How to read return on ad spend (ROAS)

“Return on ad spend: 5x.” The number looks excellent, and it may well be. It may also be covering a campaign that loses money. The difference between the two isn’t in the number itself, but in how you read it.

This article explains what return on ad spend (ROAS) measures, what it doesn’t, and how to calculate the threshold you specifically need for advertising to be profitable.

What is return on ad spend?

ROAS is the revenue attributed to ads divided by what was spent on them. If you spent SAR 10,000 and generated SAR 40,000 in attributed revenue, the return is 4: four riyals of revenue for every riyal of advertising.

Two words matter here: revenue and attributed. Revenue isn’t profit, and attributed isn’t necessarily what the ad actually created.

Why the number alone isn’t enough

Because revenue isn’t profit

From the SAR 40,000 in the example above, you must subtract product cost, delivery, payment fees, returns and perhaps discounts. If the margin left after those costs is 25%, the profit before advertising is only SAR 10,000 — exactly what you spent on the ads. In practice, the “excellent” return means breaking even.

Because credit isn’t impact

Part of the attributed revenue might have happened without the ad: existing customers, or people who searched for your brand by name. That part inflates the apparent return without adding real growth. We explain the difference between credit and impact in How to measure the impact of growth channels.

Because it ignores later value

On the other hand, the return may look weak on the first order while the customer comes back and buys again. Here the number understates the ad’s real value.

Calculate your own break-even return

The most important number to know before judging any campaign is the return at which you neither make nor lose money. It is simple to calculate:

Break-even ROAS = 1 ÷ contribution margin

The contribution margin is the share of the selling price left after direct, variable costs. Some illustrative examples:

This is why there is no general “good return” that suits every business. A return of 3 may be excellent for a high-margin store and a clear loss for a low-margin one.

When to accept a return below break-even

There are sensible cases for accepting a loss on the first order:

• When customers really do come back: if your data shows new customers buy several times, their value is higher than their first order. The yardstick becomes customer lifetime value (LTV) against customer acquisition cost (CAC) — provided it is based on real customer behavior, not an assumption.
• When it is a learning phase: testing a new channel or segment may need time before efficiency settles. Set a spend cap and a test period in advance.
• When the goal is launching a new product: provided it is a conscious decision with a defined budget, not an unintended outcome.

How to read the return correctly

1. Separate brand search from everything else. Campaigns targeting people searching for your name usually show a high return because the customer was already coming to you. Read them separately so they don’t inflate the average of the rest.
2. Separate new customers from existing ones. Retargeting people who already bought isn’t the same as acquiring new customers.
3. Read it at product or category level. An account average can hide products that make money and others that lose it.
4. Compare it with your own break-even return, not with general averages or a competitor’s figures.
5. Check the source. Compare the revenue attributed in the platform with what was actually recorded in your store or system.

A hypothetical worked example

Imagine an online store with an average order of SAR 300 and a contribution margin of 40% after direct costs — SAR 120 of profit per order before advertising. Its break-even return is 2.5.

• A campaign to acquire new customers returns 2: each order cost SAR 150 in ads, and each order earns SAR 120. The loss on the first order is SAR 30.
• But its data shows that a good share of these customers come back and buy again within a few months, with almost no extra advertising. If the average profit a new customer generates over a year, based on the behavior of real past customers, clearly exceeds SAR 150, the campaign is profitable at customer level despite losing money on the first order.
• By contrast, a campaign aimed at people who already visited the store returns 6. It looks excellent, but many of those people would have bought anyway, and the real question is how much the campaign actually added.

The example shows why return isn’t read as a single number for the whole account, but alongside margin, customer type and later behavior.

Return on ad spend versus overall marketing efficiency

Some companies also track a broader indicator alongside attributed return: total revenue divided by total marketing spend over the same period, sometimes called marketing efficiency ratio (MER). Its strength is that it doesn’t rely on platform attribution at all, so overlap and tracking limits don’t affect it. Its weakness is that it doesn’t tell you which channel made the difference. Using both gives a balanced view: attributed return to optimize campaigns, overall efficiency to judge the full picture.

Complementary indicators

• Cost to acquire a new customer: clearer than ROAS when new and existing customers are mixed.
• Profit after advertising: contribution margin minus ad spend. This is the number that pays salaries.
• Ad spend as a share of total revenue: looks at the whole picture, not just what platforms attribute.
• Repeat purchase rate: tells you whether accepting a lower return at the start makes sense.

In summary

ROAS is a useful indicator, but it means nothing until you know your margin. Calculate your own break-even return, read the numbers split by brand, new customers and products, and tie every reading to what was actually recorded in your system. Then the number becomes a tool for a real decision rather than a slogan in a report — such as the decision on when to scale your ad budget.