ROAS is not revenue.

Every ad platform will tell you it's doing a wonderful job. Meta says it returned five times your spend. Google says six. Add up what each one claims and you sometimes get more sales than the business made in total. That's the first clue.

Return on ad spend, as reported inside the platform, isn't exactly a lie. It's a number produced by the party that benefits most from it looking good. We'd treat it the way we'd treat a real estate agent's opinion of the house they're selling: worth hearing, not worth betting on.

Why the platform number flatters itself

Three reasons, mostly.

Everyone claims the same sale. Someone sees a Meta ad on Monday, searches your brand name on Wednesday, clicks a Google ad and buys. Meta counts the sale. Google counts the sale. You banked it once.

It takes credit for people who were going to buy anyway. Brand search and retargeting are the usual suspects. Show an ad to someone who already has your product in their cart, then claim the sale, and you get a terrific ROAS without much new revenue.

It counts revenue, not profit. A ROAS of 4 sounds healthy. Whether it is depends on your margins, your discounts, your shipping and your returns. The platform knows none of that.

A worked example

These numbers are made up, so swap in your own.

Say you spend $10,000 a month on ads. Meta reports $30,000 in sales and Google reports $26,000. Between them the platforms are claiming $56,000, a combined ROAS of 5.6.

Now open your store's own sales report. Total revenue for the month, from every source, was $70,000. So the platforms are claiming 80% of everything you sold, which would leave your email list, your repeat customers and everyone who simply typed in your web address with a fifth of sales between them. Possible. Not likely.

Then take it down to profit. Say 40% of that $70,000 is left after product costs, shipping and payment fees. That's $28,000. Take off the $10,000 in ad spend and you have $18,000 to cover wages, rent and everything else. That's the number your accountant cares about, and it doesn't appear on any ad dashboard.

Here's where it gets interesting. You raise spend to $15,000, and the platforms report ROAS holding steady at 5.6. Lovely. But total revenue only moves to $76,000. The extra $5,000 in spend bought $6,000 in sales, which at a 40% margin is $2,400 in gross profit.

You paid $5,000 to make $2,400. The dashboard looked the same the whole time.

What we look at instead

1. Blended return. Total revenue divided by total ad spend. It can't double count, because it only uses two numbers and both come from your own books. In the example it fell from 7 to about 5 while platform ROAS didn't move.
2. The bank statement. When spend went up, did total sales go up with it? If you can pause a campaign for a fortnight and nothing changes in the bank, that campaign was taking credit, not creating sales.
3. Contribution margin. Revenue minus cost of goods, shipping, fees and ad spend. What's left to run the business. Set your targets against this one, because it's the one that pays people.

If the ads are working, the bank statement will say so.

Platform ROAS still has a job. It's useful for comparing one ad against another inside the same platform, with the same settings, over the same period. Think of it as a steering number, not a scoreboard.

What to do Monday

• Pull last month's total revenue and total ad spend. Divide one by the other and write it down. That's your baseline.
• Add up what each platform says it drove and compare the total with real revenue. Note the gap.
• Work out contribution margin for the month. Rough is fine.
• Pick one campaign you suspect is claiming sales that would have happened anyway, usually brand search or retargeting. Plan a short pause and watch total revenue, not the dashboard.

Then do it again next month. The trend will tell you more than any single reading.