IPO filing reveals social company estimates less than 5% active users are ‘false or spam’ accounts, but it admits that could be higher
On Twitter, nobody knows if you’re a dog — or a robot, spammer or another entity that is useless to advertisers.
The social media company, in its initial public offering filing to raise up to $1 billion, disclosed that it estimates less than 5% of its monthly active user accounts are “false or spam accounts.”
But that means upwards of 10 million of Twitter’s 218.3 million average monthly users for the three months ended June 30 could be bogus.
And Twitter admits that might be low. “(O)ur estimation of false or spam accounts may not accurately represent the actual number of such accounts, and the actual number of false or spam accounts could be higher than we have currently estimated,” Twitter said in the filing.
The company also noted that it treats multiple accounts held by a single person or organization as multiple users for purposes of calculating its active user base — another factor that would diminish the total addressable audience for advertisers.
The reason this matters is that Twitter is selling advertisers promoted tweets, including its video-based Amplify ads aimed at specific groups of users. Marketers will have less confidence in Twitter’s audience numbers if it can’t reliably report what portion of those are duplicate or fake accounts. (Twitter charges only for tweets that are interacted with, but that only means bogus accounts that are retweeting those are wasted impressions.)
Previous industry estimates have pegged the proportion of Twitter’s spam accounts as high as 20%. But the company said it has taken more aggressive steps to purge fake accounts this year.
“We made an improvement in our spam detection capabilities in the second quarter of 2013 and suspended a large number of accounts,” Twitter said in the S-1 filing.
Much of the problem stems from Twitter’s philosophical approach to be a completely open platform. Twitter does offer “verified” accounts for a number of high-profile accounts (like @WarrenBuffet). But otherwise, it’s pretty much anything goes. Facebook, by contrast, requires that users to use real identities on the service, which gives it more certainty that those accounts are actual individuals; Facebook also tracks demographic data entered by its users, including gender, location and age.
Spamming activities on Twitter — which includes large numbers of unsolicited mentions of a user, duplicate tweets and misleading links (such as to malware or webpage-hijacking pages) — are prohibited by the company’s terms of service. Twitter also bans the creation of serial or bulk accounts to “artificially inflate the popularity of users seeking to promote themselves on Twitter.”
But spammers will continue to try to exploit Twitter, the company acknowledged, and fighting them will represent sizable ongoing operating costs.
“Our actions to combat spam require the diversion of significant time and focus of our engineering team from improving our products and services,” Twitter said in the S-1. “If spam increases on Twitter, this could hurt our reputation for delivering relevant content or reduce user growth and user engagement and result in continuing operational cost to us.”