MER vs ROAS vs POAS vs ROI: Which Number Actually Runs Your Store?

The problem: the dashboard says ROAS 4, the bank account says something else

The conversation I have most often with store owners starts with the same sentence: "The campaigns are excellent, the ROAS is high, but I don't see the profit." Then we open Meta Ads Manager and find a good number, we open Shopify and find a lower one, and we open the bank statement and find a third story entirely.

This isn't necessarily a fault in the platform or in tracking. In most cases it's a fault in the question. Each of these metrics — the ROAS, the MER, the POAS, the ROI — answers a different question, and anyone who uses one of them to answer another metric's question will make the wrong decision while feeling perfectly reassured.

What follows is the order I use in practice: what each number measures, where it breaks down, and which number I look at for which decision.

ROAS: an operational metric, not a profitability metric

ROAS is the revenue attributed to a given ad divided by the cost of that ad. The important word here is "attributed". The number you see inside the platform is not a measurement of what happened in your store's economics; it is the platform's opinion of its own contribution.

Hence its recurring problems:

• Double counting. If a customer sees an ad on Meta and another on TikTok, then searches for your store name on Google, all three platforms may claim the same order. Add up the attributed revenue from the three dashboards and you'll find it larger than your store's actual revenue.
• Ignoring what isn't measured. Orders coming from WhatsApp, from a friend's referral, or from a customer who returns two months later don't show up where you expect them to.
• Flattening products. ROAS doesn't know that one product has a thin margin and another has a comfortable one.

This doesn't mean scrapping ROAS. I use it daily, but as an operational metric inside the account: which campaign is outperforming its sibling? Which creative is starting to saturate? Has the new campaign exited learning? For relative comparisons within the same platform, ROAS is excellent. For judging the health of the business, it's the wrong number.

MER: the number the platform can't dress up

MER is total store revenue in a given period divided by total advertising spend in that same period. All revenue, all spend, no attribution and no allocation.

Its beauty is that it doesn't admit argument. It doesn't matter whether the Conversions API is working well, or how attribution windows behave after privacy updates. The revenue is in your system, the spend is on the platform invoices, and the ratio between them is arithmetic fact.

That's why it's the first number I ask any new client for, before I open their ad account. If MER is falling month after month while the ROAS reported in the dashboards is flat or improving, I know — before looking at a single campaign — that there is attribution double counting, or growth driven by discounts, or a growing reliance on retargeting that re-buys customers who would have bought anyway.

The drawback of MER is that it tells you nothing about why. It tells you the machine has become less efficient, not which part of it. So the right mix is: MER for judgement, ROAS for diagnosis inside the account.

POAS: when profit margin enters the equation

POAS replaces revenue with gross profit: the profit generated by the orders divided by the ad cost. This is where the number starts speaking the store owner's language rather than the media buyer's.

And the difference between it and ROAS isn't cosmetic. It can reverse the decision entirely.

A worked example

Suppose a store selling two types of products, using simplified numbers purely for illustration:

Campaign A sells a product priced at 200 SAR, with product, shipping and collection costs of 160 SAR, i.e. a 40 SAR margin. You spent 1000 SAR on it and it brought in 20 orders. Revenue is 4000 SAR, so ROAS equals 4. But gross profit is 800 SAR against 1000 SAR of spend, i.e. a POAS of 0.8. The campaign is losing money.

Campaign B sells a product also priced at 200 SAR, but with a total cost of 100 SAR, i.e. a 100 SAR margin. You spent 1000 SAR on it and it brought in 12 orders. Revenue is 2400 SAR, so ROAS is only 2.4. But gross profit is 1200 SAR against 1000 SAR of spend, i.e. a POAS of 1.2.

Whoever manages by ROAS will double the budget on campaign A and switch off campaign B. Whoever manages by POAS will do the opposite. This is exactly what I've seen happen in stores selling a wide range with varying margins, especially where cash-on-delivery costs and return rates are high, because both factors eat the margin and never show up in ROAS.

The price of POAS is that it demands data discipline: a true cost per SKU, actual shipping cost rather than an estimate, payment gateway fees, and a return rate calculated at product level. Without that you aren't calculating POAS, you're calculating ROAS multiplied by a guess.

ROI: the bigger question you don't manage day to day

ROI looks at the return on the whole investment: ads, salaries, tools, inventory, the agency or the freelancer, and everything else you paid to run the operation. By nature it's measured over a longer horizon, and it intersects with LTV because a customer's value isn't complete at the first order.

It's the right metric for questions like: is this new market worth entering? Is this product line worth continuing? But you don't run an ad account with it, because it responds slowly and doesn't distinguish one channel from another.

How I arrange the four in a single report

1. MER weekly and monthly — is the whole machine running more or less efficiently?
2. POAS at channel and product level — are we spending where the margin is?
3. ROAS inside the platform daily — which creative, which audience, which campaign, for the small fast decisions.
4. ROI quarterly — is this business as a whole worth the capital and the effort?

Add CAC and AOV to that, because they explain most of the movement in MER: if MER drops, the reason is usually either that customer acquisition cost has risen or that average order value has fallen. Together, the two numbers turn an observation into a diagnosis.

What to do next week

Don't start by changing campaigns. Start with three steps in this order:

Calculate MER for the last six months, month by month, from your system's revenue and the platform invoices rather than the dashboards. Put the numbers in one column and look at the trend. Then add up the reported ROAS from all platforms for the same month and compare the total attributed revenue with your actual revenue; the gap is the size of the double counting you've been basing your decisions on.

After that, build a true cost sheet for your best-selling products, including shipping, cash on delivery and returns, and calculate POAS per channel. You'll most likely find that your channel ranking changes.

And if this exercise produces a number that bothers you, that's the point of it. Decisions made on an uncomfortable correct number beat decisions made on a comfortable wrong one.