Measurement primer / 4 MIN READ
Read ROAS without confusing revenue with profit
ROAS can help compare paid activity, but it is only as useful as the value definition, attribution and costs behind it.
Reported revenue ROAS
Attributed revenue divided by advertising spend. State the attribution model, window, currency and whether revenue is net of cancellations, returns and taxes.
Conversion-value ROAS
A platform may use assigned conversion value rather than actual revenue. Google Ads describes ROAS in terms of conversion value per cost. A lead assigned a value is not the same as collected revenue.
Incremental ROAS
Revenue caused by the advertising, estimated against a credible counterfactual, divided by the advertising spend being evaluated. It can differ materially from attributed ROAS.
Profit after advertising
Contribution from the relevant revenue minus advertising and other relevant costs. Production, creator fees, discounts, fulfilment, refunds and overhead must be handled consistently.
Hypothetical arithmetic only. This is not a campaign result or forecast.
Same dashboard. Different economics.
Spend: 100. Attributed revenue: 400. Illustrative currency units.
Reported ROAS: 400 ÷ 100 = 4×
Contribution margin before ads: 40%
Contribution after ads: 400 × 40% − 100 = 60, before creative costs and fixed overhead.
If only 200 of the revenue is incremental, incremental revenue ROAS is 2× and incremental contribution after the same ad cost is 200 × 40% − 100 = −20.
With a 40% contribution margin and no other incremental costs, simplified revenue ROAS breaks even at 1 ÷ 0.40 = 2.5×. Different margins, costs or repeat-purchase assumptions change that threshold.
Before you trust the number
- Is the numerator actual net revenue, estimated lead value or lifetime value?
- Are multiple channels claiming the same order?
- What purchases might have occurred without the advertising?
- Have creator fees, production and offer subsidies been counted consistently?
- Does a marginal increase in spend bring profitable additional customers?
- What uncertainty remains around the lift estimate?
Test the additional effect
Where feasible, use a well-designed holdout or geographic experiment to estimate lift. Google’s Conversion Lift documentation distinguishes incremental outcomes from ordinary attributed conversions. Small samples and spillover can make results inconclusive; inconclusive is not zero effect.
Sources & scope
Google Ads: About Target ROAS bidding →Conversion-value-per-cost definition.
Google Ads: About Conversion Lift →Incrementality measurement and experiment approaches.
Google Ads: Understand conversion lift measurement data →Difference between incremental and attributed conversions.