We’ve been writing about this internally for a year, and it’s probably time to put it in public. The Amazon supplement industry runs on ACoS as its primary performance metric, and ACoS is not a profitability metric. It’s a spend-efficiency metric, which is a useful but incomplete view of the business.

The gap between the two is where supplement brands are quietly losing margin.

What ACoS actually tells you

ACoS — advertising cost of sale — is the percentage of ad-driven revenue spent on advertising. A 28% ACoS means you spent $28 in ads to drive $100 in ad-attributed revenue. It’s a clean, intuitive number. Every agency uses it. Every brand reports on it.

The problem is that ACoS doesn’t account for anything else that affects whether a sale is actually profitable. It doesn’t account for FBA fees, which vary dramatically across supplement categories and product sizes. It doesn’t account for the 15% referral fee. It doesn’t account for your COGS. It doesn’t account for the ASIN-level differences in gross margin that exist across your catalog.

A 28% ACoS on a product with 55% gross margin after fees is great. The same ACoS on a product with 22% margin after fees means you’re losing money on every ad-driven sale. Both show up the same way in your report.

How this hides problems

Most brands have a mixed catalog. Some products are high-margin heroes. Some are mid-margin bread-and-butter. Some are low-margin experiments or legacy SKUs that nobody’s reviewed in a while. When you report blended ACoS, you’re averaging all of them together. A brand with two heroes at 55% margin and three losers at 20% margin can easily show an “acceptable” blended ACoS while losing money on the losers every single day.

This is exactly what we find in a meaningful percentage of the audits we do. Not most — most brands don’t have structurally unprofitable ASINs — but enough that it’s worth flagging as a category-wide issue. The brands who get caught by this aren’t doing anything obviously wrong. They’re just measuring the wrong thing.

Why it’s gotten worse

Two things have happened in the last two years that make ACoS even less useful than it used to be.

First, Amazon’s fee structure has gotten more granular and less predictable. Supplement products have category-specific size-tier and weight-based fees that can shift significantly with packaging changes. Brands that repriced their packaging to save on FBA without updating their unit economics model found themselves in a different margin structure than they thought.

Second, PPC costs have gone up across almost every supplement sub-category. What was a 20% ACoS three years ago is a 30% ACoS now for the same product. Brands that haven’t repriced to match the new PPC reality are running thinner margins than their ACoS numbers suggest.

What to measure instead

The right metric is net margin per ASIN after all fees, including advertising. It looks something like:

Net Margin % = (Revenue − COGS − Amazon Fees − Ad Spend) / Revenue

Calculated per ASIN, per month, at minimum. This is harder to build than an ACoS report — it requires clean COGS data, up-to-date fee categorization, and reporting that isn’t organized around ad platforms — but it’s the only number that tells you whether your Amazon business is actually profitable.

Once you’re looking at net margin per ASIN, a lot of decisions get easier. Which ASINs deserve more PPC budget? The ones with the highest net margin, not the lowest ACoS. Which ASINs should you sunset? The ones that can’t get to positive net margin even with repricing and ad optimization. Which keywords should you prioritize? The ones that convert to your highest-margin products, not the ones that convert at the lowest ACoS.

What this means for how you pick an agency

If your current agency reports primarily on ACoS, ask them for net margin by ASIN. If they can produce it, you’re fine. If they can’t — either because they don’t have the data or because they’ve never been asked before — that’s information. Not necessarily a firing offense. But a signal that your performance measurement isn’t built to find the kind of problems that matter most.

At minimum, start asking the question. The number you get back tells you more about the health of your Amazon business than any ACoS report will.