How to vet a TikTok creator before you pay them

Most creator vetting stops at two numbers: follower count and a screenshot of one good video. Both are easy to make look impressive, and neither predicts much about whether a paid post will work for you.

The useful signals are public, they just take a bit of assembling. This guide covers what to check before you send a contract, which numbers mislead, and the two questions worth asking the creator directly because no external tool can answer them.

Start with engagement by views, not followers

Follower count tells you how many people once tapped follow. It does not tell you how many see a given post, because TikTok’s For You feed routinely pushes videos far past an account’s own followers.

The more useful ratio is engagement divided by views: likes, comments, and shares against the audience that actually saw the video. We measured this across a random sample of roughly 129,000 TikTok videos we track. The distribution:

Where a post lands	Engagement rate (by views)
Bottom quartile	3.3%
Median	6.5%
Top quartile	11.8%
Top 10%	18.2%

Set the bar to match their reach. Across a wider measurement of 1.37 million videos, engagement rate rises with the audience a video reaches through most of the range, and dips at both extremes. Par by band:

A typical video reaching	Par engagement rate
1,000 – 10,000 views	5.9 – 6.4%
10,000 – 100,000	6.3 – 6.6%
100,000 – 1 million	7.2 – 8.5%
1 million – 32 million	9.3 – 9.7%
32 million – 100 million	8.2%
above 100 million	6.6%

A single median is therefore the wrong yardstick for a creator whose videos travel a long way. Judge them against the band their typical video actually lands in — a creator whose videos routinely clear a million needs about 9.5% to be merely average, where a creator in the tens of thousands needs about 6.5%. Both are the same measurement; only the bar moves.

One caveat that matters when you are comparing notes with a creator or another agency: most published TikTok benchmarks divide by follower count instead, which produces figures in the 2–5% range. Those numbers are not lower versions of these — they measure something different. Make sure everyone in the conversation is dividing by the same thing. We wrote up the full methodology, including the reach bands, in what counts as a good TikTok engagement rate.

Check the rate across a creator’s last 20 or so posts, not their best one. A single outlier tells you they had one good week.

Look at the shape of the growth, not the size

Pull a creator’s follower count over the last few months and look at the curve.

Healthy growth is uneven but connected — it steps up when a post does well, then holds or drifts gently rather than collapsing. You can usually point at a specific video that caused each step.

Worth a question is a vertical jump with no corresponding post performance, or a series of flat plateaus separated by sudden steps. That pattern can have innocent explanations — a feature elsewhere, a cross-platform moment, a burst of ads — so treat it as a prompt rather than a verdict.

Be honest about the limits here. Public data shows you that an account jumped. It does not prove how. Anyone selling you certainty about fake followers from public numbers alone is inferring, not measuring.

Check consistency before you check peaks

For a paid post, a reliable floor matters more than a high ceiling. You are buying one video, and you want to know what an ordinary one does.

Two things to look at:

• Median views versus best-ever views. If the gap is enormous, their reach depends on catching lightning. Their median is closer to what you are actually buying.
• Posting cadence over the last quarter. A creator who posted twelve times in one week and then went quiet for a month has a different relationship with their audience than one who posts twice a week without fail.

Steady is usually worth more than spiky, and it is usually cheaper.

Read the comments, not just the count

Comment volume is a weak signal on its own. What the comments say is a much stronger one.

Skim the comments on a creator’s recent posts and sort roughly into three piles: genuine interest in the subject, generic filler, and criticism or controversy. High comment counts driven mostly by argument still count as engagement, but they are not an audience you want your brand adjacent to.

Look specifically for questions. Audiences that ask “where is this from?” or “does it actually work?” are audiences with purchase intent — and they are exactly the audiences where a product mention lands naturally rather than as an interruption.

Confirm the audience overlap is real

The most common failure in paid creator work is not a fake audience. It is a real audience that does not care about your category.

Look at what a creator’s best posts are actually about — not their bio, not how they describe themselves on a media kit, but the recurring subjects of the videos that performed. A fitness creator whose top posts are all comedy has a comedy audience. Your product will land in front of people who came for jokes.

Check the sounds and hashtags they use repeatedly too. Recurring, category-specific tags suggest they are established in a niche; a scatter of unrelated trending tags suggests they chase reach wherever it is.

The two questions to ask the creator directly

Some things are not public for anyone, at any price. For an account you do not own, no third-party tool can show you retention curves, audience demographics, or whether a video was boosted with ad spend. Any vendor claiming otherwise is worth a second look.

So ask, and make it part of the deal:

1. A screenshot of their native analytics for their last few posts — specifically average watch time and audience location. This is the only way to see it, and a professional creator will have it ready.
2. Whether any of their recent standout posts were boosted. Not disqualifying at all — plenty of good creators boost. But it changes what you should expect from an organic post, and a creator who volunteers it is a creator worth working with.

If either request causes friction, that itself is information.

A short checklist

Before you send the contract:

• Engagement by views across their last ~20 posts, against the band their typical video reaches
• Median views, not just their best video
• Posting cadence over the last quarter
• Follower curve shape — steps that match real posts
• Comment quality: interest, filler, or argument
• Subject matter of their top posts versus your category
• Native analytics screenshot for watch time and location
• A direct answer on paid boosting

None of this takes long once you have the numbers in one place. The part that takes long is collecting them.

How busypipe helps

Add the creators you are considering to a portfolio and busypipe tracks them from that point on: follower and engagement trend charts, per-post views, likes, comments, shares and saves, posting cadence, and the hashtags and sounds they use repeatedly. Post-level engagement is shown against a peer benchmark, so you can see whether a number is good for that kind of account rather than good in the abstract.

For posts with enough comments, we classify comment sentiment and intent, which turns the “read the comments” step above into something you can see at a glance. Portfolio Themes groups a creator’s recent posts by subject, which is the fastest way to check whether their actual content matches how they describe themselves. Everything exports to CSV for the deck.

One thing worth knowing before you start: history begins building when you add an account. If you are likely to evaluate a creator next quarter, add them now — the depth is the useful part, and it is the one thing that cannot be backfilled.

Signing up starts a 14-day Professional trial — 100 tracked accounts and a year of history, no card required. After it ends, the free plan keeps three accounts with 30 days of history.