Guides / Paid media / 7 min read / Updated 1 August 2026

What is a good ROAS?

Somebody will tell you 4x. They do not know your margins, so they are quoting folklore. Break-even ROAS is arithmetic, it takes one division, and every meaningful version of this question is downstream of it.

Break-even is one division

Return on ad spend is revenue divided by ad spend. Break-even ROAS — the point where advertising costs exactly what it brings in — is one divided by your gross margin.

• 40% gross margin → break-even ROAS 2.5
• 60% gross margin → break-even ROAS 1.67
• 80% gross margin (software, most services) → break-even ROAS 1.25
• 20% gross margin (a lot of retail) → break-even ROAS 5.0

That list is the entire reason a single "good ROAS" number cannot exist. A 3x return is a loss for one of those businesses and a triumph for another. Anybody quoting a target without asking about margin is telling you about their last client, not about yours.

Use gross margin, not revenue margin after overheads, and remember to include payment processing, shipping, packaging and your realistic return rate. A 3% return rate on a physical product moves break-even more than people expect.

The number the platform shows you

Every ad platform reports conversions it believes it caused. That belief is built from its own attribution model, its own lookback window, and increasingly from modelled estimates where it cannot see the user directly.

None of that is fraud, but it has consequences you should know before you optimise against it:

• Windows differ. A 7-day click window and a 1-day click window produce different numbers from identical reality. Compare like with like or do not compare.
• View-through counts somebody who scrolled past. Sometimes that is real influence. Sometimes it is credit for a sale that was always going to happen.
• Platforms mark their own homework. Meta and Google will both claim the same sale, and if you add their reports together you will find you sold more than you did.

Blended ROAS, and why it is boring and correct

Total revenue for the period divided by total advertising spend for the period. Sometimes called MER. It is slow, it does not tell you which ad worked, and it cannot be argued with, because both numbers come out of systems that have no stake in the answer.

Run both. Platform ROAS for deciding which ad set to cut on Tuesday; blended ROAS for deciding whether the whole channel is worth funding. When the two move in opposite directions for more than a couple of weeks, believe the blended one and go and find out what the platform is over-claiming.

Why a higher ROAS is often the worse outcome

This is the part that costs people the most money, and it is counterintuitive enough to be worth an example.

Say break-even is 2.0. Campaign A spends $5,000 at 5x: $25,000 revenue, $12,500 gross profit, $7,500 after ad spend. Campaign B spends $25,000 at 3x: $75,000 revenue, $37,500 gross profit, $12,500 after ad spend.

Campaign A has the better ratio and Campaign B has the better business. As you increase spend you buy less-qualified attention and the ratio falls — that is not a failure, it is what scaling is. The question is never "is the ROAS high" but "is the next dollar still above break-even".

The reverse mistake exists too: an account run at a punishing ROAS target will look efficient and starve, because the bid is too low to enter the auctions where the buyers are.

First-order ROAS versus what a customer is worth

If people buy again, judging a channel on the first purchase understates it, sometimes badly. A subscription or a consumable can rationally be acquired at break-even or below on order one.

This is a real argument and it is also the single most abused one in advertising. It only holds if you have measured repeat rate from your own data rather than assumed it. "We can afford it on lifetime value" without a cohort report behind it is how companies run out of cash while reporting growth.

What to actually put on the dashboard

• Break-even ROAS, computed once from real margins, written down where everybody can see it.
• Blended ROAS weekly, from your own order data and your total spend across every platform.
• Contribution after ad spend in currency, not as a ratio. This is the number the business actually keeps, and it is the one to grow.
• Platform ROAS per campaign, used for relative decisions between campaigns and never quoted to anybody as revenue.