ROAS is the most quoted metric in paid media and one of the worst used. The formula is trivial. The problem is that almost nobody defines what to compare it against, and without that, ROAS cannot tell you whether a campaign is good or bad.

The formula

ROAS = revenue generated ÷ advertising spend

Spend USD 1,000, generate USD 4,000, and your ROAS is 4 — or 400%. Both ways of expressing it are correct and both circulate in the same meeting, which is its own small source of confusion. If someone says “we’re at a ROAS of 4”, check whether they mean four times or four percent.

That’s the two-second part. Now the part that matters.

Your break-even ROAS

A ROAS of 4 can be excellent or ruinous. It depends entirely on your margin.

Break-even ROAS = 1 ÷ gross margin

  • 50% margin → you need a ROAS of 2 just to break even.
  • 25% margin → you need 4.
  • 10% margin → you need 10.

The business with a 10% margin and a ROAS of 4 is losing money on every sale, and its dashboard is showing 400%. This is the most expensive mistake I see in small accounts: nobody ever calculated the number the ROAS should be measured against.

Before you look at any campaign’s ROAS, write your break-even ROAS down somewhere. It’s a single number and it changes the entire reading.

Why ROAS alone isn’t enough

It doesn’t know your costs. ROAS looks at revenue, not profit. Shipping, payment processing, returns, cost of goods — none of it is in there. This is why more teams now look at POAS (profit on ad spend), which replaces revenue with contribution margin. It’s harder to build, because it requires product cost to reach the ad platform, and it’s far more honest.

It doesn’t know whether the customer comes back. A campaign at ROAS 2 bringing customers who purchase three more times is worth more than one at ROAS 5 bringing one-time coupon buyers. Which is why ROAS has to be read alongside LTV and CAC, not instead of them.

The last click claims it. Each platform calculates the ROAS it reports using its own attribution model, and each is generous with itself. Add up the ROAS Google, Meta and TikTok report and you will usually have “generated” more revenue than the business actually booked. See what last-touch attribution does.

It can’t tell new demand from existing demand. A brand campaign almost always shows the highest ROAS in the account. Not because it’s the best campaign, but because it’s charging you for people who were already coming to find you.

The number that fixes most of this

MER = total business revenue ÷ total media spend

Also called blended or combined ROAS. It has one enormous advantage over platform ROAS: it can’t be inflated by attribution. Total revenue comes from your billing system; total spend comes off your card. There’s no model in between deciding who gets credit.

The practical way to use it: watch MER month over month. If you increase spend by 30% and MER holds, you scaled well. If you increase spend and MER falls, you’re buying volume that doesn’t pay for itself — however good each platform’s own ROAS still looks.

It doesn’t replace campaign-level ROAS, which is what you need for decisions inside a platform. What it gives you is a judge with no stake in the verdict.

Reading all three together

A worked example. An ecommerce business with a 35% gross margin, so a break-even ROAS of about 2.9.

  • Google Ads reports ROAS 5.2. Meta reports 3.1. Combined reported revenue: USD 180,000.
  • Actual revenue in the billing system that month: USD 128,000.
  • Total media spend: USD 42,000. MER = 3.05.

Both platforms are above break-even on their own numbers. The business is at 3.05 against a 2.9 break-even, which is a thin 5% cushion before overheads — a very different conversation from “our Google account is at 5.2.” The USD 52,000 gap between reported and actual revenue is double counting, and it’s normal.

Now the decision. If you push another USD 10,000 into Google because 5.2 looks great, the question isn’t what Google will report. It’s what MER does next month. That’s the only number that answers it.

How to use it in practice

  1. Calculate your break-even ROAS from your gross margin. One number, written down.
  2. Track MER monthly to know whether the business is improving.
  3. Use campaign-level ROAS only to move budget within a platform.
  4. When a campaign’s ROAS spikes, check whether it’s eating your brand traffic — that’s the Performance Max problem in particular.
  5. Once you have margin data flowing, move toward POAS for anything with uneven product margins.

ROAS isn’t a bad metric. It’s a partial metric used as if it were total, and the difference between those two is what decides whether you scale something that works or something that merely looks like it does.

If you’re not sure the ROAS you’re seeing reflects the business, measurement is usually the first thing I audit.