Return on ad spend is revenue divided by ad spend, and whether a given figure is profitable depends entirely on gross margin. Break-even ROAS is one divided by that margin: a 40 per cent margin breaks even at 2.5×, a 20 per cent margin at 5×, and a 70 per cent margin at 1.43×. A 3× ROAS is excellent for one business and loss-making for another.
That single line is why ROAS benchmarks circulate endlessly and mean nothing. The number that matters is your own break-even, and it takes one division to find.
What does break-even look like across margins?
The relationship is a reciprocal, so it climbs sharply as margin thins.
| Gross margin | Break-even ROAS |
|---|---|
| 70% | 1.43× |
| 50% | 2.00× |
| 40% | 2.50× |
| 35% | 2.86× |
| 25% | 4.00× |
| 20% | 5.00× |
| 15% | 6.67× |
Below the break-even figure, every additional unit of ad spend loses money — which is the useful way to read a target that is being missed. It is not underperformance, it is a campaign operating at a loss.
Why should the target sit above break-even?
Because the margin figure does not include everything. Returns, payment fees, support and the fixed costs of running the business all come out after gross margin, so a campaign at exactly break-even produces zero gross profit and a net loss.
Setting an automated bidding target at break-even is worse still, because the platform optimises toward the number you gave it and lands around it. The target needs headroom — enough that hitting it leaves something after the costs the margin excludes.
What does raising the target do to volume?
It squeezes spend into the highest-converting slices of the auction, and the relationship is not linear. Volume falls away sharply past a point, so a target that looks efficient can deliver very little revenue.
Ten thousand of revenue on two thousand of spend at a 40 per cent margin is a 5× ROAS against a 2.5× break-even: 4,000 of gross profit less 2,000 of spend leaves 2,000. Raise the target to 8× and the platform might cut spend to 500 and revenue to 4,000 — 1,600 of gross profit less 500 leaves 1,100. Better ROAS, less profit. Efficiency and profit are different objectives and they diverge past the break-even point.
Why do the platform numbers not add up?
Because after privacy changes broke deterministic tracking, the sum of every channel’s claimed revenue routinely exceeds actual revenue — sometimes by a wide margin. Two platforms can each claim the same sale, and neither is lying by its own attribution rules.
MER sidesteps the argument entirely: total business revenue divided by total marketing spend, across every channel and regardless of attribution. Eighty thousand of revenue on sixteen thousand of spend is a MER of 5×, meaning marketing costs 20 per cent of revenue.
MER is blunt — it cannot tell you which channel worked — and it is the only figure that cannot be double-counted. The workable practice is to steer channels by ROAS and to judge the whole operation by MER.
The other reason MER matters is that it is the figure a finance function already understands. It has no attribution model inside it, it reconciles against the accounts, and it can be tracked month over month without anyone arguing about which platform deserves credit.
What is the most you can pay for a click?
Gross profit per conversion multiplied by conversion rate. A 45 profit at a 3 per cent conversion rate makes each click worth 1.35, and that is the ceiling before the campaign stops paying.
This one line explains why two advertisers on the same keyword can rationally pay wildly different amounts. A business with a high conversion rate on high-margin services can outbid one converting at 2 per cent on thin retail margins by a factor of twenty — and neither is bidding irrationally.
It also reframes the usual complaint about expensive keywords. A keyword is not expensive in the abstract; it is expensive relative to what a click is worth to you, and that is a number about your business rather than about the auction.
Questions people ask
Is ACoS the same as ROAS? The same relationship inverted. ACoS is ad spend over revenue and ROAS is revenue over ad spend, so a 5× ROAS is a 20 per cent ACoS. Which you see depends on the platform.
Should ROAS use revenue or gross profit? Revenue, by convention, which is exactly why the break-even calculation exists. A profit-based ROAS would break even at 1× and confuse every benchmark conversation.
How do I split a budget across channels? By weights rather than percentages, so adding a channel does not mean recalculating everything. A 10,000 budget split 50/30/15/5 gives 5,000, 3,000, 1,500 and 500.
Does new-customer ROAS differ? Substantially, and it should be measured separately. Retargeting flatters a blended figure by taking credit for customers who were going to return anyway.
What margin should I use in the break-even? Contribution margin rather than gross, if you can — it takes out fulfilment, payment fees and returns as well as cost of goods, and those are the costs an ad-funded order actually incurs.
Work out your break-even, add headroom, and judge the whole account on MER. The ROAS calculator and break-even ROAS calculator start from margin, the target ROAS calculator adds the headroom, the MER calculator removes the attribution argument, the break-even CPC calculator sets the bid ceiling, and the ad budget split calculator divides the total.