They measure different things. Trust MER for the business; ROAS for channel-level optimization.

ROAS (Return on Ad Spend): platform-reported revenue attributed to ads / ad spend. Uses last-click or platform-attributed conversions. Typically reported per-channel (Meta ROAS, Google ROAS). Subject to over-attribution: platforms count conversions that would have happened organically.
MER (Marketing Efficiency Ratio): total revenue / total marketing spend, blended across all channels. No attribution model, no platform reporting — just the bank-account math. Usually much lower than ROAS because it includes customers who would have converted anyway.

Why they diverge:

Attribution inflation. Meta and Google both claim credit for the same purchase. Platform ROAS double-counts. Multiply by ~1.5–2x across platforms and you recover the "phantom" revenue.
Organic baseline. ROAS assumes every attributed sale was caused by the ad. MER implicitly accounts for customers who'd have bought anyway.
iOS 14.5 and cookie loss. Platform ROAS has become less reliable since 2022; MER is infrastructure-independent.

What I'd tell the CEO:

"ROAS is the dashboard the paid team uses to decide where to move the next dollar within Meta or Google. MER is the number that tells us whether marketing is profitable for the business. At 1.8 MER with a 40% gross margin, marketing is contributing $0.72 in gross profit per $1 spent — which is below a healthy 1:1. Platform ROAS of 4.5 is real, but it's measuring a different question. The divergence isn't a scandal; it's expected. What matters is the trend of MER and the incrementality tests we run on paid channels."

Bonus signal: mentioning geo-based holdout tests or incrementality testing as the tie-breaker when ROAS and MER disagree.

Paid team reports ROAS = 4.5. MER is 1.8. Which matters more?

Draft your answer

Stuck? Peek at a hint

More product metrics questions