live · autopilotMomntLifecycleOct 2025 — present
Retention + discovery loop
300+ discovery interviews defined the ICP; per-session usage in Firestore plus automated win-back surveys turned a hardware + app product into 85% two-month retention.
Co-founder · Head of Product
FirebasePythonReactMeta
~/systems/momnt-retention
runningSession logsfirestore · per usesource
Inactivitypython · 7+ daystool
Auto-surveywin-back · whytool
Feature tests50-user cohortgate
Retention85% at 2 mooutput
- Session logsfirestore · per use
- Inactivitypython · 7+ days
- Auto-surveywin-back · why
- Feature tests50-user cohort
- Retention85% at 2 mo
tool agent gate / human record in flight
300+
discovery interviews
300+
units sold
85%
2-month retention
70%
6-month retention
Problem
Behavior-change products die on retention. We needed to know why people stopped, and we needed to know fast enough to fix it.
Architecture
- 01
Discover
300+ face-to-face interviews isolated weak physical enforcement and poor retention as the gaps competitors missed.
- 02
Instrument
Per-session usage logged to Firestore.
- 03
Detect + ask
Python scripts flagged users inactive 7+ days and automatically surveyed them.
- 04
Test
Fixes shipped to a 50-user subset first, then rolled out.
Design decisions
- →Ask churned users why, automatically, at the moment they churn, not in a quarterly survey.
Results
200-user campus pilot, 300+ units sold ($9K+ revenue), 85% two-month and 70% six-month retention; pitched 5 VCs and won $7K across two startup competitions.
Next system
Swoop Tutors: Outbound engine