Retention
Playerbase → Player Retention (/p/<projectId>/playerbase/retention)
Why this page exists
Retention answers: After someone joins for the first time, do they come back the next day? A week later? Use it to judge onboarding, first-session quality, and whether changes (updates, difficulty, ads) help or hurt day-1 and day-7 return rates.
How to use it
Set a date range that defines who counts as “new”: only players whose first join falls in that window enter the cohort.
Choose Combined (all servers) to see the whole network, or one server to measure a single world or hub (useful after a reset or a new season).
Pick a timeframe (first 7 / 14 / 21 / 30 days after first join). Longer windows show more of the long tail; shorter ones react faster after a change.
Read Cohort size so you know if the numbers are statistically meaningful (very small cohorts bounce around by chance).
Follow the retention curve in the chart: a steep drop after day 1 usually means the first session did not hook people; a flat tail means you keep a stable core.
Use Still active at day N and average retention as headline KPIs when comparing before vs after a release or campaign.
What to do with the insights
Low day-1 retention: improve first-time experience, loading, or clarity of goals; check if the wrong audience is joining.
Good early days but bad later: add reasons to return (events, progression, social).
Different servers, different curves: compare Combined vs single server to see if one game mode underperforms.
If the page looks empty
You need new players in the date range and session data. Widen the range, pick Combined, or confirm PlayerAtlas is running on the servers you care about.
Last updated