- Par · the baseline
- Teams chase full attribution — data-driven models, multi-touch — before they know whether the spend worked at all.
- The lie · the constraint
- Attribution tells you who gets credit. It doesn’t tell you what would have happened without the spend.
- The shot · the call
- Geo holdouts: go dark or double spend in matched U.S. markets and read the gap.
- Result
- HoldoutA CFO-level answer, not a channel-level argument.
01
Plan the route
Measurement is a strategic road trip.
Know where you’re going (the question), which towns you’ll skip (the holdouts), and what could send you on a detour (weather, seasonality, a competitor’s launch). U.S. DMAs are big and similar enough to make that trip representative.
02
Read the gap
Difference, not credit.
Compare the skipped markets to the ones still spending, against how they tracked before the test. That gap is the answer. Everything else is a debate.
03
What it changes
It decides what earns the 70%.
If you can’t say “did it work?”, it doesn’t belong in the proven bucket. Plan your own trip on the homepage.