is the trigger. The action in response is designed to be as easy as possible (frictionless), which encourages Barbara to take the next step. In this case, the next step is clicking on the photo, which takes Barbara to Pinterest, the photo-sharing platform, which happens to be entirely new to Barbara. There she receives her reward: a variety of additional tantalizing, carefully curated photos selected specifically to engage her curiosity. (Imagine a photo collection titled “Ten Best Unknown Beaches of the South Pacific.”) Finally, Pinterest asks Barbara, whom they just rewarded, to make a small investment. She may be asked to invite friends, state preferences, build virtual assets, or learn new Pinterest features.²⁹ Any of these actions will set up a new set of triggers for Barbara and others, and the cycle starts over.

In the case of Pinterest, the norms fostered by this system of behavior design have produced a body of content that is a valuable public good. Of course, behavior design is not always used to benefit participants. It can also be used as a tool of selling and manipulation—which is one reason why users themselves should be aware of how such governance mechanisms work.

As a rule, it’s desirable to have users participate in shaping the systems that govern them. Elinor Ostrom, the first woman ever to receive the Nobel Prize in Economics, observed that successful creation and policing of public goods by communities follows several regular patterns. Clearly defined boundaries delineate who is and who is not entitled to community benefits. People affected by decisions regarding how community resources will be appropriated have recognized channels they can use to influence the decision-making processes. People who monitor community member behavior are accountable to the community. Graduated sanctions apply to those who violate community rules. Members have access to low-cost dispute resolution systems. And, as community resources grow, governance should be structured in nested tiers, with certain simple issues controlled by small, local groups of users, and increasingly complex, global problems managed by larger, more formally organized groups.³⁰ Norms that emerge in successful platform communities generally follow the patterns outlined by Ostrom.

Jeff Jordan, former senior vice president at eBay, recounts the problems the company encountered when it sought to add a fixed-price sale format to the traditional auction format.³¹ The two main categories of market participants reacted quite differently to the plan. Buyers liked the idea of fixed prices, but the sellers who paid eBay’s marketplace fees feared that fixed-price listings would effectively kill the golden goose of auction-driven price escalation.

The process Jordan used to resolve the dispute paralleled several of Ostrom’s ideas. eBay used focus groups and a “Voices” program to reflect user perspectives and gauge the strength of their feelings. Jordan’s team deployed thoughtful communications to alert buyers and sellers about proposed rule changes. They tested programs on smaller groups and pulled back if changes turned out badly. Ultimately, eBay’s leadership sided with the buyers, reasoning that the sellers would ultimately remain loyal to the platform because “merchants go where the consumers are.”³² The decision succeeded. Today, “Buy it Now” fixed-price listings account for about 70 percent of eBay’s $83 billion in gross merchandise volume.

Architecture. In the world of platform businesses, “architecture” refers basically to programming code. Well-designed software systems are self-improving: they encourage and reward good behavior, thereby producing more of the same.

Online banking platforms such as peer-to-peer lending businesses are using software algorithms to displace traditional, labor-intensive, and expensive loan officers. They calculate a borrower’s likelihood of repaying using both conventional data such as credit scores and nontraditional data such as a Yelp rating (for a restaurant), the stability of the borrower’s email address, her connections on LinkedIn, and even how thoroughly she interacts with the loan assessment tools before applying.³³ As the platform architecture becomes better at predicting borrower behavior, participation risk declines, attracting more lenders. Meanwhile, the low overhead costs allow the platform to offer lower rates, which attracts more borrowers. Greater participation further improves the data flow, and the cycle repeats.

No wonder peer-to-peer lending platforms like the British firm Zopa have been enjoying such notable success. When Zopa proudly announced that it had provided more than $1 billion in loans, one of this book’s authors, Sangeet Choudary, congratulated the company leaders and politely asked, “Wouldn’t your loan default rate be a more significant measure of success?” Zopa responded by publicizing the fact that its loan default rate had fallen to 0.2 percent from 0.6 percent three years earlier.³⁴ Such is the power of well-designed platform architecture.

Architecture can also be used to prevent and correct market failures. Recall the middlemen on eBay who took advantage of seller misspellings. Although one might lament the lost opportunity for the hapless sellers to complete the deal, these middlemen provided market liquidity (“thickness” in Alvin Roth’s formulation) through the process known as arbitrage. If no one bids on misspelled items, the interaction never happens—so arbitrageurs can be viewed as providing a valuable service. Yet the existence of arbitrage opportunities also highlights market inefficiencies. eBay now uses automated systems to provide spelling assistance, so sellers can have more confidence that they’ll receive what their items are worth. In a case like this, wise governance may disenfranchise a specific group of stakeholders, such as arbitrageurs, in order to increase the overall health of the ecosystem.

High-speed trading on the New York Stock Exchange offers another example. Firms like Goldman Sachs use supercomputers to determine when an order placed in one market will spill over to another market. Then they swoop in to intercept the deal, buying low, selling high, and skimming the margin. This methodology gives a few market participants who can afford massive computing power an unfair advantage over others.³⁵ Such asymmetric market power risks driving away players who feel cheated. To solve this problem, competing exchanges, such as the alternative trading system IEX, are using their own supercomputers to precisely time the order of bids, thereby eliminating the advantages of a Goldman Sachs.³⁶ Architecture can level the playing field, making markets more competitive and fair for all.

One of the most innovative forms of architectural control ever invented made its appearance in 2008, when an anonymous coding genius known as Satoshi Nakamoto published a paper on the Cryptography mailing list defining the Bitcoin digital currency and the so-called blockchain protocol governing it. Although Bitcoin is notable as the world’s first unforgeable digital currency that cannot be controlled by a government, bank, or individual, the blockchain is truly revolutionary. It makes possible fully decentralized, completely trustworthy interactions without any need for escrow payments or other guarantees.

The blockchain is a distributed public ledger that enables storage of data in a container (the block) affixed to other containers (the chain).³⁷ The data can be anything: dated proof of an invention, a title to a car, or digital coins. Anyone can verify that you placed data in the container because it has your public signature, but only your private key can open it to see or transfer the contents. Like your home address, a blockchain container is publicly, verifiably yours, but only people you authorize have a key that permits entry.³⁸

The blockchain protocol makes decentralized governance possible. Normally, when you sign a contract, you must either trust the other party to honor the terms or rely on a central authority such as the state, or on an escrow service like eBay, to enforce the deal. Public blockchain ownership empowers us to write self-enforcing smart contracts that automatically reassign ownership once contract terms are triggered. Neither party can back out because the code, running in a decentralized public fashion, is not under anyone’s control. It simply executes. These smart, autonomous contracts can even pay people for the output of their work—in effect, machines hiring people, not the other way around.

For example, imagine a smart contract between a wedding photographer and a couple planning their nuptials. The blockchain-stored contract could specify that payment of the final installment of the photographer’s fee will be made promptly when the edited photo files are delivered electronically to the newlyweds. The automatic digital trigger ensures that the photographer is incentivized to deliver the photos promptly, while it also relieves her of any concern that her clients might fail to pay.

Nakamoto’s invention has given birth to a new kind of platform—one with open architecture and a governance model but no central authority. Having no need for gatekeepers, it will put serious pressure on existing platforms that rely on costly gatekeepers. Financial services that claim 2–4 percent of transactions simply for passing them may in the future be hard pressed to justify their rake.

Furthermore, while most platforms address the problem of the market power of particular participants, Nakamoto’s platform addresses the problem of monopoly power of the platform itself. Not even Nakamoto, whose real identity remains a mystery, can remake the rules of the open source code to favor one participant over another.

Markets. Markets can govern behavior through the use of mechanism design and various incentives—not money alone, but the trifecta of human motivations that may be summarized as fun, fame, and fortune. In fact, on many platforms, money is far less important than the more intangible, subjective form of value known as social currency.

The idea behind social currency is to give something in order to get something. If you give fun in a photo, you can get people to share it. Social currency, measured as the economic value of a relationship, includes favorites and shares.³⁹ It also includes the reputation a person builds up for good interactions on eBay, good news posts on Reddit, or good answers on Stack Overflow. It includes the number of followers a user attracts on Twitter and the number of skill endorsements she garners on LinkedIn.

iStockphoto evolved a useful market mechanism based on social currency to manage exchange of photos. Every photo download cost the downloader one credit and earned one credit for the person who’d originally uploaded the photo.⁴⁰ Credits could also be purchased for 25 cents each, and photographers received cash payment for accumulated credits valued at $100 or more. This system created a fair social exchange that allowed professional photographers and non-photographers to participate in the same market. The mechanism simultaneously encouraged supply and market “thickness,” giving birth to the micro stock photo industry.

Social currencies have a number of remarkable and underappreciated properties. We can even use them to answer Brad Burnham’s interesting question about the “monetary policy” of a platform.

The enterprise management platform company SAP uses a social currency like that of iStockphoto or Stack Overflow to motivate developers to answer one another’s questions. Points earned when the employee of a development company answers a question are credited to a company account; when the account reaches a specified level, SAP makes a generous contribution to a charity of the company’s choice. The system has saved SAP $6–8 million in tech support costs, generated numerous new product and service ideas, and reduced average response time to thirty minutes from the one business day that SAP promises.⁴¹ SAP estimates that knowledge spillovers from these activities account for a half-million-dollar gain in annual productivity for a typical enterprise software partner.⁴²

Even more interesting, SAP has used the social currency supply to stimulate its developer economy in the same way as the Federal Reserve uses the money supply to stimulate the U.S. economy. When SAP introduced a new customer relationship management (CRM) product, it offered double points on any answer, code, or white paper relating to CRM. During the two-month duration of this “monetary expansion” policy, developers found gaps in the software and devised new features at a vastly higher rate.⁴³ Used as a money supply, the increased flow of social currency caused overall economic output to rise. In effect, SAP employed an expansionary monetary policy to stimulate growth—and it worked.

In addition to promoting economic growth, well-designed market mechanisms can incentivize the creation and sharing of intellectual property and reduce the riskiness of interactions on the platform.

Beautiful, useful ideas are public goods. Which raises the question: what is the optimal intellectual property policy for a platform business? If a developer working on a platform invents a valuable idea, who should own it, the developer or the platform? It’s possible to imagine arguments on both sides of the issue. Giving ownership to the developer provides incentives for idea creation. Giving ownership to the platform facilitates standardization and sharing, and enriches the platform ecosystem as a whole. State-mandated laws regarding patents and other forms of intellectual property protection are clumsy and expensive to enforce. A more elegant platform-based solution is required.

SAP has tackled this problem through two practices. First, it publishes an 18–24-month advance road map indicating what new products and services it intends to build to enhance its offerings to its corporate clients. This not only tells SAP’s outside developers what digital real estate will be available for their own innovations but also gives them up to two years before they face competition. The two-year window thus serves as a metaphorical patent period.⁴⁴ Second, SAP has made a policy of partnering with developers financially or buying them out at a fair price. This assures developers that they will be fairly compensated for their work, reduces partner risk, and encourages outside investment in the SAP platform.

The issue of risk reduction on a platform is a perennial one. History shows that platform owners generally seek to avoid responsibility for the risks faced by platform participants, especially in the early days of the platform. For example, in the 1960s, credit card companies, which host the two-sided merchant and cardholder platform, resisted insuring cardholders against fraud on their cards. They argued that insurance would promote fraud as consumers would become careless with their cards, and that banks forced to absorb more risk would become more reluctant to extend credit, hurting low-income consumers.

Over the vigorous objections of major banks, the Fair Credit Reporting Act (1970) and a subsequent amendment required fraud insurance, imposing a limit of $50 on consumer liability for fraudulent use of a credit card. The disaster predicted by the credit card companies did not occur. Freed from the fear of fraud, consumers used their cards so much more often that the increase in interaction volume more than offset the increase in fraud. The business benefit from fraud insurance is so powerful that, in order to encourage adoption and use, many banks now waive the $50 charge if consumers report a lost or stolen card within twenty-four hours.⁴⁵

In recent years, new platform businesses have made the same mistake as credit card companies did in the 1960s. Initially, Airbnb refused to indemnify hosts against bad guest behavior, and Uber refused to insure riders against bad driver behavior.⁴⁶ Eventually, both companies realized that this refusal was hurting the growth of their platforms. Today, as we’ve noted, Airbnb offers its host members $1 million in homeowners’ protection, and Uber is partnering with insurance firms to create new types of policies to protect its drivers.⁴⁷

Rather than seeking to minimize their own risk, platforms should use market mechanisms such as risk pooling and insurance to reduce risk for their participants and thereby maximize overall value creation. Good governance means looking after the health of one’s ecosystem partners.

Kings and conquerors like to make the rules; they don’t always like to abide by them. Yet results improve when smart rules of governance are applied to platform companies themselves as well as to platform partners and participants.

The first big principle of smart self-governance for platforms is internal transparency. In platform companies, as in virtually all organizations, there’s a tendency for divisions or departments to become “siloed”—to develop unique perspectives, languages, systems, processes, and tools that are difficult for outsiders to understand, even those in another department of the same company. This makes it extremely hard to solve complex, large-scale problems that span two or more divisions, since it means that members of different work teams lack a shared vocabulary and tool set. It also makes it much harder for outsiders—including platform users and developers—to work effectively with the platform management team.

To avoid this kind of dysfunction, platform managers should strive to give all their business divisions a clear view across the entire platform. Such transparency promotes consistency, helps others develop and use key resources, and facilitates growth to scale.

The so-called Yegge Rant, executive Steve Yegge’s attempt to summarize a mandate issued by Amazon’s Jeff Bezos, captures the spirit of this principle very effectively. Bezos insisted that all members of the Amazon team must learn to communicate with one another using “service interfaces”—data communication tools specifically designed to be clear, understandable, and useful to everyone in the organization as well as to outside users and partners. The idea is to treat everyone you do business with—including your colleagues in other departments and divisions of the organization—as customers with legitimate and important information needs that you are responsible to meet. Hence the seven rules presented in the Yegge Rant:

1. All teams will henceforth expose their data and functionality through service interfaces.

3. There will be no other form of interprocess communication allowed: no direct linking, no direct reads of another team’s data store, no shared-memory model, no back-doors whatsoever. The only communication allowed is via service interface calls over the network.

4. It doesn’t matter what technology they use. HTTP, Corba, Pubsub, custom protocols—doesn’t matter. Bezos doesn’t care.

5. All service interfaces, without exception, must be designed from the ground up to be externalizable. That is to say, the team must plan and design to be able to expose the interface to developers in the outside world. No exceptions.

Astute application of this principle of transparency underlies the success of Amazon Web Services (AWS), the platform’s giant cloud services company. Andrew Jassy, Amazon’s vice president of technology, had observed how different divisions of Amazon kept having to develop web service operations to store, search, and communicate data.⁴⁸ Jassy urged that these varied projects should be combined into a single operation with one clear, flexible, and universally comprehensible set of protocols. Doing this would make all of Amazon’s vast body of data accessible and useful to everyone in the organization.

Even more important, Jassy recognized that solving this problem for Amazon could have broader external applications. He reasoned that if multiple business units within Amazon had to solve this problem, then a reliable data management service that had solved the problem effectively might well be useful to outside firms with similar needs. AWS was born—one of the first businesses to offer cloud-based information storage and management services and expertise to companies with massive data challenges. Thanks to Jassy’s vision, AWS today has more market capacity than the next dozen cloud services combined.⁴⁹

By contrast, companies that limit their own ability to see across business units are likely to fail to establish viable platforms or to build their platform businesses to scale.

Sony Corporation provides a sobering example. The Sony Walkman had dominated portable music since the 1970s. In 2007, when the Apple iPhone was introduced, Sony’s dominance of the world of electronic devices seemed unshakeable. Sony had a world-class MP3 player, it had a pioneering e-reader, and it made some of the best cameras. In the fall of that year, Sony introduced the next-generation PlayStation Portable (PSP), the best gaming device in the world. Sony even owned Time Warner movie and television studios, giving it the opportunity to offer unique content. Yet despite these piecemeal advantages, Sony never developed the vision to offer a platform. Instead, the firm embraced separate product lines and focused on individual systems.

Sony’s siloed business vision prevented it from creating a unified platform ecosystem. Within a few years, Apple’s iPhone and the apps built on its growing platform had swept the field. Two years after the crash of 2008, Sony’s stock price was still almost one-third below its prior value, while Apple’s price had soared to historic highs.

The second big principle of platform self-governance is participation. It’s crucial for platform managers to give external partners and stakeholders a voice in internal decision processes equal to that of internal stakeholders. Otherwise, the decisions made will inevitably tend to favor the platform itself, which will eventually alienate outside partners and cause them to abandon the platform.

In their book Platform Leadership, Annabelle Gawer and Michael A. Cusumano offer a vivid example of how giving partners a voice can be great platform governance. The ecosystem built around the universal serial bus (USB), promoted by Intel, was one of the first standards to facilitate data and power transfer between peripherals—keyboards, memory devices, monitors, cameras, network connectors, and so on—and computers. Yet peripherals were outside Intel’s core microchip business.⁵⁰ This meant that Intel faced a particularly acute version of the chicken-or-egg launch problem that we discussed in chapter 5. No one wants to produce peripherals for a computer standard no one else has adopted, but no one wants to buy a computer for which no one else has made peripherals. And potential hardware partners were reluctant to link with Intel since, as the owner of the standard, Intel would have the option of making future changes to the standard that could render competing products incompatible, thereby capturing all the long-term value of partner investments.

Intel cracked the chicken-or-egg conundrum by entrusting USB to its Intel Architecture Labs (IAL) division. As a new business unit, IAL did not fall under the authority of any internal product line. Its job was to serve as a neutral negotiator between ecosystem partners and internal business units, which it could only achieve through independence. IAL earned the trust of partners by advocating and enacting policies that advanced the health of the ecosystem even at the occasional expense of Intel business units. Over the course of a year, the IAL team visited over fifty companies, inviting them to help define the standard and design licenses that made them comfortable. Through IAL, Intel also committed not to trample partner markets. Intel used both reputation and contracts to limit its own future behavior. (See the sidebar here for a summary of IAL’s self-governance principles.)

These efforts paid off. A consortium of seven companies—Compaq, DEC, IBM, Intel, Microsoft, NEC, and Nortel—united behind USB, producing an ecosystem standard that has evolved successfully for more than a decade.

This returns us to the deep design principle we introduced early in this chapter: Just and fair governance can create wealth. We saw this principal operate in the story of Singapore’s rise; we see it here in the story of IAL and the launch of the USB standard.

Fairness helps create wealth in two ways.⁵² First, if you treat people fairly, they are more likely to share their ideas. Having more ideas creates more opportunity to mix, match, and remake them into new innovations.

Second, one of this book’s authors (Marshall Van Alstyne) has shown formally that fair governance leads participants in a market to allocate their resources more wisely and productively.⁵³ Consider the USB standard. If each of the seven firms involved in creating the standard is sure to get a fair share of the value created, then each will willingly participate. By contrast, if five of the firms could gang up to steal the value of the other two—and those two know this could happen—the two may never join the coalition. This fracturing caused by the possibility of unfairness could have split the USB standard into competing standards, or, worse, prevented the development of any standard at all.

This is not to say that fairness always creates wealth or that wealth can never be created without it. Keurig, Apple, Facebook, and others have all, at times, treated their communities badly yet thrived financially. But in the long run, designing fair participation into ecosystem governance prompts users to create more wealth than if the rules grant a platform owner the ability to make arbitrary decisions without accountability. Many platform managers choose governance principles that favor themselves over their users. Yet platforms that respect their users more can expect more from their users—with benefits ultimately accruing to all.

Governance will always be imperfect. Whatever the rules, partners will find new forms of private advantage. There will always be information asymmetries and externalities. Interactions lead to complications, which lead to interventions, which lead to new complications. Indeed, if good governance allows third parties to innovate, then, as they create new sources of value, they will simultaneously create new struggles to control that value.

When such conflicts arise, governance decisions should favor the greatest sources of new value or the direction where the market is headed, not where it used to be. Firms that choose only to guard their aging assets, as Microsoft has done, stagnate. The governance mechanism must therefore be self-healing and promote evolution. Sophisticated governance achieves efficiency at the level of “design for self-design”—that is, it encourages platform members to collaborate freely and experiment fearlessly in order to update the rules as necessary. Governance should not be static. When signs of change appear on the horizon—such as new behaviors by platform users, unanticipated conflicts among them, or encroachments by new competitors—information about the change should spread rapidly through the organization, encouraging creative conversations about how the governance system may need to evolve in response.

No matter what kind of business or social ecosystem your platform inhabits, it will always include both fast-moving parts and slow-moving parts. Smart governance systems are flexible enough to respond to both.⁵⁴

TAKEAWAYS FROM CHAPTER EIGHT

Governance is necessary because absolutely free markets are prone to failures.

Market failures are generally caused by information asymmetry, externalities, monopoly power, and risk. Good governance helps prevent and mitigate market failures.

The basic tools for platform governance include laws, norms, architecture, and markets. Each must be designed and implemented with care in order to encourage platform participants to engage in positive behaviors, incentivize good interactions, and discourage bad interactions.

Self-governance is also crucial to effective platform management. Well-run platforms govern their own activities following the principles of transparency and participation.