Marketplace advertising consoles report advertising cost of sales (ACoS): ad spend divided by the sales attributed to ads. It is easy to read and easy to misread. Attribution credits ads with sales that would often have happened anyway, and ACoS says nothing about margin.
Three measures that tell you more
| Measure | Calculation | What it shows |
|---|---|---|
| TACoS | Ad spend ÷ total sales (advertised and organic) | How heavily the business leans on advertising overall |
| Contribution after advertising | Contribution margin before ads − ad spend | Whether a product makes money once advertising is paid |
| Incremental return | Extra contribution caused by ads ÷ ad spend | Whether the spend creates profit you would not otherwise have |
Reading the patterns
- Falling ACoS, rising TACoS: ads are capturing sales that organic search used to deliver. Spend is rising without growth.
- Stable TACoS, rising total sales: advertising is supporting growth in proportion. Usually healthy.
- Negative contribution after advertising: the product loses money as advertised. Fix price, cost or targeting, or reduce spend.
Measuring incrementality without a data science team
- Pause tests. Switch off a campaign or keyword group for two to four weeks on a stable product, and compare total sales with a similar product left running.
- Branded keyword checks. Customers searching for your brand often find you organically; test whether branded campaigns add sales or only cost.
- Geographic or time splits. Where a platform allows, run spend in one market or time window and hold it back in another.
Rule for budget moves. Shift budget towards products with positive contribution after advertising and measured lift, and away from campaigns whose sales largely reappear organically when paused.
Stop, hold or scale
Write the rule before the month starts. For example: scale campaigns with incremental return above an agreed threshold; hold those near it while testing; stop those below it for two consecutive reviews. Rules written in advance prevent budgets from following whichever dashboard looks best that week.
Weighing a decision like this?
Discuss it with usMore research