What is a good D7 retention rate for mobile games?

Across 11,600 mobile games in GameAnalytics' 2025 benchmark report, the median D7 retention is 3.4–3.9% — and the top quartile of games reaches 7–8%. Install-weighted industry averages, which lean toward bigger polished titles, put D7 near 9%, with genre leaders like match games in the low teens. So the honest answer to "what's a good D7?" is: above ~4% you're beating half the market; above ~8% you're in the top quarter; low teens is genre-leader territory.

If those numbers look shockingly lower than the "20% D7 is good" advice you've read — that gap is the most useful thing on this page. Let's unpack it.

Why do published retention numbers disagree so much?

Because three different kinds of "benchmark" get quoted interchangeably:

1. Folk targets — the classic "D1 40% / D7 20% / D30 10%" rule of thumb. It describes chart-topping titles, and it predates the last few years of retention decline. Useful as an aspiration; useless as a health check for an indie release.
2. Install-weighted averages — like AppsFlyer's widely-cited figures (D1 29.5%, D7 8.7%, D30 3.2%). These average over installs, so games with millions of downloads — polished, UA-funded, live-ops-heavy — dominate the number.
3. Per-game medians — like GameAnalytics' benchmarks (median D7 3.4–3.9% across 11,600 games). Every game counts once, including the long tail of small titles. This is the honest comparison set for a small studio, because it's the population you're actually in.

Same industry, three very different bars. Pick the one that matches your question: "how do I compare to the median game?" → per-game median. "What do big successful titles achieve?" → install-weighted averages.

The numbers, by source

GameAnalytics — 2025 Mobile Gaming Benchmarks Report (11,600 games, 1.48B MAU, classic retention, published February 2025):

AppsFlyer — retention by genre (install-weighted averages, D7, Q3 2022 — older data, but still the most-cited genre breakdown):

Two structural patterns worth knowing on top of the raw numbers:

• Genre flips over time. Arcade and hyper-casual games win D1 but fall off hardest by D7; board, card, puzzle, and casino games — the "classic" genres — hold players longest. A strategy game with 8% D7 and a hyper-casual game with 8% D7 are in very different shape.
• Platform gaps are real. iOS retains meaningfully better than Android at every checkpoint (AppsFlyer Q3 2023: D7 of 12.6% vs 7.5%). If your install mix is Android-heavy, benchmark accordingly.

Why D7 is the number to watch

D1 measures your first impression. D7 measures whether a habit formed — a player who comes back a week after install has found a durable reason to play. That's why D7 correlates with long-term value far better than D1, why soft-launch go/no-go decisions usually key on it, and why UA campaigns evaluated on D1 alone routinely overvalue sources that deliver curious-but-uncommitted installs.

The decay between the checkpoints is as informative as the levels: healthy games typically keep roughly a third of their D1 players at D7. Retaining 30% D1 → 10% D7 is a normal curve; 30% D1 → 3% D7 means the first impression writes cheques the second session doesn't cash.

How do you actually improve D7?

The honest short version — each of these is measurable with the events you already have:

• Fix day one first. Most D7 loss is inherited from day-one churn — you can't retain at day 7 someone who never came back for day 1. We wrote a full guide on finding your first-session cliff: Why F2P games lose most players on day one.
• Give the first week a spine. Goals in flight, an unfolding unlock schedule, a streak worth keeping — the player should always close a session knowing what tomorrow's session is for.
• Find the walls. One overtuned level or one unclear mechanic in the first week shows up as a drop-off cliff in your funnel. Retries and time-on-step tell you whether it's difficulty or confusion.
• Watch the fork between returners and churners. Whatever behaviour separates week-one keepers from leavers in your game is your real retention lever — more precise than any genre benchmark. This is exactly the pattern churn prediction automates.

Frequently asked questions

Is 20% D7 realistic?

For most games, no — that target describes chart-toppers. Median D7 across 11,600 games is 3.4–3.9%; above 8% beats three-quarters of the market.

What's the difference between classic and rolling retention?

Classic D7 counts players who returned exactly on day 7; rolling counts day 7 or later, so it's always higher. Most published benchmarks (including GameAnalytics') are classic — measure yours the same way before comparing. Definitions for every metric on this page live in our game analytics glossary.

My D7 is 2% — is my game dead?

No, but you're below the cross-genre median, so retention deserves attention before UA spend does. Check genre context first (hyper-casual sits naturally lower), then hunt the specific moments players leave — it's usually a few fixable cliffs, not general disinterest.

Why does D7 matter more than D1?

D1 is a first impression; D7 is a habit. Habits predict long-term value.

Sources

• GameAnalytics 2025 Mobile Gaming Benchmarks Report — summary by GameDev Reports (11,600 games, 1.48B MAU, February 2025)
• AppsFlyer retention benchmarks, as compiled in Mistplay's benchmark roundup (genre data Q3 2022, platform data Q3 2023)

A note on our own numbers: PlayGenus computes genre benchmarks from cross-game telemetry as part of the product, and once our founding cohort is large enough to publish defensible aggregates, this page will carry our own medians — anonymised, and refreshed quarterly. Until then, we'd rather cite others' data than dress up a small sample as an industry number. If you want your game benchmarked properly in the meantime, the demo dashboard at app.playgenus.com/auth/demo shows how we do it, and we're onboarding a founding cohort of live studios for free — email us.