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

# Retention

> See how many members keep coming back to your community week over week.

Retention measures whether your active members return. You can have strong growth numbers, but if no one comes back after their first day, you're not building a community — you're building a leaky bucket.

<Card title="Open in Apetopia →" href="https://apetopia.co/dashboard/analytics/retention">
  Go directly to this feature in your dashboard.
</Card>

<Note>
  Retention data requires at least two weeks of message tracking by the [Discord Bot](https://www.apetopia.co/dashboard/discord) to be meaningful.
</Note>

## Stat cards

| Card                    | What it shows                                                |
| ----------------------- | ------------------------------------------------------------ |
| **Retention Rate (D7)** | % of users active last week who also sent messages this week |
| **Churn Rate**          | 100% minus the retention rate — members who did not return   |
| **Returning Users**     | Estimated count of retained users out of total active users  |

## How Retention is Calculated

The retention rate is a **7-day rolling metric**: it compares the set of users who were active in the previous 7-day window to those who were active in the current 7-day window. Users who appear in both windows are counted as retained.

* **Above 50%** — healthy. More than half your active members return each week.
* **40–50%** — average. There's room to improve engagement and habit-forming.
* **Below 40%** — worth investigating. Members may be checking in once and not finding a reason to return.

## Retention Breakdown

The bottom section shows two progress bars — Retention Rate and Churn Rate — and a count of retained vs. churned users for the period.

## Period selector

The period selector (**7d / 30d / 90d**) controls the data window fed into the retention calculation. Periods with insufficient data are automatically disabled.

<Tip>
  If churn is high, the most common fixes are: a consistent posting schedule so members have a reason to return, and pinned or highlighted content that makes the server feel alive even on quiet days.
</Tip>