On May 9, 2008, Facebook won the Internet. Or at least it should have.
On that date, in an open letter to the developer community, Dave Morin, then head of Facebook’s platform efforts, announced the release of Facebook Connect:
Facebook Connect is the next iteration of Facebook Platform that allows users to “connect” their Facebook identity, friends and privacy to any site. This will now enable third party websites to implement and offer even more features of Facebook Platform off of Facebook.
Facebook Connect was a radical evolution of Facebook’s API. It allowed third-party developers to use Facebook as the identity and login system for their sites and apps. For developers, there is often no greater challenge than getting users to sign up and invite their friends. Integrating Facebook Connect solved that problem. Why recreate your contact lists by hand when you could just sign in and let Facebook do it? Facebook Connect was easy for users, it was easy for developers, and it could have made Facebook as central to the app ecosystem as Google is to the Web.
To this day, Connect is the smartest business move Facebook has ever made. Over the years, though, it’s become the biggest opportunity Facebook has ever squandered.
Facebook failed to leverage Connect into de facto distribution across the Internet. It failed not because it underestimated the power of Connect or the upside of achieving platform ubiquity. It failed because it overplayed its hand in the short run: alienating developers and insisting on the primacy of the central Facebook over the distributed Facebook. It treated Connect like a value-add rather than an existential necessity. In doing so, Facebook risked selling short its own future.
Looking back, we can see why the company felt everything was fine inside the walled garden. By late 2007, on the heels of Microsoft’s $240 million advertising investment, the company had an informal valuation exceeding $15 billion. It would grow to more than $40 billion by 2010, by which time Facebook had more than 500 million active users. Highly active users—more than 50 percent of them logged into Facebook every day. The average user spent more than 700 minutes a month on the site and posted at least three times a day. Collectively, users shared 30 billion links, posts, and pictures each month.
And all of this was happening on Facebook’s site, within the confines of the Facebook experience. That heavy usage pattern—implying Facebook’s apparent position as the “home page” of the Internet—instilled a heroic degree of overconfidence in Facebook’s managers and investors. But it should have scared the hell out of them. It placed 500 million eggs—soon to be more than a billion—in the same basket, because everything was taking place inside Facebook: on its site or on its mobile app. That meant that any threats to any aspect of Facebook’s user experience would be threats to all of Facebook.
But nobody thought that was much of a problem. The more common critique of Facebook, especially in the runup to its initial public offering in May 2012, was to point out its weaknesses in mobile—mobile advertising in particular. By 2012, Facebook had become an advertising company. Analysts were scared that mobile users had a lower tolerance than Web users for seeing ads and less of a reason to click on them. (Facebook’s reticence about the health of its mobile advertising business seemed to validate those fears.)
But Facebook shouldn’t have had those weaknesses in the first place. More than half a million developers had begun to use Facebook’s API by 2010. By the time the IPO came around, Facebook should have been integrated into every site and app around. It wasn’t then, and it’s not today, and that’s largely Facebook’s fault.
In the beginning, Facebook made its API widely available to developers, offering them all the benefits of Facebook’s social graph and generally setting off a wave of excitement for the platform. After the smashing success of early developers like Zynga—powered entirely through Facebook’s traffic—everyone wanted to get a piece of the action. Developers shifted their efforts to “social apps”; at the time, the “social” category meant Facebook, and Facebook alone. It was a hell of a leg up for the platform.
The excitement was short-lived. For one thing, Facebook set sky-high transaction fees to developers seeking to make money off of its user base. It took a standard 30 percent cut of all revenue from developers monetizing on the Facebook payments platform. That was a spectacularly myopic demand, particularly in the mobile space. Consider the plight of an iOS games publisher: It has to pay 30 percent of sales to Apple, and if it chooes to monetize on Facebook, it has to pay 30 percent there too. Many developers have come to see Apple as the smarter choice; it's a more lucrative distribution channel at the same revenue cut.* As Bill Gurley, a general partner at venture capital firm Benchmark, put it, “The bottom line is that the entire gaming industry has lost some of its enthusiasm for the Facebook platform, and it will be difficult for Facebook to recreate the magic and momentum they once had.”
That’s no small issue, especially when you consider that games account for the lion’s share of mobile application revenues and that the category is still growing at 66 percent per year. To fall out of favor with the mobile gaming community is, effectively, to fall out of favor with the mobile community. That’s something Facebook can’t afford to do.
And then there are the developers Facebook has shut down or blocked out. It still makes its APIs widely available, but if you’re growing big enough to look threatening, it’ll decide either to buy you or break you. Its motives for breaking an app—either cutting off API access for non-Facebook apps using Connect or turning off Facebook apps altogether—have been opaque. In 2011, when Facebook shut down the popular apps Goodreads, Photo Effect, and Social Interview without notice, it offered scant explanation, either to those apps’ developers or to their millions of users. (Goodreads CEO Otis Chandler was later able to restore his app’s access; as he told GigaOm, rather ominously, the experience was a “great reminder of the power Facebook has over all of us developers.”)
Facebook’s suzerainty over its platform partners neither makes a lot of friends nor influences people. Centuries of economic practice have shown us that a distributed, free ecosystem of entrepreneurs, hustlers, and inventors can out-innovate and out-perform a centrally planned economy. Facebook chose to be a central power. It thought it knew better than its developers. As a result, most of the innovation that should have accrued to Facebook’s benefit accrued, instead, to the iOS and Android platforms. More and more developers found working with Facebook to be less necessary, and more onerous, than seemed to be the case in the heady days of 2008.
But Facebook is wrestling with an even bigger, more philosophical problem: the question of identity itself. What does it mean to be a Facebook user? What do Facebook users really care about when they use Facebook? From all indications, it’s not Facebook they care about; it’s one another. Facebook has done little to strengthen its brand or to draw psychological ties between the Facebook experience and your social circle. There is no “Facebook lifestyle” the way there’s an Apple lifestyle or even, in some respects, a Google lifestyle. Facebook is a part of its users’ lives, but it’s treated more like a means to an end than an end in itself. It’s a set of pipes. If the brief history of social networking has shown us anything, it’s that there’s always a faster, better, or cooler set of pipes waiting around the corner. Users just want to connect and share with one another, and they’re a lot less picky about where they do their sharing and connecting than Facebook has generally assumed. In Facebook’s case, that miscalculation has cost the company its teenage users. Teens are the trendsetters of the digital world, and losing their favor is often a marker of slow decline.
Facebook knows this all too well. It has paid billions to maintain engagement with the young users who’ve been tuning it out. It spent $1 billion on Instagram (a deal that looks increasingly like the steal of the century), and recently, $19 billion on WhatsApp (a deal that has many of us scratching our heads). From a defensive standpoint, these moves may have been necessary. But their benefits will quickly bleed away if Facebook can’t make the Facebook identity mean more.
In a move that got buried by the WhatsApp story, Facebook recently shut down its email service. For the last three years, Facebook offered its users the ability to create “@facebook.com” forwarding addresses. Few people knew about this, and evidently, few enough used the service to make it worth the company’s while to maintain. But the shutdown is very telling. Had Facebook played its cards correctly, we’d all be using @facebook.com right now. Facebook would own our identities online, and we’d use those identities everywhere. We’d use them on Instagram, we’d use them in WhatsApp, and we’d even use them for payments. They’d be our credit cards, our PayPal accounts, our driver’s licenses, and our contact information. Facebook’s central site and experience might have taken a hit, but Facebook’s platform would have eaten the world.
It’s not too late for Facebook to recover. But the company has never been more strategically vulnerable. Sooner or later, it’ll run out of the tens of billions of dollars necessary to keep buying every emerging threat to its user base. And sooner or later, Mark Zuckerberg will emerge—not the Mark Zuckerberg who runs Facebook today, but the next Mark Zuckerberg, an entrepreneur smart, cocky, and reckless enough not to take a Facebook buyout. When that happens, Facebook will wish it had made less of “Facebook” and more of “Connect.”
Correction, Feb. 28, 2014: This article originally stated that Facebook takes a 30 percent cut of sales from developers who use the Facebook platform, even for revenues that come from sources outside of Facebook. Although this has been the case with at least one prominent developer in the past, it is not the case today. Facebook takes a 30 percent cut only from developers using Facebook's desktop payments engine, for apps running on Facebook itself. (Return.)