91% of Marketers Don't Trust the Numbers Their Ad Platforms Show Them. They Are Still Spending Anyway.
Affinity Solutions research shows 91% of marketers believe their platform-reported results are overstated. The cost of optimising to the wrong number is operational, not just analytical.
If your platform is grading its own homework, you should not be surprised when the report card comes back full of As.
Nearly 91% of marketers think the numbers their ad platforms report are overstated. New research from Affinity Solutions, drawn from real consumer purchase data, says billions of dollars are being optimised against figures that do not reflect what actually drives sales.
This is not a small gap. The platforms report conversions through last-click, view-through windows that stretch 24 hours past an impression and modelled estimates where direct measurement is impossible. The actual purchase data tells a different story.
Meta, Google, TikTok and Amazon all sell the same impression, count the same conversion and take credit for the sale. If a customer sees a Meta ad, clicks a Google search ad and then buys, both platforms will report the conversion. Marketers know this. They have known it for years. The new finding is how widespread the awareness has become, and how little it has changed behaviour.
Almost 91% of marketers believe their platform-reported results are overstated, yet most still optimise toward those numbers
The Australian context makes this sharper. Local advertisers spend a higher share of paid media budgets on Meta and Google than the global average. When those platforms inflate reported ROAS, the budget reallocation downstream is wrong by default. You are not just measuring badly. You are choosing badly.
Why it matters
The cost is not just analytical. It is operational.
Budget moves toward whatever channel reports best. If a platform overstates by 30%, more money flows there next quarter. Less money goes to channels that report honestly but lower, like organic search or email. The platforms with the most flattering measurement win the budget war. The channels with the truest measurement lose it.
This compounds. Brand-building, retention and word-of-mouth are systematically underfunded because their attribution windows are shorter and their measurement honest. Paid social and search are systematically overfunded because their windows are long and their counting generous.
The result over five years is a marketing mix that has more in common with the platforms' commercial interests than with what actually drives growth.
What to do about it
Run a holdout test on your highest-spending paid channel. Pause it in one geography for four weeks. Compare total revenue change to the platform's reported contribution. The gap is the inflation.
Stop optimising to platform-reported ROAS as the only metric. Combine it with media mix modelling, marketing mix analysis or matched-market testing.
Bring in independent measurement where the budget justifies it. Affinity Solutions, Nielsen and MMA Australia all offer this.
Build a measurement hierarchy. Platform reports for in-flight optimisation, MMM for quarterly budget allocation, holdouts for validation.
Stop letting platform account managers walk you through performance reviews unchallenged. They are not neutral.
Most marketers know the numbers are inflated. The brave ones are starting to act on it.