Break-even ROAS for subscription apps: the formula with store fees
How to work out break-even ROAS with store fees, VAT and ad levies, a worked 15% vs 30% example, and why day-7 targets mislead for subscription apps.
Key takeaways
- Break-even ROAS is the return on ad spend at which the money you keep equals what the ads cost. With every cut measured on the same revenue: break-even = 1 / (1 − store fee − other cuts).
- Take sales tax and VAT out of revenue first. Apple and Google charge their fees on the price without tax.
- A levy on ad spend, such as Meta's location fees, raises the cost of the spend: multiply break-even by (1 + levy). Income tax does not belong in it.
- With a 15% store fee and 1% of other cuts, break-even is 1.19×. At 30% it is 1.45×.
- A day-7 ROAS target only works if you know what share of a year's payments your cohorts collect by day 7. For subscription apps that share varies a lot.
What break-even ROAS means
Break-even ROAS is the return on ad spend at which an ad pays for itself and nothing more. ROAS is revenue divided by ad spend, so a ROAS of 1.5× means $1.50 of revenue for every $1 spent.
Revenue is not what you keep, though. The store takes a fee, governments take tax, and other tools and partners take a share. Break-even ROAS is the revenue multiple that leaves exactly $1 for each $1 of spend once all of that is gone. Above it the ads make money. Below it they lose money.
The formula
Break-even ROAS is 1 divided by what is left of each revenue dollar after the store fee and other cuts, as long as every cut is a share of the same revenue you use for ROAS. In a formula:
break-even ROAS = 1 / (1 − store fee − other cuts)The store fee is the commission Apple or Google takes. Other cuts are anything else charged as a share of revenue: a revenue-based tool, a payment processor's fee on web purchases, or a partner's revenue share. RevenueCat, for example, charges 1% of tracked revenue once you pass $2,500 a month (RevenueCat).
With a 15% store fee and 1% of other cuts, you keep 84 cents of each revenue dollar. Break-even is 1 / 0.84, or 1.19×.
The cuts add up only when they share a base. If a partner takes 10% of proceeds rather than of revenue, convert it first: 10% of what is left after a 15% fee is 8.5% of revenue.
Break-even ROAS on revenue without tax
Taxes: take VAT and sales tax out first
Measure revenue without sales tax or VAT before you apply the formula. Apple charges its commission on prices net of the taxes it collects (Apple), and Google's service fee is exclusive of transaction taxes (Google).
This matters because the revenue many dashboards show includes tax. RevenueCat's default Revenue includes taxes and the store's commission (RevenueCat). In a country with 20% VAT, a $11.99 sale holds about $2.00 of tax. Compare that gross figure with a break-even that ignores tax, and every campaign looks better than it is.
You have two clean options.
- Use revenue without tax, such as RevenueCat's Revenue (net of taxes), and apply the formula as it is.
- Use proceeds, which already have tax and the store fee taken out. Then break-even is 1 / (1 − other cuts), usually just above 1.0×.
Income tax is different. It is charged on profit, and at break-even there is no profit, so it does not belong in the formula. It matters later, when you decide how much profit a campaign must make to be worth running.
Ad levies: count them as spend
A levy on ad spend raises the cost side, so it multiplies break-even instead of joining the cuts. Meta, for example, adds location fees to ads delivered in some countries: 2% in the United Kingdom, 3% in France, Italy and Spain, and 5% in Austria and Türkiye. The fee is billed on top of your budget and does not appear in Ads Manager's reporting (Meta).
break-even ROAS = (1 + levy) / (1 − store fee − other cuts)A campaign that delivers only in Italy pays a 3% fee. At a 15% store fee and 1% of other cuts, its break-even rises from 1.19× to 1.03 / 0.84, or 1.23×. For a campaign spread across countries, use the levy weighted by where its impressions were delivered.
The same logic covers VAT on your ad invoices. If your business can reclaim it, leave it out. If it cannot, treat it as a levy.
A worked example: 15% vs 30%
The same campaign can make money at a 15% store fee and lose money at 30%. This fictional one spent $10,000, and the customers it brought paid $13,000 without tax: a ROAS of 1.30×, with other cuts of 1% of revenue.
| 15% store fee | 30% store fee | |
|---|---|---|
| Revenue without tax | $13,000 | $13,000 |
| Store fee | −$1,950 | −$3,900 |
| Other cuts, 1% | −$130 | −$130 |
| Kept | $10,920 | $8,970 |
| Ad spend | −$10,000 | −$10,000 |
| Profit | $920 | −$1,030 |
| Break-even ROAS | 1.19× | 1.45× |
| Break-even with a 3% levy | 1.23× | 1.49× |
So the first thing to get right is which rate applies, and that depends on the store, the program, the country and how long each subscriber has paid.
| Store and case | Fee | Source |
|---|---|---|
| App Store, standard commission | 30% | Apple |
| App Store, auto-renewable subscription after a subscriber's first year of paid service | 15% | Apple |
| App Store Small Business Program: up to $1 million in proceeds in the previous and the current calendar year, enrollment required | 15% from the first payment | Apple |
| App Store in the EU with Apple's in-app purchase, under the terms in effect from October 1, 2026 for developers who accept them | 26%, or 15% after a subscriber's first year and for Small Business Program members | Apple |
| Google Play, auto-renewing subscriptions, in markets still on the earlier fees | 15% | |
| Google Play billing in the US, EEA and UK from June 30, 2026, and in Australia and Japan from September 30, 2026 | 10% plus a 5% billing fee |
For the subscribers your ads bring in, the first-year rate is the one that matters. For most apps that is 15% on Google Play, and on the App Store 15% inside the Small Business Program or 30% outside it, or 26% in the EU.
Other storefronts have their own rates now. Japan, Brazil, mainland China and Taiwan all changed theirs in the past year (Japan, Brazil, China, Taiwan). In the United States, App Store apps may link to purchases on the web, and in December 2025 an appeals court ruled that Apple may not charge a commission on them until a court approves a fee (Ninth Circuit). These terms keep changing, so check the linked pages before you settle on a number.
How far through a year's payments are early cohorts?
It depends on the plans you sell. A cohort on annual plans has paid almost all of its first year once its trials convert, while weekly and monthly cohorts can take months to get most of the way. A cohort here means the customers who started in the same week.
The table shows three fictional apps whose cohorts end their first year at the same ROAS. Only the timing differs. "By day 7" means the first seven days, day 0 to day 6, and refunds are left out.
| Fictional app | Plans and trial | Day 7 | Day 30 | Day 90 | Day 180 |
|---|---|---|---|---|---|
| App A | Weekly plan, no trial, 85% renew each week | 15% | 56% | 88% | 99% |
| App B | Annual plan, 7-day trial, 10% buy without the trial | 10% | 100% | 100% | 100% |
| App C | Monthly plan, 3-day trial, 80% renew each month | 21% | 21% | 52% | 78% |
So a 30-day-old cohort of App C has paid only about a fifth of its first year, while App B's has paid nearly all of it.
Why day-7 ROAS targets mislead for subscriptions
A day-7 ROAS target assumes every app collects the same share of its year in the first week. It does not: with the same first-year return, the three apps above show day-7 ROAS from 0.16× to 0.34×.
| Fictional app | First-year ROAS | Day-7 ROAS | With a 0.25× day-7 target |
|---|---|---|---|
| App A, weekly | 1.60× | 0.24× | Cut |
| App B, annual with a trial | 1.60× | 0.16× | Cut |
| App C, monthly with a trial | 1.60× | 0.34× | Scale |
All three campaigns make the same money in a year, above break-even at either fee. A single target cuts two of them and scales one. Copy a target from another app, or from an app with a different plan mix, and the error is built in.
Early revenue is still useful. It can be a fair signal of where a cohort will land, but only measured against your own history, not against a fixed number. A day-7 target also breaks when you change a price, a trial length or the paywall, because the share collected in the first week changes with them.
A better rule: forecast the year, then compare
Turn early revenue into a forecast with your own cohorts, then compare the forecast with break-even. In a spreadsheet:
- Collect mature cohorts. For each weekly cohort at least a year old, note its revenue without tax by day 7 and by day 365. If your app is younger, use the oldest day you have, such as day 90, and say so.
- Compute each cohort's multiplier. Divide its day-365 revenue by its day-7 revenue.
- Take the median of recent cohorts. Use cohorts with your current prices and trial, and note the lowest and highest multipliers.
- Forecast new cohorts. Multiply a new cohort's day-7 ROAS by the median multiplier. The lowest and highest multipliers give a range.
- Compare with break-even. If even the low end clears break-even, the campaign is likely profitable. If the high end does not, it is not. Anything in between needs more time.
- Refresh after every change. A new price, trial or paywall changes the multiplier. Until cohorts with the new setup have aged, treat forecasts as rough and widen the range.
How INTARC does this
INTARC does this for each app from read-only keys, with no SDK. It applies each app's store fee and the taxes and levies you set, forecasts what each week's new customers will pay in a year with a range, and sets that against break-even after the store's cut. Every forecast is later scored against what really happened. You can look around the demo or create an account to run it on your own apps.
Frequently asked questions
What is a good ROAS for a subscription app?
Any ROAS above your break-even, measured over the period that matters to you, usually the first year. On revenue without tax, break-even is about 1.19× with a 15% store fee and 1% of other cuts, and about 1.45× at 30%.
Should I use a 15% or a 30% store fee?
Use the rate your new subscribers actually pay. On Google Play, subscriptions pay 15% in most markets. On the App Store, Small Business Program members pay 15%, and other developers usually pay 30% in a subscriber's first year and 15% after it. If you sell on both stores, weight the two rates by revenue.
Do taxes change break-even ROAS?
Sales tax and VAT do, if your revenue figure includes them, so take them out first. Income tax does not, because it is charged on profit and there is no profit at break-even.
Is day-7 ROAS useless for subscription apps?
No. Early revenue can be a fair signal when you know your own ratio between day 7 and day 365. What misleads is a fixed day-7 target copied from another app, or from an older version of your own.
How do refunds affect break-even?
Use revenue after refunds. RevenueCat's Revenue already deducts them, on the day each refund is processed. If you work from gross revenue before refunds, add your refund rate to the other cuts.
Store fees, tax rules and ad levies change often. We checked every linked source on November 9, 2026. Examples use fictional apps and numbers.