Few metrics are as quoted — and as misused — in digital marketing as ROAS. It is on every dashboard, it is the number taken to the meeting and, often, the one that decides where the next euro of budget ends up. And yet it is surprisingly easy to look at a high ROAS and make exactly the wrong decision.
ROAS stands for Return on Ad Spend. In its simplest form, it answers a direct question: for every euro spent on ads, how much came back in revenue? It is a useful metric, but only when we understand what it measures — and, more importantly, what it leaves out.
This article explains how to calculate ROAS, how it differs from ROI, how to find the point at which a campaign turns a profit, and which mistakes make this number look better than reality.
What ROAS is and how to calculate it
The formula could not be simpler:
ROAS = Revenue generated by ads / Ad spend
If a campaign generated €10,000 in revenue and cost €2,000 in ads, ROAS is 5 — or, as a percentage, 500%. For every euro invested, five came back in revenue. Platforms such as Google Ads and Meta Ads calculate and display this value automatically, which helps explain its popularity.
Notice one word: revenue. ROAS measures turnover, not profit. And that is exactly where the misunderstandings begin.
ROAS vs ROI: why they are not the same thing
It is tempting to treat ROAS and ROI as synonyms, but they measure different things.
- ROAS compares revenue with the cost of the ads. That is all.
- ROI (return on investment) compares profit with the total cost — ads, but also the cost of the product, shipping, commissions, payment fees, team salaries.
An example makes the difference clear. Imagine a ROAS of 4: €4 of revenue for every euro of ad spend, which looks excellent. But if the product margin is only 20%, those €4 of revenue bring €0.80 of gross margin — less than the euro spent to generate them. ROAS says "win"; ROI says "we are losing money". Anyone looking only at ROAS may be happily scaling a campaign that destroys value.
Break-even ROAS: where margin comes in
From this comes a far more useful concept than ROAS on its own: the break-even ROAS, the value above which the campaign stops making a loss. It is calculated from the margin:
Break-even ROAS = 1 / contribution margin
With a 20% margin, the break-even ROAS is 1 / 0.20 = 5. That is, this company only starts making money above a ROAS of 5 — a ROAS of 4 is, in fact, a loss. A business with a 60% margin, on the other hand, breaks even at a ROAS of just 1.67. This is why there is no universal "good ROAS": the same number can be excellent for one company and ruinous for another.
So what ROAS is good?
The honest answer is: it depends on your margin, your customer lifetime value and the goal of the campaign. A few reference points help you think:
- If you sell low-margin products, you need a high ROAS just to avoid losing money.
- If a first-time customer tends to buy again, you can accept a low — or even negative — ROAS on the first sale, because the profit comes from the following ones. This is where ROAS talks to customer lifetime value (LTV).
- Awareness campaigns should not be judged by ROAS at all: the return does not show up in the immediate sale.
Setting a target ROAS tied to margin, rather than a round number picked by eye, is the first step to using it well.
Aggregate ROAS vs ROAS by channel
A healthy overall ROAS can hide channels that are burning money. If Google Ads returns a ROAS of 8 and a second platform returns 1.5, the average may look comfortable while half the budget is wasted. Looking only at the aggregate number is one of the most common ways to leave money on the table.
It is worth breaking ROAS down by channel, by campaign and, where possible, by audience segment — but with a caveat, which is the subject of the next section.
Attribution and windows: the ROAS you see depends on what you count
The ROAS numerator — the revenue attributed to ads — is not an objective fact; it is the result of an attribution choice. A sale that passed through an ad, an email and a search is credited to whom? And for how long after the click does it still "count"?
The attribution window changes everything. A 30-day post-click window always shows a higher ROAS than a 1-day one, because it captures more sales. Two platforms, each claiming credit for the same sale, can add up to a ROAS that is impossible on paper. Before comparing ROAS across channels, you have to make sure they speak the same attribution language.
Incrementality: the ROAS that misleads
This is the most expensive mistake of all. ROAS assumes that all the revenue attributed to ads would not have happened without them — and that is almost never true. Some of those customers were going to buy anyway; the ad merely took the credit for a sale that was already guaranteed.
The most classic case is brand ads (the customer who searches the company name and clicks the ad instead of the organic result just below). ROAS looks fantastic, but much of that revenue would have happened all the same. The right question is not "how much revenue did this ad receive?" but "how much extra revenue did this ad generate?". Answering that requires incrementality tests — switching the campaign off in one region and comparing with another — not staring at the dashboard.
Mini-case: when a high ROAS hid a problem
An online shop was proud of an aggregate ROAS of 6. Breaking it down by channel, the team found that brand ads — which accounted for almost half the attributed revenue — had a ROAS of 15, while prospecting for new customers ran at around 2.5. The conclusion seemed obvious: cut prospecting and pour money into brand.
Before acting, they ran a test: they switched off brand ads for two weeks in half the country. Revenue barely fell — customers kept arriving through the organic result. The ROAS of 15 was, to a large extent, an attribution illusion. They redirected the budget to prospecting, the only engine bringing in new customers, and real growth accelerated, even as the aggregate ROAS dropped to 4. A lower number, a healthier business.
In practice
ROAS is a useful compass, as long as you know what it points to. Always use it alongside margin, so you know where your break-even sits; never confuse it with profit; and be suspicious of any ROAS that is too good without first asking how much of that revenue is incremental.
In practice, ROAS works better as a comparison tool — this channel against that one, this week against the last — than as an absolute verdict on a campaign. Combined with ROI, with customer lifetime value and with a healthy dose of incrementality testing, it stops being a pretty number on the dashboard and becomes what it should have been all along: a help in deciding where to invest the next euro.
