Explain an ACoS change before changing bids.

The question

ACoS moved from 25% to 37.5%. The first reaction is often to lower bids. Before doing that, compare the cost of a click, the rate at which clicks become purchases and the attributed sales per purchase. Any of those can change the ratio.

Work through the numbers

These illustrative periods are equal in length and use the same settled, click-attributed definitions. Purchases are orders, not units.

ACoS is CPC divided by the product of purchase rate and sales per purchase. Earlier: $1 ÷ (10% × $40) = 25%. Later: $1.20 ÷ (8% × $40) = 37.5%. Clicks became more expensive while the purchase rate fell.

The increase is 12.5 percentage points, or 50% relative to the original 25%. Those are different descriptions of the same change. Be explicit about which one you mean in a weekly report.

ACoS Change Explainer allocates the observed increase by averaging all six possible orders of replacing the three factors. Here it assigns 5.625 percentage points to CPC, 6.875 to purchase rate and zero to sales per purchase. They sum to 12.5.

Why average the orders? Replacing CPC first gives a different intermediate effect than replacing purchase rate first. Averaging shares that interaction rather than treating your chosen order as the explanation. The result describes the arithmetic; it does not identify what caused shoppers or auctions to behave differently.

A published case to examine

My home-goods case describes responding to seasonal CPC and query mix, promoting terms weekly, and reviewing bids and placements alongside conversion and inventory. The published August 19–September 18 snapshot reports $734,138.50 in attributed sales, $73,109.25 in spend and a 9.96% ACoS.

This provides a practical list of things to inspect when the arithmetic changes. The public case has one period and no factor-level comparison, so it cannot prove which factor improved or quantify a before-and-after lift. The year is not stated in the published snapshot.

Read the home-goods case.

Try it on your numbers

1. Open ACoS Change Explainer. Choose two equal-length periods with settled attribution and comparable marketplace, account and product scope.
2. Enter clicks, spend, click-attributed purchases and attributed sales for each period. Add dates and currency; confirm the comparison basis.
3. Alternatively, import one CSV, TSV or XLSX per period. Check column mapping, full attribution windows and the absence of subtotal rows or overlapping exports.
4. Read the factor breakdown. For CPC, inspect auctions and placement mix. For purchase rate, inspect traffic relevance, availability and listing changes. For sales per purchase, inspect product mix and promotions.
5. Download the review and choose one investigation. Use Search Term Report Checker if you need the terms behind the account averages.

Open ACoS Change Explainer

Inspect search term rows with Search Term Report Checker.

What still needs your judgment

The bridge includes product, campaign and traffic mix inside the averages. It does not isolate them or establish causality. A fall in average purchase rate could reflect more low-converting traffic even if each product’s own rate stayed stable.

Zero sales makes ACoS undefined; zero purchases prevents the three-factor breakdown. If report definitions differ, correct the comparison before interpreting it. A controlled test or more detailed evidence is needed to connect a change to a cause.

Explain it to your team

“The ACoS increase reflects both higher CPC and a lower purchase rate. I will inspect placement and query mix before deciding whether bids or the listing need attention.”