ChurnClock

ChurnClock is a web app (with optional Slack alerts) that forecasts customer churn and revenue risk for subscription businesses using time-based behavioral signals. It connects to Stripe, Chargebee, HubSpot, and product analytics (Segment/Amplitude) to build a simple “next 30/60/90 days” churn and MRR forecast, plus trend explanations you can actually act on. Instead of generic ML dashboards, it focuses on temporal patterns: declining usage velocity, billing friction sequences, support-ticket timing, and cohort aging effects. The MVP outputs a weekly forecast, confidence bands, and a prioritized list of accounts likely to churn with the top 2–3 leading indicators per account. It’s not magic: forecasts will be wrong sometimes, especially with low data volume, but it gives teams a measurable, repeatable way to spot churn earlier than lagging metrics like logo churn reports.