Comparing real influence and fake followers

Image: New York Times

The New York Times published an expose of how “influencers” pay to boost their follower counts on Twitter, using services like Devumi. Real influence comes from ideas that spread and generate change in how people think. You can’t measure that from follower counts.

The Times article “The Follower Factory” is worth a read. According to the Times, Devumi has built an army of 3.5 million automated Twitter accounts, which, for a price, will follow and retweet whatever you want. Among those who have bought followers, according to the article, are Dell CEO Michael Dell, model Kathy Ireland, Randy Bryce (who is running against Paul Ryan for Congress), Treasury Secretary Steven Mnuchin’s wife Louise Linton, and celebrity baker Paul Hollywood.

Do fake followers matter? The easiest (and least accurate) way to measure influence is to count followers. Twitter’s algorithms also boost posts with many retweets; it’s not clear if they adjust that algorithm based on retweets from robots. So follower count does matter in rising above the noise.

You can actually measure the integrity of an account’s followers using TwitterAudit. The TwitterAudit algorithm is unsophisticated, but it is a rough measurement. For example, Robert Scoble has 429,000 followers, of which 99% are real, according to TwitterAudit. This is because he actively blocks fake followers. Jeremiah Owyang is at 98% real, I’m at 92%, and Natalie Petouhoff is at 61%. But being less than pure on TwitterAudit is not necessarily a sign of buying fake followers. For example, according to TwitterAudit, only 43% of Lady Gaga‘s 76 million followers are real, as are 67% of Donald Trump‘s 47 million followers, and 64% of Hillary Clinton‘s 22 million. In these cases, it seems likely that fake accounts like Devumi’s automatically follow these famous people to appear less suspicious. Short of assiduously blocking fake accounts like Scoble does, there’s no way to prevent fake accounts from following you.

In an ideal world, Twitter would police and deactivate these accounts. Mark Cuban has called on Twitter and Facebook to work harder to verify that accounts are associated with real people. But anonymity has benefits, allowing people who could be targeted for unpopular or dangerous beliefs to have a say. The problem of fake social media accounts isn’t going away any time soon.

If followers can be fake, how can you measure influence?

I believe that raw measures of influence, like follower counts and Klout scores, are corrupt anyway. While they’re easy to calculate, they don’t accurately represent the true value of influence.

Let’s take an example. Imagine two thinkers, say, experts in the auto industry.

The first expert is Sally Ubik. Sally assiduously collects followers; she follows back every one. She finds and posts links to every possible article about cars, including every article on Jalopnik. She retweets everything Elon Musk tweets. She regularly posts and reposts funny or ironic pictures — memes — which spread far and wide. As a result of her activity, she has 150,000 followers on Twitter, and similarly large followings on Instagram and Facebook.

The second expert is Alistair Wiseman. Alistair has developed a comprehensive theory about the auto industry, based his quantitative analysis of 25 years of auto sales. His predictions in the last five years have been 80% true. He gives talks at industry conferences and regularly meets with top executives in the business. He blogs about his ideas and distributes them in a subscription-only newsletter. He gets quoted in articles in mainstream media because of his expertise. He also tweets occasionally and has 11,000 Twitter followers, including Elon Musk, the CEO’s of the ten largest auto companies, and the most prominent business and trade reporters.

Who has more influence, Sally or Alistair?

When someone in a position of power needs to make a decision, who are they more likely to consult?

When there is a shift in the way the industry runs, whose ideas are more likely to influence that shift?

If there is a shift in the value of auto industry valuations and stock prices, reflecting a widespread change in investors’ concept of where the value in the industry comes from, whose ideas are more likely to be a cause of that shift?

Who are people more likely to talk about or disagree with?

Alistair has influence. Sally has only followers.

How to measure influence

Influence, in my view, is the ability to create change.

Influence comes from ideas. In the absence of ideas, there is no influence.

Rachael Denhollander, the brave woman who first came forward to accuse Larry Nassar of sexual abuse of generations of gymnasts, has influence. Her testimony will forever change gymnastics, as well as how we think aboaut sexual assault by people in authority. She has no Twitter account, but she had an idea of how the world needed to be different, and now we all know about it.

The writer Steven Johnson has provocative ideas about Bitcoin. He has influence.

Diane Hessan, who writes about interviews with her panel of voters and what they think of current political events, has influence.

Oprah Winfrey, who made a notable speech about race, sex, and truth at the Golden Globes, has influence.

Here are a few things that help define the influence these people have.

  • They have a clear idea what they believe the truth is.
  • They seek ways to spread that truth.
  • People in a position to create change listen to them. This includes both masses of everyday people and smaller numbers of people in power.
  • After interacting with them, you may think about the world differently.
  • Over time, you can see their ideas growing and evolving to reflect changing conditions in the world.
  • You can disagree with them, but you cannot in good conscience ignore them.

So stop counting followers and start understanding where real influence comes from. If you seek that influence, start by building and testing your ideas. If your ideas are strong enough, a moderate amount of effort will spread them. If they’re not, no amount of effort will gain you actual influence.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.


  1. Quality matters more than quantity. Unfortunately, I fear that I hold the minority view. You reminded me that I need to purge some of my fake followers. 95% is too low for me.

  2. Great post. I think the market will shift towards your view anyway. One reason why quantity mattered so much was because Influencers were often in the audience selling business (as opposed to the actual influence business), primarily as an alternative form of distribution. So, for example, if I was a Publisher of funny clickbait, I could either pay Facebook to promote my posts, or I could pay George Takei to repost my content to his followers at a discount “organically”. Sort of like taking the local roads to avoid a toll booth on the highway. The Social Networks are shutting that model down in several different ways.

  3. Nice post, Josh! I shared it with my Principles of PR class at UNC-Chapel Hill. BTW…I believe tech and internet startups from my experience are notorious for buying friends and followers. A vanity metric at best; bullshit to use your word at worst.

  4. Hi,

    Hope you are doing great!

    I am Alishba, works for an online digital marketing agency Futuristic Artists
    I came across your site and got impressed with the frequent updates on it
    Wondering if you accept sponsored blog posts on your blog?Kindly let me know price per post
    Also, let me know price per in-content text link placement on an existing blog post of your site


  5. Pingback: Razum i uticajnost