The Average Views Lie: What We Learned From 675 TikTok Creator Videos

Creator marketing isn’t an experiment anymore. The IAB estimated that $37 billion would go to creator advertising in the US in 2025, growing about four times faster than media overall. But when brands vet creators, they still tend to look at how many followers someone has, scroll through their recent posts, and then check one number from the media kit: “average views per video”.

How reliable is that average? We pulled the most recent videos from 52 TikTok creators in skincare, fitness, cooking and personal finance, from accounts with ~11k followers to accounts with more than 4m, and worked out what you’d actually be paying for.

Short answer: the average is often the least useful number in the media kit. Here’s what we found, plus a vetting process that avoids the most costly mistakes.

How we did the study

In October 2026, we used Social Fetch’s API to grab the latest videos from 52 individual TikTok creators (no brands, shops or repost accounts), split into three tiers:

For each creator we took their latest videos, leaving out pinned posts and anything less than three days old (those are still picking up views). That came to 675 videos, a median of 14 per creator. Then we calculated each creator’s typical (median) views, their average (mean) views, their engagement rates, and how much of their views came from a single post.

It’s a snapshot, not a census. The sample is small, it comes from TikTok search results, and it covers a few weeks of activity. But the patterns were consistent enough to change how we’d approach any creator deal.

Finding 1: For most creators, the average isn’t the typical video

When a creator says “my videos average 100k views”, most people hear “a typical video gets about 100k views”. Our data says that’s often not true.

Across the 52 creators:

• For 73%, average views were at least 1.5x their median (typical) video
• For 52%, the average was at least double the median
• For 37%, it was at least 3x the median
• For 21%, it was at least 5x the median

The worst case was a cooking account with ~30k followers. Its median video got 617 views, but one viral video pushed its average above 92k. That’s about 150x its typical video. If you’d paid based on the average, you’d have expected roughly 150 times the views you were likely to get.

Mid-tier creators had the biggest gap. Their average was typically 3.7x their median, compared with 1.8x for micro creators and 1.6x for macro.

Lesson: always ask for median views, or work out the median from the last 15–20 videos yourself. The median tells you what your sponsored post is likely to get. The average tells you what happened once.

Finding 2: It might all be one viral video

The reason for the gap is concentration. For the typical creator in our sample, their single best video made up 36% of all their recent views. In 71% of accounts the top video was at least 30% of the total, and in a third of accounts (17 of 52) it was more than half.

Mid-tier creators were the most hit-dependent. Their best video was typically ~23x bigger than their typical one, and it made up a median 58% of their views.

That’s not a knock on them. It’s how short-form video works: every so often the algorithm picks up a video and pushes it way past the creator’s followers. But you can’t buy a viral video. You’re buying a typical one, with a small chance of a hit.

The good news is that below those hits, most creators were pretty consistent. The typical creator had almost no videos below a quarter of their own median. So once you know a creator’s median, it’s a reliable baseline.

Finding 3: Bigger accounts reach a smaller share of their followers

Followers measure past interest, not current reach. In our sample, the share of followers a typical video reached dropped as accounts got bigger:

Engagement = likes + comments + shares. For each creator we took the median of their videos’ engagement rates. Tier figures are the median across creators.

Under-delivery was common, and most common for the biggest accounts. Almost a third of all creators (16 of 52) had a typical video that reached fewer than 5% of their followers. For macro creators it was nearly half (8 of 17). One finance creator with 1.74m followers had a median of ~4.4k views per video, or 0.25% of their audience. Only one creator in the whole sample, a skincare creator with just under 100k followers, had a typical video that reached more people than their follower count.

So a big follower count doesn’t mean a big audience for your post. Pay for the reach a creator actually delivers, not the reach their follower count implies.

Finding 4: Engagement per view barely changes with size

Look at the last column in that table. Measured against followers, engagement drops as accounts grow, from 1.58% for micro to 0.28% for macro. Measured against views, it barely changes. People who watched a macro creator’s video engaged at about the same rate as people who watched a micro creator’s video, and slightly more on average.

So the familiar line that “micro-influencers have higher engagement” is partly down to the formula. Smaller creators reach a bigger share of their followers, which makes follower-based engagement look much better. But per person reached, the bigger creators in our sample did just as well.

This matters when you compare creators, and when you read industry benchmarks. Published TikTok engagement rates range from a few percent to well over 8%, while our own follower-based median was under 1%. That’s because everyone uses different formulas: dividing by followers or by views, using the mean or the median, counting saves or not, measuring brand accounts or creator accounts. Before you compare a creator with a benchmark, make sure both use the same method.

We recommend using both measures, for different questions:

• Median views ÷ followers: what share of their audience does this creator still reach?
• Engagement ÷ views: do the people who see the content actually care about it?

The five numbers to check before you sign

Based on those findings, here’s what we’d look at for any creator.

1. Median views

Take their last 15–20 non-pinned videos (skipping anything posted in the past few days) and find the middle one. It’s the most important number in any deal, because it’s your realistic expectation for a sponsored post.

Social Fetch’s free TikTok engagement rate calculator gives you a per-video breakdown for any public handle in seconds, with no signup. It’s an easy way to see the spread between a creator’s typical videos and their hits, without copying numbers out by hand.

2. Mean-to-median ratio

Divide average views by median views. Under 1.5 means steady performance. Over 3 means the average is being bumped up by hits, so don’t trust any media-kit number based on the average.

3. Reach rate

Divide median views by followers. In our sample, the middle values were ~23% for micro creators, 9% for mid-tier and 5.5% for macro. If a creator is way below their tier, especially under 5%, their audience is probably mostly inactive or no longer seeing their content. Ask why before you pay for followers you won’t reach.

4. Engagement per view

For each video, divide likes + comments + shares by views, then take the median. Ours came out at ~6–8%. Much lower could mean passive viewing. Unusually high, especially with thin comments, is worth digging into.

5. Comment quality

Numbers can be pumped up. Conversations are much harder to fake. The typical creator in our sample got around 1 comment per 1k views, so even a smaller creator should have enough comments to read. Look at 50–100 comments across their recent posts and ask:

• Do people respond to what was actually said, or is it all generic praise and emojis?
• Do followers ask where to buy things, or for recommendations?
• Does the creator reply?
• Do the same few accounts comment on every post? That could be an engagement pod.

Red flags worth pausing for

Some patterns should stop a deal until you’ve heard a good reason:

• A typical video reaching under 5% of followers, especially on a mid-tier account.
• A media kit that quotes average views but no median, with one or two huge hits on the profile.
• Big follower jumps with no viral video to explain them. Track the count for a few weeks before signing, or ask for growth screenshots.
• High engagement that’s empty: lots of likes, and comments that are all emojis from accounts with no profile picture.
• A feed dominated by sponsored posts. The audience may be tuning the ads out.
• Reluctance to share first-party analytics. Professional creators expect the request.

Fake audiences are still a big issue. In Influencer Marketing Hub’s 2026 benchmark survey, fake or bot followers were by far the most common quality issue (56.5% of the issues respondents ticked), way ahead of templated comments and bought engagement.

Ask for first-party data, then check it

You can learn a lot from public numbers, but only the creator can show you their audience demographics: country, age and gender. It’s getting easier to ask for that data in a verifiable way:

• TikTok One, TikTok’s creator marketing hub, lets brands search for creators and see engagement and audience information.
• Instagram’s Creator Marketplace got an API in 2025, and Meta announced that its Creator Insights endpoint would return audience demographics and average engagement rate for marketplace creators.
• YouTube started inviting creators in 2025 to opt in to sharing channel and audience data with brands.

Where you can, ask for platform-sourced data instead of screenshots, and check it against what you can see publicly yourself. If a creator says their audience is mostly in the US, but their comments are mostly in another language and they post at 3am US time, ask about it. Independent data helps too. Social Fetch, for example, can return estimated audience location for a public TikTok profile, so you get a second opinion on where a creator’s followers actually are before you see their own numbers.

Set up a repeatable vetting process

Vetting one creator by hand takes ~20 minutes. Vetting 50 for a campaign takes days. Teams that do it well tend to split the work into stages:

1. Discovery. Build a long list from hashtags, keyword searches, competitors’ partnerships and creators your customers already follow.
2. Automated screening. Pull recent videos for every candidate, calculate the five numbers above and filter out anyone below your thresholds.
3. Human review. Read comments, watch the content and judge brand fit, for the shortlist only. This is where your team’s judgment adds the most value.
4. Contract and brief. Agree what you’re paying for, how it will be measured and what data the creator will share.
5. Measurement. Compare each sponsored post with that creator’s own median.

Stage 2 is where automation really helps, and it’s the stage we automated for this study. With Social Fetch’s influencer data API, screening goes from days of copying numbers to a task that runs in minutes:

• All the major creator platforms, in one format. Profiles and recent posts from TikTok, Instagram, YouTube, X and more come back in the same structure, so you can use the same screening rules for every creator, whatever platform they’re on. Social Fetch covers 25 platforms in total.
• The numbers in this article, ready to go. Engagement audits for TikTok and Instagram profiles summarise a creator’s recent posts, and TikTok audience location data helps you check that a creator’s followers are in your market.
• Built-in campaign tracking. Monitors can check a creator’s account on a schedule and send new posts into your systems by webhook, so you can measure sponsored posts against each creator’s median without checking by hand.
• Works with the tools you already use. Pull results into a spreadsheet or CRM, or hook Social Fetch up to an AI assistant like Claude or ChatGPT and ask things like “which of these 40 creators reach more than 10% of their followers?”
• Built for real campaigns. You start with 100 free credits (no card needed), pay-as-you-go credits never expire between campaigns, and you don’t get charged when a lookup fails on Social Fetch’s side.

We grabbed all 675 videos for this study in minutes, and it cost well under $1 in credits. If you’ve only got a few creators to vet, a spreadsheet is fine. But once you’re screening dozens each month, automating the data collection pays for itself many times over.

A simple screening scorecard

Negotiate on median views

Data isn’t just good for picking creators. It changes the price conversation too.

Say two creators both quote $2k for a video. One has a median of 25k views. The other has a median of 8k views, but an average of 60k thanks to one viral post. Priced on median views, the first creator costs $80 per thousand views. The second costs $250, more than 3x as much, even though the second creator’s media kit looks better.

Some ways to handle it:

• Anchor offers on cost per thousand median views, then adjust for engagement quality and audience fit.
• For hit-dependent creators, offer a lower base fee plus a performance bonus if the post beats their median by an agreed margin. They get the upside if it goes viral, and you’re not paying upfront for that possibility.
• Buy the strong platform, not the bundle. If a creator’s TikTok works well but their Instagram only reaches 2% of followers, pay for TikTok.
• Add data-sharing terms to the contract, such as post-campaign analytics screenshots or platform-sourced reporting within a set number of days.

Most professional creators are fine with this approach. A brand that understands their numbers is easier to work with than one that only looks at follower counts.

Don’t forget compliance

Buying fake influence is now explicitly illegal in the US. The FTC’s rule on consumer reviews and testimonials took effect on 21 October 2024. It prohibits selling or buying fake indicators of social media influence, such as followers or views generated by bots or hijacked accounts, where the buyer knew (or should have known) they were fake and used them to misrepresent influence for commercial purposes. Knowing violations can bring civil penalties of more than $50k each. Vetting creators carefully protects your budget and your brand’s reputation, and it shows you aren’t knowingly paying for fake reach.

Lots of creators still ignore disclosure. When the European Commission checked 576 influencers in 2023, 97% had posted commercial content, but only 20% consistently disclosed it as advertising. In the UK, the ASA’s 2025 monitoring found only ~57% of influencer ads on Instagram and TikTok were likely to be properly disclosed. The FTC’s endorsement guidance is clear that “ad” or “sponsored” works, while vague tags like “collab” don’t cut it. Put the disclosure requirement in your brief, and check every post before you pay.

After the campaign, measure against the creator’s own baseline

Compare each sponsored post with that creator’s median, not with their best video, and not with another creator. Did it hit their median? Did engagement per view hold up, or did viewers scroll past the ad? Did your tracking links or discount codes convert?

Over a few campaigns, you’ll build up an internal record of which creators deliver for your brand, and which types of creator (by tier, niche and platform) work best for your product. That’ll be worth more than any external benchmark.

The takeaway

Our 675-video sample pointed to five practical rules:

1. Use median views, not averages. For half the creators we studied, the average was at least double the typical video.
2. Expect one hit to dominate. For the typical creator, over a third of recent views came from one video.
3. Don’t pay for followers. Macro creators’ typical videos reached ~5.5% of their followers, and almost half reached under 5%.
4. Measure engagement per view. Per person reached, bigger creators engaged their viewers just as well as smaller ones.
5. Verify, then measure. Ask for first-party data, check it against public signals, and judge each post against the creator’s own baseline.

Creator marketing deserves the same rigour as any other paid media. Most of the data you need is already public. You just have to look at the right numbers.

---

Methodology: Data collected with Social Fetch on 17 October 2026 from 52 public TikTok accounts (18 micro, 17 mid-tier, 17 macro) across skincare, fitness, cooking and personal finance, found through TikTok search. We used up to 30 of each creator’s most recent videos, excluding pinned videos and videos less than three days old (675 videos in total). Engagement means likes plus comments plus shares. Tier figures are medians across creators. This is a snapshot of recent activity, not a representative survey of TikTok.