ROAS Return on Ad Spend
Return on ad spend divides the revenue a campaign produced by what it cost to run. Spend $1,000, take $4,000, and the ROAS is 4, usually written 4x. That is the whole calculation, which is much of why it spread so far.
One figure, four relatives
The confusing part is not the arithmetic. It is that half a dozen metrics sit around ROAS measuring almost the same thing, and account managers swap between them mid-sentence.
| Metric | What sits on top | Reads as |
|---|---|---|
| ROAS | ad revenue | multiple, higher is better |
| ACOS | ad spend, over revenue | percentage, lower is better |
| POAS | gross profit | multiple, break-even at 1 |
| ROMI | gross profit, minus the spend | percentage, break-even at 0 |
| MER | total revenue, over total spend | multiple, whole business |
ACOS is ROAS upside down. POAS and ROMI are the same quantity in two notations. MER is the only one that leaves the campaign layer entirely.
The margin is the missing half
A 4x ROAS is not good or bad on its own. At a 50% gross margin you break even at 2x, so 4x is comfortable. Drop to an 18% margin and break-even moves to 5.6x, so the same 4x quietly loses money on every order.
Which is the practical point. The platform reports a number that looks like performance, and performance is not what it measures. It measures revenue efficiency. The break-even ROAS turns your own margin into the threshold that number has to clear.
What the reported figure has stopped being
Since Apple’s tracking prompt, attribution on the major platforms has drifted. Purchases get counted twice, or credited to ads that a customer would have found anyway, and view-through windows sweep in people who never clicked. The direction of the error is consistent: reported ROAS runs high.
None of that makes the metric useless. It does mean the number in the ads dashboard and the number that lands in your bank account have stopped agreeing with each other, and that the gap between them widens the more heavily you lean on retargeting. Teams that got burned by this tend to keep a second figure that no platform can inflate — total revenue over total marketing spend, whatever the attribution says.
Where it still earns its place
Same product, same margin, two campaigns: ROAS ranks them instantly and correctly. That is a real job and it does it well.
The failure mode is comparing across products with different margins, or reading the campaign figure as a business result. For the first, ROMI does the work. The second has no fix at the campaign layer at all.