> ## Documentation Index
> Fetch the complete documentation index at: https://docs.manalens.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# How to measure retention

> Build a cohort retention chart with D1, D7 and D30 or weekly retention, compare cohorts after a release, and find the people at risk of churn.

Retention shows how many people come back. This guide builds a retention chart, reads it, and finds the people who stop coming back.

## Part 1: Build a retention chart

**1. Add a retention chart**

Open **Dashboards**, open a dashboard and click **Add chart**. Pick the **User retention** example, or click **Blank chart** and set the **Chart type** to **Retention**.

**2. Choose the entry event**

The entry event defines the cohort. Keep **Any event**, or choose an event such as `signup_completed` or `game_open`.

**3. Choose the returning event**

A user is retained in a period when they do this event. For a game, `session_start` is a good choice.

**4. Choose the type and period**

**Rolling** (default) counts a user who returns in that period or later. **Recurring** counts only a return in exactly that period. Choose **Weekly** (weeks 0 to 12) or **Daily** (days 0 to 30, with D1, D7 and D30).

**5. Save**

Click **Save Chart**.

## Part 2: Read the chart

* **Read a row** to follow one cohort over time.
* **Read a column** to see where people stop. A large drop from period 0 to period 1 often points to onboarding.
* **Compare rows** before and after a release. Newer rows that stay higher mean that the change helped.
* **Look for a flat tail.** When the curve flattens, you have found the people who stay.

Period 0 is always 100%. Periods that are not complete show a dash, not a low number.

> [!TIP]
>
> For a game, use the **App** filter group to compare builds or storefronts, for example **Build is 571**.

## Part 3: Compare groups

* Choose a saved [segment](https://docs.manalens.dev/guides/segmentation.md) in **Users** to count only those people.
* Add a dashboard filter, such as **Channel is Paid Search**, to compare acquisition channels.

## Part 4: Find people at risk

Open **Users** and click **Audience insights**. **User Lifecycle** counts the people who are New, Returning, At Risk, Churned or Resurrected. See [audience insights](https://docs.manalens.dev/features/audience-insights.md).

## Related

* [Retention](https://docs.manalens.dev/features/retention.md): how people and periods are counted.
* [How to track DAU, WAU and MAU](https://docs.manalens.dev/guides/dau-tracking.md)
* [Metrics](https://docs.manalens.dev/data/metrics.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.