Study: Congressional Twitter accounts are plagued by phony followers

Congress is followed by a lot of phonies: a new study shows a large percentage of accounts following legislators on Twitter are fake.

Jon Tilton, the general manager for digital marketing firm Advocacy Media, ran a follower check last weekend on every member of Congress using StatusPeople, a tool designed specifically to check for fake followers on Twitter. He found that an average of 38 percent of accounts following representatives on Twitter and 42 percent of those following senators are a combination of fake and inactive accounts.

ADVERTISEMENT
The averages held true within a few percentage points when broken down by party.

Overall, Rep. Jared Polis (D-Colo.), Nancy Pelosi (D-Calif.) and Sen. John McCain (R-Ariz.) had the highest level of phonies following them. Polis’s followers are 82 percent fake, according to the study, a higher number than any other Democrat or Republican. Most others had a higher percentage of inactive accounts beefing up their follower numbers, but Tilton clarified that fake and inactive accounts tend to mean the same thing: there is no human being on the other side of that account.

Pelosi’s analysis showed 24 percent of her followers are fake and 49 percent inactive, an identical result to McCain’s.

The study is bad news for members of Congress who judge their social media success by their follower count. New-media competitions held by House Democrats and House Republicans each year aim to encourage best practices in congressional offices’ use of social media, but the competition is judged based on new Twitter followers, YouTube subscribers and Facebook “likes” gained during the allotted time. Polis won the House Democrats’ competition this year.

President Obama has been known to joke about how many Twitter followers he has, comparing his follower numbers to other high-profile tweeters more than once, but many digital strategists scoff at the idea that more followers means more effective when it comes to social media.

“The value of a platform like Twitter is not so much the reach as it is engagement,” Tilton said. “At the end of a day you want to earn a follower. ... These legislators don’t know who these people are.”

Fake followers do not necessarily signal intentional deceit by the user. Automated Internet bots scan anything digital looking for high-profile names to latch onto to look more legitimate.

But accusations of buying fake followers have become increasingly common in politics, especially in campaigns.

Mitt Romney’s campaign was forced to deny accusations of buying fake followers last month when his follower count spiked over one weekend. An independent firm later found the percentage of Romney’s followers who are fake wasn’t any higher than average, and it turns out, according to StatusPeople, that Obama’s account has more fake followers than Romney’s does.

But Newt Gingrich’s presidential campaign dealt with it first, facing charges last year during the GOP primaries that most of his 1.3 million followers were fake. Social media analytics service PeekYou found last August that 92 percent of Gingrich’s Twitter followers were fake — meaning at the time his actual follower count was closer to 106,055, noted ABC News — but his campaign denied it had ever used an outside agency to inflate the former House Speaker’s numbers.

“Twitter is run amuck with phony accounts,” Tilton said, noting that the problem raises questions about Twitter's legitimacy as a communication tool. His advice is for congressional members and campaigns to invest in regular attempts to scrub their account of phonies, recommending campaigns and offices dedicate resources now in order to avoid wasting them later.

Otherwise, he said, “This chips away at the validity of Twitter as a platform of real people communicating their thoughts with their [representative] on an issue.”

For accuracy, Tilton cross-checked StatusPeople using his personal Twitter account before and after buying 1,500 fake followers from a marketing company. StatusPeople works by grabbing a sample of any account’s followers, which means that testing an account with more than 10,000 is more prone to error. But Tilton’s testing indicates certainty to within a few percentage points. Similarly, a test run using StatusPeople on Romney’s follower count last week mirrored the results of a study by security firm Barracuda Labs the week before.