In the beginning, there was Instagram — fresh, new, and ripe for spam.
The photo-sharing site was riddled with fake accounts that liked, followed and hashtagged out the wazoo. Then came the great “Instagram Rapture,” Instagram’s pledge last December to clean out the fake accounts and do “everything possible to keep Instagram free from the fake and spammy accounts that plague much of the web.” “The Purge” cost famous grammers thousands, sometimes millions, of followers. Users bemoaned the dropping follower account, the diminished likes and the sense of lessened popularity.
Only it didn’t work.
According to the Wall Street Journal, a team of Italian security researchers say up to 8 percent of Instagram accounts are still “computer-generated.” The study looked at activity from 10.2 million Instagram accounts, many of which followed high-profile grammers like Taylor Swift (@taylorswift) and National Geographic magazine (@natgeo). These were the accounts that supposedly lost millions of followers in the Rapture — apparently, not millions enough.
William Howard knows a lot about these “computer-generated” accounts — he’s a spokesperson for Rantic, a company that sells followers, likes, views and more on social networks like Facebook, Twitter, even Google Plus.
And Instagram, of course.
“Before the whole shenanigans that happened with the purge and the rapture, whatever people called it, Instagram had no spam filters,” Howard says. “It was a fresh site when they were getting more users. Their spam filter was very basic. There wasn’t a lot of protection versus people who create bots and stuff.”
Companies like Rantic.com profit off these fake accounts — they own bots that generate thousands to millions of faux Instagrams and then sell them to celebrities, corporations and anyone else looking to grow an account quickly and expediently.
After the rapture, Instagram nixed the fake accounts and updated its algorithm — and then bots like Rantic’s had to get smarter.
“Smarter” fake accounts aren’t just photoless usernames that plague the Instawebs with random likes and comments with links. Instead, they’re fully-fleshed accounts with biographies and avatars and posts and followers of their own. Many of them even comment on other accounts.
The most important part, however, is that the accounts stay active. The Italian research team estimated that 30 percent of the accounts in its sample were inactive, meaning they didn’t post at all over the course of the month-long study.
According to Instagram, the goal of the purge in December was to ensure that these fake accounts were deactivated and then no longer counted as followers.
Rantic gets its fake accounts from a bot built by Russian programmers (Howard says it’s “one of the best”). Prior to the Rapture, Howard says the Rantic bot was generating 30 million fake accounts — and after the purge, about 10 million of those were deleted.
“Selling followers is a multi-million dollars business,” Howard says. “And a bot is worth gold.”
Howard says most of Rantic’s clients spend somewhere around $9.50 for 500 new followers that are “active” — and not easily spotted as fake. If it’s a celebrity, an athlete or someone else with a higher following demand, they’ll probably buy more, both to protect their reputations as well as their endorsement deals. That means the bots have to be smarter than ever — and companies like Rantic have to move quickly after the next Instagram purge to bounce back with better fake accounts.
So: can we ever escape the bots? On some places, maybe — the business of buying Facebook likes, for example, is “almost over,” according to Howard.
Instagram maintains it is committed to fighting spam aggressively well into the future — even as bots remain a fact of Internet life.
So in the end, there will still be Instagram — and there will probably still be bots, too. This is the great trial of the Internet age, and it always has been: it will always get harder and harder to tell what is real, and what is fake, and who is real, and who is fake. That isn’t changing any time soon.
And ultimately, you’re not going to be able to tell the difference — and maybe Instagram won’t, either.
Like this? Try these:
- Twitter has a huge problem — and it’s all in your head.
- I created the Caitlyn Jenner bot @she_not_he. This is what I learned.
- Always click the first Google result? You might want to stop doing that.