Add to Chrome

Views Guarantee Calculator

Get the views guarantee to write into the deal, plus your odds of hitting it. Point it at a creator and it simulates 10,000 buys by resampling that channel's real 30-day view counts, then returns the smallest video bundle that clears 80%.

Use in Claude Code Set up your agent

A views guarantee is a bet on a creator’s next few videos. This one comes with the odds.

Where the 80% comes from

The likelihood isn’t a confidence you’re asked to take on faith. Every recent video on the channel is measured at the same age — its view count between day 28 and day 32, averaged out of the snapshot history we’ve been recording since June 2018 — and those real totals get resampled 10,000 times to simulate the buy you’re actually considering. Nothing is fitted and no curve is assumed; the simulation can only ever draw from views this channel has genuinely produced. Set that against the number you’d otherwise negotiate from: a lifetime average that flatters one old viral hit, or whatever figure the creator put in their media kit.

Sizing the bundle is the other half. Because the simulation prices a one-video buy and a four-video buy separately, you can see where the guarantee stops being a stretch — and when a creator is too volatile to clear 80% at any size, the skill says so and reports the floor with the likelihood the data really supports. That’s a negotiation position, not a failure: a guarantee you’re comfortable taking at 70% can still be the right deal, since a miss usually just means the creator owes you a make-good video. Judgement like that comes from placing sponsorships since 2018, across 174,345 recorded deals — the same record your buys will join.

Try this prompt
> What views guarantee should I push for with MKBHD, and how many videos should I buy?

Four lines back: the channel, the bundle size, the guarantee, and your odds of hitting it.

To Claude

How it works

  1. Resolve
    Give it a creator's name or a channel ID and it finds the channel record.
  2. Sample
    Take the videos published 30–90 days ago — recent enough to reflect the channel as it is now.
  3. Measure
    Read each one's view count at the same age, between day 28 and day 32.
  4. Simulate
    Resample those totals 10,000 times per bundle size, from a single video up to four.
  5. Pick
    Return the smallest bundle whose guarantee still clears an 80% likelihood.