How we estimate sponsorship income
Our estimates give a rough sense of a sponsorship’s value. They are basic calculations, not verified payments or quotes from a creator.
The current method
We multiply a video’s recorded views by a provisional range of US$15–US$50 per 1,000 views and calculate its midpoint. We then apply a US$100 floor to the midpoint and both ends of the range before rounding for readability.
The US$100 minimum per integration is a provisional model assumption, not evidence that real sponsorships always pay at least that much. We have not calibrated this floor or the view-based rates against a representative set of real deals.
The radar and video cards use the same method. Creator totals on sponsor profiles add the estimates for every distinct sponsored video we have found with that brand, with the floor applied to each integration before summing and before rounding. A total covers our saved examples, not the creator’s complete earnings or the brand’s verified spending.
| Recorded views | Estimated range | Displayed estimate |
|---|---|---|
| 1,000 | $100 | $100 |
| 10,000 | $150–$500 | $330 |
| 100,000 | $1,500–$5,000 | $3,300 |
| 1,000,000 | $15,000–$50,000 | $33,000 |
What the calculation misses
Views are cumulative snapshots, not the audience expected when the deal was signed. Actual fees can depend on audience, format, negotiation, usage rights, exclusivity, package deals, commissions and product compensation. A single agreement may cover several videos; some videos have multiple sponsors.
Missing or unusable views produce no estimate. A partial total states how many videos it covers. Subscriber counts do not stand in for missing views, and the range is not a statistical confidence interval.
Creator reports are separate
When we quote a creator’s own income disclosure, we link to their source and keep its reporting period and currency context. Those amounts are self-reports, not outputs of this estimator or independently verified payments.
How the estimates can improve
As we gather more source-backed fees and matching video data, we can test these assumptions and revise the method. More records alone do not make an estimate more accurate: we need to compare predictions with disclosed deals. This page will describe the current method as it changes.