Player Retention

The Retention page shows a cohort curve: for every group of players who joined in the same week, what percentage came back on day 1, day 7, and day 30. This is the single most reliable signal for whether your server is healthy.

Cohort definition

Players are grouped into weekly cohorts by their first recorded session. A player who first joined during the week of 5 May 2026 belongs to the "May 5" cohort regardless of when you started tracking them.

A player is counted as retained in a period if they have at least one session that starts within that period.

Granularity options

The chart and table support four granularity settings:

Reading the chart and table

Each row in the table represents one weekly cohort. The first column shows the cohort size (total players who joined that week). Subsequent columns show the retention percentage at each interval: D1 means the player returned at least once during the first day after joining, D7 means within the first week, D30 within the first month.

The chart plots each cohort as a separate line so you can see whether recent cohorts retain better or worse than older ones. A cohort that curves sharply downward after D1 suggests your first-session experience needs work. A cohort that holds steady through D7 indicates players are finding reasons to come back.

What retention tells you about your server

A Minecraft server with strong retention is one where players have reasons to log in again. D1 retention is largely about first impressions: spawn design, onboarding flow, and whether there is something to do immediately. D7 retention reflects whether the core gameplay loop is compelling enough to pull players back across multiple sessions. D30 retention is a proxy for community strength: players who return after a month typically have social ties to other players, a base they care about, or long-term progression goals. If D1 is high but D7 is low, focus on the second and third sessions rather than the first.