DAU
—
—distinct users, trailing 24h
WAU
—
rolling seven days
MAU
—
calendar month (UTC)
Stickiness (DAU/MAU)
—
daily actives ÷ monthly actives
Median Session (min)
—
median, trailing week
Sessions / User / Week
—
trailing week
Feature Usage & D30 Retained Lift
first-party events| Feature | Usage Rate | D30 Lift (pp) | Signal |
|---|---|---|---|
| — | |||
Usage = distinct actives using the feature ÷ all distinct actives, trailing month. Lift = D30 retention of first-week feature users minus non-users, from first-party events alone. Pearson correlation heatmap (G7-4) deferred.
Account Health Distribution
usage + payment
Healthy (70+)
—
accounts
At Risk (40–69)
—
accounts
Critical (<40)
—
accounts
Avg health score: — / 100 · — active accounts scored
Score = usage (active days, trailing month) + payment status, equally weighted. NPS and support components are absent from our data — honestly excluded, not faked (G7-5).
At-Risk Accounts (score <40)
— accounts
— MRR at risk
Aggregate only — the named at-risk account list (G7-6) is deferred pending the row-level-lists decision (D-1).
Power Users — Top 10% by Engagement
aggregate bandPower users (top decile)
—
Tracked users
—
Avg sessions/week (power band)
—
Composite score: sessions/week ×0.4 + features used ×0.4 + tenure ×0.2 (G7-7). The PRD's named power-user list ships only if row-level lists are approved (D-1) — aggregate bands only, no user identifiers.