What is a hypothesis ledger?

The written record of every creative, audience, channel, or strategy hypothesis tested in a Dynamic Ad engagement. Each row pairs the question with the kill criterion fired at launch and the disposition (promoted / retired / iterated) at close.

Why does a hypothesis ledger matter for buyers?

Because the buyer ends the engagement with an auditable library of what was tested and what was learned, not a single 'this is what worked' sentence. The ledger compounds across cycles: cycle two starts with cycle one's winners as the new baseline. Without the ledger, the learning resets every cycle.

Is the hypothesis ledger published?

Yes, as part of the operating record at cycle close. Operators have right of redaction over identifying detail before publication; the ledger structure itself is not optional. Two operators countersign before the record ships.

Hypothesis Ledger

The written record of every creative or strategy hypothesis tested in a Dynamic Ad engagement, with the kill criterion and the read disposition.

Common questions

What is a hypothesis ledger?: The written record of every creative, audience, channel, or strategy hypothesis tested in a Dynamic Ad engagement. Each row pairs the question with the kill criterion fired at launch and the disposition (promoted / retired / iterated) at close.
Why does a hypothesis ledger matter for buyers?: Because the buyer ends the engagement with an auditable library of what was tested and what was learned, not a single 'this is what worked' sentence. The ledger compounds across cycles: cycle two starts with cycle one's winners as the new baseline. Without the ledger, the learning resets every cycle.
Is the hypothesis ledger published?: Yes, as part of the operating record at cycle close. Operators have right of redaction over identifying detail before publication; the ledger structure itself is not optional. Two operators countersign before the record ships.

Most agencies ship creative without a written hypothesis. The brief says 'we are testing carousel against single-image'; the result says 'carousel won.' Nothing in between. When the next cycle starts, the operator cannot remember what was being tested.

The hypothesis ledger replaces this with a structured row per test. Each row carries: the question being asked, the variant or strategy under test, the kill criterion fired at launch (impressions floor, CTR ceiling, ROAS floor), the actual reading at the kill point, and the disposition (promoted, retired, iterated). Every row in the ledger contributes to the angle library.

The ledger is part of the operating record. Operating records that publish at cycle close show the hypothesis ledger structure, anonymized where the operator requires.

Example

A Lead OS engagement closes its first 30-day cycle. The hypothesis ledger shows 47 variants tested: 12 carousel-versus-single-image creative tests (six in healthcare-language, six in operator-language), 8 audience-shape tests (named buying centers vs broad title targeting), 4 channel-reweight tests, 23 lifecycle-stage tests. Of those 47 rows, 12 winners promote to next cycle, 4 are iterated, 31 are retired. Cycle two starts with the winners as the new baseline.

Always on · Algorithm-led · Human-approved

Walk this operation against your account.

Thirty focused minutes with one of the founders. We map Hypothesis Ledger against your stack and tell you which entry point fits your stage.

Book a 30-minute call →Or write us