Free answer · verified July 2026
Can I trust my ads dashboard's ROAS?
Not at face value. Ad platforms measure their own performance, and peer-reviewed studies show the overstatement is systematic — industry measurements put it anywhere from 20 to 60% depending on the study. A dashboard that says 3x may be closer to 1.2–2.4x once you measure what the ads actually caused.
Why the dashboard flatters itself
The platform reporting your ROAS is the same platform selling you the ads. Its attribution model gives itself credit for conversions that would have happened anyway — the customer who searched your name, clicked the ad out of habit, and bought exactly as they planned to. Academic studies comparing platform-reported results against randomized experiments keep finding the same direction: self-reported returns run high, and not by a rounding error.
What to do instead
Decide before you spend which numbers you’ll believe. The working rule the map uses: incrementality is the arbiter — the only question that matters is what the ads caused that wouldn’t have happened without them. You don’t need enterprise tooling for a first pass: a simple holdout (pause a region or an audience for two weeks and watch what actually changes) settles more arguments than any dashboard.
Write the kill/scale decision down in advance: “if the holdout says the true return is below X, we stop; above Y, we scale.” That one page of writing is the difference between a measurement and a mood.
Where this comes from
This is a small slice of the free chapter of The Modern Marketing Map — the data and measurement chapter is free in full, no signup, with every claim dated and sourced. A real excerpt of the evidence ledger behind it ispublic here.