What is ROAS?

Return on ad spend is revenue divided by the spend that produced it: a ROAS of 4 means four units of revenue per unit of spend. The arithmetic is simple. What is not simple is the numerator, because "the revenue this spend produced" can be counted in very different ways.

Platform ROAS is what Google Ads or Meta report: revenue from conversions the platform saw and claimed, divided by spend on that platform. It is the best tool for optimizing inside the platform, and every platform counts in its own way, so the numbers cannot be compared across platforms or added up.

Attributed ROAS comes from a rule such as last click. It credits the touchpoint before the purchase, so branded search and retargeting look strong and TV looks like nothing.

Incremental ROAS comes from a marketing mix model: the sales a channel caused above the baseline, divided by its spend, with every channel measured against the same actual sales. It is usually lower than platform ROAS for bottom-funnel channels and higher for upper-funnel ones, because it only counts sales that would not have happened anyway.

Even incremental ROAS comes in two forms. Average ROAS is the channel's total incremental sales over its total spend for the period: what it earned per unit on average. Marginal ROAS is what the next unit would earn at today's spend, read off the slope of the saturation curve. On any channel with diminishing returns the marginal figure is lower than the average, and it is the marginal figure that budget decisions belong on. Scaling a channel because its average ROAS is high is the most common way to waste the increase.

In Odins both are shown per channel, next to the platform's own figure, so the three can be compared and the difference understood.

• Marginal ROAS: What the next unit of spend in a channel earns at today's level. The number a budget change actually asks for.
• Incremental sales: The sales a channel caused that would not have happened without it. The only kind of sales worth paying for.
• Saturation curve: The response curve per channel: how sales respond to spend, and where the next unit stops paying.
• Attribution: Crediting conversions to the touchpoints that preceded them. Great inside a platform, not a currency across them.
• Last-click attribution: All the credit to the final click. Simple, free, and systematically wrong about which channels create demand.

Go one level deeper: Why Google Analytics and Meta report different numbers. Or go back to all terms in the glossary.