What Is a Good ROAS for Mobile Apps? D7 & D30 Targets by Category (2026)

ROAS — return on ad spend — looks like the simplest metric in mobile growth: revenue divided by spend. The confusion starts when someone asks whether theirs is good. A 20% D7 ROAS is excellent for one app and a slow-motion disaster for another, because the answer depends on three things the metric itself does not contain: your margin after store fees, the shape of your revenue curve, and how long you can afford to wait for payback.

This guide gives working 2026 ranges by category as context, shows how to compute the only ROAS target that actually matters — your own break-even line — and names the failure modes that make "we are above benchmark" a dangerous sentence. For definitions, lean on primary sources like the AppsFlyer ROAS glossary and measure cohorts in an MMP view such as Adjust's cohort analysis, not only in ad-platform UI.

First, agree what you are measuring

Half of all ROAS confusion is definitional. Day-N ROAS is cohort revenue in the first N days divided by the spend that acquired the cohort. Beyond the window, three choices change the number materially:

1. 01

   Gross vs net

   — store fees (see Apple's Small Business Program and Google's Play service fees), refunds, and taxes you do not keep.

2. 02

   Platform-reported vs MMP/internal

   — SKAdNetwork windows and modelled conversions often flatter early numbers.

3. 03

   Purchase-only vs purchase + ad revenue

   — hybrid apps that omit ad revenue look artificially weak; hybrid apps that double-count look artificially strong.

Pick one definition, write it down, and never compare numbers across definitions. A "good" gross platform ROAS can be a bad net internal one. If two people in the Tuesday meeting cannot recite the same definition, you are not ready to debate targets.

Why "good ROAS" is the wrong first question

"Good" compared to what? Category peers? Last quarter? The ROAS that clears your payback deadline? Only the last one decides whether you can scale. Category ranges answer a different question: "is our market position sane?" Mixing those questions produces two expensive errors — celebrating cheap, low-quality ROAS that never pays back, and panicking at expensive ROAS that pays back fast because you selected better users.

Public hubs like Business of Apps benchmarks help you sense-check whether you are in a weird zip code. They do not set your bid. Your payback maths does. In Exp(G) operating experience, the fastest way to waste a quarter is to manage to a peer ROAS screenshot while your cash trough deepens.

Working D7 / D30 revenue ROAS ranges in 2026

With net revenue as the basis, these are Exp(G) operating ranges we see across accounts, used as orientation — not universal targets, and not a substitute for your break-even line:

1. 01

   Subscription (consumer, weekly/monthly plans)

   D7 often roughly 5–15%, D30 roughly 20–40%. Revenue builds slowly; the model lives or dies on renewals landing later.

2. 02

   Subscription with strong annual-plan mix

   D7 can sit much higher (often roughly 25–60%+), D30 roughly 50–100%+, because annual cash pulls revenue forward.

3. 03

   Hybrid IAP + ads

   D7 roughly 10–25%, D30 roughly 30–60%, with ad revenue stabilising the early curve.

4. 04

   Casual games

   D7 roughly 10–20%, D30 roughly 25–50%; hyper-casual often looks stronger on D7 because monetisation is front-loaded.

5. 05

   Mid-core / RPG

   D7 roughly 5–12%, D30 roughly 20–40%, with long whale-driven tails that make D90+ the honest judgment point.

If your number sits below these bands, diagnose before panicking: it is usually a CPI problem (see 2026 CPI benchmarks) or an early-monetisation problem, and the fixes are completely different. Cross-check CPI and retention context in the benchmarks tool.

The only target that matters: your break-even line

Category ranges are context, not strategy. Derive the target you manage to:

1. 01

   Start from your payback deadline

   — the month by which a cohort must cover its cost (payback period guide).

2. 02

   Work out the maximum CPI

   that deadline allows on your revenue curve — the Payback Engine solves this backwards.

3. 03

   Divide typical D7 (or D30) revenue per user by that CPI

   to get the ROAS your creatives and channels must clear — the kill line that turns weekly reviews from debate into arithmetic.

Done this way, "is my ROAS good?" stops being a comparison question and becomes a solvency question: above the line you can scale; below it you are renting growth.

Worked example: 15% D7 can be winning or losing

App X: max CPI for month-6 payback is $2.80; typical D7 net revenue per install is $0.42. Break-even D7 ROAS = 0.42 / 2.80 = 15%. Hitting 15% means the creative is on track. App Y: max CPI for month-6 payback is $1.90 with the same $0.42 D7 revenue. Break-even D7 ROAS = 22%. The same 15% D7 that looked "fine" on a blog benchmark table is underwater for App Y. Same headline metric, opposite decision — which is why we refuse to answer "is 15% good?" without the payback model open.

Run the projection yourself in the ROAS calculator: CPI, retention-shaped revenue, and fees in; D7/D30/D90 and break-even CPI out. If the calculator says you are solvent only on heroic renewal assumptions, believe the calculator, not the category meme. Also stress-test the line: what happens to break-even D7 if refunds rise two points, if annual attach rate falls, or if CPI in your top geo rises 20%? The teams that scale are not the ones with the prettiest current ROAS — they are the ones who know which assumptions the line is standing on.

Three honest ways to raise ROAS (and one dishonest one)

1. 01

   Fix the creative before the funnel

   Creative quality drives CPI more than most bidding changes. Score scripts with the hook analyzer before spending another cycle on media.

2. 02

   Pull revenue forward

   Annual-plan placement, onboarding paywalls, and starter offers move revenue earlier — same long-run LTV can produce dramatically better early ROAS and payback.

3. 03

   Cut losers faster

   Reallocating spend from below-kill-line creatives to winners raises blended ROAS without a single new asset.

4. 04

   The dishonest one

   switching the dashboard to platform-reported gross ROAS because it looks better. It also decides nothing, because it does not map to cash.

Creative craft still matters for the honest path. Platform creative guides are useful for making better ads — they are not ROAS targets. Use them to raise the numerator and lower CPI; use payback maths to decide whether the resulting ROAS is enough.

One more operator habit that pays for itself: keep a two-column weekly table — platform D7 ROAS vs MMP/net D7 ROAS — for each major channel. The gap is not a moral failing; it is a calibration input. When the gap widens after an iOS change or a new modelled conversion setup, fix measurement before you rewrite creative strategy. In Exp(G) operating experience, many "ROAS crises" are definition crises wearing a performance costume.

Failure modes and measurement traps

Comparing iOS SKAN ROAS to Android MMP ROAS as if they were the same species. Optimising to D1 purchase ROAS on a trial product where D1 is mostly credit-card holds. Scaling a channel that wins D7 and loses D90 because you never looked at retention-weighted LTV. Averaging brand search ROAS into cold social and declaring the programme healthy. And moving the goalposts after every bad week until "good" means "whatever we hit".

Also watch definitional drift after a monetisation change. Launch an annual plan and D7 ROAS jumps — correctly — but if you keep the old kill line without recomputing payback, you will either kill winners or scale trash. Rebuild the break-even line whenever pricing, trial length, or fee tier changes.

What to do Monday morning

Write your ROAS definition in one sentence and put it in the weekly doc so the argument cannot restart from scratch. Compute break-even D7 and D30 from payback, and show both next to the category ranges so newcomers understand context versus target. Kill below the line; iterate near it; scale above it. Review the line whenever pricing, trial length, fee tier, or retention assumptions change — stale break-even lines are how last quarter's winners become this quarter's silent losses. If you want a second set of eyes on whether your "good" ROAS is actually solvent, book a discovery call — finding a clean break-even line is usually a one-meeting exercise when the data is available.

Sources & further reading

• AppsFlyer — ROAS glossary
• Adjust — Cohort analysis
• Apple — SKAdNetwork
• Apple — App Store Small Business Program
• Google Play — service fees
• Business of Apps — app benchmarks
• Meta — Ads Guide
• TikTok — Creative best practices