A dated report puts a number on Instagram influencer fraud

The persona

No single persona is measured in this entry. The subject is the measurement backdrop behind any virtual-influencer audience claim: how much of the wider influencer market is estimated to run on bot or fake-follower activity, and who is doing the estimating.

What the documents establish

HypeAuditor's "State of Influencer Marketing 2022" report, dated 25 January 2022 and covering 2021 activity, states that "49% of Instagram influencers were impacted by fraud in 2021," a figure it says had fallen 14.3% since 2019. Its methodology section states the analysis "checked and combined data from 248k Instagram influencer accounts, 19.6k YouTube channels, and 14.6k TikTok accounts from the UK" held in HypeAuditor's internal database, with fraud calculated using the company's own machine-learning models rather than a disclosed manual audit. A separate, earlier claim from Points North Group, a fraud-measurement vendor archived from its now-defunct site as it stood in November 2019, stated more broadly that "the followers of top influencers can be over 50% purchased bots," without citing a specific sample or period on that page.

Character versus company

Neither document names a virtual persona; both describe the human-dominated influencer market generally, using each company's own detection tooling. That distinction matters for this site's purposes: a claim about "influencer fraud" made by a measurement vendor is that vendor's company-level finding about a market it sells services into, not a statement about any specific virtual persona's account. Applying either number to a particular persona's follower count, without that persona's own audited figure, substitutes an industry backdrop for a fact the operator or platform has not disclosed for that account.

What to watch

HypeAuditor's cited post does not define what counts as an account being "impacted by fraud," beyond stating it draws on audience, content, subscription and activity signals, and its sample is UK-specific rather than global. Points North Group's own site no longer resolves, which is worth noting when a figure like this circulates without the original page still live to check. Both companies sell the kind of fraud-detection or measurement tooling their reports describe the need for, a commercial interest worth naming rather than ignoring.

• Does the fraud percentage a report cites come from the same market, platform and period as the persona whose numbers a reader is trying to check?
• What does the provider disclose about how it distinguishes a "fraudulent" account from an inactive or automated one?
• Has a provider's own commercial interest in selling fraud detection been named alongside its published estimate?

Read narrowly, these documents establish that at least two measurement vendors have published double-digit fraud estimates for the influencer market at large; they do not establish a fraud rate for any specific persona this site covers.