The short answer
Ask every creator for screenshots of their audience location and age, and look at the median views on their last ten posts, not the best one. Compare their engagement with others in the same tier on the same platform. Read their comments, and check how their sponsored posts perform next to their normal ones. A creator who passes those checks at 20K followers is a better buy than one who fails them at 200K.
Why follower count misleads
Follower count measures how many people chose to follow a creator at some point. It doesn’t measure who those people are, where they live or whether they still watch. In SushiVid’s data on Malaysian creators, engagement among the middle half of micro creators on Instagram ranges from 0.51% to 5.03%. Two creators priced the same can differ tenfold in how many people engage.
The eight checks
-
Audience location
Ask for a screenshot of the audience section of their insights, showing top countries and cities. For a product sold only in Malaysia, most of the audience should be here. If you have a single outlet in Penang, a creator whose followers are concentrated in the Klang Valley won’t bring in walk-ins.
-
Audience age and gender
Ask for the same screenshot’s age and gender breakdown. A creator whose content looks youthful may have a much older audience, or the other way round.
-
Median views, not the best post
Write down the views on their last ten posts and take the middle value. One viral video can carry a profile for months. The median tells you what to expect from your post.
-
Engagement against their tier
Engagement rate is likes, comments, shares and saves divided by followers (or by views, if that’s how you compare across creators). Use the same method for everyone, and compare a creator only with others of similar size on the same platform. The benchmarks are below.
-
What the comments say
Read thirty or so comments across recent posts. Real audiences ask questions, tag friends and argue, in Bahasa, English, Mandarin, Tamil or a mix of them. Rows of emoji or “nice post” from accounts with no profile photo are a warning sign.
-
Sponsored posts against organic ones
Find their last few paid partnerships and compare their views with the creator’s normal posts. Some creators hold their audience’s attention through an ad. Others lose half their views as soon as a brand shows up.
-
Who they’ve worked with
Check whether they promoted a direct competitor recently. If they did, their audience has just heard the opposite pitch, and you may need exclusivity terms.
-
How they work
Note how quickly they reply, whether their rate card is clear and whether they ask sensible questions about the brief. A creator who is hard to deal with before you pay is unlikely to improve after.
Engagement benchmarks
| Tier | TikTok | |
|---|---|---|
| Nano1K–10K | 3.08% (0.89–7.64%) | 8.95% (2.73–27.63%) |
| Micro10K–50K | 1.79% (0.51–5.03%) | 8.22% (1.88–26.67%) |
| Mid50K–500K | 1.83% (0.55–5.29%) | 6.76% (1.68–19.7%) |
| Macro500K–1M | 1.17% (0.31–3.26%) | 4.84% (1.44–14.28%) |
| Mega1M+ | 0.85% (0.26–2.46%) | 3.9% (1.28–10.58%) |
Source: SushiVid, 22,579 Malaysian creator profiles, October 2026.
Engagement falls as follower counts rise, on both platforms. TikTok figures run three to five times Instagram’s at every tier because the platforms count differently, so never compare a TikTok rate with an Instagram one. Below the bracketed range for the tier, ask questions. Far above it, look for one viral post or bought engagement.
Signs of bought followers
- A sudden jump of thousands of followers in a day or two with no viral post to explain it. Free tracking tools show follower history.
- High follower count, low views. A 100K-follower account whose posts get 1,500 views is not reaching 100K people.
- A local creator whose audience screenshot shows most followers in countries where they don’t post or sell.
- Comments that could appear under any post, posted within minutes of each other.
- A creator who won’t share insights screenshots, or sends them cropped.
None of these proves fraud on its own. Two or more together is reason to move on. The Content Forum’s 2026 draft guidelines specifically name fake followers and bought comments, so this may soon be a compliance issue as well as a waste of money.
A scorecard you can copy
Score each creator from 1 to 3 on each line, multiply by the weight and add up. Adjust the weights to your goal. For a store opening, location might count three times.
| Check | Weight | 1 | 3 |
|---|---|---|---|
| Audience location | ×3 | Under half in your market | Mostly in your market |
| Audience age and gender | ×2 | Mostly outside your buyer | Matches your buyer |
| Median views (last 10) | ×2 | Under 10% of followers | Over 30% of followers |
| Engagement against tier | ×2 | Below the bracket | Above the median |
| Comment quality | ×1 | Mostly generic | Questions and conversation |
| Sponsored against organic | ×1 | Sponsored far lower | About the same |
| Content fit and quality | ×2 | Off-brand or rough | You’d run it as an ad |
The view thresholds are our working rule of thumb, not a published benchmark. TikTok views often exceed follower counts, so for TikTok compare a creator with others of similar size.
Many small creators or a few big ones
For most launches, combine them. A few larger creators give the campaign reach and a moment people notice. A wider group of nano and micro creators gives you more videos, more angles and a better chance of a breakout. In one launch SushiVid reported, 3 of 125 posts drove about two-thirds of the views. You can’t pick the breakout in advance, but you can book enough creators to get one, and then put ad spend behind it.
Our rate guide shows what each tier costs, and the budget calculator lets you test different mixes.