How to tell if someone bought followers or likes on Instagram

Ask any influencer-vetting tool whether an account has fake engagement and it will answer with a score. Ask it how it knows and the answer is always some version of the same thing: it looked at the public profile, counted likes and comments, compared them to the follower count, and inferred the rest.

That is the best anyone can do from outside. It is also why the answers vary so much between tools — they are estimating the same hidden quantity with different guesses.

If you have an account’s follower count, likes and reach in front of you, the fake follower check runs the tests below for you. The rest of this post explains why they work.

Why can’t you tell from the outside?

Because the two numbers that would settle it are not public. Reach — the count of unique accounts Instagram showed the post to — is served only to the account holder. So are saves and shares, so are profile visits, and so is every figure describing what somebody did after seeing a post.

Bought likes are cheap; bought reach is not a product anybody sells, because reach is what Instagram did, not what an audience did. That asymmetry is the whole test — it is invisible from outside the account and arithmetic from inside it.

What bought engagement looks like against reach

Five patterns, in rough order of how hard they are to argue with. The first three are the ones that cannot be explained away.

1. Likes exceeding reach. A like requires having seen the post, and reach counts everybody who did. More likes than reach is not an unusual week; it is impossible without something outside Instagram’s own delivery.
2. A like rate that no audience produces. Likes over reach, medianed across the window. Above roughly 45% on an account with a real-sized audience, the question stops being whether it is unusual.
3. Followers arriving while reach stays flat. If the follower count climbs sharply over a month while daily reach does not move, the growth did not come from anyone seeing anything.
4. A comment-to-like ratio that a few posts share and the rest do not. The shape an engagement pod leaves: comments clustered on the posts that were submitted, absent everywhere else.
5. Likes that do not track reach day to day. In an organic account, the days that reached more people collected more likes. Engagement that ignores delivery is engagement that came from somewhere else.

What a real account looks like

Unremarkable, and consistent. Likes well under reach on every post. A like rate in the single digits or low tens. Follower movement that goes the same direction as reach. Comments that scale with likes across the whole window rather than on three posts.

None of those are impressive numbers. That is the point: an integrity check is a check for the absence of something, and passing it should look boring. A grading product that turns it into a headline is selling suspicion.

What this method cannot catch

Being honest about the limits is what separates a check from a scare. Three things get past it.

• Engagement pods. Real accounts, really looking at the post, really commenting. The behaviour is coordinated rather than fake, and only the comment-ratio pattern picks it up — and only when the pod is inconsistent.
• Bought followers with no engagement attached. They depress the rates rather than inflating them, so the arithmetic above reads them as a weak account, not a dishonest one. The follower-against-reach pattern is what catches those.
• Small samples. Eight measured posts is the floor below which an anomaly cannot be told apart from a quiet fortnight. Any verdict on fewer is noise wearing a percentage sign.

What a brand should ask for

Not a screenshot of the Insights tab. A screenshot is a number with no sample, no window and nothing stopping it being edited, and asking for one puts the brand in the position of trusting an image.

Ask instead for something computed from the account’s own data that states what it was drawn from: how many posts, over what window, and what share of the score had no data behind it. A figure that arrives with its sample attached is a figure you can argue with, which is the only kind worth having.

That is what the account grade publishes — the five checks above run on every grade, a hard failure caps the letter whatever else scored well, and the whole thing recomputes when the link is opened. The rate underneath it is engagement rate by reach, for the same reason: it is the one an outside tool cannot compute.