Your ACoS improved. Did your Amazon business actually make more money?

A lower advertising cost of sales can be a good result. It can also mean you stopped reaching new shoppers, sold more of a low-margin product, or benefited from demand someone else created.

The point is not to discard ACoS. It is to give the number a job it can actually do. Amazon defines ACoS as ad spend divided by attributed sales, expressed as a percentage. That makes it an advertising efficiency measure, not a complete profit calculation. Amazon's ACoS guide.

For an established brand, the useful question is: after product costs, Amazon costs, advertising and the other costs of serving the order, how much contribution is left?

Start with the product economics

Build a simple per-unit model for each commercially important SKU:

• Net realized selling price, with discounts and refunds treated consistently.
• Landed product cost.
• Referral and fulfillment costs.
• A reasonable allowance for returns and other variable costs.
• Contribution before advertising.

Keep fixed overhead separate unless you deliberately allocate it. Do not subtract a refund twice by reducing revenue and also counting the same refund as a cost.

Here is a hypothetical example, in US dollars:

• Net realized revenue per unit: $40.
• Landed product cost: $10.
• Referral and fulfillment costs: $12.
• Returns allowance and other variable costs: $2.
• Contribution before advertising: $16, or 40% of revenue.

In this simplified one-product model, spending $16 to generate one genuinely additional $40 order leaves zero contribution before fixed overhead. That is the origin of a 40% break-even advertising ratio.

It is not a universal 40% target. If you require $6 of contribution per order, only $10 remains for acquisition. The corresponding planning ratio is 25%.

Why the ad dashboard will not equal that model

The illustration assumes aligned revenue, costs and orders. Real reports are messier.

Amazon's attribution documentation notes differences between attributed and retail sales, including reporting by the ad-interaction date and differences in how discounts are reflected. Attribution rules also vary by campaign and advertiser type. Match the actual report fields and windows rather than applying one generic delay to everything. Amazon's attribution guidance.

Other complications come from the business itself:

• A campaign can sell products with different margins.
• A price promotion can change revenue per order before costs change.
• Returns may arrive after the advertising report looks complete.
• Attributed orders are not automatically orders caused by the ad.

Use a product-level estimate for decisions, then reconcile it against settled operating results. Calling an estimate “profit” does not make it accounting truth.

A lower ACoS can produce less contribution

Consider two hypothetical periods with identical product economics and no other changes.

Period A generates $100,000 of sales attributed to ads. Contribution before ads is 40%, or $40,000. Ad spend is $25,000. Estimated contribution after ads is $15,000.

Period B generates $60,000 of attributed sales. Contribution before ads is $24,000. Ad spend falls to $12,000. Estimated contribution after ads is $12,000.

ACoS improved from 25% to 20%, while the simplified contribution estimate fell by $3,000.

This does not prove Period A was better. Cash constraints, inventory scarcity or a deliberate harvest strategy may justify Period B. It demonstrates why the objective must be stated before judging the metric.

Use TACoS as context, not proof of incrementality

Teams often use TACoS to mean total ad spend divided by total Amazon sales. Write that definition into the report because naming conventions vary.

It is useful for seeing advertising intensity across the business. It cannot prove that advertising created organic sales or that a lower ratio was caused by better campaign management.

If total sales rise after an external promotion while ad spend stays flat, TACoS falls mechanically. If you run out of a popular product, both the numerator and denominator can move in ways that obscure the real problem.

Review the ratio alongside product mix, price, availability, promotions and comparable demand periods.

Choose targets by business role

One account-wide target is rarely enough for a mature catalog.

Classify important products by the decision you are making:

1. Established, well-stocked products: protect contribution while testing additional demand.
2. Inventory-constrained products: preserve the stock plan and avoid buying demand you cannot fulfill.
3. New products: agree a bounded learning budget and a review date.
4. Aging stock: judge the plan against the realistic cost of holding or disposing of inventory.

These categories do not authorize unlimited spending. Each needs an owner, a monetary boundary and a reason to stop.

Bring four numbers to the next review

For the products that matter most, show net revenue, contribution before advertising, ad spend and estimated contribution after advertising. Put ACoS beside them, not above them.

Then ask: what would change our decision? A lower supplier cost? A revised replenishment date? Evidence that the next dollar buys weaker demand?

That conversation is more valuable than celebrating a percentage without understanding the business beneath it.

For the next layer, read how to allocate a constrained advertising budget and how inventory changes PPC decisions.

If your reports stop at ACoS, request an account review focused on the decisions behind the numbers.

‍