TL;DR
- An influencer is somebody whose recommendation changes what other people buy, which makes purchase behaviour the test rather than audience size.
- GMV per follower ranged from $58.21 to $0.52 across the US top ten TikTok Shop creators in a single month, a spread of 112 times.
- The largest account in that group placed ninth by revenue, with 3.23m followers producing $1.69m while a 25,770 follower account produced $1.50m.
- Average views and follower counts drift apart, with one 2.5m follower creator averaging 315 views a video while leading the revenue table.
- Tiers describe audience size while the outcome moves independently, so a brand hiring on follower bands is paying for the wrong variable.
Dictionaries define an influencer by audience, usually as a person with the power to affect what others buy.
The second half of that sentence is the part worth keeping, because it points at an outcome, while the first half points at a vanity number that has stopped predicting anything.
Cruvalytics publishes the US creator leaderboard each week, and for the 30 days to 17 November 2026 the ten creators driving the most TikTok Shop revenue carried audiences ranging from 25,770 followers to 3.23m.
This piece defines the term through that data, sets out the three tests a brand can actually run, and explains what the follower tiers still tell you.
What does the sales data say influence is?
Something closer to trust per person than to reach.
Divide each creator's revenue by their followers and the table inverts, since @lznn2424glg turned 25,770 followers into $58.21 of GMV each while @plugbettertech turned 3.23m followers into 52 cents each.
Four of the five highest figures belong to accounts under 160,000 followers, which is the pattern our post on nano influencers describes from the recruiting side.
Reading the same number the other way is useful for budgeting, because a brand paying by audience size is paying 112 times more per unit of demonstrated influence at the top of that table than at the bottom.
| Creator | Followers | Revenue, 30 days | GMV per follower |
|---|---|---|---|
| @lznn2424glg | 25,770 | $1.50m | $58.21 |
| @hannahbentley | 75,410 | $1.39m | $18.43 |
| @reedsdeals | 138,980 | $1.60m | $11.51 |
| @elfyeah | 2.50m | $2.04m | $0.82 |
| @plugbettertech | 3.23m | $1.69m | $0.52 |
Does audience size predict anything?
Reach, and very little else that a shop can bank.
Put the biggest and smallest accounts in that leaderboard side by side and the audience gap is 125 times while the revenue gap is about 13%, which is the clearest statement of the problem available.
Average views travel separately again, since @elfyeah led the revenue table on 315 average views a video, a figure that would read as a failing account on any engagement dashboard.
So the honest summary is that followers describe potential exposure, views describe realised exposure, and both stay silent on whether anybody bought, which our study of creator concentration traces through a whole programme.
Category carries part of every result, which our energy drink study in the UK makes visible by holding one category still and letting the creators vary.
What are the three tests of influence?
Three questions, and a brand can answer all of them before sending a message.
Did their last videos sell? Attributed GMV over the past 30 days answers whether the audience responds to recommendations at all, and our metric glossary defines how each figure is counted.
Did they sell this category? A creator with a record in beauty tells you little about a power station, which our food creator study shows while pricing the value of a category match.
Do they repeat? A creator who posts once and leaves is an advert, while a creator who posts six times is a channel, and that gap decides the cost of every video after the first.
Our lookalike search exists to run those three tests across a market at once, starting from a creator or product that already works for you.
Where do follower tiers still matter?
Three of them, and every one is about economics rather than influence.
Rates follow tiers, since published 2026 benchmarks price a nano creator video far below a micro creator video, which makes tier a budgeting input.
Willingness follows tiers too, because smaller accounts accept product-only collaborations at much higher rates than accounts above 25,000 followers.
Reach ceilings are real, so a launch needing broad simultaneous exposure buys audience size on purpose, which is a media decision rather than a creator-selection one.
Beyond those three, tier is a weak proxy, and our guide to becoming a TikTok influencer sets out why the creators who build selling audiences optimise for something other than follower growth.
What this definition leaves out
Three limits, stated plainly.
Leaderboards publish the top ten, so these figures describe creators who already win rather than the full distribution.
Attributed revenue counts affiliate content alone, so a creator who sends somebody to search and buy later gets credited with less than they earned, and our content insights are where that wider effect becomes visible.
Brand accounts sit in the same table as individuals, so @elfyeah and @comfrt behave as shop storefronts rather than as people, which is worth separating before comparing them to a creator.
Conclusion
Define an influencer by what their recommendation does to purchase behaviour.
Then test it on sales over the past 30 days, on category fit, and on whether they publish repeatedly.
Followers describe reach, which prices the deal rather than predicting the result.
We build Cruva so those three tests run before anybody sends a message.