Daily life is basically a constellation of microscopic choices. Let’s say you choose to read some Ayn Rand on the bus home from work today. Then let’s say you play Pandora’s country hits while you make dinner. Next week, when the polls open up for the 2014 midterms, you’ll weigh your options and choose to vote for Republican candidates.
To you, each of these choices is discrete, isolated — even random. But to Facebook, the 1.35-billion-user behemoth where we increasingly record our daily choices, no person, comment or casual thumbs-up is isolated: Everything is part of a larger pattern.
And in this particular instance, according to data Facebook released Monday, your pattern played out precisely as the social network might have predicted it would: Generally, Facebook fans of Republican candidates are far more likely to also be fans of Blake Shelton and Ayn Rand.
There are more trifling insights where that came from, of course: Democrats disproportionately listen to The Beatles; Republicans tend to watch “Duck Dynasty”; the Jersey Shore is the one vacation spot all Americans agree on. (Jersey, you can have that mention for free.)
But while these tidbits are interesting — funny, even! — they’re just whispers of a much larger operation. Facebook is inferring millions of such correlations, based on billions of disparate data points, predicting any number of user behaviors. In many cases, it’s also selling those data points back to advertisers, political campaigns and other people who are interested in manipulating all the tiny daily choices you make. In an update made to Facebook’s ad platform in May, the site released a trove of new, anonymized data to advertisers, showing them — among other things — the demographics and “liked” pages of potential customers, the better to help companies learn their “interests and behaviors.”
“Help people decide,” the network crows in its materials targeting political campaigns. “Facebook is one of the most effective marketing platforms on the planet.”
That’s true, in large part, because of how much Facebook knows about us: age, family, location, education level, on- and 0ff-site behavior. (Advertisers can target all these things.) But given all that objective, demographic data, it’s amazing how much Facebook can still infer from our subjective likes. The like, after all, is the single most thoughtless, fleeting gesture in all of Internetdom — it’s a single click, an almost Pavlovian response to things we find vaguely interesting or noteworthy or at least worth acknowledging. It’s “the wordless nod of support in a loud room,” the blogger Elan Morgan wrote in August as part of her experiment not to “like” anything on Facebook for two weeks. It’s just another way to signal we exist amidst the constant onslaught of Internet noise.
Reading through your own liked pages should confirm as much: Mine amount to little more than a senseless jumble of local bars and restaurants I once wanted to try, random companies I followed to unlock sales or specials, and the long-forgotten side hustles of entrepreneurial friends. Facebook itself hasn’t even had the feature for a full five years.
And yet, simply by virtue of its scale, Facebook can vacuum up all those senseless data points — all those clicks, all those tiny choices — and parlay them into something like an insight about your identity. These are just observations and correlations, mind you, but they’re still pretty striking: Oh, you liked the campaign page for New York Gov. Andrew Cuomo? You might also like “The Great Gatsby.” Did you fan the page of Hello Kitty or Flo the Progressive Girl at one point? You, for whatever reason, are more likely to vote.
Tell its data team what pages you’ve liked, and Facebook will tell you who you are.