I think ACoS is the most damaging metric in Amazon advertising.
By michael-erickson-facchin · September 8, 2026 · Curated by George's Blog
I think ACoS is the most damaging metric in Amazon advertising.
Not the least useful. The most damaging. Those are different claims, and I mean the stronger one.
Here's why.
A few years ago we took on an account with a single digit ACoS. By every conventional standard, an excellent account. The kind you'd screenshot.
It was one of the most underperforming accounts I'd seen.
Single digit ACoS on a growth stage product isn't efficiency. It's underinvestment wearing efficiency as a costume.
When we broke the account apart, individual keywords sat at 1%, 3%, 8%, 14% ACoS. Others were running past 200%. The blended number described none of them. It was an average of things that had nothing to do with each other, and the team was making decisions off it every week.
So we rebuilt it. 86% of spend into manual campaigns, 14% left in auto for discovery. Raised bids where the data supported it, cut where it didn't.
Then we nearly doubled the ad spend. Up 96%.
CTR rose 21% in the first month. Revenue grew 448% through the holidays, and held at 50% growth after.
And total ACoS went up by 2 percentage points.
That's the whole post right there.
A manager optimizing for lowest ACoS would have called that a failure and reversed it. I've watched people do exactly that. They kill the thing that's working because the dashboard turned slightly redder.
What I've said for years and will keep saying:
Most of the time, people want more revenue at a target ACoS, not less revenue at a lower ACoS.
ACoS is a control dial. It is not a scoreboard.
You can drive ACoS to 8% tomorrow by pausing most of your campaigns. Your dashboard will look spectacular. Your revenue will collapse and your organic rank will follow it down within a quarter, and by then nobody will connect the two.
The questions that actually matter:
1) What's my break-even, calculated from real margin per ASIN?
2) What target below that serves my growth stage?
3) How much profitable volume can I buy at that target?
Answer those three and ACoS becomes what it should be. An input you set, not a grade you receive.
If your PPC reporting leads with ACoS and buries revenue, you're measuring the dial and ignoring the room temperature.
I'd genuinely like to hear the counterargument, because I've been making this one for a long time and I'd rather be challenged than agreed with.