How to calculate ROAS: formula and examples

To calculate ROAS, divide the revenue your ads brought in by what the ads cost. That is the whole formula, and it takes ten seconds. The slow part is deciding which revenue and which spend go into it, because VAT, returns and what you count as ad spend each move the answer.

Below, one month of a shop’s ads is worked through in euros. Depending on what you count, it reads 4.84, 4.0 or 3.8, and each product is then judged against its own break-even. The ROAS glossary entry has the short definition, and the break-even ROAS calculator takes your own figures.

What is the ROAS formula?

ROAS (return on ad spend) is revenue from ads divided by the ad spend that produced it. Say a shop spends €2,000.00 on ads in a month and records €8,000.00 of revenue from them: €8,000.00 ÷ €2,000.00 = 4.0. Every €1.00 of ads brought in €4.00 of revenue.

The same result is written three ways, depending on the tool:

Google Ads Help describes target ROAS as a percentage: sales ÷ ad spend × 100%.

Count revenue excl. VAT and after returns

The ROAS formula takes whatever revenue you give it. Two common mistakes make the number look better than it is, and neither shows up as an error.

VAT. Your margin and break-even ROAS are worked out on prices excl. VAT, because the VAT goes to the tax office, not to you. So ROAS has to use revenue excl. VAT too. If Google Ads records conversion values incl. VAT, the same month looks better than it was: at 21% VAT, €8,000.00 of revenue shows as €9,680.00, and €2,000.00 of ad spend reads as a ROAS of 4.84 instead of 4.

Returns. Google Ads counts an order as a conversion when it is placed, and keeps counting it if the customer sends it back. Take the refunds and cancellations for the period off the revenue: if €400.00 of the €8,000.00 came back, ROAS is €7,600.00 ÷ €2,000.00 = 3.8.

With both corrections the month is a ROAS of 3.8, while the ad account can show 4.84.

Return rates differ by category, so the gap between reported and kept ROAS differs by product too. A product group that comes back in the post a lot will report a ROAS it never keeps. Ecommerce return rate shows how to work out yours and what it does to break-even.

Count only the spend behind that revenue

Ad spend is the media cost of the campaigns whose revenue you counted, for the same dates. Divide Shopping revenue by the cost of Shopping, Performance Max and brand search together and you get a number that describes none of them.

Agency fees and software don’t belong in the ROAS of a campaign. If you want the full cost of running ads, add them and call the result ROAS after fees. Label which one you report, so next month compares like with like.

Orders also arrive days after the click, so the last few days of any period under-report. Compare full weeks rather than yesterday.

An account ROAS hides its products

An account ROAS is an average, and averages are good at keeping secrets. Split the same €2,000.00 evenly over two products. Product A earned €6,000.00 from €1,000.00 (a ROAS of 6.0) and Product B earned €2,000.00 from €1,000.00 (2.0). The account shows €8,000.00 ÷ €2,000.00 = 4.0, a figure neither product comes near.

With only a handful of clicks, one order can move a product’s ROAS from 0 to 10, so judge the products that have enough clicks and group the rest until they do.

For one product, divide its conversion value by its own ad cost. Doing the Product A and Product B sum for thousands of products is slow by hand, and Product Segmentation runs it for every product: Product Metrics ML places each one in one of six segments by its ad clicks and its return, where return is ROAS or POAS, working from the revenue and spend your ad account reports.

Read every ROAS against its own break-even

A ROAS is only good or bad against the product’s break-even ROAS, which is 1 ÷ contribution margin. Contribution margin is the share of the price (excl. VAT) left after cost of goods and the variable costs of the order. A 25% margin gives 1 ÷ 0.25 = 4.0. The margin calculator gives a product’s contribution margin from its cost, price and order costs.

In profit terms, POAS = ROAS × contribution margin, and break-even POAS is 1.0. POAS vs ROAS explains when to use which.

Back to the two products. Product A has a 40% margin, so it breaks even at 2.5 and its 6.0 clears the line easily. Product B has a 20% margin and needs 5.0, so its 2.0 loses money. In euros, A keeps €2,400.00 on €1,000.00 of ads (+€1,400.00) and B keeps €400.00 (−€600.00). The account keeps +€800.00 and reads 4.0 against a blended break-even of 2.9. It looks healthy, and €600.00 of that month’s loss sits inside it.

Product B has two ways out. Its ROAS has to reach 5.0, or its margin has to reach 50%, because at a ROAS of 2.0 break-even is 1 ÷ 2.0. A better price or lower variable costs move the margin, and lowering its priority limits the loss while you work on it.

Products above break-even are candidates to increase priority. What is a good ROAS answers “is 4 good?” for different margins, and break-even ROAS per product covers the four steps.

The VAT basis moves break-even as well as ROAS, so a 4.84 incl. VAT and a 4.0 excl. VAT can describe the same product on the same day:

The break-even ROAS calculator takes revenue incl. VAT with a VAT rate and works out ROAS and break-even on that basis. If your conversion values exclude VAT, enter revenue excl. VAT and set VAT to 0%.

ROAS, POAS and ROI on the same account

All three divide by ad spend; only the top line changes. ROAS counts revenue, POAS counts contribution margin, and ROI on ad spend counts what is left after the ads. Here they are for the two-product account above, at a blended 35% margin: €8,000.00 of revenue excl. VAT is €2,800.00 of contribution margin.

Illustrative data. POAS equals ROAS × margin (4 × 0.35 = 1.4), and ROI equals POAS − 1. The marketing ROI entry explains the profit-based measure. Total revenue over total marketing spend is a different ratio again: see marketing efficiency ratio vs ROAS.

Four inputs that skew the number

When a ROAS looks wrong, check these before you question the maths:

ROAS also counts only the revenue your ad account reports. It can’t show which of those sales would have happened without the ads, which is what incrementality measures. Incrementality testing shows how to measure it with a holdout or geo split.

Calculate yours in five steps

Do the sum once per product you care about, then once for the account. I’d start with the best sellers: they take most of the budget, so a wrong reading costs most there.

If the result is under break-even, check the VAT basis, the returns and the spend you counted before you touch a campaign.

Finish with profit per euro of ads

Comparing a product’s ROAS with its break-even tells you whether it makes money. To see how much, multiply its ROAS by its contribution margin. The result is POAS, profit on ad spend: the contribution profit each €1.00 of ads brings back.

• Product A: 6 × 40% = a POAS of 2.4. Every €1.00 of ads brings €2.40 of contribution profit, €1.40 more than it cost.
• Product B: 2 × 20% = a POAS of 0.4. Every €1.00 of ads brings €0.40 back, so it loses €0.60.

A POAS above 1 makes money and below 1 loses it, whatever the product’s margin, so one number works for every product. POAS vs ROAS explains when to steer by POAS instead of ROAS. The break-even ROAS calculator handles one product at a time, and Product Segmentation shows every product at once: it works on ROAS straight away and switches to POAS once you connect margins.

Run that sum for your ten biggest spenders before you change a single target.