Glossary - Acquisition and attribution
ROAS
ROAS (return on ad spend) is the revenue an ad platform attributes to a campaign divided by the cost of that campaign, expressed as a multiple such as 3.2x. Ecommerce operators use ROAS to compare campaigns and channels inside ad platforms, judged against a margin-based break-even ROAS floor rather than in isolation.
| Attributed revenue | Revenue the ad platform or attribution model credits to the campaign over the period. |
| Ad spend | The media cost of that campaign over the same period. |
Worked example
At 3.2x against a 2.5x floor, the campaign clears its variable costs - subject to how accurate the attributed revenue in the numerator really is.
What is a good ROAS?
There is no universal good ROAS, because the floor is set by your margin structure, not by your platform or your category. Break-even ROAS is 1 divided by contribution margin before marketing, so the same arithmetic gives a 4.0x floor at a 25% contribution margin, 2.5x at 40%, and about 1.43x at 70%. A ROAS that looks strong on one margin structure loses money on another.
What counts as good therefore depends on your contribution margin before marketing, the share of new versus returning customers in the attributed revenue, and how far the platform's attribution overstates what it actually caused. Compute your own floor with the free break-even ROAS calculator.
ROAS vs related metrics
| Metric | What it measures | How it differs from ROAS |
|---|---|---|
| Break-even ROAS | The minimum ROAS at which a campaign stops losing money: 1 ÷ contribution margin. | Not a performance number - the floor a reported ROAS has to clear. |
| MER | Total store revenue ÷ total marketing spend, whole account. | No attribution involved; it cannot judge one campaign and cannot be inflated by platform claims. |
| POAS | Contribution margin from attributed orders ÷ ad spend. | Swaps revenue for margin in the numerator, so break-even is always 1.0. |
| Blended CAC | Total sales and marketing spend ÷ all new customers. | Prices the customer rather than the revenue, and needs no attribution model. |
Common mistakes
- Judging ROAS without a margin floor. A 3.5x means nothing until you know whether your break-even is 1.6x or 4.0x.
- Treating platform-reported ROAS as ground truth. Attribution windows overlap, so several platforms can claim credit for the same order.
- Optimising revenue ROAS when order margins vary. A discount-heavy 4.0x campaign can carry less profit than a full-price 2.5x one.
- Mixing new and returning customers. Retargeting existing buyers inflates ROAS with revenue that may have happened anyway.
- Comparing ROAS across platforms with different attribution windows and models as if it were one consistent number.
FAQ
Not necessarily: at a 25% contribution margin a 4.0x ROAS is exactly break-even, so the campaign makes nothing. Whether any ROAS is good depends on the margin left on the sale before marketing, not on the multiple itself.
ROAS measures one campaign using attributed revenue; MER divides total store revenue by total marketing spend with no attribution at all. ROAS answers which campaign, MER answers whether marketing as a whole is paying for itself.
Because each platform counts the revenue it claims, not the revenue it caused, and attribution windows overlap. One order can appear as a conversion in several dashboards at once, which is why per-platform ROAS figures cannot be added together.
Related
- Break-even ROAS
- MER (marketing efficiency ratio)
- POAS
- Contribution margin
- Attribution model
- Break-even ROAS calculator
- All 14 free tools →
In Blufire, S5 Acquisition & Attribution shows attribution models compared side by side, so you can see how much of a campaign's reported ROAS survives a different lens, and S6 Marketing & Channels tracks CM1-MER per channel.
Updated July 2026