TikTok fake followers: 6 checks you can run with public data

No tool can see TikTok fake followers directly. Only TikTok knows which followers are bots, so every checker estimates from public signals, and you can read those signals yourself. We drew red-flag lines from 58 real creators' recent posts: 40 tripped none, 16 tripped one and only 2 tripped two. One flag is normal. Two or more means look closer.

Below are the six checks and the numbers behind each line, then the results for three accounts we checked on 30 September 2026.

What a TikTok fake follower checker can and can't see

TikTok doesn't show anyone outside the company who follows an account or which of those followers are real. A public profile gives you followers, following, total likes and the posts. Checkers such as ViralMango say they read engagement, growth patterns and comment authenticity, then turn them into a score. Treat that score as an opinion and check the inputs.

TikTok polices this side too. Its Community Guidelines ban fake engagement, including buying followers, likes or views, and its enforcement reports count the fake followers and likes it removes each quarter. In the US, the FTC's 2024 rule on fake reviews also bans buying or selling fake followers or views to look more influential for business. When TikTok removes fake accounts, follower counts drop (check 6).

The 6-point fake follower checklist

Each line comes from our benchmark sample: 58 creators TikTok search surfaced in six niches, with 10 recent posts each. "Normal" is the middle half of those creators. The red flag sits below almost all of them.

Data: 58 creators surfaced by TikTok search for fitness, cooking, personal finance, skincare, small business and comedy, 572 posts with views, pulled 21 September 2026. Lines and counts derived 30 September 2026. Benchmarks snapshot.

1. Views per follower

Fake followers don't watch, so views fall behind the follower count. Size matters, though. The typical creator in our sample reached 30.04% of followers per post at 10K to 100K followers, 6.96% at 100K to 1M and 3.53% past 1M. Big, older accounts sink lower: 8 of the 29 accounts over 1M sat under 0.5%, TV brand accounts among them. That's why the line is tier-based.

Also check that views vary. The typical creator's biggest recent post had 8.2 times their median views; only 4 of 58 stayed under twice the median. Every post landing on the same number is odd.

2. Likes per view

The median creator got likes on 4.53% of views, and only 3 of 58 fell under 2%. It catches bought views and reposted clips people scroll past more than bought followers. Either way, a brand gets little attention. More in the engagement rate calculator.

3. Comment quality, not comment count

Don't flag an account for few comments: 12 of our 58 real creators had under 0.2 comments per 1,000 views. Read them instead. A generic comment fits under any video ("Wow", "So nice 🔥"); a specific one mentions the dog, the product or asks a real question. On our healthy example, 41 of 50 were specific, 9 generic and none emoji only. We set the flag at more than half generic or emoji only, well above that. Repetition alone isn't proof: read what the repeats say (Account A below).

4. Lifetime likes per follower

Divide the total likes on the profile by the follower count. The median creator had 13.2 likes per follower. Five of 58 sat under 3, but all five had 12,400 followers or fewer. On a bigger account, under 3 means many followers never liked anything.

5. Following count

Follow-for-follow accounts follow thousands to get follows back, and those followers rarely watch. The median creator in our sample followed 69 accounts and none followed more than 2,365. In a fresh pull of 40 home decor and dog grooming accounts, one followed 9,999.

6. Sudden follower jumps

Public data shows today's count only, so we couldn't see these either. Use a follower-history site such as Social Blade, or ask the creator for their TikTok Studio follower chart. A jump that lines up with a viral video is normal. Thousands of new followers in a day with no post behind them, or a sharp drop, is the flag.

More than a quarter of real creators trip one flag, usually a big account with low views per follower or a small shop with few lifetime likes. Only 2 of 58 tripped two. Judge on the pattern, never on one number.

We ran the checklist on three accounts

We searched TikTok for "home decor" and "dog grooming" accounts and picked three. The two with weak numbers are anonymised, because public signals can't prove anyone bought followers.

Data: TikTok account search, recent posts and comments via TokConnect, pulled 30 September 2026. Follower counts are exact figures from account search. Posts snapshot, comments snapshot, account search snapshot.

What healthy looks like

The dog groomer's 20 posts from 11 August to 28 September got between 14,200 and 2,400,000 views, with a median of 68,500: about a third of her followers per post. Likes ran at 11.0% of views, more than twice the sample median. Her comments talk about the dogs by name, and 4 of the 50 we read were people asking her price, her location or which brushes she uses. Nothing to flag.

Low numbers, real buyers

Account A sells home goods and has posted 1,299 videos. Its median post reached 3.6% of followers, likes ran at 2.2% of views and it had 3.62 lifetime likes per follower. Weak, but above every line. The comments settle it: on its top recent post (196,700 views), 46 of the first 50 were orders or "still available?" questions in Tunisian Arabic, mostly "nheb nechri", "I want to buy". Buyers often skip the like button and ask how to order instead. The same two words, repeated again and again by different accounts, looked like bots until we read them.

Two flags: look closer

Account B follows nobody. Its last 10 clips went up between 3 and 10 September and reuse four headlines. Its views look normal at 24.2% of followers. But only 0.9% of viewers liked a post and it has 2.97 lifetime likes per follower, two flags. Its best post had 91,800 views and under 10 comments; of the 5 we got, 3 were real reactions in three languages and 2 were emoji only. That doesn't prove fake followers: reposted clips often get watched and ignored. It does tell a brand the audience watches without engaging, so ask for the creator's own analytics before paying.

Before you pay a creator: what to ask for

Public checks narrow the list. The creator's own analytics finish the job, because TikTok Studio shows them what nobody else can see (TikTok Help Center).

1. Check the last 20 posts, not the pinned ones. Pinned videos are old hits and flatter every ratio.
2. Ask for three TikTok Studio screenshots. Follower growth over the last 60 days, audience locations and age, and the audience breakdown for one recent post. If most viewers are outside your market, the followers may be real and still useless to you.
3. Compare against the niche. A comedy account and a small shop have different normal ranges. Pull five peers and see where the creator sits. Our guide to finding TikTok influencers covers building that shortlist.
4. Start small. Pay for one post with a tracked link or code before a bundle. Real audiences click.

Creators: brands run these checks, and bought followers drag down every ratio above. See how to get brand deals for what they look for, or the account analyzer guide for a wider look at an account.

How we got these numbers

The red-flag lines come from our September 2026 benchmark sample: the first 10 accounts TikTok search returned for six niche words, with each account's 10 most recent posts (58 creators, 572 posts with views, pulled 21 September). We took each creator's median across posts, set each line below almost all of them and counted who crossed it. The comment line is our judgement.

On 30 September 2026 we added two account searches (20 accounts each for "home decor" and "dog grooming"), up to 20 recent posts for three accounts and the first 50 comments on one video per account, which we labelled by hand. Most of the sample has over 100K followers, so small creators are thin in it, and no public check sees a follower list. Ask TokConnect "does @[handle] look like it has fake followers?" and it runs the same pulls on any handle you're vetting.