Platform Revolution

Not long ago, one of the authors of this book—Marshall Van Alstyne—was approached by a pair of company founders on their way to a meeting with a group of venture capitalists. They’d created a new platform business that, for the purposes of this story, we’ll give the fictitious name of Ad World. They were hoping to impress the assembled VCs with an astute business plan and thereby win a significant offer for funding.

“Here’s the idea,” one of the founders explained to Marshall. “Our new platform Ad World will provide a listing service for firms to find ad agencies. We’ll make it easy for companies that are ready to begin new advertising campaigns to post requests for bids, and for ad agencies to post proposals and offers that companies can look at and respond to. It’s just like 99designs, which lets graphic artists link up with consumers who want help with artistic projects, only in the B2B space instead of the B2C space.”

“Here’s what we want to know,” the founder replied. “We’re pretty sure Ad World is going to provide value to our users and attract a fair amount of interest. But we’re wondering how to generate revenues from it. Should we charge the ad agencies to join the platform and put up their profiles? Should we charge the firms that are seeking services? Or should we charge for individual project listings? Or maybe for all three?”

“And we need an answer fast,” interjected his partner. “We need to figure out our strategy so we can run the numbers and make our business case to the VCs.”

The aspiring platform moguls gazed at Marshall with such sincerity that he almost hated to burst their bubble. But he had to do it. As gently as he could, Marshall replied, “You’ve listed three possible ways to monetize Ad World, and you’ve asked me to pick one of them—or maybe all three. My answer is, none of the above.”

The two founders in this story were smart, talented, thoughtful business leaders. They’d done a lot of homework on the nature of platform ecosystems. They did understand, in a general way, how platform businesses work and the challenges of attracting both sides of the market to make a robust set of interactions possible. But when it came to monetization, they were asking the wrong questions.

These founders shouldn’t charge either side to be listed on their platform. Doing so would put terrible friction on entry into the ecosystem, discouraging many potential participants from becoming users. To charge for posting a deal simply means people post fewer deals. That’s bad. It reduces potential interaction volume, not to mention realized interactions. As a result, it also reduces the volume of data available to the platform—data desperately needed to enable the platform to forge powerful matches between consumers and producers.

In fact, rather than charging users to join the platform, the founders should be subsidizing their participation—perhaps by providing tools and services to make it easy, fast, and effective for them to complete their profiles.

This wasn’t a complete surprise to the founders. They had partly intuited this, as Marshall could tell by the fact that they’d used “scrapers”—automated software tools for collecting data from the Internet—to produce user profiles. They understood that building a base of users was their first and biggest challenge, and that creating friction around the process by charging money for it would be a serious mistake.

How, then, could the founders monetize their platform model? The answer: They can charge users for the value they accrue from the ecosystem, but the charge should be levied on deal completion, not at the time of listing. They would make it possible for firms to post a deal risk-free by charging a fee only once the firms get what they need. The fee becomes performance-based, and it feels negligible because it simply skims off a small fraction of a transaction that’s occurring anyway.

What’s more, the best strategy may be one the founders didn’t even consider. Why not charge ad agencies for a service that helps them do a postmortem to discover why they lost a deal? Not only would this fee create no friction on a deal, but it would reflect the value of the feedback provided; it would be recurring revenue rather than one-time revenue; and it could help ad agencies improve the quality of their offerings, thereby encouraging a rise in the value of interactions over time.

The story of this fledgling platform and the strategic challenge its founders faced illustrates some of the complexity of platform businesses—as well as the creative thinking platform managers need to practice if they are going to realize the full value-creating potential of the ecosystems they are building. Monetization, in fact, is one of the most difficult—and fascinating—issues that any platform company must address.

As we’ve explained, the inherent value of a platform business lies chiefly in the network effects it creates. But monetizing network effects poses a unique challenge. Network effects make a platform attractive by creating self-reinforcing feedback loops that grow the user base, often with minimal effort or investment by the platform manager. Higher value creation by producers on the platform attracts more consumers, who, in turn, attract more producers and further value creation.

Yet, ironically, this powerfully positive growth dynamic makes monetization very tricky. Any charge levied on users is likely to discourage them from participating on the platform. Charging for access may lead people to avoid the platform altogether; charging for usage may inhibit frequent participation; charging for production reduces value creation, making the platform less attractive to consumers; and charging for consumption reduces consumption, making the platform less attractive to producers. This was the precise dilemma that the founders of Ad World were grappling with.

How, then, do you monetize a platform without damaging or even destroying the network effects you’ve worked so hard to create?

Some students of platform businesses leap to the assumption that the collaborative nature of value creation on the Internet means that the natural price for goods and services distributed online must be free. But of course a business that charges nothing for the benefits it offers is unlikely to survive for very long, since it will not produce the resources needed to maintain or enhance it, and investors will have no incentive to provide the capital needed for its growth.

Some element of free pricing can be useful in building network effects for a platform business. But it’s important to understand the different models in which partially free pricing can power growth. As every business student learns, the model behind the safety razor business that entrepreneur King Gillette founded in 1901 involved distributing the razors for free—or at a very low, subsidized cost—while charging for the blades.

As it happens, research by Randal C. Picker of the University of Chicago Law School has called into question the traditional story of Gillette and the razors-and-blades pricing strategy. Picker found that the timing of price changes for Gillette razors and blades, as well as the expiration date on the patent covering Gillette’s unique razor design, seems to undermine the notion that the company was employing the razors-and-blades strategy as commonly understood.¹ Nonetheless, the familiar story remains a handy symbol for a strategy that has been used in a number of markets, including, for example, the printer market, where sales of high-priced toner cartridges generate profits that the relatively cheap printers themselves don’t produce.

Another version of this strategy is the freemium model, in which a free layer of service attracts users who eventually pay for an enhanced version. Many online service platforms, including Dropbox and MailChimp, work this way. Both the razors-and-blades model and the freemium model monetize the same user base, or portions thereof.

Platforms may also offer free or subsidized pricing to one user base while charging full price to an entirely different user base. This makes the design of monetization models more complex, since the platform must ensure that the value it gives away to one side can be used to capture value on the other side. Significant scholarly work has been done in this area. Two of this book’s authors (Geoff Parker and Marshall Van Alstyne) were among the first scholars to lay out the theory of two-sided market pricing.² And the theory was mentioned as part of the 2014 Nobel Prize awarded to one of the other originators of two-sided market economics, Jean Tirole.³

Achieving the right balance among the complex factors involved in two-sided market pricing isn’t easy. Netscape, one of the pioneers of the Internet era, gave away browsers for free in hopes of selling web servers. Unfortunately, there was no proprietary connection between browsers and servers that Netscape could reliably control. Anyone could just as easily use Microsoft’s web server or the free Apache web server, which meant that Netscape was never able to monetize the other side of its free browser business. As this example illustrates, platform businesses that intend to use free pricing as part of their strategy need to ensure that the value they create and hope to ultimately monetize is fully controlled by the platform.

Meeting the monetization challenge must begin with an analysis of the value created on the platform. Traditional non-platform businesses—pipelines—deliver value to their customers in the form of a product or service. They may charge for ownership of the product, as Whirlpool does when it sells a dishwasher, or for utilization of the product, as GE Aviation charges for installation and regular servicing of its aircraft engines.

Like Whirlpool and GE, platform companies are engaged in designing and building technology. But rather than putting the technology in the hands of customers in exchange for a fee, they invite users to join the platform—and then they seek to monetize the platform by charging for the value that the platform technology creates for those users. This value falls into four broad categories:

• For consumers: Access to value created on the platform. Video viewers find the videos on YouTube valuable; Android users find value in the various activities made possible by the apps; students on Skillshare find value in the courses made available through the site.

• For producers or third-party providers: Access to a community or market. Airbnb is valuable for hosts because it provides access to a global market of travelers. Company recruiters find LinkedIn valuable because it enables them to connect to potential job-seekers. Merchants find Alibaba valuable because it enables them to sell their products to customers around the world.

• For both consumers and producers: Access to tools and services that facilitate interaction. Platforms create value by reducing the friction and barriers that prevent producers and consumers from interacting. Kickstarter helps creative entrepreneurs raise money for new projects. eBay, combined with PayPal, allows anyone to start an online store that serves customers anywhere in the world. YouTube allows musicians to provide their fans with performance videos, without having to produce physical products (CDs or DVDs) and without having to sell through intermediary retailers.

• For both consumers and producers: Access to curation mechanisms that enhance the quality of interactions. Consumers value access to high-quality goods and services that address their specific needs and interests, while producers value access to consumers who want their offerings and are willing to pay a fair price for them. Well-run platforms build and maintain curation systems that connect the right consumers with the right producers quickly and easily.

These four forms of value wouldn’t exist without the platform, and so they may be described as sources of excess value that the platform generates. Most well-designed platforms create far more value than they directly capture—which is why they attract large numbers of users, who are happy to enjoy the benefits of all the “free” value provided by the platform. A smart monetization strategy begins by considering all four forms of value, then determines which sources of excess value can be exploited by the platform without inhibiting the continued growth of network effects.

Founded in 2005 by Ethan Stock, Zvents was originally an online guide to local events in the Bay Area. It grew rapidly, expanding beyond California and becoming the largest site of its kind, serving hundreds of markets and attracting over 14 million visitors every month. It was a hit with both producers—the local event organizers who posted their concerts, shows, fairs, festivals, and other activities on the site—and consumers, who logged on to Zvents to find something fun to do after work and on the weekend.

Stock seemed to be living the Silicon Valley dream. Having built a platform that millions of people had come to rely on, his only remaining challenge was to monetize it. But that proved to be far from simple.

“Once we reached critical mass,” Stock recalls, “and it was clear we were becoming the market leader, we expected event organizers would start paying. … But there is a fatal flaw in some businesses that can hogtie their ability to make money—the expectation of completeness.”

The problem was that consumers who visited Zvents expected to find a comprehensive listing of local events. If only a few of the available options were included, the interest of users would quickly evaporate. Which meant that Zvents didn’t have much leverage over the event organizers they hoped to charge. If they’d threatened organizers with having their listings pulled, the threats would have no teeth, since the whole value of Zvents lay in the completeness of its listings. Charging producers for access to the platform wasn’t going to work.

Zvents experimented with another monetization technique that we’ll discuss later in this chapter—charging producers for enhanced access. They did manage to get a few organizers to pay for more prominent listings, but the value of the enhancements in terms of improved attendance or ticket sales proved small. Zvents ended up with a tiny trickle of revenue rather than the gusher they’d hoped for. In June 2013, with his hopes of building a lucrative platform empire to rival Google or Facebook dashed, Stock sold his company to eBay, which now uses Zvents as a bulletin board for arts and entertainment events in conjunction with its StubHub ticket resale platform.

The lesson? Network effects as measured by numbers of visitors alone don’t necessarily reflect the monetary value of a platform. The interactions facilitated must generate a significant amount of excess value that can be captured by the platform without producing a negative impact on network effects. When that’s not the case, monetization may not be possible.

The paradoxical relationship between network size and monetization potential doesn’t stop there. In some cases, the ability to monetize a platform may actually increase dramatically when the number of users declines—reflecting the power of negative network effects to impact the value of a platform.

Meetup was launched in 2002 as a way for groups to organize offline meetings (“meetups”) by connecting online. Its cofounder, Scott Heiferman, said he was inspired by the way people in New York came together as a community following the 9/11 terror attacks.

Meetup gained traction as a free platform, but the dot-com bust of the late 1990s served as a constant reminder to its managers of the need to develop a credible monetization model. They first tried to generate income by using lead generation, charging fees to offline venues such as restaurants and bars for the number of users that came in during a meetup. However, in a pre-smartphone world, this monetization model didn’t work very well. The number of people who actually showed up for events was different from the number who signed up, and Meetup had no way of counting heads to determine an appropriate fee.

Meetup abandoned the lead generation model and experimented with other ways of monetizing its service. It tested advertising, but failed to attract significant numbers of advertisers. It tried offering a premium product called Meetup Plus, but the additional value provided generated little interest. (Perhaps understandably: years later, when asked in an interview to explain the extra service included with Meetup Plus, Heiferman responded with a laugh, “God, I don’t even remember what it got you. It got you some kind of features where you would be able to … I don’t know. I can’t remember.”) Meetup even tested charging fees to political organizations, which were a growing portion of the platform’s user base, but this, too, produced only modest revenue. Meetup’s options were running out.

Meanwhile, Meetup found itself confronted with another problem—one that, paradoxically, would help save the company. This was the growth of negative network effects. As the platform grew, with low barriers to planning a meetup, many meetups were being started without a clear purpose or adequate planning. There was a lot of noise on the platform, leading to disappointing experiences for users, who would sign up for a meetup only to discover that attendance was small and activity minimal.

Meetup’s leaders made a risky decision. They decided to start charging meetup organizers, despite the potential for drastically diminishing the scale of the platform and weakening its network effects. They reasoned that charging organizers would help them solve their monetization problem while weeding out organizers who weren’t serious about their goals. A letter went out to all organizers informing them that they would henceforth be asked to pay $19 per month for the right to keep using Meetup’s service.

The backlash was huge. After Businessweek ran an article on Meetup’s new strategy, the magazine received emails from countless platform users predicting the service’s demise. One user from London wrote, “I think it’s fair to say most organisers were shocked, and most of the ones I’ve spoken to will simply cease organising for their groups … There isn’t anything Meetup is doing these days that users can’t simply do on their own and more effectively, and there’s plenty of open source software to make use of and create your own website.”⁴

But despite the backlash, the strategy worked. The number of meetups promoted on the site fell drastically, but their quality, and therefore the quality of the interactions generated, improved significantly. As Heiferman explained in an interview five years later, “The big headline, by the way, in going from free to fee for us is: Yeah, we lost 95% of our activity but now we have much, much, more going on than we ever did before and half the Meetups are successful, as opposed to 1–2% being successful.”⁵

As we’ve discussed, a platform’s goal is not simply to pump up the numbers of participants and interactions. It must also take steps to encourage desirable interactions and discourage undesirable ones. Meetup’s monetization model helped it achieve exactly that. By discouraging organizers who weren’t serious about their objectives, the pricing mechanism created a culture of quality on the platform.

It’s a mistake to assume that network effects can always be optimized by simply refraining from charging users. A better approach to analyzing the monetization challenge is to ask these questions: How can we generate revenues without reducing our positive network effects? Can we devise a pricing strategy that strengthens our positive network effects while reducing our negative network effects? Can we create a strategy that encourages desirable interactions and discourages undesirable ones?

To begin exploring some of the ways that an effective monetization strategy can be developed, let’s look back at the four forms of excess value created by platforms—access to value creation, access to the market, access to tools, and curation. All four culminate in some kind of interaction. In many cases, this interaction involves an exchange of money, as when an Uber customer pays a driver for a ride, an eBay buyer pays a seller for a product, or a company using Upwork pays a freelancer for a completed project. Platforms that facilitate such monetary transactions can monetize the value created by charging a transaction fee, which may be calculated as either a percentage of the transaction price or a fixed fee per transaction. The latter system, which is simpler to administer, is particularly appealing when a high frequency of transactions is expected without a significant variation in the transaction size.

Charging a transaction fee is a powerful way of monetizing the value created by the platform without hampering the growth of network effects. Because buyers and sellers are charged only when an actual transaction occurs, they are not discouraged from joining the platform and becoming part of the network. Of course, if the transaction fee is excessive, it may discourage transactions. Platform managers may need to experiment with various levels of fee to find the rate that captures a fair percentage of the value created without driving users away.

A more serious and persistent challenge is to capture on the platform itself all the interactions facilitated by the platform. Buyers and sellers who find each other on the platform are naturally incentivized to take the interaction off the platform if they can, in order to avoid paying the transaction fee.

This problem is especially rampant with platforms that connect service providers with service consumers. With the rise of the freelancer economy and the spread of the online sharing economy, platform businesses from Airbnb and Uber to TaskRabbit and Upwork have sprung up to facilitate service interactions. However, most of them are faced with the challenge of capturing the interaction on-platform. In most cases, the interaction can’t occur until the producer (in this case, the service provider) and the consumer (the purchaser of the service) agree on the terms of the service, which usually requires the two to interact directly. Furthermore, the actual exchange of money often follows the delivery of the service, which also requires the two participants to interact directly. These direct interactions weaken the platform’s ability to capture value by creating an opportunity for the parties to make a deal off-platform. As a result of avoiding the transaction fee, the consumer can obtain a discount on the service, while the provider gets to keep more of the total service charge. The only loser is the platform company itself.

Platforms like Fiverr, Groupon, and Airbnb solve this problem by temporarily preventing participants from connecting. These platforms try to provide all the information a consumer needs to make an interaction decision, without connecting the consumer directly with the producer. Groupon does this by featuring services that are largely standardized, while the less-standardized Airbnb and Fiverr provide rating mechanisms and other social metrics that indicate the reliability of a service provider, making direct contact between the parties less necessary.

Sometimes strategies like these are insufficient. This is especially the case with platforms that create a market for professional services, which often require discussions, exchanges, and workflow management before and during the provision of services. As a result, it may not be possible for the platform to retain control of all communications between the producer and the consumer, and charging the consumer ahead of the interaction may not be an option.

In cases like these, the platform must extend its role as an interaction facilitator to include more value-creating activities. For example, Upwork provides tools for monitoring the service provider remotely. This enables consumers of professional services to monitor projects and make payments based on actual delivery of work.

The Clarity platform for connecting advice seekers with experts retains control of the interaction through a similar mechanism. In the past, expert-matching platforms connected the two sides, charged a lead generation fee, and allowed the transaction to occur off-platform. Clarity provides additional call management and invoicing capabilities that serve to capture the interaction on the platform. To benefit producers, Clarity offers integrated payments and invoicing, making it simple for advice givers to generate income through small, one-off engagements. To benefit consumers, the call management software provides per-minute billing, which gives them the option to opt out of a call that isn’t proving useful. Both sides receive enough additional value to keep them connected to Clarity, with minimal incentive to take their interaction off-platform.

As these examples illustrate, service provider platforms that want to capture and monetize interactions must create tools and services that benefit both parties by removing friction, mitigate risk, and otherwise facilitate interactions.

However, additional benefits like these may not be sufficient to enable all service provider platforms to thrive. Local service platforms that connect consumers with relatively simple services like plumbing and house painting continue to struggle with the challenge of owning the interaction. The risks involved in such interactions are lower than when hiring a professional freelancer: the two sides get to meet in person, the work is simpler, its quality is less variable, and, since the work itself happens off-platform, usually under the direct supervision of the consumer, the consumer can monitor the service provider without having to rely on software tools. Some of these local service platforms may need to move to the enhanced access monetization model, which we’ll describe in a later section of this chapter.

In some cases, it’s possible to monetize a platform by charging producers for access to a community of users who have joined the platform not in order to interact with producers but for other, unrelated reasons.

Dribbble has rapidly gained prominence in the design community as a high-quality platform for designers—artists, illustrators, logo creators, graphic designers, typographers, and others—to show off their work, thereby gaining exposure, credibility, and valuable feedback from their peers. Adapting the jargon of basketball, Dribbble users call new images “shots,” image groups “buckets,” and reposts of favorite images “rebounds.” This unique lingo has helped foster a highly engaged community that includes many of today’s best designers.

Dribbble’s managers are eager to protect the long-term value of this specialized community. For this reason, they do not charge members for access to the platform, which could weaken network effects. They’ve also opted not to permit sponsored images that provide enhanced access to the community (for example, by popping up, uninvited, on users’ home pages), since these could reduce the prestige of the site and its perceived value to its users. (We’ll describe the enhanced access strategy in more detail a bit later in this chapter.) So to monetize the site, Dribbble has invited third parties to pay for access to the community. In this case, companies looking for designers are charged to post employment listings on the site’s jobs board.

This form of monetization creates interactions that benefit both parties. Designers are motivated to put up their best work on Dribbble, since these may generate leads to new gigs, while companies get access to top-flight designers whose portfolios have already been curated by the creative community.

Dribbble’s monetization method can be described by the simple shorthand term “advertising.” But notice that, unlike most advertising, Dribbble’s highly targeted job listings generate value for the community, enhance the core interaction, and strengthen network effects rather than adding noise and depleting value.

In a similar fashion, LinkedIn allows recruiters to present job opportunities to its members and offers companies the ability to compare and target professionals based on their résumés and professional brands. LinkedIn’s power as a recruiting platform encourages users to update their profiles more often, thereby keeping the platform active and healthy.

As we’ve observed throughout this chapter, a monetization model is sustainable only when it strengthens network effects (rather than weakening them). Charging third-party producers for community access is effective if and only if the newly added contents—such as the job listings on Dribbble—enhance the value of the platform to its users.

Sometimes a platform that facilitates a monetary transaction may be unable to own, and hence to monetize, the transaction. Such platforms may instead charge producers for enhanced access to consumers. This refers to the provision of tools that enable a producer to stand out above the crowd and be noticed on a two-sided platform, despite an abundance of rival producers and the resulting intense competition to attract consumer attention. Platforms that charge producers fees for better targeted messages, more attractive presentations, or interactions with particularly valuable users are using enhanced access as a monetization technique.

The system of monetizing enhanced access generally doesn’t harm network effects, since all producers and consumers are permitted to participate in the platform on an open, non-enhanced basis. But those for whom the additional value of enhanced access is particularly great can pay for that extra value—allowing a portion of that value to be captured by the platform business.

The traditional classified advertisement model, for example, supported local newspapers for decades through paid placement of commercial messages. Today, online platforms use a similar model, asking producers to pay for more prominent placement of their messages. Yelp, for example, offers restaurants enhanced visibility and better branding on its platform by charging for a premium listing in search results. Restaurants pay for this service, since it makes it easier for them to break through the noise and attract the attention of the most valuable potential consumers.

Google search may also be seen through a similar lens. Every website publisher can achieve higher placement for its site through search engine optimization, a self-managed website design and coding process that produces no revenue for Google. However, some publishers choose to buy premium placement through Google Adwords. In a similar fashion, Tumblr, a micro-blogging platform acquired by Yahoo in 2013, allows users to promote their posts to a larger audience for a fee. Twitter, too, promotes sponsored content at the top of its feed.

Another way of monetizing enhanced access is by charging users for lowering barriers that otherwise exist between users. Dating websites, for example, often allow men to see profiles of women without revealing identifying details. Users who pay a subscription fee are allowed access to additional information that enables them to connect directly with other users who interest them.

Monetizing enhanced access to users needs to be done with care. If not done right, it can increase the noise level on the platform and decrease the relevance of content for consumers, leading to negative network effects, as described in chapter 2.

One important principle is to ensure that consumers can easily distinguish between content that has been elevated or highlighted as part of a paid access program and content whose high ranking or prominence is organic. Premium listings on Yelp and ads associated with Google’s search results look different from organic results, creating a sense of transparency that enhances user trust. Most search engines before Google that failed to follow this principle ended up confusing and annoying users and damaging the value of their platforms. So-called native advertising techniques, where paid content on the Internet is designed to resemble unpaid content, run the risk of appearing deceptive and alienating users.

Platform managers must also be careful not to allow monetization of enhanced access to create the impression that user access is being restricted. As the world’s largest social network, Facebook creates tremendous value for brands that want to engage with current and potential customers. Some consumer brands have grown massive followings on Facebook. However, during 2014 and 2015, Facebook was widely criticized for making curation changes that limit the reach of brands on the platform—except for those that pay extra for access to a wider audience. The perception was that Facebook was reducing services available to platform participants in order to facilitate a revenue grab. Facebook’s massive size and the powerful network effects it generates have enabled it to shrug off these complaints—at least, so far. But few other platforms would be able to get away with actions like these.

Finally, platform managers must ensure that their usual curation principles are applied rigorously to content from producers who pay for enhanced access. Facebook’s value is based on the relevance of its news feed, and a torrent of sponsored posts that lack such relevance could eventually drive consumers away from the platform.

When we think of network effects, we often assume that more is better. However, as we noted in chapters 2 and 3, positive network effects are driven not simply by quantity but also by quality. When the quantity of content on a platform becomes overwhelmingly great, consumers may find it increasingly difficult to find the high-quality content they want, thereby reducing the platform’s value for them. When this happens, consumers may be willing to pay for access to guaranteed quality—in other words, for enhanced curation.

Sittercity, a platform we’ve mentioned elsewhere in this book, charges parents to access the platform. To ensure quality and choice, it performs rigorous curation and screening of the babysitters who get access to the platform—a source of significant additional value to parents who worry about the well-being of their children. This extra value allows Sittercity to charge a subscription fee to parents instead of the transaction fee usually associated with a service provider platform.

As an advisor at Skillshare, an education platform, Sangeet Choudary has helped the platform transition from a transaction-fee-only model to a model that provides enhanced value upon payment of a subscription charge. Skillshare allows students to pay separately for each course taken. But once the platform managers had curated a sizeable amount of high-quality courses, Skillshare began to allow students access to multiple courses via a monthly subscription fee. Teachers are paid “royalties” based on the number of subscription-paying students who sign up for their classes. The growing number of students who choose this model get better value per course consumed, while generating recurring revenues for the platform.

A typical platform supports multiple types of users performing multiple roles. Given the differences among the users, their economic status, their motivations, their objectives, their incentives, and the differing forms and amounts of value they derive from the platform, decisions about whom you should charge and whom you should not can be complex—especially since every decision you make about one user category impacts others in ways that may not be obvious.

However, the general goal of encouraging positive interactions that will create value for all participants—together with observations from the histories of successful platform businesses—enables us to develop a few useful heuristics about when particular pricing choices are appropriate and when they are not.

• Charging all users. As we’ve noted, platform businesses rarely charge all their users the way pipeline businesses generally do. Charging all users would, in most cases, discourage participation, thereby reducing or destroying network effects. However, in a few cases, charging all users actually enhances network effects. In the offline world, for instance, prestigious membership organizations like country clubs charge all members. High membership dues (along with vetting processes such as requiring recommendations from existing members) serve as curating techniques to guarantee member quality. Some online platforms use this model—for example, Carbon NYC, a platform for multimillionaire residents of New York City. However, in many social and business settings, “willingness to pay” and “quality” are far from synonymous, so this pricing system must be used very carefully and selectively.

• Charging one side while subsidizing another. Some platforms are able to charge members of one category of users (call them A) provided they allow members of another category of users (B) to participate for free—or even subsidize or incentivize them. This works when users from category A highly value the opportunity to make contact with users from category B—but the feeling is not equally reciprocated. As we’ve noted, bars and pubs in the offline world have long used this strategy by offering women free or discounted drinks on Ladies’ Nights. Many online dating websites follow a similar strategy, incentivizing memberships for women as a way of attracting male members who will pay full freight.

• Charging most users full price while subsidizing stars. Certain platforms choose to subsidize or incentivize stars—super-users whose presence attracts large numbers of other users. In offline business, malls have been known to offer attractive lease terms to popular large retailers like Target, whose presence guarantees the customer traffic that other mall occupants will readily pay a premium for. In a similar way, online platforms like Skillshare and Indiegogo go to great lengths to court celebrity teachers and campaign creators, whose star power attracts other producers as well as large numbers of curious consumers. Microsoft learned this lesson when creating its Xbox gaming platform. Its initial monetization strategy paid game developers (producers) a one-time purchase fee while channeling ongoing user fees to Microsoft. But superstar game developer Electronic Arts refused to work on these terms and threatened to develop for Sony instead. Microsoft eventually had to succumb and agree to special terms for EA, though the details have not been publicly disclosed.

• Charging some users full price while subsidizing users who are price-sensitive. The category of users that is most sensitive to pricing is more likely to abandon the platform when charged, killing the network effect. Thus, it generally makes sense to discount or subsidize users who are price-sensitive while charging others the full freight. Real-world experience shows that it can be difficult to predict which side of a platform market is likely to be more price-sensitive. During the 1990s, the Denver real estate market experienced a glut of properties, making property owners desperate for rentals. Real estate agents charged owners broker fees, while tenants paid nothing. By contrast, during the same period, Boston had a scarcity of available properties, making would-be renters desperate to find places to stay; real estate agents charged potential tenants, while allowing owners to list their properties without paying a fee.

As you can see, deciding whom to charge is a delicate balancing act. The need to monetize the platform must be carefully weighed against the friction invariably produced by imposing a cost. Deciding precisely where the system can afford to create some friction—and how much friction can be tolerated without crippling the growth of network effects—is no easy matter.

Sometimes, a less-than-optimal monetization strategy can be made workable through ingenuity. During its early years, Alibaba, the e-commerce platform company that has been called China’s eBay and Amazon rolled into one, was unable to charge transaction fees simply because its primitive software had trouble tracking the flow of online deals. CEO Jack Ma was forced to charge membership fees instead—an option he would have preferred to avoid because of the entry friction it creates. Alibaba managed to overcome this problem by offering sizeable commissions to salespeople who convinced others to sign on to the platform. When word got around that some Alibaba sales agents were earning commissions in excess of a million Chinese yuan, equivalent to well over $100,000, the drive to enlist members shifted into high gear despite the friction caused by the entry fees. To this day, Alibaba charges no transaction fees, having managed to monetize the platform through advertising—as if a company that facilitated transactions, like Amazon and eBay, were to generate its profits by selling ads like Google.

As many of the cases mentioned in this chapter suggest—and as plenty of familiar real-world examples illustrate—the imperative to create and grow network effects often leads platform founders to begin by offering their services for free. Creating value for users while asking nothing in return is often a great way to attract members and encourage participation. “Users first, monetization later,” as the slogan goes. Or, in a variant that we heard from the executive in charge of platform strategy for Haier Group, a Chinese manufacturing company, “You never take first money.” In other words, only after a value unit has been created and exchanged with results that are satisfactory to both the producer and the consumer should the platform business itself seek to capture a share of that value.

A number of highly promising platforms have foundered because they ignored this rule and instead rushed to monetize their offerings prematurely. Sean Percival, the former vice president of online marketing for the failed social network Myspace, recalls the dire effect of financial pressures following the acquisition of the platform by Rupert Murdoch’s News Corporation. The “last nail in the coffin,” Percival says, was when Murdoch promised stock analysts in an earnings call that Myspace would generate a billion dollars in revenue that year—at a time when the true revenue figure was a tenth of that. As a result, Myspace managers began scrambling to sign up any program or service that someone was willing to sponsor—no matter how inane or annoying it might be. This was one of the factors that ultimately led users to abandon Myspace in favor of Facebook.⁶

As we’ve observed, there are a number of ways to shift later to a monetization model that lets the platform business capture some fraction of the value created. However, the transition is often fraught with difficulty. These key principles of platform design help ensure that the transition to monetization—from free to fee, as Meetup’s Heiferman put it—can be managed successfully.

• If possible, avoid charging for value that users previously received for free. People naturally resent being told that they have to pay for a good or service they’ve previously received for free—as we saw in the case of Meetup. And not all platform businesses have been as successful as Meetup at navigating such a transition. Some, like Zvents, either died or were forced to dramatically alter the nature of their offering.

• Also, avoid reducing access to value that users have become accustomed to receiving. As we noted, Facebook was offering tremendous value for free and actually needed to cut down on that organic value when it decided to provide premium content promotion to paying producers. This led to complaints from both producers and consumers. Facebook’s enormous network effects enabled it to survive this course correction, but for many lesser platforms it might have been fatal.

• Instead, when transitioning from free to fee, strive to create new, additional value that justifies the charge. Of course, you must ensure that, if you charge for enhanced quality, you control for it and guarantee it. Critics have assailed Uber for charging a Safe Rides fee to pay for drivers’ background checks and other safety measures while apparently cutting corners on those same steps.

• Consider potential monetization strategies when making your initial platform design choices. From the time of launch, a platform should be architected in a manner that affords it control over possible sources of monetization. This directly impacts how open or closed the platform is. For example, if platform managers hope to monetize through a transaction fee, they need to ensure that the platform design makes it possible for them to capture control of transactions. If platform managers hope to monetize by charging for access to their user base, the platform should be designed to control the avenues through which content reaches the users as well as the flow of data about users.

As this chapter has shown, monetization is a complicated challenge—and a crucial one that may determine the ultimate viability of a platform. Those seeking to launch a successful platform can’t afford to ignore the issues surrounding monetization or defer thinking about them until after network effects have been established. Instead, platform managers should be thinking about potential monetization strategies from day one, and plan their design decisions so as to keep as many monetization options as possible open for as long as possible.

TAKEAWAYS FROM CHAPTER SIX

A well-managed platform can create excess value in four ways: access to value creation, access to the market, access to tools, and curation. Monetization is about capturing a portion of the excess value created.

Techniques for monetizing a platform include charging a transaction fee, charging users for enhanced access, charging third-party producers for access to a community, and charging a subscription fee for enhanced curation.

One of the most crucial monetization choices is deciding whom to charge, since the difference in roles played by various platform users means that charging them can have widely differing network effects.

Given the complexity of the monetization challenge, platform managers should take potential monetization strategies into account in every decision they make regarding platform design.