POAS (Profit On Ad Spend) is gross profit divided by ad spend, where ROAS is revenue divided by ad spend. POAS accounts for cost of goods, fees, and returns, so it reflects money actually kept rather than revenue generated. For businesses with variable or thin margins, optimizing on ROAS can scale unprofitable sales; POAS optimizes on the number that keeps the lights on.
As the person who reconciles the books, I can tell you the most dangerous number in marketing is a healthy-looking ROAS. It feels like proof the account works — revenue divided by spend, big multiple, everyone nods. But ROAS counts the top line before a single cost comes out. The business doesn’t bank revenue; it banks profit, and those two numbers can point in opposite directions. For more on improving your UX, consider the impact of a fast landing page.
POAS — profit on ad spend — closes that gap. It’s the same ratio with an honest numerator, and for any business where margins vary, it’s the difference between scaling growth and scaling losses.
What POAS measures that ROAS hides
The two metrics differ in one place — the numerator — but that one change rewires every decision built on top of it. For more on query control, see our approach to negative keywords.
| ROAS | POAS | |
|---|---|---|
| Numerator | Revenue | Gross profit |
| Subtracts COGS | No | Yes |
| Subtracts fees & returns | No | Yes |
| Reflects cash kept | No | Yes |
Why a great ROAS can still lose money
Picture two products at the same 5× ROAS. One carries a 60% margin; the other, 15%. On ROAS they look identical, so the algorithm scales both. But the thin-margin product may be selling at a loss once shipping and returns are counted, and you’ve just told the machine to find more of it. ROAS made a losing product look like a winner.
Same 5× ROAS, very different profit
Illustrative — margin changes everything.
How to put POAS to work
The practical move is to feed profit, not revenue, into the platform as the conversion value — through the tag or, better, a server-side feed tied to real margin data. From there smart bidding optimizes toward POAS automatically. Reporting follows the same logic: every campaign is judged on the profit it produced, reconciled against the books, not on the revenue the platform claimed.
Should every business switch to POAS?
If all your products carry identical margins, ROAS is a fine proxy. The moment margins vary — and for almost every real catalog they do — ROAS misallocates, and POAS becomes the metric that actually protects the business.
ROAS answers a marketing question; POAS answers a survival question. The businesses that last are the ones that bid, report, and decide on the profit they keep — not the revenue they can show in a slide.
This article is a spoke node connected to our core technical hubs. To explore the broader architecture, visit our primary pillar pages: