CPA vs ROAS: Which Metric to Use When

Both metrics describe the same underlying ad performance, just from opposite directions — one in dollars spent per sale, the other in revenue multiple per dollar spent. Neither means anything for profitability without your margin as context.

By Marginory team · Online sellers with hands-on experience across Etsy, Shopify & PODUpdated Fee data verified against official platform documentation

Two views of the same number

MetricFormulaExpressed as
CPAAd spend ÷ Number of conversionsDollar amount per sale
ROASRevenue ÷ Ad spendMultiple (e.g., 4x)

When CPA is more useful

If you already know your margin in dollars per unit (say, $12 profit per sale before ad spend), CPA lets you set a direct, intuitive target: keep cost-per-sale below $12 to stay profitable. This dollar-to-dollar comparison is often easier to reason about than converting margin into a ROAS multiple first.

Calculate your target CPA →

When ROAS is more useful

ROAS is the default metric shown in most ad platform dashboards, and it scales naturally across products at different price points — a 4x ROAS target applies whether your product is $20 or $200, whereas a CPA target needs recalculating for every different price point.

Calculate your target ROAS →

Both need margin as the reference point

Neither metric tells you anything about profitability in isolation. A campaign can hit an impressive ROAS or a low-looking CPA and still be unprofitable if margin isn't part of the calculation. Always translate whichever metric you're watching back into a profit-per-sale or margin check before deciding to scale ad spend.

Frequently Asked Questions

What's the difference between CPA and ROAS?
CPA (cost per acquisition) measures ad spend per individual sale/conversion in dollars. ROAS (return on ad spend) measures revenue generated as a multiple of ad spend. They describe the same campaign performance from different angles.
Which metric is more useful?
Neither is universally better — CPA is more intuitive when comparing against a known target cost-per-sale (like your margin per unit in dollars), while ROAS is more common in ad platform reporting and easier to compare across campaigns with different price points.
Can CPA and ROAS give conflicting signals?
They shouldn't mathematically conflict since they're derived from the same underlying numbers, but they can be easier or harder to misinterpret depending on which one you default to — ROAS especially, since a high multiple feels good regardless of your actual margin.