Feature adoption breadth decline
Horizon: 50 days before churn. Number of distinct features used per user drops by 30%+. Signals a narrowing use case — the account is reducing dependence before deciding to cancel.
How to detect it
Track distinct feature events per user per 30-day window (PostHog / Mixpanel / product event stream). If the user's feature breadth drops below 70% of baseline, fire. Must be measured per-user then rolled up to account, not account-wide.
Recommended intervention playbook
Deploy a targeted feature-reintroduction campaign within 21 days. Identify which features stopped firing and why: (1) lost the team member who used them, (2) replaced by competitor, (3) broken by a recent release. Do not upsell; repair the core use case first.
Industries where this signal is strongest
- Dev Tools
- BI/Data
- Martech
- CRM/Sales
- Health Tech
Other universal churn signals
- Login frequency decline (horizon: 45d)
- Champion / sponsor departure (horizon: 75d)
- Payment failures or delays (horizon: 10d)
- Support ticket volume spike (horizon: 21d)
Track this signal with ChurnDefense
ChurnDefense ingests your product + CRM + billing data and fires this signal automatically when the detection thresholds trip. No manual dashboard auditing. Every alert is explainable — you see which threshold fired, when, and which account.