Web users have knocked Facebook's willingness to use their information to sell ads. But the company's practices resemble more accepted forms of database marketing
When you get home tonight, you might find a pile of direct mail on the kitchen counter. The phone may ring with a local police group seeking a donation. E-mails from L.L. Bean or Hilton Hotels could fill your inbox. If you've visited a Lexus (TM) dealer about that special promotion you got in the mail, computers will crunch your financial profile to see if you're worthy of a loan. All this is driven by data, the virtual bread crumbs you leave behind as you fill out warranty cards, subscribe to magazines, or buy on credit. And you're cool with that. So why do people freak out about social networks using data the same way? Part of the answer has to do with the immediacy of Internet advertising, which often customizes ads to reflect, eerily, your most recent online activity. If you peruse an automaker's website, for instance, chances are good you'll soon see banner ads for the exact same model you were researching elsewhere online. This practice, called "retargeting," tags your computer with a snippet of code so you're served follow-up ads for a product. That's probably O.K. for a Ford (F) Mustang, but perhaps embarrassing if you'd been shopping for Viagra. Yet the consumer benefits of wider availability of social data on the Web could be compelling. Facebook users' outcry and congressmen's complaints about the site's difficult-to-use privacy settings led the company to unveil on May 26 what Chief Executive Mark Zuckerberg called in a blog post "a simpler way to control your information." Selling Data to Outsiders
I have a more radical recommendation for Facebook: Go ahead, Mr. Zuckerberg, take your marketing practices a step further and sell that social data the company is collecting. Facebook is poised to become the biggest consumer database in history. Instead of using that knowledge to show users tiny ads, a more potent business model could be to sell information about users to outside marketers. To be sure, Facebook has no plans to sell consumers' data. I asked Facebook spokesman Barry Schnitt about my idea, and he said in an e-mail that "people share so freely on Facebook, in part, because they trust that we're not going to sell their information. We don't have any interest in violating that trust." Let's play a mental game, though, and see what the social networking world might be like if Facebook and other sites sold data that could make the coupons that arrive in consumers' mailboxes more relevant. Marketers maintain lists, or sets of data about you, that couple your name with various descriptors. Consumers walk through life leaving clues that let marketers figure out who they are and what they're likely to buy. Your driver's license, voting registration records, product purchases, and periodical subscriptions are all fair game. Patterns That Predict
Get Highlights magazine in the mail? Marketers know you have children. Retailers resell data from many credit card swipes to big database marketing firms that include Experian (EXPGY), TransUnion, and Equifax (EFX). The resulting lists help Omaha Steaks, for example, predict whether you'll buy beef by mail.
Consumers have grown comfortable being tracked in databases, because, frankly, it's often useful. We like catalogs tailored to our interests; we appreciate knowing American Express (AXP) will call us if someone steals our credit card and makes a wild purchase that's out of sync with our behavior. What's missing from marketing lists is the ability to connect consumers with others who have similar tastes and habits. Marketers try to group like-minded consumers with such tools as the Nielsen Prizm system, a set of 66 demographic descriptions that give each U.S. household colorful such names as Young Digerati (young, tech-savvy, live in a hip condo, and go to a fitness club), or Park Bench Seniors (retired, modest incomes, and top viewers of TV game shows). Preferences Among Friends
If marketers could group people by their declared Facebook connections, it would unleash the power of "homophily," the tendency of friends to like the same things. In 2004, AT&T (T) Labs Research studied telephone "network neighbors," or people who call each other frequently. The company found something amazing: If your friend purchases something, you're three to five times more likely than average to want the same product. If marketers could find your friends, they'd be able to pinpoint your desires more accurately. Who has the best data on networks of friends? Hmm. Facebook's data set is growing stronger. The site collects users' gender, location, birthdays, sexual orientation, and connections. It also scores the frequency and value of who communicates with each user, which is why Facebook's news feeds show only certain posts. Behind the scenes, Facebook is evaluating every post for affinity, importance, and timeliness. Facebook's "Like" button seeds other websites with the ability for users to reveal even more of their preferences. That means Facebook has pulled off a difficult marketing feat: matching observed data about users' individual preferences with a map of the connections between those people. Check Your Junk Mail