What Your Amazon ACoS Is Actually Telling You

“What's a good ACoS?” is the most common question in marketplace advertising, and the least answerable as asked. ACoS is not a score. It is four different signals compressed into one number — and reading it as a score is how brands cut the spend that was building their rank while keeping the spend that merely defends it.

There is no good ACoS. There is only an ACoS your economics can justify for a purpose you have named. Launch spend on a new product should run hot — it is buying data, reviews and rank. Defensive spend on your own brand terms should run cheap. Harvesting spend on proven keywords should pay its way every week. A single blended number hides all three jobs.

The corrective is to tag campaigns by intent and judge each against its own standard: launch and rank-building against visibility gained, defence against its low ceiling, harvest against contribution. Total advertising cost of sales — ad spend against total revenue, not just ad revenue — then tells you whether the flywheel is learning to turn on its own.

If ad-attributed sales hold while total sales grow, organic rank is compounding — the system works. If every incremental sale must be bought, the flywheel never engaged, and more spend will only rent the illusion of growth.

ACoS is a question, not an answer: is this spend building something, or renting something? Every campaign should know which.