Snapshot active-player counts hide ~40% decay that cohort curves expose 60 days early

Snapshot active-player counts hide ~40% decay that cohort curves expose 60 days early

Measurement-tool comparison: a snapshot 'active players' number versus a churn-cohort curve. They can tell opposite stories about the same account, and only one warns you in time.

The distortion:
— A snapshot (active players this month) stays flat or rises as long as new signups replace churned ones. It looks healthy right up until acquisition slows — then it falls off a cliff with no warning, because new blood was masking the bleed.
— A churn cohort (track one signup batch's survival month by month) exposes the real retention shape immediately. Across tracked betting cohorts, month-1-to-month-3 survival commonly drops ~40% — invisible in the snapshot, obvious in the cohort.

Why it changes decisions: snapshot metrics let you scale acquisition into a leaky bucket because the topline looks fine. Cohort curves show you the leak is structural ~60 days before the snapshot would, so you fix retention before you've over-spent filling the bucket.

When a snapshot suffices: stable, mature traffic where you only need a pulse check. When you need cohorts: any growth phase, any new GEO, any retention decision.

Benchmark of the day: snapshots hide ~40% month-1-to-3 decay until acquisition stalls; cohort curves expose it ~60 days earlier — never make a retention call off a snapshot.