Mark Zuckerberg’s favorite phrase is “social graph.” In nearly every public appearance, the Facebook founder describes his company as a kind of cartographer on an endless mission to build an ever larger, more comprehensive, and more accurate map of humanity. A social graph is a digitized map of connections between people in the real world—the lines that connect you to your friends, your friends to their friends, and on and on until we’re all ultimately joined to Kevin Bacon. At the moment, no corporation has a more comprehensive graph of these connections than Facebook, a fact that will surely bring the social network insane profits. Yet before Facebook, there was another way that people mapped their world online. It’s called e-mail—and without it, Facebook would be nowhere.
Every day, hundreds of thousands of people around the world join Facebook. Right after they do, the site asks for access to their inboxes. While it’s permissible to decline, most people give up their addresses. For Facebook, this is essential: access to our e-mail provides an unmatched look at our social connections. The site analyzes your inbox to determine people you might know on Facebook. Even better, e-mail provides Facebook with an endless stream of potential new recruits—when the site finds addresses it doesn’t recognize, it asks for your permission to fire off a plea to those hold-outs to join the Facebook Borg. Some of those people succumb, which results in new members, and new e-mail addresses, and more new members, and so on and so forth until Facebook has signed up everyone in the known universe.
Facebook didn’t invent the practice of mining data from one social network in order to build another. It’s an old tactic online, and it’s one that confers a lot of benefits on Facebook’s users. After all, if you already have a database of contacts in Gmail, it’s a pain to rebuild the same graph when you go somewhere else.
For that reason, Facebook has long insisted that other sites should make it easy for their users to transfer data to their social networking profiles. But Facebook isn’t big on reciprocity. Using both technical and legal maneuvers, Facebook unilaterally decides what data you can remove from the site, and where you’re allowed to take it. Facebook, in other words, is a roach motel for your social graph—your data checks in, but unless Mark Zuckerberg approves, it can never leave.
This week, Facebook’s one-way philosophy erupted into an ugly spat with its frenemy Google. The dispute began when Google added a clause to its terms of service that requires any application that accesses a Gmail address book to provide a reciprocal export feature for its users’ contacts—a clause designed specifically to force Facebook to make its own user data portable. Facebook decided it didn’t want to do that; instead, it found a way around Google’s new terms by creating a tool that extracted people’s Gmail data through a Google feature that lets users download their contacts for their own use. This process is somewhat cumbersome, but it gives Facebook the same trove of data without having to abide by Google’s new terms.
Though Google decided it wouldn’t be wise to block a feature that lets people download their data for their own use, the search company returned fire nevertheless. On Wednesday, it put up a page warning Gmail users who export their contacts to Facebook that they may be doing something regrettable. “Hold on a second,” the page asks. “Are you super sure you want to import your contact information for your friends into a service that won’t let you get it out?” If you do indeed decide to export your contacts to Facebook, you’ve got to click a check box that says, “I recognize that I won’t be able to export it back out.” Take that, Zuck!
I don’t think Google did everything right here, but it does clearly have the moral high ground. The search company has long been a champion of users’ rights to transfer their data elsewhere. It even has an internal group, the Data Liberation Front, whose mission is to help all of Google’s product teams build easy-to-use export features. As a result, it’s now easy to get your data out of nearly every Google service—Gmail, Google Calendar, Google Docs, Google Maps, Google Health, Google Reader, Orkut, and Blogger, among others. Google’s change to its terms of service to prevent people from taking their data to Facebook was out of step with this history—if Google believes users should have the right to take their data anywhere, that has to include places that won’t let the data back out. (There’s also the question of Google’s timing—why change the terms now, after years of not seeming to care about its users’ taking their data to Facebook? Nobody knows for sure, but many in the tech world speculate that the move was a precursor to a new Google social network, whose path to success would be smoother if it was easy to export data from Facebook.)
Still, Facebook’s position is much less defensible. In a statement that Facebook engineer Mike Vernal posted on Techcrunch, the company argues that the export rules for a social network and an e-mail program should be completely different. In Gmail, your friends’ e-mail addresses belong to you—they’re part of your account. But on Facebook, Vernal says, your friends’ e-mail addresses belong to your friends: “Each person owns her friends list, but not her friends’ information,” Vernal wrote. “A person has no more right to mass export all of her friends’ private e-mail addresses than she does to mass export all of her friends’ private photo albums.”
This is bogus. For one thing, how did Facebook arrive at its convenient distinction between who owns e-mail addresses in an e-mail program and who owns those addresses in a social network? It’s the same information on either platform. If Facebook believes I have the right to export Tom’s, Dick’s, and Harry’s e-mail addresses from Gmail, why is it skittish about my exporting those same addresses out of Facebook?
But Facebook’s hypocrisy goes deeper. As Search Engine Land’s Danny Sullivan points out, Facebook has made what seems like a special exception for Yahoo Mail and Hotmail—Facebook lets users of those two sites pull e-mail addresses from Facebook and into those e-mail programs. In other words, Facebook will let you export your friends’ e-mail addresses, so long as you’re not exporting them to Gmail.
Sullivan points out another deception, too. In his statement to TechCrunch, Vernal argues that Facebook offers its users portability because it allows them to “export friends lists.” He’s right: In October, Facebook added a feature that lets people download much of their data from the site. This was a salutary move—it’s nice to be able to keep a local copy of all your posts, pictures, and other Facebook stuff in case you ever want to delete it from the Web. But this feature is not a substitute for data portability. The file that you download from Facebook contains just a single plain-text list of your friends on the site—there are no e-mail addresses and no links to their Facebook profiles. This makes the data useless for reconstituting your social network somewhere else—if some other, better social network comes along in the future and you’d like to set up shop there, Facebook won’t let you take your friends with you.
Facebook understandably doesn’t want to make it easy for its users to leave en masse for the next Friendster or MySpace. Nevertheless, there’s a legitimate business reason for the site to change its data-hording ways. Brian Fitzpatrick, * the Google engineer who started the Data Liberation Front, told me earlier this year that one of the reasons to let people export their data is to force Google engineers to innovate. “What keeps people using Google search?” Fitzpatrick asks. “We can’t lock them in. The only thing we can do is put a ton of engineering resources into it to keep up search quality.” The easier it is for your customers to leave, the more you’ve got to do to keep them around.
Sure, Facebook has an interest in locking people in now. But that’s a short-term interest; over the long run, a huge captive audience breeds complacency. Facebook has more than 500 million active users, and soon it will have 600 million, then 700 million, then even more. Soon, it will have enough users that Zuckerberg and company will come to think of themselves as essentially unbeatable—and they’ll get lazy. And that’s when, out of nowhere, someone will come along and eat Facebook’s lunch. Hoarding our information won’t make that reality any less likely. It will just make the stampede away from Facebook a little more annoying.
Correction, Nov. 11, 2010: This article originally misidentified a Google engineer. He is Brian Fitzpatrick, not Brad Fitzpatrick. (Return to the corrected sentence.)