How to spot fake Instagram followers: seven checks in ten minutes

The seven checks

None of these is conclusive alone. Three or more together, in the same direction, is enough to remove a creator from a shortlist before any money is discussed.

• Follower-growth curve — gradual slope, or vertical steps.
• Engagement rate against the indicative band for the tier.
• Comment quality — sentences about the place, or emoji strings and “nice pic”.
• Audience geography against the markets you actually sell to.
• Story views as a proportion of followers.
• Saves and shares relative to likes.
• Reach against impressions, and the follower share of reach.

The growth curve is the fastest tell

Real accounts grow unevenly but continuously: a slow base, occasional steps when a post travels, then a return to the underlying slope. Bought followers produce a different shape — a flat line, a vertical jump of several thousand over two or three days, then flat again, with no corresponding rise in likes or comments.

The second pattern to look for is decay. Purchased accounts get removed in platform sweeps, so a bought-heavy profile shows sawtooth movement: sharp gains followed by sharp unexplained losses. A genuine viral moment leaves a visible trail — the post that caused it is still there, and engagement rose with the followers.

Ask the creator to walk you through any step you cannot explain. A legitimate answer is usually immediate and specific: a feature by a large account, a reel that reached beyond followers, a press mention.

Engagement rate, and the bands that matter

Calculate engagement on feed posts as likes plus comments divided by followers, averaged across the last ten to twelve posts, ignoring outliers at both ends. For reels, use views and saves instead; measuring a reel against followers understates it badly because reels reach non-followers by design.

Indicative bands for European travel accounts, and they are indicative rather than rules:

• Nano, 1,000–10,000 followers — 4–8%.
• Micro, 10,000–100,000 — 3–6%.
• Mid, 100,000–500,000 — 1.5–3.5%.
• Macro, 500,000+ — 1–2%.

Why both ends of the range are suspicious

Engagement far below the band suggests followers who are not watching, whether bought or simply lapsed. Engagement far above it is equally worth investigating, because engagement pods and comment groups produce inflated interaction on accounts with genuine but inattentive audiences.

The signature of a pod is timing and composition: a burst of comments within the first fifteen minutes, from the same twenty or thirty accounts, on every post, all of them creators of similar size. Open three posts from different months and compare the commenter lists. Overlap of most names across all three is the answer.

Comments, and what a real one looks like

A real travel audience asks operational questions. Which room is that, was the water warm in May, how far is the walk from the port, did you hire a car. Those comments are the single best qualitative indicator that the audience is planning something rather than scrolling.

Bought engagement reads generically because it has to be generic to be sold at scale: “amazing”, “so beautiful”, flame emoji, three-word compliments with no reference to the location. A feed where a photograph of Amalfi and a photograph of Cappadocia attract the same twelve comments is not a feed with an audience.

Check whether the creator replies. Consistent replies signal an account that is worked rather than harvested, and replies are also where you see whether the audience is in your language.

Audience geography, the check most brands skip

This is the most commercially important number on the whole list and it is the one least often requested. A 60,000-follower creator posting in English about the Aegean can have an audience that is 70% in markets that will never fly to your property.

Two patterns should stop a deal. The first is a geography split with no plausible relationship to the content: a German-language account about the Alps with a plurality of followers in a country where nobody speaks German and few travel to Europe. The second is extreme fragmentation, where no single market exceeds a few per cent, which usually indicates followers acquired in bulk rather than earned by topic.

Set a threshold before you look. If your guests come from the UK, Germany and the Netherlands, require that those three account for a meaningful share — 40% is a reasonable floor — and treat everything else as decoration.

Stories, saves and the reach ratio

Story views typically run at 5–15% of follower count for an active account. Below 5% consistently, the follower number is doing very little work, whether because the followers are fake or because they stopped watching years ago. From a campaign’s point of view the distinction hardly matters.

Saves and shares are the metrics that actually predict travel intent. A save is somebody filing your property for a trip they have not booked yet; a share is a recommendation to a specific person. Both are difficult and pointless to fake, which is exactly why they are worth asking for. A post with 4,000 likes and eleven saves is decorative.

Finally compare reach with impressions and look at the follower share of reach. Healthy accounts show reach comfortably below impressions, meaning people watched more than once, and a meaningful proportion of reach coming from non-followers on reels. Reach that barely exceeds a small fraction of followers is the same story as low story views.

Why a screenshot is not evidence

Every check above can be run from a screenshot of an Insights panel, and every one of them can be defeated by one. A screenshot shows a single panel, on a single day, over a date range the creator chose, and it can be altered in a browser in under a minute. It is a claim about metrics, not the metrics.

The alternative is to read the same figures through the platform’s own API under read-only OAuth, where the creator authorises access at Instagram, TikTok or YouTube and the numbers come back from the platform rather than from an image. Follower history, reach, impressions, saves and audience geography arrive as data with a timestamp, and they refresh.

This is how ETC handles creator metrics: official OAuth only, never collecting or storing passwords, read-only minimum scopes, tokens encrypted in a vault held separately from the application database, one-click disconnect that revokes access at the platform, and no scraping or cookie sharing at any point. The badge that results reads “Metrics Verified ✓” rather than “Connected”, because the claim being made is about the numbers, not about a login.

The practical effect for a brand is that the first two hours of diligence disappear. You still have to judge whether the audience is right for your property, which is a commercial decision. You no longer have to work out whether the numbers in front of you are real.