Playbook replay
Authenticity and fake-follower flagging
Reads engagement consistency across recent posts and scores the risk that an audience was bought.
Run for Lumira, a demo brand we invented for this page. No customer data. Nothing here runs against a live account.
Every session here was run once and saved, against a demo brand or public data. On your own account Naise works from your brand voice, your campaigns and your connected channels, so it reasons about your data instead of a stand-in.
Bought followers are invisible in a follower count and obvious in a ratio. Likes can be purchased cheaply, comments cannot be purchased convincingly, so the ratio between them is the signal worth reading.
Reading the last 30 posts for the shape of the engagement rather than its level, and checking the comment-to-like ratio, which is harder to fake than either number alone.
Risk score: high. Three signals, two of them strong.
Signal 1, follower growth with no post to explain it (strong).
| Window | Followers gained | Posts in window | Best post views |
|---|---|---|---|
| Weeks 1 to 4 | +2,100 | 9 | 41k |
| Week 5 | +38,000 | 1 | 12k |
| Weeks 6 to 9 | +1,400 | 11 | 38k |
A 38,000-follower week off a 12,000-view post is not organic. Nothing in that week reached enough people to produce that growth.
Signal 2, comment-to-like ratio collapsed after week 5 (strong).
Before: roughly 1 comment per 40 likes, which is normal for this category. After: roughly 1 per 310. The likes kept scaling with the new follower count and the comments did not move at all.
Signal 3, comment quality (weak, informational).
A visible share of post-week-5 comments are single emoji or generic four-word compliments. That pattern is consistent with engagement pods, but plenty of real accounts look like this too, so I would not score on it alone.
What I would do: ask for a screenshot of their audience-country and age breakdown from their own analytics before contracting. If the followers were bought, the country mix is usually the tell, and it is the one thing you cannot see from outside.
What this is not. This is a heuristic read of public posts, not a forensic audience audit. I can see the shape of the engagement, not the accounts behind it. Everything above is consistent with bought followers and also consistent with one week of paid follower ads that under-delivered on quality. High risk means worth checking, not proven.
Naise created an artifact
Three signals with the evidence behind each, and an explicit statement of what a public-post heuristic cannot see.
Signals
| # | Signal | Strength |
|---|---|---|
| 1 | +38k followers in a week off a 12k-view post | Strong |
| 2 | Comment-to-like ratio fell from 1:40 to 1:310 | Strong |
| 3 | Generic single-emoji comments | Weak |
Next step
Ask for their audience country and age breakdown before contracting.
Limit: heuristic read of public posts, not a forensic audience audit. High risk means worth checking, not proven.
That was a sample. Run it on your own brand.
Naise reads your brand, your channels, and your market, then does this against your real data.
Run this playbook free