What is POAS, and why it matters more than ROAS

YYour 4x ROAS might be the most expensive way to lose money.

Sounds absurd? Bear with me. This is a typical case you see all the time. A small-appliance store. Eight months with an agency. The weekly reports showed 3.8x ROAS. Spotless. Praised in every meeting. Good performance, congratulations.

You walk into the bank, ask for a statement, sit down with it. You have lost money. The ROAS had climbed. The business was quietly dying. Nobody had told you.

What good is a beautiful report if it does not pay the bills? That is the question everyone avoids.

That is why POAS exists

ROAS measures what you billed. POAS measures what you kept. Between the two sits everything that matters: cost of goods, shipping, returns, platform commission, payment processing.

This is not an accounting technicality. It is the difference between staying in business and going bankrupt elegantly, with beautiful reports on the table.

We got this wrong too. We took over accounts where we looked at ROAS first. That is exactly why we know where others go wrong. POAS taught us the lesson.

The typical case of the store that wanted to quit

Picture the scenario, it is one you see all the time. A small-appliance store. Eight months with an agency, 3.8x ROAS, a loss somewhere around RON 42,000. The owner's conclusion: I want to quit.

The mistake is not the business. It is the metric. You open the account and you see something that repeats endlessly.

The products had a net margin of 12%. High cost of goods, expensive shipping, frequent returns. Out of those 3.8 lei that came back per leu spent on ads, only 0.45 lei was profit. Subtract the leu spent on ads. You are left with minus 0.55 lei.

On every leu. Eight months. In silence.

Did the agency know? If they did and said nothing, that is serious. If they did not, they were incompetent. The owner pays dearly either way.

The formula. Plain.

ROAS is the easier cousin. Revenue divided by ad spend. It does not care whether the product you sell for EUR 50 costs you EUR 5 or EUR 48. It only counts transactions.

That is why it is dangerous. It gives you a number that sounds good. And can, at the same time, drive you to bankruptcy.

Two founders, same ROAS, different movie

What good is speed if you are headed the wrong way?

Take two online stores. Both spend EUR 10,000 a month on Meta. Both bring in EUR 40,000. ROAS 4x. Identical report. Twins. Until you open the calculator.

Store A bleeds EUR 4,000 every month. Store B banks EUR 8,000 in clean profit. Same ROAS. Two different worlds.

Imagine you are the founder of Store A. Your agency praises you every week with 4x ROAS. You feel good. Until the balance sheet.

IN, AT, and ABOVE the ads: three levels of looking at POAS

IN the ads. This is where most people look. Budget, audience, creative, CPC, CPM. Pure execution. At this level, POAS is a number you calculate after the campaign runs. Reaction, not action.

AT the strategy level. Here you decide which product gets budget, which platform, at what margin. POAS becomes a filter. Before you start the campaign, you already know whether the math works.

ABOVE everything. Here POAS dictates the business model itself. You change suppliers. You renegotiate prices. You drop products. You do not optimize the ad. You optimize the company. And that is the difference between a media buyer and someone who actually scales you.

How to calculate POAS on your business

1. 01Take the ad-attributed revenue for last month, straight from Meta Ads Manager or GA4. No estimates, no assumptions. Only what the platform says it brought.
2. 02Calculate net margin on your dominant product. Sale price minus cost of goods, shipping, packaging, platform fees, payment processing. Divided by sale price. That is the real margin, not the one in your head.
3. 03Multiply revenue by margin. That is the gross profit your ads produced, before subtracting ad costs.
4. 04Divide by ad spend. What comes out is POAS. Above 1, you are profitable. Below 1, you are losing.

When big revenue hides a loss

Take a common case. A store with big ad revenue, 2.8x ROAS reported by the old agency. Sounds decent. The owner is unhappy because the money does not add up anywhere.

You open the account. You do the math the way it should be done.

Big revenue, hundreds of orders, 2.8x ROAS. But the real net margin on the dominant product is thin. High variable costs, above-average returns, platform commission.

Real POAS comes out below 1. For every leu spent on ads, you lose money out of pocket.

Here you do not optimize audiences. You do not test another creative. You stop the bleeding and work on what causes it.

Then you work on the offer. Restructure products, kill the SKUs that consume budget with no profit, negotiate a volume discount with the supplier, adjust price where the market bears it.

In a few months, POAS climbs above 1. Roughly the same budget, sometimes fewer orders, but now each order brings profit, not loss.

Notice something important. Revenue can drop. Order count can drop. ROAS can drop. But if POAS goes up, the business turns profitable.

That is the difference between optimizing for pretty numbers and optimizing for real money. Put plainly: better fewer profitable orders than many that drag you into loss.

Where POAS can become a lie

Fair warning. POAS can be weaponized too.

If an agency tells you POAS 3x without showing you the sheet where they calculated margin, they probably used a round number that sounds good. Margin has to be net of every variable cost. Not gross, not optimistic, not rounded up.

What goes into variable cost, non-negotiable:

• Cost of goods. The real price at which the product lands in your warehouse, transport from supplier included. Not the quote price. The final price, with everything added.
• Shipping to customer. If you offer free shipping, the cost still comes out of margin. It does not disappear. It just hides.
• Packaging plus estimated returns on that category. If 5% of products come back, you add 5% to cost. Returned products either break or go into secondary stock at a discount.
• Platform commission, if you sell on Amazon, eMAG, Allegro. On eMAG you can hit 15 to 20%. On Amazon 12 to 15%. Real money leaving your profit.
• Payment processing. Stripe, PayU, bank fees. Between 1.9% and 3% per transaction. At EUR 100,000 in sales, that is EUR 2,000 to 3,000 you forget if you do not put it in the calculation.

Skip even one, and you are lying to yourself. POAS comes out higher than it is. A few months in, at year-end, you wonder where the money went.

It went into costs you never counted.

POAS varies per campaign

In the same account you can run a cold-audience campaign at 2.5x POAS on premium products, and a retargeting campaign at 0.4x POAS on cheap accessories.

ROAS blends them into a pleasant average. POAS separates them and shows you exactly what to scale and what to kill today.

From a real account I audited: cold traffic campaign on products at RON 300 to 500, margin 40%, POAS 2.1x. Same month, retargeting campaign on accessories at RON 30 to 50, margin 12%, POAS 0.3x.

Combined ROAS: 2.4x. Sounds good. Combined POAS: 0.85x. Losing money. If you do not separate by POAS, you do not know one campaign is destroying you.

What to do when POAS is below 1 but you want to scale

Common situation. POAS 0.8x. You are losing money, but the business is growing. What do you do?

It depends. If the growth brings customers who repurchase, maybe it is worth losing on the first sale to win on the second and third. That is called investing in LTV.

But not all businesses have LTV. If you sell TVs, the customer is not coming back in 3 months. If you sell shampoo, they are.

What is a customer worth over 12 months? If LTV covers the loss on the first sale, POAS below 1 on acquisition is temporarily acceptable. If it does not, it is not.

Short recap

• ROAS. Gross revenue divided by ad spend. The number on the report. Beautiful, incomplete, sometimes dangerous.
• POAS. Profit after margin divided by ad spend. The number in the bank. Ugly when below 1, the only one that matters when above.
• Two campaigns with identical ROAS can have opposite POAS. Same report, different movie.
• Net margin is calculated honestly, with every variable cost included. No rounding, no optimism, no excuses.
• Healthy e-commerce POAS sits between 1.5 and 2.5, sustainable over six months.