POAS vs ROAS: Why Profit on Ad Spend Should Run Your Account

Two ecommerce brands both report a 3.0× ROAS this month. One of them made money. The other one paid for the privilege of selling its own stock. Same metric, same number, opposite outcomes - and ROAS cannot tell you which brand is which.

That's the whole case for POAS in one paragraph. ROAS measures revenue per pound of ad spend. POAS measures profit per pound of ad spend. Since no business banks revenue, the second number is the one your P&L actually feels.

What ROAS measures - and what it hides

ROAS (return on ad spend) is attributed revenue divided by ad spend. A 4.0× ROAS means £4 of tracked revenue for every £1 spent on ads. It's the default metric in every platform dashboard, and it isn't useless - it's a fast, directional read on whether a campaign is finding buyers.

But ROAS is blind to three things:

1. Your margin. ROAS treats a £100 sale of 20%-margin product exactly the same as a £100 sale of 70%-margin product. The cash outcome is wildly different.
2. Attribution inflation. Each platform marks its own homework. Meta and Google will happily both claim the same purchase, so channel ROAS routinely adds up to more revenue than you actually made. (That's why the account-level check is MER, not summed ROAS.)
3. Costs that scale with sales. Shipping, payment fees, returns - all invisible to ROAS, all very visible to your bank balance.

What POAS measures

POAS (profit on ad spend) swaps revenue for contribution margin - what's left of each sale after COGS, shipping, fees and returns:

POAS = (revenue − COGS − shipping − returns) ÷ ad spend

A POAS of 1.0 means each pound of ad spend returned exactly one pound of profit - break-even. Above 1.0 your ads are genuinely making money; below it, they're an expensive way to move inventory.

Worked example: £40,000 of attributed revenue on £10,000 of ad spend is a 4.0× ROAS - looks great. But at a 46% contribution margin, that revenue carries £18,400 of profit, so POAS is 1.84×. Still good - but now you know how good, and how much room you have before scaling stops paying. Drop the margin to 25% and the exact same "4× ROAS" account earns a 1.0× POAS: running flat out to stand still.

You can run your own numbers in our free POAS calculator.

The same 3× ROAS, five different businesses

Here's what a reported 3.0× ROAS actually earns at different contribution margins:

Two takeaways. First, there is no such thing as a universally "good ROAS" - the same figure is a disaster or a triumph depending on margin. Second, the ROAS number you should actually care about is your break-even ROAS (1 ÷ contribution margin) - the floor below which ads lose money. Find yours with the break-even ROAS calculator.

When ROAS still earns its place

POAS advocates sometimes oversell the case, so let's be fair to ROAS. It's still useful when:

• Margins are uniform. If everything you sell carries roughly the same contribution margin, ROAS and POAS move together and ROAS is the simpler proxy.
• You need speed. Platforms report ROAS in near real time; profit data usually lags. For in-flight budget calls within a channel, ROAS against a margin-aware target works fine.
• You're diagnosing a channel. ROAS is a reasonable within-platform comparison of campaigns and creative, as long as you never mistake it for a business outcome.

The failure mode isn't using ROAS - it's optimising the business to it. ROAS is a diagnostic. POAS (per channel) and MER (blended) are the decision metrics.

How to move your account to POAS: three steps

1. Get contribution margin per product, not per business. A blended average hides the story. Pull COGS, shipping cost, payment fees and return rate per SKU or per category. Most brands discover their ad account is heavily weighted toward their worst-margin products, because those are often the easiest to sell.

2. Set margin-aware targets per campaign. Convert each campaign's product mix into a break-even ROAS, then set targets above it. A campaign selling 60%-margin bundles can profitably run at a ROAS that would sink a campaign selling 25%-margin singles.

3. Feed profit to the algorithms. In Google Ads, send margin-adjusted conversion values (profit, not revenue) so Smart Bidding's tROAS effectively becomes tPOAS - the algorithm starts chasing your best-margin buyers instead of your biggest baskets. On Meta, the equivalent is passing profit-adjusted values through the Conversions API. This is the single highest-leverage change: it doesn't just report profit, it optimises toward it.

Why POAS falls as you scale - and why that's fine

Expect blended POAS to decline as spend rises. Your first pounds of budget reach the cheapest, most likely buyers; every additional pound reaches colder, more expensive ones. That's not failure, it's marginal economics. The question that should gate scaling is not "is POAS falling?" but "is marginal POAS still above 1.0?" - does the last increment of spend still return more than a pound of profit? Keep scaling while it does. Brands that judge spend increases on blended POAS alone stop scaling too early and leave profitable growth on the table.

POAS, MER and break-even ROAS: how the three fit together

These three metrics answer three different questions, and a well-run account watches all of them:

• Break-even ROAS - what's my floor? A fact about your margins: the ROAS below which ads lose money.
• POAS - is this channel making profit? Contribution margin ÷ channel spend.
• MER - is the whole engine efficient? Total revenue ÷ total marketing spend, reconciled to real revenue so no platform can inflate it.

(All three, plus the rest of the acronym soup, are defined in plain English in our glossary.)

The next blind spot: profit metrics still can't see your stock

One warning as you get religious about POAS: profit per pound of spend is still only half the picture, because it ignores inventory. A 2.0× POAS product with six days of stock left shouldn't be scaled - you'll sell out mid-flight and hand the algorithm a dead signal. A 0.9× POAS product with 400 days of cover might deserve spend, because ageing stock has a carrying cost too. The framework for combining profit with days of cover - scale, maintain, ease or liquidate - is covered in our guide to inventory-aware advertising.

The bottom line

ROAS tells you your ads generated revenue. POAS tells you whether that was worth doing. If you only make one measurement change this quarter, calculate your contribution margin properly and re-read your "best" campaigns through a POAS lens - most accounts find at least one celebrated campaign that's been quietly losing money, and at least one unloved one that deserves more budget.

That's the shift: from ads that look good in a dashboard, to ads judged on the profit they leave behind. It's the same standard we hold our own work to - we build creative from proof, and we judge it on profit.