70% of Matt Cutts’s Twitter Followers are Fake

posted August 27th, 2012

Read this NY Times article about buying fake Twitter followers last week and wondered whether any influential SEOs bought Twitter followers. So I ran the top ten most influential SEOs on Twitter according to wefollow in the Status People Fake Follower Check , and was surprised at the results.

It appears as though someone’s playing a joke on Matt Cutts, as 70% of his followers are fake, according to the tool. Unless you believe that the head of spam at Google bought about 175k of his 250k Twitter followers, that is.

Same goes for Search Engine Land Editor Danny Sullivan, who has 36k fake followers according to the tool.

Obviously Danny’s not buying followers, so these are probably spam bots for fake profiles.

The rest of the top ten appears to either not be buying followers or to be hiding it very well, as all but three have fake follower percentages in the single digits.

I used my own account and the account of the top followed SEO in the wefollow directory as a control here, as I know I’ve never bought followers and I’m not familiar with this SEO that has almost half a million followers.

Seems to be pretty accurate, but according to the methodology, it’s based on a sample of 1,000 followers and gets less accurate the more followers that you have.

Two takeaways:

1. Follower counts aren’t a true measure of influence in our industry, as percentages of fake accounts are increased by spam bots for fake profiles, negative SEO, buying followers or all of the above.

2. It’s impossible to assume that a high percentage of fake followers means that those followers were purchased by the person holding the account, even though in some cases it’s pretty obvious. I’m sure Mitt Romney, Barack Obama, and to a lesser extent Matt Cutts will be pleased by this news. Thank you, anonymous blackhat buying followers for Matt Cutts for teaching us this valuable lesson. :)

Data and screenshots are below.







Related Posts:

  • No Related Posts
  • 1 Comment »

Leave a Reply

One Response to “70% of Matt Cutts’s Twitter Followers are Fake”

  1. Good catch Bryson and, yes, Matt is influential.

    But I was wondering about the overall membership in that “WeFollow” group. Do you have to join WeFollow in order to be on the “influential” list?

    @webceo, with 4550 followers, has always had the policy of blocking fake accounts because we predicted that the search engines would get around to judging the quality of followers. So we’ve got less than 1% of fake followers according to that measuring website and I went ahead and deleted that 1% for good measure two weeks ago.

    Now I’m realizing that we “could have gotten away” with having 8,000 followers and we probably could have kept that number by following only 4000 back.

    We’d look so much more popular if we had done that.

    Bing, however, seems to have rewarded us for doing something right.

    Webceo.com is in the top 5 for “Online SEO Tools” on Bing while it’s only around #9 for that keyword on Google.

    We’ve wondered, since Google has traditionally been best to us in terms of scoring in the top 10 for SEO software related keywords, why is Bing now seeing us in better terms?

    The only answer I can come up with is that Bing sees that our Twitter activity is high, engaged and we keep the best of our followers (and customers) by not unfollowing them to make ourselves look good.

    Bing also sees that we actively seek out quality people to follow and we follow them with a high percentage of them following back.

    Google apparently doesn’t reward us for this type of behavior. We haven’t been losing too much ground with them, but we did fall from #2 to #7 for “SEO Software” over the past year. I don’t see many of those “influencers” beating us for our most important keywords, but I get the feeling Google doesn’t care that we have 4550 non-fake followers and maybe Google slightly dings us for following slightly more than follow us (dings us for practicing what used to be normal marketing activity before the age of Web 2.0 social media snobbery).

    Apparently the Google algorithm doesn’t search for what I call Snob Spam. Snob Spam is when someone unfollows a pile of people to make it look like they are more interesting than they really are…but where the smarter of the unfollowees just unfollow themselves, which ends up diluting the quality of the snob’s following. The spam comes in the form of unfollow messages.

    This should be measured more because some people believe that, if you mass unfollow, only bots will unfollow back, leaving you with a higher quality following.

    I believe the opposite. If we unfollowed 1000 people, we’d probably lose a few thousand dollars per month ($100,000 per year) in sales by insulting maybe a few dozen real life paying customers who would unfollow back and find another supplier.

    Just ask Southwest Airlines or United Airlines if they think it would be a good idea to unfollow 100,000 customers with the hope of retaining their business while only ticking off the bots.

    I’ve been asking Matt Cutts for over a year if Google had anything to say at all about this topic of people acting on the belief that Google will reward them for having a high follower to follow ratio. I keep writing that search engines should not reward this the way they do a high backlink to link ratio for a website. It’s not the same thing.

    It’s too easy to mass unfollow to make oneself look good (while losing the pros who get unfollow reports) and it’s too easy not to block fake followers.

    The next Penguin update might take a look at this issue.