Operating Record
The auditable cycle-by-cycle log of every prediction, decision, reallocation, and result on a Dynamic Ad engagement, published at cycle close.
Common questions
Common questions
- What is an operating record?
- The auditable, cycle-by-cycle log of every prediction, decision, reallocation, variant launch, kill criterion fire, and signed-off result on a 90-day Dynamic Ad engagement. Published at cycle close as the canonical artifact of the engagement.
- How is an operating record different from a case study?
- A traditional case study presents a curated narrative of what worked. An operating record publishes the full ledger: every prediction, every decision, every result, including the ones that did not work. Case studies sell; operating records substantiate.
- Can clients redact information from the operating record?
- Clients have right of redaction over identifying detail (company name, vertical specifics, individual figures they consider confidential) before publication. They do not have the right to block the existence of the record. The structural facts and the make-good cycle status always publish.
Most marketing engagements close with a deck. The operating record closes with a written log. Every prediction the system made, every decision the operator approved, every variant that launched, every reallocation, every kill criterion that fired, every result measured against the prediction, all written into one continuous ledger across the 90-day cycle.
The record is the artifact of an engineered operation, not the summary of one. It is the structural counterpart to the make-good guarantee: the make-good claim is only defensible because the record exists to back it. Published case studies are operating records, not testimonials.
Operators have right of redaction over identifying detail before publication, but not over the structural existence of the record itself. The record is countersigned by two operators before it ships.
Example
A Lead OS cycle closes. The operating record published on /case-studies contains 47 weekly entries: 12 strategy reweights with the data that triggered each, 38 creative variants launched with the hypothesis and kill criteria, 14 reallocations across channels with the signal and the operator approval timestamp, the final cycle-close numbers measured against the frozen trailing-90 baseline, and the names of the two operators who countersigned the close.
Always on · Algorithm-led · Human-approved
Walk this operation against your account.
Thirty focused minutes with one of the founders. We map Operating Record against your stack and tell you which entry point fits your stage.