Break-even ROAS, then the truth behind it.
Meta reports what Meta can see. This desk computes the ROAS you need to break even, the ROAS you need to hit a profit target, and the size of the gap between the platform’s reported figure and what actually cleared your bank.
How much overstatement can you absorb?
22%Working with post-iOS attribution
Why the reported figure drifts
App Tracking Transparency removed the device identifier that once let Meta observe a purchase directly. What replaced it is modelling: Aggregated Event Measurement reports conversions in delayed, privacy-thresholded buckets, and the platform fills the gaps statistically. Modelling is not fraud, but it is directional, and it is tuned by a party with an interest in the answer.
Three effects push the reported figure above reality, and they stack:
- View-through credit. A one-day view window claims orders from people who scrolled past a video and later searched your brand name.
- Cross-platform double counting. Meta, Google and TikTok each claim the same order. Add the three reported revenues together and you will often exceed your actual takings.
- Organic cannibalisation. Retargeting and branded prospecting frequently buy customers who were already arriving.
Estimating your own overstatement
Do not guess twice. Measure it once, then re-measure quarterly.
- Take a fortnight of total store revenue and total Meta spend. Compute blended MER for Meta-dominant periods.
- Compare that against Meta’s own reported revenue over the same window.
- The ratio between them approximates your overstatement. A reported $240,000 against a realistic $190,000 share implies roughly 21%.
- Validate with a geo holdout or a scheduled spend pause. A week dark in two matched regions is crude but honest.
Most direct-to-consumer accounts land between 15 and 30%. Heavy retargeting and short purchase cycles push it higher. Long consideration windows and strong organic brand demand push it higher still.
Break-even ROAS is a margin fact, not a benchmark
Everything that scales with an order comes out first: goods, freight, payment fees and the expected value of returns. What remains, as a share of revenue, is your contribution margin. Break-even ROAS is its inverse.
Break-even ROAS = 1 ÷ contribution marginTarget ROAS then adds the profit you want to keep.
Target ROAS = 1 ÷ (contribution margin − target margin)If contribution margin is 45% and you want to keep 15 points of it, target ROAS is one divided by 0.30, which is 3.33. Setting a bid cap against any other number is arithmetic with extra steps.
What to do when adjusted ROAS sits below break-even
- Shorten the attribution window in Ads Manager to seven-day click and one-day view. Comparability suffers for a fortnight; honesty improves permanently.
- Split prospecting from retargeting in reporting. Retargeting almost always carries the heaviest overstatement.
- Raise average order value before cutting spend. Bundles and free-shipping thresholds move break-even ROAS down without touching the media plan.
- Check freight last. It is the cost most likely to have drifted since you last modelled it.
Common questions
What ROAS do I actually need on Meta?
Whatever one divided by your contribution margin comes to, plus a buffer for attribution overstatement. A store with 45% contribution margin and 20% overstatement needs a reported ROAS of roughly 2.78 to genuinely break even, because the adjusted figure lands at 2.22, which is the true break-even point.
Is a 2× ROAS good or bad?
Unanswerable without margin. At 70% contribution margin, 2× is strongly profitable. At 35%, it is a loss on every order. This is why ROAS benchmarks shared between brands are close to useless.
Should I use one-day click or seven-day click?
Seven-day click reports higher and optimises on more signal, which usually improves delivery. One-day click reports lower and tracks blended MER more closely. Run seven-day click for optimisation and judge the account on blended MER, rather than trying to make the reported figure honest.
Does the Conversions API fix attribution?
It improves signal quality and event match rates, which helps delivery. It does not remove view-through credit or cross-platform double counting, so overstatement persists. Expect better performance, not a truer number.
Why is my break-even CPA more useful than ROAS?
Because it is a dollar figure you can set a bid against. Break-even CPA is simply contribution per order before advertising. If each order leaves $38 of contribution, any acquisition cost under $38 is profitable, whatever ratio the platform prints.