What is a good Meta ads ROAS for ecommerce?

Return on ad spend is revenue divided by spend, and revenue is not profit. Two stores can both report the same figure and one is making money while the other is going backwards, because their gross margins are different.

So the first number you need is gross margin as a share of revenue: price less cost of goods, payment fees, packaging and shipping you absorb. Break-even return on ad spend is one divided by that share. A store keeping a third of every pound needs roughly three in revenue for every pound spent just to stand still. A store keeping two thirds needs half as much.

That is the floor, not the target. Above the floor you still have fixed costs, staff, software and the profit you are actually in business for. So set a working target above break-even with enough room to cover those, and treat anything below break-even as buying revenue at a loss regardless of how healthy the number looks in the dashboard.

Once you have your own floor, other people's targets become irrelevant, which is the point.

Meta reports conversions it can attribute, using its own attribution window and its own view of who saw an ad. That is a legitimate way to measure, and it is not the same as revenue in your accounts.

Three things drive the gap. Attribution windows: a click today and a purchase in five days is counted by Meta and may be attributed elsewhere in your analytics. View through attribution: someone who saw an ad and later searched for your brand may be counted by Meta and by Google at the same time. And modelling: where signal is missing, some conversions are estimated rather than observed.

None of that means the platform number is useless. It means it is a steering number, good for comparing ad sets and creatives against each other, and unreliable as a measure of what your ads earned overall.

The number to run the business on is blended: total store revenue divided by total advertising spend across every channel. It is crude and it cannot be gamed. Watch the platform figure for decisions inside the account and the blended figure for decisions about the account.

When the two disagree in direction over a month, believe the blended one.

A Meta account can report a comfortable figure while doing almost nothing useful, if most of its conversions are retargeting people who were going to buy anyway. Retargeting reliably shows a high return because it takes credit for demand created elsewhere.

So split the reporting. Look at prospecting on its own, and at retargeting on its own, and hold different expectations for each. Prospecting is buying new customers and will show a lower return. Retargeting will show a high one and cannot be scaled, because the audience is capped by how many people visited your site.

Better still, look at return on new customers. Most stores can flag first orders, and comparing spend against revenue from first time buyers tells you what the account is actually contributing. If prospecting sits below break-even on the first order, that can still be fine, but only if you know your repeat rate and can say how long the payback takes.

The common mistake is reading a healthy blended account number, scaling the budget, and watching it collapse. It collapses because the retargeting pool was fixed and the new money all went into prospecting, where the true number always lived.

Chasing a high number is not the same as making more money. If you tighten targeting and cut budget until the return looks excellent, you will end up with a small, very efficient account that contributes very little profit.

The question is total gross profit after advertising, not the ratio. Spending more at a lower return can produce more profit in absolute terms, right up to the point where the marginal order costs more than it earns. That point, not the account average, is where to stop.

This is why the account average is a poor decision tool when scaling. What matters is what the next increment of spend returns. Increase the budget in steps, wait for the results to settle, and watch whether total gross profit after ad spend rises or falls. When it stops rising, you have found your ceiling for now.

Repeat purchases change the maths as well. If customers reliably buy again, you can afford a lower return on the first order, because the second one costs nothing to acquire. Work out your typical repeat behaviour before deciding how much of a loss on order one you can carry, and be conservative, because a repeat rate you hope for is not a repeat rate you have.

Write down four numbers and keep them somewhere everyone can see. Gross margin as a share of revenue. Break-even return, which is one divided by that share. Your target return, which covers break-even plus overheads and profit. And your blended target across all channels.

Then decide the reporting window. Judging a Meta account on a single day is noise, especially at modest budgets. A week is usually the shortest honest window and a month is better, because delivery, creative fatigue and weekday patterns all move within a week.

Build one simple weekly view: total store revenue, total ad spend across every channel, blended return, gross profit after ad spend, and new customer count. Five numbers. If those move in the right direction, the account is working, whatever any individual campaign says.

And re-derive the target whenever costs change. A supplier price increase, a shipping rate rise or a change in your product mix moves break-even, and a target set last year against last year's margin can have you scaling something that is now losing money on every order.