I Am Not Sgt. Sperm Donor: Why Facebook Should Open the Social Graph

At this point, Facebook owns your social life. The site, which just overtook Yahoo to become the second most visited Web destination in the U.S., has had six years to record all the photo tags, friend requests, and pokes made by its 400 million users. The result is a well-developed social graph—a tangled ball of twine that describes how we connect to one another and what we care about.

Facebook intends to keep that information to itself for as long as it can. The company has been compared to AOL circa 1997, a room with locked doors and shuttered blinds. Only a narrow sliver of light escapes to be indexed by search engines. Facebook's tweetlike status updates aren't shared, either—in stark contrast to Twitter itself, which funnels its data to anyone who asks. Facebook is the Club Med of the Internet—an all-inclusive resort that keeps the hostel dwellers at bay.

This is good for Facebook. It made a reported $550 million in 2009, most of it from serving targeted ads. But it's bad for just about everyone else.

That's not a new insight, but I was reminded of that fact last week when Pete Warden, a former Apple engineer, unveiled his latest project. Warden spent the last six months crawling Facebook's public pages, in effect reconstructing as much of the social graph as he legally could. He then performed some rudimentary analysis on the data, building a heat map of America's obsessions. The results won't shock anyone. Utah, which is overwhelmingly conservative and Mormon, loves Glenn Beck and Stephenie Meyer, who are both conservative and Mormon. But Warden had access to just a thread of Facebook's data. Imagine what he or a smart team of researchers could do with the whole ball of twine.

Warden isn't the only barbarian battering at Facebook's gates. Google's botched Buzz debut may have been a privacy and PR nightmare, but it signaled how badly the search juggernaut wants to recreate the social graph for its own use. Partly this is for ideological reasons. As a Gmail engineer put it, "One of our core philosophies at Google is that users' data should never be held hostage." But Google is no Atticus Finch, standing on principle alone. It's in this fight for the money. Facebook is able to extract half a billion dollars a year from the social graph, and they're not even very good at mining it. (According to Facebook's ad-serving algorithm, I'm a military veteran interested in joining the FBI and a likely sperm donor.) Google must know it can do better.

If competition breeds innovation, closed systems kill it. During the dotcom boom, a succession of companies—Lycos, AltaVista, Yahoo, Ask Jeeves—one-upped each other trying to build a better search engine. Google won the day not because it herded the Web into a proprietary corral, but because it built a better technology. Today, there's no war over who can better mine the social graph. That's because Facebook holds the only key, and for now, we're all locked inside.