Competitive Claim Verification Gate (QA)

Competitive Claim Verification Gate (QA)

1. 01

   Freeze the evidence set at a stated collection date before any claim is drafted

   Record the pull date and the lookback window on the cover of the deliverable. A pitch that cites a competitor's ad spend from an unknown week is not defensible when the client's CMO asks how the number was derived.

2. 02

   Confirm each competitor figure traces to a named source and a stated method

   Ad spend estimates, traffic totals, and award values come from different collection methods and carry different error bars. Label which numbers are measured, which are modeled, and which are directional.

3. 03

   Re-pull the three highest-stakes numbers within 48 hours of the client readout

   Competitor budgets and placements move weekly. Anything older than a quarter should be flagged as historical context rather than current position.

4. 04

   Separate observed data from interpretation in the written deliverable

   Put the raw finding in one column and the agency's read in another. Clients pay for the read, but they need to see where the data stops and judgment begins.

5. 05

   Test every claim against a second source before it enters the deck

   Cross-check traffic, keyword, and procurement figures against an independent dataset. A single-source claim that collapses in the room costs more than the extra hour of verification.

6. 06

   Log the confidence level and known gaps for each claim in an appendix

   State what the data cannot show, such as private ad spend or unreported tenders. Naming the blind spots preempts the objection and protects the retainer.

7. 07

   Route the final claim set through one reviewer who did not build it

   A second reader catches overstated causality, stale comparisons, and numbers that drifted between draft versions.