Mobile Game Retention: D1, D7 and D30 Explained, and How Ads Affect Them

Retention is the share of players who come back. It is the most important number in a free-to-play game because every other number depends on it: a player who leaves on day two cannot buy anything on day ten, watch an ad on day twenty or tell a friend on day thirty. It is also the number most often blamed on the game alone, when part of it is decided before the player ever opens the app — by the ad that persuaded them to install.

This guide explains how day-N retention is calculated, how to read it by cohort, what each checkpoint tells you about the game, and how acquisition creative shapes it.

What D1, D7 and D30 retention mean

Day-N retention is the percentage of players who installed on a given day and opened the game again exactly N days later. Install day is day 0.

• D1 retention: came back the day after installing. The verdict on the first session — the tutorial, the first minutes, and whether the game matched what the player expected.
• D7 retention: came back a week later. The verdict on the core loop — whether it stays satisfying once the novelty is gone.
• D30 retention: came back a month later. The verdict on the meta game — progression, collection, social features, events: reasons to keep playing for weeks.

How to calculate retention

An illustration with invented numbers: 2,000 players install on Monday. On Tuesday, 700 of them open the game, so D1 retention is 35%. The following Monday, 260 of the original 2,000 are active, so D7 retention is 13%. The denominator never changes — it is always the original cohort — so the numbers can only be compared within the same definition.

Classic versus rolling retention

Analytics tools define retention in two common ways, and mixing them produces nonsense comparisons.

• Classic (or day-N) retention counts a player only if they were active on day N itself.
• Rolling (or unbounded) retention counts a player if they were active on day N or any day after it. It is always higher than classic retention for the same cohort.

Also check what counts as a day — calendar days in one time zone, or 24-hour windows from the install time — and what counts as active, usually opening the app. Write the definition down and keep it fixed.

Cohort analysis: reading the retention table

A cohort table puts install days or weeks in rows and day N in columns, with retention in each cell. Reading it well is mostly about which direction you read.

• Read across a row to see one cohort's curve — how fast players leave, and where it flattens.
• Read down a column to see whether a given checkpoint is improving across cohorts — whether last week's update helped.
• Split rows by source: organic and paid, network, campaign, and where possible creative. Different sources often have very different curves.
• Split by platform and country, because device performance and market habits move retention on their own.

The shape matters as much as any single point. A curve that drops steeply and then flattens has found a core audience; one that keeps falling steadily has not, and late features will not save it.

Finding where players leave

Retention tells you that players left; funnel events tell you where. Log the steps of the first session — loading complete, tutorial steps, first level started and finished, first reward collected — and look at the share of each install cohort reaching each one.

1. A large drop before the first level starts usually means loading time, crashes or a confusing first screen.
2. A drop inside the tutorial means it is too long, too slow or teaches the wrong thing first.
3. A drop after the first few levels points at difficulty, pacing or a core loop that does not hold up.
4. A drop between D7 and D30 usually means the game runs out of goals: the meta layer is too thin.

How to improve retention

• Shorten time to fun. Put the player in the core loop in the first seconds and teach through play, not text.
• Tune early difficulty. The first levels should be winnable and feel earned; a hard wall before the player is invested loses them.
• Give a reason to return tomorrow: daily rewards, timed energy, a streak, an event, an unfinished goal.
• Build the meta: collections, upgrades, progression and social features carry D30.
• Respect the player's time with ads: a premature or too-frequent interstitial is one of the quickest ways to lower D1.
• Fix performance on low-end devices, where loading time and crashes quietly remove whole segments.

Common retention mistakes

• Reading blended retention only. Organic players usually behave differently from paid ones, and mixing them hides what the campaigns are doing.
• Comparing figures from tools that define a day or an active user differently.
• Judging a change on a cohort too small to be stable, or before the cohort has reached the checkpoint.
• Treating D1 as the whole story. A game can hold its first-day players and still lose them all by the end of the first week.

How ads affect retention

Retention is usually treated as a game-design problem, but the acquisition creative decides who is in the cohort in the first place. Every player arrives with an expectation set by the ad they saw, and the first session either confirms it or breaks it.

When the ad shows a mechanic, a theme or a level of spectacle the game does not have, the players it attracts open the app looking for something that is not there. Many leave during the first session, and D1 drops for reasons no tutorial change can fix. Comparing retention by creative, where your attribution allows it, often shows exactly this: two ads with similar install costs bringing players with very different curves.

The opposite also holds. An ad that honestly demonstrates the core loop selects for players who enjoy that loop. They arrive knowing what to do and wanting more of it, which helps D1, and because they chose the loop rather than a fantasy version of it, they are more likely to still be playing at D7. That is the practical case for showing the real game: it is a retention lever as well as a policy requirement, as misleading ad policy explains.

Where playable creative fits

A playable ad goes one step further than honest footage: the viewer actually plays the loop before installing. The ones who install have already tested whether they enjoy it, which is the closest an ad can get to a first session. Read playable cohorts on retention and ROAS rather than CPI alone — see ROAS and LTV for mobile games and how to lower CPI without buying worse users.

Playable Ads Maker builds playables from your real game's art and mechanic, in the browser, from prebuilt Game Templates or from scratch, and exports them for every major network. Keeping the demonstrated loop faithful to the game is a design choice the tool leaves with you — and it is the one that pays back in the cohort table.