When strong ROAS hides weak profitability

Return on ad spend is the most quoted number in eCommerce marketing, and one of the most misread. A campaign reporting a ROAS of 3 sounds healthy. Spend one dollar, get three back. But revenue isn’t profit, and the gap between the two is where many growing brands quietly get into trouble.

We’ve written before about why we manage to MER rather than platform ROAS. This note goes one level deeper: even when the revenue number is right, it can still hide an unprofitable business.

What ROAS leaves out

ROAS compares revenue to ad spend. Everything else that happens to that revenue is invisible to it:

• Product cost. What the goods themselves cost to make or buy.
• Shipping and fulfilment. Pick, pack, postage and packaging.
• Payment fees. A small percentage of every transaction.
• Returns and refunds. Revenue that comes back out, plus the cost of handling it.
• Discounts. Often already netted out of revenue, sometimes not, depending on how reporting is set up.
• Repeat customers. Platforms frequently take credit for sales from people who were already going to buy again.

A ROAS of 3 on a product with thin margins can mean losing money on every order. The same ROAS on a high-margin product can be very profitable. The number alone tells you almost nothing.

Finding the break-even ROAS

The useful question is: at what ROAS does an order stop making money?

Start with the contribution margin before marketing, meaning the share of each sale left after product cost, shipping, fees and expected returns. Break-even ROAS is roughly one divided by that margin.

An illustrative example: if 40% of revenue is left after those costs, break-even ROAS is 1 ÷ 0.40, or 2.5. Anything below 2.5 loses money on the first order. A campaign at 3 is only modestly profitable, not the comfortable win the headline suggests. If the margin is 25%, break-even is 4, and that same ROAS of 3 is losing money on every sale.

Every brand should know its break-even number by product. It changes what “good performance” means.

When losing money on the first order is fine

Sometimes a brand should accept a first order that loses money. If customers reliably come back, the profit on later orders can pay back what it cost to acquire them. Subscription and consumable products often work this way.

The key word is reliably. That decision should rest on real repeat-purchase data, not optimism. And it comes with a cost that ROAS never shows: cash. Paying today for profit that arrives over the next six months means funding the gap, including the inventory to serve those customers. Growth that looks profitable on paper can still run a business out of cash.

What we look at instead

Alongside ROAS, we track a small set of numbers that tell the real story:

• Contribution margin after marketing, per order and in total.
• New-customer acquisition cost against first-order contribution.
• Payback period: how long until a new customer has repaid what it cost to win them.
• Blended efficiency across all channels, so platform over-reporting doesn’t flatter the picture.

When these line up, a strong ROAS is genuinely good news. When they don’t, it’s a warning sign dressed up as a success.

The takeaway

ROAS is a useful steering tool inside an ad account. It’s a poor scorecard for a business. Know your margins, know your break-even, and judge growth on what each customer actually leaves behind. The brands that do this can spend with confidence. The ones that don’t sometimes find out the hard way that they were buying revenue at a loss.