Fake engagement isn't usually invisible — it's just never been checked against a baseline. Run the account's numbers against real creator benchmarks and the fakes tend to stick out immediately.
The direct answer
Pull the account's engagement rate, then compare it to two things: the median for its follower tier, and the ratio of likes to comments. Real engagement clusters in predictable ranges. Bought engagement usually breaks one of them — either the rate is implausibly high, or the comments look disconnected from the like count.
Red flag 1: engagement rate way above the top quartile
Across 17,000+ real creator profiles, here's where genuine engagement sits:
| Percentile | Engagement rate |
|---|---|
| Bottom 25% | 0.72% |
| Median | 1.87% |
| Top 25% | 5.56% |
A rate sitting comfortably in or below that top-quartile range is plausible. A rate that's 2-3x the top quartile — 15%, 20%, higher — on an account with a normal-looking follower count is not a "great creator," it's usually engagement pods, bots, or a follower purge that shrank the denominator without touching old bought likes.
Red flag 2: engagement rate that ignores follower-tier gravity
Engagement rate falls as follower count rises for real accounts — bigger audiences are less personal by nature. Our tiered data:
| Follower tier | Median ER |
|---|---|
| Under 10K | 2.23% |
| 10K–50K | 1.52% |
| 50K–500K | 1.93% |
| 500K+ | 2.46% |
If an account with 300K+ followers is posting 8-10% engagement — well outside even the 500K+ tier's realistic range — that's the single clearest tell. Real accounts don't skip the gravity that pulls every other large account down.
Red flag 3: comments that don't match the likes
Real engagement has a rough shape: our dataset's median is 1,127 likes to 36 comments per post. That ratio holds reasonably steady across tiers — comments scale with likes because both come from the same real audience. Fake engagement breaks this shape in one of two ways:
- Comment-starved: thousands of likes, a handful of comments. Bought likes and bot likes rarely bother generating comments.
- Comment-heavy but generic: a wall of comments that are all short, repetitive, or emoji-only ("🔥🔥", "Amazing!", "Love this"). This is the classic engagement-pod or comment-farm pattern — real audiences comment with more specific reactions.
Neither shows up in the raw engagement rate. You have to look at likes and comments separately, not just the combined percentage.
Red flag 4: engagement that doesn't move with content quality
This one takes a few posts, not just one number. Real audiences respond unevenly — some posts flop, some spike. An account where every single post lands within a tight 0.2% band of the same engagement rate, regardless of what's actually posted, is a sign of scheduled/bought engagement rather than an audience actually reacting.
How to actually check
- Run the account through the free engagement checker to get its real rate from actual recent posts.
- Compare that rate against the account's niche benchmark and follower tier — not the global median, since "good" varies enormously by niche and size.
- If the rate clears the top quartile by a wide margin, scroll the comments on a couple of recent posts and check whether they read like real reactions or generic filler.
None of these signals alone proves fake engagement — a viral post can spike a normal account's numbers for a week. But an account that's consistently far outside its tier's realistic range, post after post, is the pattern worth walking away from before you pay for a partnership.