Twice a week, The Hype Index takes one claim leaders are being asked to act on and scores it against its own source. You get the verdict, the method behind the score, and a dated call we grade in public.
Every claim we score stays here with its verdict, its component scores, and the date its call resolves. The archive is the product.
| # | Claim | Verdict | Index | Call resolves |
|---|---|---|---|---|
| 001 | 95% of enterprise AI pilots fail. July 26, 2026 |
Overstated | 74 | July 2027 |
Same skeleton every edition. New claim each time. Built to be read in under three minutes and to hold up when someone checks it.
One statement leaders are being asked to act on, quoted exactly, with the source attached.
A named verdict and a number from zero to one hundred, stated before any argument.
Source quality, sample and method, independence, replication, and drift. Each scored and published.
The part of the research worth acting on, separated from the headline that traveled.
A dated prediction with stated resolution criteria. It goes on the record whether it lands or misses.
Three things worth doing before the next board meeting.
Somewhere in that chain the caveats fall off, the scope widens, and a narrow finding becomes a verdict on an entire category. By the time it reaches a board meeting, the person quoting it usually cannot say where it came from. The Hype Index scores that chain, not just the research at the start of it.
Two editions a week. Tuesday and Thursday, 1:00 PM Central. Every call is dated and graded in public, including the ones we get wrong.