Platform-reported ROAS is a flattering number. It takes credit for purchases that would have happened anyway, double-counts cross-device conversions and ignores how much of your revenue came from remarketing people who were already going to buy.
If you are scaling spend based on the ROAS number inside Ads Manager, you are almost certainly over-crediting Meta and under-crediting everything else.
The three ways in-platform ROAS misleads you
1. It claims organic buyers. A customer sees your ad, ignores it, then buys three days later after a friend's recommendation. Meta's attribution window claims that sale.
2. It loves remarketing too much. Remarketing audiences convert at high rates because they were already interested. High ROAS on remarketing is often just harvesting demand you created elsewhere.
3. Cross-device double counting. The same customer counted on mobile and desktop inflates conversion totals - especially for brands with longer consideration cycles.
The question is never "what ROAS does the platform report?" It is "how much revenue would we lose if we turned this campaign off?"
What to measure instead
- MER (Marketing Efficiency Ratio) - total revenue divided by total ad spend, across all channels. It cannot be gamed by attribution windows.
- New customer ROAS - strip out remarketing and returning customers. This tells you if prospecting actually works.
- Incrementality tests - turn campaigns off for a holdout region or period and watch what actually happens to revenue.
Here is a short walkthrough of how we set this reporting up for clients:
Building the truth-telling report
Start with a simple weekly sheet: total revenue, total spend, MER, new customer count, and blended CAC. Add platform-reported numbers as a secondary column - useful for optimization decisions inside each platform, never for budget decisions across platforms.
Once MER and new-customer ROAS become your primary scaling signals, budget conversations get much simpler - and much more honest.