Facebook: Big-Data Beast of the Future?

By Derek Thompson

Facebook's IPO flop is a fascinating story. But in the long-run, the last seven days are probably irrelevant to the success of the company, which depends on whether Facebook builds a business model that does not exist right now. Our valuation source Espen Robak put it perfectly last week: "If Facebook's profit model stays the same, this [stock price] doesn't make any sense."

The fact that Facebook is still trading at 60-times its earnings (for perspective: Google trades closer to 15) suggests the market consensus is that Facebook's profit model will change. But how? I've spoken with a few people from the social media world who say the company is such an essential component of the Web architecture and our online experience that it could simply start bullying consumers into paying -- e.g.: $1 for each human to access the site and $100 per company to use "like" buttons. Facebook has 900 million users. If each paid a dollar this year, that would nearly equal last year's profit.

Okay, so those aren't so much business ideas as illustrations of the kind of power we're meant to believe Facebook has. (They are, in fact, horrible ideas. Can you imagine the popular uproar if Facebook announced that it would charge subscriptions? Or the freak-out on Wall Street that Facebook was so desperate to show it could make money it was actually thinking about a social media pay-wall? Disastrous.) But this might be a more reasonable illustration: Companies, investors, and individuals using Facebook's ginormous pool of social information to make investment decisions:

Using status updates from U.S. users, Facebook's data team has constructed a daily measure of America's mood--an index of gross national happiness. There is a long literature on how people's mood affects the market (yes, sunnier days in New York City are better for stocks), so Yigitcan Karabulut, a graduate student at Goethe University Frankfurt, decided to examine the link between Facebook's GNH index and share prices.

What Mr. Karabulut found was that GNH didn't seem to merely move up and down with the stock market, but to predict it: A bump up in people's sentiment one day tends to lead to a statistically significant increase in returns the following day.

Today, Facebook makes more than 90 percent of its revenue from online ads and Zynga games. It's conceivable that, as online ad revenue might never quadruple on a per-user basis, Facebook builds a big-data platform that skims aggregated information off its user base and sells the results as a financial service. That's a vision of Facebook, big-data service provider. I don't know if it makes any sense. But it's a vision.

This article available online at:

http://www.theatlantic.com/business/archive/2012/05/facebook-big-data-beast-of-the-future/257690/