How to spot bought engagement before you pay for it

Follower count is the easiest number in this industry to buy, and engagement rate is the second easiest. Both are still the two numbers most shortlists are built on. Here is a manual audit that takes about ten minutes per creator and catches the overwhelming majority of problems.

1. Look at the shape of the growth

Ask for a screenshot of the follower graph from the creator's professional dashboard, over 12 months. Organic growth is lumpy but continuous — steps up after a post lands, plateaus in between. Bought growth is a vertical line followed by a flat plateau, or a sawtooth as purchased accounts get purged. One vertical step can be a viral post; three identical vertical steps cannot.

2. Read twenty comments, not the count

Real comments reference something specific to the post — the room, the joke, the third product on the shelf. Bought comments are generic, complimentary, emoji-heavy and arrive in a tight cluster within minutes of posting. Two tells that are nearly conclusive:

• Comment timestamps clustered in the first few minutes, then near-silence.
• The same handles commenting on every post with interchangeable phrases — this is an engagement pod, which is not fraud exactly, but it is not audience either.

3. Check audience geography against your delivery map

Ask for the audience-location breakdown. Two failure modes: a large share of followers in countries the creator has no connection to (a classic purchased-follower signature), or a legitimate audience concentrated in cities you do not serve. The second is not fraud, but it costs you the same amount of money.

4. Use the ratio that is hard to fake

Saves and shares per thousand views. Both require intent, both are visible to the creator in their own analytics, and neither is a standard product in the engagement-purchase market. A creator with a strong like rate and near-zero saves has approval without interest, which does not sell anything.

5. Compare reels views to follower count

For most healthy Indian creators, median reel views land somewhere between 20% and 150% of follower count. Consistently far below 10% suggests either purchased followers or an audience that has stopped watching — practically, the same problem. Consistently far above suggests the account grows on discovery rather than following, which is fine and often excellent, but it means the follower number is not what you are buying.

6. Check the last twenty posts for conflicts

Competitor collaborations inside your exclusivity window, categories you cannot be adjacent to, and anything politically or communally charged. This is the check that is skipped most often and costs the most when it is wrong.

What we do not bother with

Third-party "authenticity score" tools are useful for triage across hundreds of accounts and unreliable as a decision. They cannot see the creator's own analytics, so they infer from what is public, which is exactly the layer that is manipulated. Use them to sort a longlist, never to approve a booking.