The Influencer-Fraud Stats Everyone Cites Are Made Up. Here Are the Real Ones.

The short version: The two most-quoted influencer-fraud stats — "37% of the average influencer's followers are fake" and "$4.6–4.8 billion lost to fraud" — trace back to unverifiable SEO blogs citing each other. The defensible numbers are lower and better-sourced. And even the real ones can't tell you what actually matters: whether this creator is faking it.

Key Takeaways

• Avoid the viral figures: "37% fake followers" and "$4.6–4.8B in fraud" trace to content-farm blogs with no primary source.
• Defensible instead: on average ~22% of an influencer's followers are suspicious, and roughly half of Instagram influencers have engaged in some follower fraud (HypeAuditor, audits).
• The credible dollar anchor is $1.3B lost to influencer fraud (2019), from a university-backed study (CHEQ × University of Baltimore, 2019).
• An industry average can't vet your creator. Run a free receipt to measure the specific one you're paying.

Open any "influencer fraud statistics" listicle and you'll meet the same two numbers: 37% of followers are fake and fraud costs brands $4.6 billion a year. They're everywhere, they sound authoritative, and they're essentially made up.

We went looking for the primary sources. There aren't any.

The stats to stop citing

The "37% fake" and "$4.6–4.8B" figures trace back through a chain of SEO blogs that cite each other, none of which lead to an actual study, dataset, or methodology. They're the marketing-stat equivalent of a rumor that got a citation because it got repeated. If you can't click through to a method, don't put it in your deck.

The numbers that hold up

Here's what's actually defensible, side by side:

Notice these are smaller than the viral versions. That's usually how it goes: the honest number is less dramatic than the one built to be shared. The fraud problem is real and serious — you don't need to inflate it, and inflating it just gets you quoted next to the fakes.

Why even the real average is the wrong tool

Here's the part that matters most. Say the true average is 22% suspicious followers. You're not hiring the average. You're hiring one specific creator, and her number could be 3% or 63%. An industry statistic tells you the category has a problem; it tells you nothing about the person you're about to pay.

This is the quiet reason fraud stats get inflated in the first place: a scary average is easier to sell than the boring truth, which is that averages are useless at the point of decision. The only number that helps you is this creator's number, measured now. That's the entire reason a receipt exists — it replaces "the industry is ~22% fake" with "this account is X% fake," which is the only figure you can actually make a call on. For what that looks like in practice, see how to spot bought likes and how the reported engagement rate hides fraud.

Run a free receipt on any creator and trade the industry average for the real number on the person you're hiring.

Because a made-up statistic and a fake audience have the same problem: nobody checked the source.

Run a free receipt on any creator — just your email. No card, no trial.