TikTok views percentile calculator, checked against 222 real videos

The calculator

Use your median, not your average. One viral video moves an average and moves a median not at all, which is exactly why brands who know what they are doing ask for the median. If you paste your last ten view counts into the benchmarks page tool, it works the median out for you and hands you back a block you can paste into a media kit.

The ladder it checks against

The ladder is the view count a video has to clear to sit at each percentile of the 222 video export. Reading up, half of the account's videos sit below 583 views, and only 2 cleared 100,000.

Between rungs the calculator interpolates linearly, and it says so rather than pretending the ladder has more resolution than 222 videos give it. Below the bottom rung it interpolates down to zero, and above the top rung it runs out of data at the account's best video, 286,400 views, and tells you that instead of inventing a number.

Counts above a threshold

The same distribution read the other way, how many of the 222 videos cleared each round number. Useful when a brief says something like "videos that do at least a thousand".

Engagement percentiles, same account

Views are only half of what a brand reads. The export also carries likes, comments, reposts and saves per video, so the same percentiles exist for engagement rate per view. The account's median is 2.77 percent, and the ladder around it

If your median views land high but your engagement lands at the bottom of this ladder, that combination is worth understanding before a brand does. High reach with bottom decile engagement usually means the audience is bought, incidental or in the wrong country.

What one account can and cannot tell you

• It can give you a real, checkable reference point instead of a round industry average like "500 to 1,000 views is normal for small accounts", with the entire table published so you can verify every rung.
• It can show you what a small account's spread actually looks like. The mean here is 5.8 times the median, because two videos carry 61.2 percent of all views. Any page quoting an average for accounts this size is quoting the outliers.
• It cannot tell you your niche percentile. Travel short form is not the same distribution as comedy or finance, and this is one account in one niche.
• It cannot rank you against accounts of a different size. Views per follower is printed alongside your percentile precisely because raw views without an audience size next to them mean nothing.
• It will not flatter you. If your median is under 144 views it says you are below the bottom decile of this account, and that is the honest answer.

Methodology

How the performance numbers were measured. The public TikTok channel was exported with yt-dlp on 4 September 2026, giving one row per video with a publish date, view count, like count, comment count, repost count, save count and duration. 222 of the 224 videos the channel reports came back, and the export does not say why the other two were absent, so no reason is claimed for them. Every per video figure divides by 222.

The table is published at /data/creator-dataset.json and the site recomputes from it at build time, so a page and its data cannot drift. Figures were last recomputed on 7 September 2026.

What is a measurement and what is a decision. Prices are a commercial decision and are labelled that way everywhere they appear. Nothing on this page claims a rate was derived from data. The cost per thousand views is the one bridge between the two, and it divides a chosen price by a measured median.

Where the eleven comparison sources came from. Each page was fetched on 7 September 2026 and its figure, its tier definition and its own statement about sourcing were read off the page. Two of the thirteen attempted could not be fetched and are excluded rather than summarised from a snippet. The sourcing column is deliberately verbatim, because what a page says about where its number came from turned out to be more informative than the number.

What the calculator will not do. It reads one export of one account, so a percentile here describes this account's history, not a platform wide distribution. It rounds no ladders and substitutes no averages. If a video is missing from the export, the ladder simply does not know about it, which is why the export date sits above the tool rather than in a footnote.

Limits. 222 videos is a real sample but a small one. Ranks near the top of the ladder move with every new upload, and the spread between the 90th and 95th rung is narrow enough that a single viral video can move a video several rungs. The ladder is honest about that instability rather than smoothing it.

Questions