What is ROAS?

The question is not just what your ROAS is, but which ROAS you need to break even. That figure, your break-even ROAS, is set by the margin, not by an industry rule of thumb:

• Work out the gross margin. If you sell for 1,000 kronor and the item costs you 600 to deliver, the margin is 400 kronor, or 40 percent.
• Break-even ROAS is one divided by the margin. At a 40 percent margin you break even at a ROAS of 2.5: below that the advertising loses money, above it earns.
• Low-margin businesses need a high ROAS, high-margin businesses get by on low. A ROAS of 3 is excellent for a low-margin e-commerce store and a loss for a service with an 80 percent margin that needed a ROAS of 1.25 to turn a profit.

This is why a bare ROAS figure does not tell you whether the advertising pays off. Two companies with the same ROAS can be on opposite sides of profitability, entirely depending on the margin behind it. The first thing we do in a review is work out your break-even, because without it ROAS is just a number with no threshold to compare against.

Even with the right break-even, ROAS is an incomplete metric. Four things it does not see can turn a seemingly profitable campaign into a loss:

• Margin. ROAS counts revenue, not profit. A high ROAS on products with a thin margin can earn less than a lower ROAS on products with a fat one.
• Returns. In e-commerce, a share of what is sold gets returned. ROAS books the sale, not the returns, so the real return is lower than the figure shows.
• Lifetime value. A customer who buys again is worth more than the first purchase. A campaign with a low ROAS on the first purchase can be highly profitable over time, which ROAS never captures.
• Attribution. Who gets the credit for the purchase? The buyer may have seen a display ad, then searched for you and bought after a newsletter. Credit it all to the last click and you overstate some channels and understate others.

None of these objections mean ROAS is worthless. They mean ROAS is one of several metrics, not the verdict. Decision support, not status: a good basis puts ROAS next to margin, return rate and lifetime value, so you see whether the advertising actually makes money, not just whether a figure looks high.

The most common question about ROAS is also the one with the least universal answer. "What is a good ROAS?" implies there is a global verdict. There is not, and here is why:

• A good ROAS is higher than your break-even, full stop. If break-even is 2.5, a ROAS of 3 is good and a ROAS of 2 is a loss, regardless of what the industry says.
• Rules of thumb like ROAS 4 are meaningless without the margin. The same four can be brilliant or catastrophic depending on the business behind it.
• The customer journey matters. Advertising that captures sharp purchase intent shows a higher ROAS than advertising that creates demand early. Comparing their ROAS directly punishes the work that builds future purchases.
• Scale matters. The first krona of advertising reaches the most purchase-ready and gives the highest ROAS. Scaling up usually lowers ROAS, but can still pay off in absolute kronor.

Benchmark against your own break-even and your own development over time, not against a number someone mentioned at a conference. A ROAS of 2 can be excellent for one business and poor for another. Without break-even as a threshold and margin as context, the chase for a higher ROAS figure is easy to win and easy to lose money on at the same time.

A ROAS figure is never better than the tracking behind it. Before you trust it, or trust someone who promises to raise it, three things are worth checking:

• Does the tracking measure correctly? A misconfigured pixel can count the same purchase several times or miss conversions entirely. An impressive ROAS on broken measurement is fiction.
• Whose is the data? Ad accounts, pixel and measurement setup should be yours. Change agency and the figures and history should come with you, not be held hostage.
• What is promised in advance? No one honest guarantees a particular ROAS before the campaign has met reality. A guaranteed ROAS is a hope disguised as a promise.

What we do is set up the measurement correctly and read ROAS in the light of margin and business, not report a figure that looks good: decision support, not status. An experienced hand quickly sees whether a high ROAS is real or a measurement error, and that is often the difference between scaling a profitable channel and scaling a loss. To find out what your advertising actually returns, see how Memorise works with advertising.