featured contributor

By Punkmorten (Own work) [Public domain], via Wikimedia Commons

They seem to come in waves: articles about the preponderance of fake followers on celebrity accounts (Lady Gaga, Barack Obama), or about the proliferation of fake Twitter accounts. And the evidence oftentimes comes from one of the automated fake follower tools on the web, most notably from StatusPeople or SocialBakers.

But how accurately are these tools in identifying fake accounts? Probably not as accurate as you might think.

What’s in a number?

At first glance, there’ s a lot of consistency between the tools. Consider two assessments of the @leaderswest Twitter account:

Statuspeople leaderswest

Social Bakers fake follower check leaerswest

Each tool determined that of my 20,000 followers, about 600 may be “fake.” But the Social Bakers tool goes one step further and proposes “fake” accounts to block:

SocialBakers block list

To SocialBaker’s credit, they list the criteria for inclusion in this list and say specifically:

“We understand that these criteria, number 6 in particular, don’t necessarily define fake followers. However these kinds of followers can be considered empty or inactive and therefore not helpful to you in terms of reach.”

But let’s look at the first “fake” follower @tweetcaroline:

@tweetcaroline

Caroline Barry Klout

@tweetcaroline Faker breakdown

Caroline appears to have been miscategorized as a “fake” account because she uses the paper.li content aggregator. The standard paper.li tweet repeats the same phrase in multiple tweets. But @tweetcaroline most certainly isn’t a fake account. And this is a common reason that some legitimate accounts were identified as “fake” with the SocialBakers tool.

Assumptions for fake Twitter followers

Anytime you make assumptions about a large population, slight imprecision can be hugely inaccurate. A recent (more serious) example of this were the Excel errors in the calculations for European austerity measures by two renowned economists. The austerity measures that have been implemented over the last few years were based on small miscalculations that produced huge errors.

Percentage of fake Twitter followers will never approach the seriousness of austerity, but the same principle applies. Assumptions about a group of two-hundred million people, no matter how slight can result in some miscalculation, so while these tools are helpful in a broad sense they aren’t as precise as they sometimes get credit for.

What do you think? Are these tools useful? Are they accurate?

Photo by Punkmorten (Own work) [Public domain], via Wikimedia Commons
Jim Dougherty

Jim Dougherty

Writer and chief of miscellany at leaderswest.com
I aspire to give people something to think about rather than tell them what to do. My favorite Google Alert is "social media research," I am increasingly compelled by Gen Z, and I appreciate good writers agnostic of where they write. At one time I was Kred's 12th most influential social media blogger and Klout's most influential person on the topic of David Hasselhoff. Transplant from Seattle living in Cincinnati. Haven't entirely adopted the local sports teams yet.
Jim Dougherty
Jim Dougherty
Jim Dougherty
Fey1IsleofSkye

I don’t use those two services. I do, however, use who.unfollowedme.com and justunfollow.com, and I would agree, that you have to be careful in unfollowing peeps that they deem as bots. A friend of mine came up as a fake follower because she had been offline for a while. The way I determine whether or not to unfollow based on these recommendations is a careful scrutiny of their timeline first. That seems to work for me.

April 25, 2013, 12:30 PM
jimdougherty

Great insight Sidney! I use justunfollow and am always a little concerned about the accuracy of their inactive #s.

April 26, 2013, 12:01 AM
Tommy Landry

I use untweeps from time to time, but they also come down hard on any API-based auto tweets like Paper.li. By the way, is it @tweetcaroline or @carolinetweets?

April 26, 2013, 7:09 AM
jimdougherty

Great catch, Tommy – now corrected. And great insight on the auto tweets – I think automated responses are most likely to be misconstrued by apps like this.

April 26, 2013, 8:09 AM
Jan Rezab

We are fixing the Paper.li thing – but repeating tweets is normally a thing of fake profiles only, and its not even possible through the Twitter website.

April 27, 2013, 8:50 AM
jimdougherty

Thanks Jan! Of all of the talk about what constitutes a “fake” account, Social Bakers is most forthright about the criteria, and that really helps to put perspective to the capabilities and limitations of these tools are. Thank you!

April 27, 2013, 3:19 AM

This is always an interesting debate.

I remember when this debate over fake followers came up last year. A few good friends of mine were being called out for having “fake followers” and therefore condemning them for getting certain things because they had a following. Most of them had no idea about these “fake followers” though.
The problem with these tools is that I can’t really figure out what they’re used for other than calling people out.
I understand that it’s one thing to buy a bunch of followers to raise your count by thousands in the matter of an hour or so (Mitt Romney, anyone?). However, I know for a fact that I’ve never purchased followers, but that doesn’t mean I don’t have a bunch of fake accounts following me. I have no control over who chooses to follow me, fake or not. (To be honest, I can’t even figure out why non-fake people follow me in the first place). And, the more popular I (or anyone) becomes, the more fake accounts I get following me.
For people like me, this tool makes no sense. You’re going to call me out because some random fake accounts decided to follow me? That makes no sense. It would be one thing if you saw my follower numbers go up a lot in a short time, then you can make the accusation that I purchased some followers, but to bring attention to the fact that people have accounts following them means really absolutely nothing.

I just don’t see the value of these tools or what they possible could be useful for.

Cheers,
Sheldon, community manager for Marketwired

April 29, 2013, 4:34 PM
jimdougherty

Thanks Sheldon,

There is definitely something to the social proof of a big following, real or not. It’s always a temptation to out some of these “gurus” whose raison d’etre is predicated on a bunch of computer generated followers. In practice it’s harmless enough, but it’s doubtful that they’re recommending a 50K follower boost on Twitter to any business they’re advising.

There’s another aspect when it comes to celebrities and these bots. In order to obfuscate bot accounts a common practice is to follow celebrity accounts. I think the rationale is that a celebrity (Mitt Romney being a good example) might have fans that join Twitter just to follow them, so it makes it difficult for Twitter to say that an account is a fan versus spam. I imagine that’s why these tools sometimes come up with false positives. What I like about the SocialBakers tool is that it’s completely transparent about the algorithm they’re using and that they may return accounts that aren’t valuable as opposed to fake accounts.

April 30, 2013, 10:09 AM

I can be a little fake at times, but my Twitter account is real! On a serious note, I am glad to read they are fixing the paper.li glitch!

May 16, 2013, 1:41 AM
jimdougherty

OMG – you’re REAL???? :)

May 16, 2013, 2:17 AM

Share your thoughts.

Current ye@r *