ROAS (Return on Ad Spend)

Amazon & D2C glossary · Harpy Media

ROAS (Return on Advertising Spend) is advertising revenue divided by advertising spend, expressed as a ratio: a 4.0 means every unit of currency spent on ads produced four back in sales. It is the mirror image of advertising cost of sales — a 25% figure there equals a 4.0 here — and the two describe the same campaign from opposite ends.

What is ROAS?

ROAS (Return on Advertising Spend) is advertising revenue divided by advertising spend, expressed as a ratio: a 4.0 means every unit of currency spent on ads produced four back in sales. It is the mirror image of advertising cost of sales — a 25% figure there equals a 4.0 here — and the two describe the same campaign from opposite ends.

Agencies tend to prefer ROAS because the numbers look more encouraging; operators tend to prefer cost of sales because the percentage lines up directly against margin. Which lens you use matters less than applying it consistently and interpreting it against what the product can actually afford.

The number is meaningless without a threshold

A ROAS on its own says nothing about whether a campaign is working. If a product carries a 40% margin before advertising, roughly two and a half back per unit spent leaves the business intact — so a 4.0 is comfortable and a 2.0 would be losing money. On a thinner product, the same 4.0 might be the difference between profitable and not.

So the useful practice is to calculate the break-even ratio first, from the product’s own unit economics, and treat every campaign result relative to it. That threshold differs by product, changes with price, and moves when fulfilment costs change — which is why a single target ROAS applied across an entire catalogue is a good way to subsidise the products that cannot afford it.

Using it to allocate rather than to judge

Read at campaign level, the metric becomes a routing tool: spend concentrates where the returns are strongest and gets pruned where they are not. Read at account level, it flatters — strong performers mask loss-makers inside the average.

The nuance that keeps it honest: a lower ROAS is often acceptable on terms that support organic ranking or defend a position, because the value includes the ranking benefit rather than only the immediate sale. And the inverse trap is mistaking high revenue for a good result. A campaign that returns ten times its spend on a product with a razor-thin margin can still be a worse use of capital than one returning four times on a product with real contribution.

In practice

A seller spends a thousand on sponsored-product campaigns and generates four thousand in ad-attributed sales — a clean 4.0. With a pre-advertising margin of 40%, that return comfortably covers its own cost and leaves a healthy contribution, so the campaign is scaled deliberately rather than left running on default bids.

How Harpy Media helps

Advertising efficiency is part of our growth work: break-even thresholds calculated per product, campaigns allocated against them, and the ranking value of spend accounted for rather than assumed.

ROAS FAQ

What is a good ROAS on Amazon?

There is no universal figure — it depends on your margin. Work out the break-even ratio for each product from its unit economics, then treat anything above that as profitable and judge how far above by how much contribution it leaves.

How does ROAS relate to ACOS?

They are inverses: a 25% advertising cost of sales is a 4.0 return on ad spend. The same information, expressed from the opposite direction — pick one and keep it consistent across reporting.

Should I pause campaigns with a low ROAS?

Not automatically. Low returns can be acceptable where the spend defends a ranking or holds a position. Pause only after checking whether the campaign is buying visibility you actually need, or simply buying clicks that do not convert.