Facebook Targets 76 Million Fake Users In War On Bogus Accounts

Facebook has not gotten its fake account problem under control — yet — but it appears to be making progress.

In its annual report, Facebook said that the number of duplicate accounts on its network had risen to 5 percent of all accounts in Q4 2012, up from 4.8 percent in Q2 2012.

However, the number of accidentally misclassified accounts and the number of abusive accounts (run by spammers and the like) has declined.

Overall, the total percentage of fake accounts declined from 8.7 percent to 7.2 percent. In whole numbers, the number of duplicate accounts went up from about 45.8 million to 52.8 million; but the total number of fake accounts (which includes abusive and misclassified accounts) declined from 83 million to 76 million over the last six months:

Facebook has therefore deleted or otherwise shut down about 7 million bogus accounts of one sort or another in the last six months, bringing the fake population down from about 83 million in Q2 to 76 million in Q4.

The improvement came after a campaign again fakes, vp/global marketing solutions Carolyn Everson told us back in December. Facebook began its purge of fake accounts and the likes they generated, back in August. After the fake purge began, Lady Gaga lost 65,505 fans, and Facebook's own Facebook page lost 124,919 fake likes, according to data from Likester, the Facebook marketing data company.

The progress is important to Facebook's reputation, even though the actual rate of questionable activity on Facebook is a small percentage of its entire traffic. Advertisers use likes and fan data to target campaigns — they don't want to waste money advertising on fake accounts.

Facebook has been sued over invalid clicks in the past (the company denies the allegations). There are persistent reports of fake likes coming from the accounts of deceased friends. The system has had flaws in it in the past that have inflated the number of likes on a page. And advertisers on Facebook have grumbled about paying for traffic they believe comes from fakes.

The company is going after bad actors first — which is why abusive accounts declined. It isn't concerned so much with the majority of duplicate accounts, such as those made for people's pet dogs and cats. So Fido is safe ... for now.