epROAS

epROAS stands for expected Profit Return on Ad Spend. It measures the profit a campaign returns per unit of marketing spend, after accounting for the returns those orders are expected to generate. It is Dema's formulation of the metric more widely known as POAS or pROAS.

The formula

epROAS = Net Gross Profit 2 ÷ Marketing Spend

Each part of the name maps to a component of that calculation:

• e — expected, meaning the expected return rate of the products sold is deducted rather than ignored.
• P — profit, specifically Gross Profit 2, which is revenue after cost of goods and fulfilment costs.
• ROAS — the ratio of that figure to marketing spend.

Gross Profit 2 adjusted for expected returns is Net Gross Profit 2, which is why that is the numerator.

Why it differs from ROAS

Standard ROAS divides revenue by spend. Because it uses revenue rather than profit, it treats every euro of sales as equally valuable — which they rarely are. Two campaigns can post identical ROAS while one sells high-margin products that customers keep and the other sells low-margin products with a high return rate. The second is considerably less profitable, and ROAS cannot show that.

epROAS closes both gaps at once. Deducting cost of goods and fulfilment accounts for margin differences between products, and deducting expected returns accounts for revenue that will not survive the return window. The result is a figure that can be compared meaningfully across campaigns selling different product mixes.

Reading the number

Because the numerator is profit rather than revenue, epROAS values are lower than the ROAS figures for the same campaigns, and they should not be compared against ROAS targets. An epROAS of 1.0 means a campaign returned exactly its spend in profit — breaking even after goods, fulfilment and expected returns, before overheads. Your own threshold depends on what overhead the contribution needs to cover.

Where it fits

epROAS is a channel and campaign efficiency metric, calculated from your own order and cost data in real time. It tells you how profitably spend converted, not whether that spend was incremental. For the causal question — whether the sales would have happened anyway — use incrementality testing.

Further reading: why ROAS can't be trusted, and what to use instead.