dick-costolo-twitter

Conventional wisdom says that Facebook is better for advertisers than Twitter. After all, Zuckerberg’s billion-strong user base willingly gives up a host of data points upon signing up, from gender to job to age.

Twitter, on the other hand, requires only a name and email address to join the service, and the name provided doesn’t even have to be real. That’s not exactly the kind of extensive data advertisers are salivating over.

But according to Twitter, the service knows a lot more about you than you might think. Over at Boing Boing, Glenn Fleishman writes that Twitter claims it knows if a user is male or female with 90 percent accuracy. How? Fleishman explains:

The service analyses our tweets and uses word choice, proximity, and other factors to make a guess. According to a 2012 post on its Advertising blog, the company relies on multiple signals to assign confidence to a gender selection. For matches with a high confidence level, Twitter tested its results against a global panel of humans found its approach 90 percent accurate.

The benefit of knowing a user’s gender is obvious. Like nearly all free products, a big chunk of Twitter’s revenue comes from advertisers. And the more advertisers know about a user the better they can target ads.

Gender is only the beginning. A company called Leadsift recently launched a new product that claims to analyze tweets to predict a user’s age, salary, and buying habits.

“We extract over a hundred attributes by modeling on a user,” Leadsift CEO Tukan Das tells me. “Let’s say you check into an airport more than three times per month. We’ll automatically learn that you are a frequent flyer and a businessperson.”

Granted, the more demographic information you try to determine about a user, the less likely you are to be right. Determining a simple binary like “male or female” is easier than saying “This is an 18-24 year old male with disposable income who is interested in buying a Honda.” But the data is still valuable. Das says clickthrough rates on ads that used Leadsift’s technology to target users went up 40 percent, and time-on-site went up five times.

Even using the most sophisticated language analysis, the advertising profile of a Twitter user will likely never be as rich as that of a Facebook user, where detailed demographic information is explicitly requested upon signing up. But that doesn’t mean Twitter can’t offer something unique to advertisers.

People interact with one another differently on Twitter than on Facebook. Because most users are public, there’s more conversation between folks that don’t necessarily follow one another. That’s why Twitter raced ahead of Facebook when it comes to celebrity engagement, though Facebook has made up ground there of late. Twitter is also a place where users interact with experts they don’t necessarily know in person, making it a great network for asking the crowd questions like, “Which laptop or vacuum cleaner should I buy?” And while these updates might get lost in Facebook’s News Feed algorithm, Twitter’s raw feed doesn’t (yet) discriminate against posts, making the likelihood that someone from a user’s network actually responds that much greater.

Some people might find it a little disconcerting to discover that Twitter knows more about them than what it’s been explicitly told. But it’s an inconvenient truth that all users of free social networks must come to grips with. To keep the lights on and the tweets flowing, Twitter needs money. And if it doesn’t get that cash from users than it needs to get it from advertisers, to whom all your personal data, implied or explicitly-known, is a big tasty carrot at the end of Twitter’s stick.

[illustration by Brad Jonas]