Lowering ACoS without killing sales: a structured PPC audit

A lower ACoS is not enough to call an audit a success. You also need to know what happened to sales and the money left after costs. Use the worked example below to see why. Then check your own search terms, margins and reporting windows before changing bids.

Start with the formulas

• ACoS = ad spend ÷ ad-attributed sales × 100. This follows Amazon Ads' ACoS guide.
• TACoS = total ad spend ÷ total sales × 100. It includes all sales in the chosen scope, as explained in Amazon Ads' TACoS walkthrough.
• Contribution after ads = total sales × pre-ad contribution margin − ad spend. This simplified check shows what remains before fixed costs and tax. It is not net profit.

Calculate the pre-ad margin after product costs, Amazon fees and other variable costs. In a matched, single-margin example, a 30% pre-ad margin means a 30% break-even ACoS on attributed sales before fixed costs. Your target may need to be lower. Mixed products need their own cost checks.

Worked example: the same ACoS, different outcomes

Illustrative example only — not client results. These are three fictional 30-day scenarios, not a measured before-and-after test. All amounts are USD. Assume the same product group, aligned sales records, mature attribution and a constant 30% pre-ad contribution margin after variable costs. The scenarios do not predict the effect of a bid change.

For the baseline: $1,200 ÷ $4,000 × 100 = 30% ACoS. Using total sales instead gives $1,200 ÷ $8,000 × 100 = 15% TACoS. The contribution check is $8,000 × 30% − $1,200 = $1,200. Ad-attributed sales are already included in total sales; do not add them again.

Both alternative scenarios reach 20% ACoS. Yet the lower-sales scenario leaves $900 after ads, below the baseline's $1,200. Even its TACoS improves while contribution falls. The higher-sales scenario leaves $2,000 under the same margin assumption. Neither ratio replaces a sales-and-cost review.

Run the audit in five steps

1. Match the reporting scope. Use the same marketplace, currency, products and comparable periods. Record the attribution window. Check Amazon's ad campaign attribution documentation; recent sales can still be credited after a click. Do not treat incomplete recent data as a final result.
2. Check costs and listing health. Refresh fees, discounts, stock availability and margin assumptions. Check listing suppression and your account-health routine before blaming a bid for falling conversion.
3. Review actual search terms. Use the Sponsored Products search term report. Sort by spend, then inspect intent, clicks, orders and attributed sales. A target keyword and the shopping query are not always the same.
4. Separate poor intent from thin evidence. Consider negative targeting for terms that do not meet the goal. Use Amazon's Sponsored Products targeting guide to choose keyword or product exclusions. Zero sales alone does not prove waste when clicks are scarce or attribution is incomplete.
5. Test a specific change. Set the budget, review point and success criteria first. Adjust selected targets or bids, record what changed, and compare ACoS, TACoS, total sales and contribution. A low ACoS is a reason to investigate capacity, not proof that raising a bid will scale profitably.

Read the signals without overclaiming

When attributed sales are zero, ACoS is undefined, not 0%. Review spend and the reporting context instead. When total sales are zero, TACoS is also undefined. Do not average keyword percentages to find account ACoS: divide summed spend by summed attributed sales within a consistent, non-overlapping scope.

A falling TACoS does not prove that ads caused organic growth. Prices, stock, seasonality and other changes can affect sales. Real ad reports and business reports can also use different timing or sales definitions. Reconcile them before using the simplified example on your own account.

There is no fixed improvement timeline here. Wait for the relevant attribution window and enough evidence for the decision. Separate campaigns or ad groups when goals and margins differ, not just to create more reporting rows. For help reviewing the account, see our PPC management service.

References checked on 3 September 2026. Public Amazon Ads guidance supports the formulas and targeting overview. The linked help pages required JavaScript; account-specific report settings were not verified. The table and chart are Dastan's fictional teaching example, not Amazon benchmarks or measured client performance.