What your D1, D7, and D30 retention numbers are telling you

Retention is the metric studios check most often and understand least precisely. D1, D7, and D30 appear on every dashboard, show up in investor updates, and generate internal debates about whether a game is healthy or not. What they rarely generate is a specific action.

Part of the problem is that the three numbers are often reported together, as if they're measuring the same thing. They're not. Each milestone captures a different player behavior and points at a different problem when it's low. Improving D1 requires different work than improving D30, and fixing the wrong one based on misreading the number is a reliable way to spend a quarter on something that doesn't move your actual situation.

What D1 is actually measuring

D1 retention — the share of players who return the day after install — is measuring whether your first session gave players a reason to come back.

A low D1 number is almost always a first-session problem, not a second-session problem. Players who don't return on day one were not engaged enough during their initial session to form a habit or feel unfinished business with the game. They left, their phone showed them something else, and the game never crossed their mind again.

The diagnosis work for low D1 is focused entirely on the install → first session → end of first session sequence. Where do new players drop off in the first session? What does the engagement curve look like within that first play — how many players quit before reaching any meaningful progression? What is the game showing players at the moment they're most likely to close it?

D1 is also sensitive to traffic source. Players who came in from a paid campaign targeting the wrong audience will show low D1 even if the core game is solid. Comparing D1 across acquisition channels — organic versus paid, different creative types, different platform sources — often reveals that the install base is heterogeneous in ways the aggregate number hides.

What D7 is measuring

D7 is measuring whether the game established a habit during the first week. By day seven, players have seen enough to form an opinion. They know whether they're the kind of person who plays this game, whether it fits into their routine, and whether the game rewards continued engagement.

Low D7 relative to D1 — a steep drop between the two numbers — indicates the first week experience is failing. Players who came back after day one lost the thread somewhere between days two and six. The most common causes are content depletion (players burned through the first-session content and had nothing new to discover), unclear progression (no obvious reason to open the game again), or a poor session-restart experience (the game didn't make it easy to pick up where they left off).

D7 is where daily engagement mechanics, streak systems, and content gating become visible in the data. If a game has daily quests and D7 is low, the quests aren't working as designed. If the game uses episode-style content and D7 drops sharply after the first episode, the second episode isn't creating sufficient pull.

What D30 is measuring

D30 is measuring whether the game became part of a player's recurring routine. A player retained at day 30 has crossed from "trying this game" to "playing this game." The number is smaller than D7 by design — most players settle into their long-term habits well before day 30 — but it's the number that determines lifetime value and the composition of a game's spender base.

The players retained at D30 are disproportionately likely to be payers. The relationship between D30 retention and total game revenue is tighter than the relationship between DAU and revenue, because D30 tracks engagement depth, not just volume.

Low D30 relative to D7 can have several causes: content that runs out at a certain depth (players reach the end of available content and disengage), progression walls that require grinding without reward (players plateau and lose motivation before spending), or a missing social layer (for games where community is part of the long-term value, players without connections disengage faster).

Why genre benchmarks mislead more than they help

Studios frequently compare their D1/D7/D30 numbers against published genre benchmarks. The instinct makes sense — you want to know whether you're in a normal range. The problem is that published benchmarks aggregate across player populations with different install sources, different game quality distributions, and different measurement conventions. They're interesting context, not actionable targets.

Your own cohort data is more useful than any benchmark. Is D1 improving month over month after a first-session redesign? Is D7 increasing after adding a daily streak mechanic? Are D30 numbers better for players acquired through organic versus paid channels? The trend within your own data, tied to specific changes you made, is where you get something you can act on.

Benchmarks also vary significantly by subgenre, monetization model, and player geography. A game with heavy social features in Southeast Asia and a mid-core puzzle game with a Western player base will show very different retention curves even if both are performing well for their category. Chasing a number that came from the wrong game type creates the wrong priorities.

The shape of the curve matters more than any single number

When D1, D7, and D30 are read as a set, the curve shape tells you where the biggest loss is happening.

A high D1 paired with a disproportionately low D7 says the first session is working but the first week isn't. A healthy D7 paired with a very low D30 says week one is fine but the game isn't creating long-term habits. A low D1 regardless of what follows tells you the install-to-first-session transition is broken and everything downstream is compromised by the funnel shape.

The ratio between consecutive milestones is often more diagnostic than the absolute numbers. A game that drops 60% from D1 to D7 and then holds reasonably well from D7 to D30 has a different problem than a game that holds from D1 to D7 and then loses most players before day 30. Both can look similar in aggregate DAU reports and yet require completely different interventions.

What to do when the number is low

Each milestone has a different fix set.

Low D1 is worked on in the onboarding and first-session design — tutorial length, speed to the core loop, clarity of the value proposition, the hook that sends players away wanting to return. Monetization has minimal impact here; the player hasn't engaged deeply enough for purchase timing to matter.

Low D7 is worked on through daily engagement design — what draws players back on day two, day four, day six. Streak mechanics, daily missions, social pressure from friends, content drip, and push notification strategy all operate in this window.

Low D30 is worked on through progression depth, content volume, social systems, and the relationship between free and paid progression. This is also the window where offer design starts to matter — players retained at D30 are the most likely buyers, and the offers they see in weeks two through four shape their long-term spending behavior.

See how we track player retention and identify where engagement breaks →

None of these are quick fixes. But knowing which milestone is the problem tells you which part of the game to invest in — which is more than most studios get from the number alone.

Talk to us about diagnosing your retention numbers →