The most common mistake with ROAS is reading it as profit. Owners divide the revenue an ad platform reports by what they spent, see a number that looks healthy, and conclude the ads are working. That number only says how much revenue came back per dollar of spend. It says nothing about whether you kept any of it. What works is a short chain: calculate ROAS, compare it to your own break-even ROAS, and check that the revenue in the formula is real. This page walks through each step.
The formula is one line; the trouble is what goes into it
ROAS stands for return on ad spend. You divide the revenue attributed to your ads by the amount you spent on those ads over the same period. If the result is two, each dollar of spend brought back two dollars of revenue. Anyone can do the division. Most of the errors come from the two inputs: which revenue counts, and which costs count.
Matching the two inputs to each other
Both numbers must cover the same campaigns and the same dates. A common slip is to take a full month of sales from your point-of-sale system and divide it by a single campaign's spend, which inflates the result. The reverse happens too: total spend goes against revenue from one channel, and the ratio looks worse than it is. Pick a campaign or a channel, pull its spend, then pull only the revenue that came from it.
Decide up front whether spend includes agency fees, creative costs, and tools. The platform's ROAS column usually counts only media cost. That is fine for judging the campaign, but it flatters the result if you are judging the whole program.
Break-even ROAS: the number that gives yours a meaning
A ROAS figure is not good or bad on its own. It is good or bad relative to your margin. To find your break-even, take your gross margin on the product or service, meaning the share of each sale left after the direct cost of delivering it. Then flip it: break-even ROAS is the reciprocal of that margin, or revenue divided by gross profit. A business with thin margins needs a much higher ROAS just to avoid losing money than one with fat margins.
| Measure | How you calculate it | The question it answers |
|---|---|---|
| ROAS | Attributed revenue divided by ad spend | How much revenue does each ad dollar bring back? |
| Break-even ROAS | Revenue divided by gross profit (the reciprocal of margin) | What is the lowest ROAS that avoids a loss? |
| Profit on ad spend | Gross profit from ad-driven sales minus ad spend | Did the ads leave money behind? |
| Blended ROAS | All revenue divided by all ad spend | Is the business healthy overall, whatever the channel says? |
Check where the revenue number comes from
Ad platforms report revenue from their own tracking, and it can differ from your books. Common reasons: the same sale counted in two platforms, credit taken for customers who would have bought anyway, refunds not subtracted, and tax or shipping included in the total. If purchases are not recorded properly, the ROAS column is guesswork with a decimal point.
For a service business, the sale often happens offline, after a call or a quote. In that case the platform only sees leads, and any revenue figure is a value you assigned. Feed real closed-job revenue back into the platform if you can. Our guide on [setting up conversion tracking that stays working](/guides/setting-up-conversion-tracking) covers how to keep that signal from silently breaking.
Why a wobbly search and an expensive click raise the stakes
About 1540 searches a month across 12 related queries point to owners trying to work this out on their own. The main phrase itself swung from 1600 a month in September 2025 to 880 in May 2026. Interest peaks around planning and review seasons, which is when budgets are set and last quarter's ads are judged. That is a good time to sort out your inputs, before the numbers drive a spending decision.
Click cost matters here too. In competitive categories, the highest top-of-page bids reach $100.51. You will rarely pay that, but it shows that a click can be costly. A ROAS calculated on bad revenue data can keep you spending at prices you cannot afford.
A pass through your own numbers this week
Start with your margin: write down what a typical sale earns after direct costs, then turn it into a break-even ROAS. Next, pull a recent stretch of ad spend and the revenue your own records tie to it, not only the platform's figure. Compare the two versions of revenue. If they are far apart, fix tracking before you touch budgets. If they are close, set a simple rule for yourself: which ROAS gets more budget, which gets a review, and which gets paused.
Look at the result by campaign, not just in total. One strong campaign often hides a weak one, and branded search tends to look better than it deserves because those buyers were already looking for you.
If the numbers still won't reconcile
Sometimes the platform, your analytics, and your books all disagree, and no amount of spreadsheet work settles it. An outside team can audit your [conversion tracking](/services/conversion-tracking-audit) to find where revenue is lost or double counted, or take over [Google Ads management](/services/google-ads) once the measurement is trustworthy. You can also stay in-house: the steps above take an afternoon and cost nothing.