12,634checks behind this list
A list like this is only worth reading if you can check it. Here is how.
In short: every view count and subscriber count here is the figure the platform's own API returned, stored with the date it was read. None is estimated. A newsletter that publishes its own count is shown with that count, labelled self-reported.
Where the numbers come from
Every audience figure is read from the platform's own API, plus two free official sources that matter: YouTube's public RSS feed, which Google publishes for every channel, and Bluesky's public API. Every YouTube average is checked against that RSS feed. It is the only check here that can catch the data source being wrong rather than the creator.
No audience figure comes from a language model, a growth model or an industry benchmark. Where a platform publishes nothing, the page says so and stops there. The one number on this site that we work out ourselves is the price estimate, and it is labelled, sourced and explained below.
The one exception is an independent newsletter that no platform reports on. Its list size is the count printed on its own advertise or sign-up page, read by hand on the date shown. That figure is labelled self-reported wherever it appears, links to the page it was read from, keeps the newsletter in the review tier rather than a pass, sorts below every platform-measured creator, and is left out of the largest-creator rankings on topic pages. It is still never estimated.
The bar for being listed
A creator has to average 20,000 views a post on every platform that publishes a view count, reach a software audience, be based in a target market, and have posted in the last six months. Two of those bars are caveats rather than exclusions. A creator outside the target markets is still listed, and their page says so. A creator below the reach bar is listed only when the platform's own count shows at least 1,000 followers or subscribers, and their page says they are below the audience minimum. Neither applies to a creator who failed anything else or raised an integrity flag. The list is ordered by the median of those same posts, not the average, because one viral hit can carry an average a thousandfold past what a typical post gets, and a sponsor buys a typical post. Where the average runs three times or more past the median, the row prints the typical figure under it. On YouTube, Shorts are left out of the average. A channel averaging 20,000 views on 40 second clips is a different buy from one averaging 20,000 on tutorials.
What the check marks mean
| Required | Something the brief demands. If it cannot be verified, the creator is held back rather than passed. |
| Integrity | Patterns that look like an inflated average or bought engagement: one viral hit carrying the mean, samples years apart, reach falling over time. The creator stays listed with the flag showing, so you weigh the figure yourself. Flags for a human, never accusations. |
| Not published | The platform published nothing to test. That is about the platform, not the creator. |
| Preference | A soft signal. Recorded, never disqualifying. |
What the sponsorship labels mean
| Confirmed | They take sponsors, and something says so outright: a public rate, a sponsorship page we checked by hand, or a brand they named as a sponsor in their own post. |
| Runs paid posts | The platform's own paid-promotion flag is on recent posts, but no sponsor is named. They run paid placements; who bought them is not on record. |
| Likely open | Their bio, a link they publish or the address they list invites sponsorship. Nobody has confirmed it, so treat it as a lead and not as a yes. |
| Declines | They publicly refuse sponsorship, with the page that says so linked. This beats every other signal, however old. |
| Not known | Nothing on record either way. It is not a no, and it is never inferred from an absent rate card. |
Every label carries the one sentence of evidence it was read from, printed on the creator's own page.
Where prices come from
Every price cites a page you can open, with the date it was read and the audience size it was quoted against, so a stale price is visibly stale.
A price the creator publishes always comes first. It is quoted as they wrote it, with the page it was read from linked beside it. Nothing outranks a number the creator stands behind.
Underneath it, on five platforms, the page also shows a figure labelled Estimate. That one is ours: the creator's own median views a video, or their follower count, multiplied by a rate a named third party publishes. The sum is printed next to it in words, and the benchmark is linked and dated so you can go and read it. An estimate is not a quote. The creator has not seen it and has agreed to nothing. On every other platform there is no benchmark we would repeat, so the page says that instead of guessing.
Audience figures are still never estimated. A view count or a follower count either came from the platform or is not shown at all. The estimate multiplies those figures; it never replaces one.
| platform | figure used | rate | benchmark |
| YouTube | median views a video | $15 to $30 per 1,000 views, one integrated mid-roll | Influencer Marketing Hub, YouTube influencer rates ↗, dated 31 Aug 2026 |
| followers | $10 per 1,000 followers, one sponsored post | Sprout Social, Influencer pricing ↗, dated 30 May 2024 | |
| TikTok | followers | $10 per 1,000 followers, one sponsored post | Sprout Social, Influencer pricing ↗, dated 30 May 2024 |
| X | followers | $2 per 1,000 followers, one sponsored post | Sprout Social, Influencer pricing ↗, dated 30 May 2024 |
| Substack | subscribers | $10 to $75 per 1,000 subscribers, one dedicated send | beehiiv, Newsletter sponsorship cost ↗, dated 29 May 2026 |
Read by hand on 7 Sep 2026. The date on each row is the date the source page itself shows.
What is on record
| Creators listed | 500 |
| On record, including held back | 504 |
| Accounts measured | 630 |
| Checks stored | 12,634 |
| Sponsorships spotted | 81 |
| Prices with a source | 2 |
| Figures last read | Sep 4, 2026 |
Known limits
Instagram, Threads and Reddit publish no view count, so creators there are measured on engagement and cannot be compared like for like against a YouTube average. Region is read only from a field a platform actually publishes, so many creators have no region rather than a guessed one. And an audience figure is a follower count, not a reach guarantee. The two are not the same thing.