TikTok Creator Vetting Before an Influencer Deal

A creator's media kit shows a follower count and a handful of screenshots they picked themselves. Neither tells a brand what actually happens when their next post goes out, or whether the audience behind that follower count is real. Before a contract and a payment go out, the two things worth checking directly are whether recent posts are genuinely performing and whether a sample of the follower list looks like real accounts.

If you're an agent (or building one) reading this rather than a human, the full machine-readable schema for every endpoint below lives at mindcase.co/skills.md.

What does a media kit not show you?

A creator's own numbers are self-reported and self-selected. Three things a media kit or a screenshot won't surface:

• Real per-post engagement. A screenshot of one high-performing video says nothing about whether that post was typical or an outlier.
• How saturated their feed already is with sponsored content. A creator who ran five brand deals last month has a different audience reaction waiting than one who hasn't posted an ad in weeks.
• Whether the follower count is made of real accounts. A large follower number can include a meaningful share of empty or inactive accounts — nothing in a public profile page flags that.

How do you check whether recent posts are actually performing?

Say a skincare brand is vetting glow.by.leah for a paid post before signing anything. TikTok Posts API pulls a creator's own recent videos by profile handle and returns 27 fields per post, including views, likes, comments, shares, saves, and isSponsored.

curl -X POST "https://api.mindcase.co/v1/data/tiktok/posts/run?wait=true" \
-H "Authorization: Bearer $MINDCASE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
  "params": {
    "profiles": "glow.by.leah",
    "resultsPerPage": 12
  }
}'

# $0.002 per post

Averaging (likes + comments + shares + saves) / views across the last dozen posts gives an actual engagement rate instead of trusting a screenshot of the best one. The isSponsored and isAd flags on each row also show how many of those recent posts were already paid content — a feed with three sponsored posts in the last twelve reaches its audience differently than one with zero.

How do you sample their followers for authenticity signals?

TikTok Followers API returns the actual follower list of a public account, not just a count — each row is a full profile with handle, bio, followers, totalLikes, and videos for that follower.

curl -X POST "https://api.mindcase.co/v1/data/tiktok/followers/run?wait=true" \
-H "Authorization: Bearer $MINDCASE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
  "params": {
    "profiles": "glow.by.leah",
    "maxFollowersPerProfile": 200
  }
}'

# $0.001 per profile

A 200-follower sample isn't a fraud audit, but the share of accounts with zero videos and zero followers of their own is a real signal — a normal audience has a mix of active and lurking accounts, while a follower list padded with empty ones skews heavily toward both counts sitting at zero.

How do you turn both pulls into a single vetting check?

Both endpoints return the current state of the account — there's nothing to schedule here, since this is a one-time check before a specific deal, not an ongoing watch:

Before signing a creator:
1. Pull their last 10-15 posts (as above) and compute the average
   engagement rate, plus the count of isSponsored/isAd posts among them.
2. Pull a 200-follower sample (as above) and compute the share with
   zero videos and zero followers of their own.
3. Weigh both against the rate the brand is being asked to pay — a low
   engagement rate or a heavily padded-looking follower sample is
   grounds to renegotiate or walk away, not just a number to note.

Neither check proves anything on its own — a real account can have a quiet month, and a real follower can genuinely post nothing. Together, though, they replace a media kit's cherry-picked screenshot with two numbers pulled from the account's actual current activity.

What can't you get from Mindcase?

Two cases.

Historical performance. Both endpoints return the account as it looks right now. There's no way to pull what a creator's engagement rate or follower list looked like six months ago — a vetting check only reflects the account's current state, not its trajectory.

A fraud verdict. Nothing in either response is labeled "fake" or "real." The follower-authenticity check above produces a share of accounts matching a pattern worth flagging, not a confirmed bot count — TikTok doesn't publish that distinction, and neither does Mindcase.

Which endpoint should you use for which job?

FAQ