Cohort dashboards

Cohort dashboards show how groups of users behave over time. Trend analysis, not snapshots.

Key views

• Retention by cohort
• Revenue per user by cohort
• Feature adoption by cohort
• LTV cumulative by cohort

What to look for

• Improvement cohort-over-cohort
• Anomalies (sudden drop)
• Segment differences

What to do with this

• Track cohort retention, not aggregate MAU, the aggregate number hides declining cohort quality for 6-12 months
• Compare cohort curves across months, if newer cohorts retain worse, investigate product or acquisition channel drift
• Segment cohorts by acquisition source, organic cohorts often retain 2-3x better than paid, budget decisions follow
• Review cohort analysis weekly, not quarterly, cohort drift is a leading indicator that becomes terminal if ignored
• Build cohort charts into your standing dashboard, if cohort analysis requires a SQL query nobody runs, nobody will see the drift