Skip to content

Latest commit

 

History

3 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 

Repository files navigation

SMB Churn Detection Framework

A Product Manager portfolio project demonstrating how partner retention can be reframed from partner-level to SMB-level early detection.

The Core Insight

Partner churn is a downstream symptom of SMB disengagement. By the time a partner churns, the real signal happened 30-60 days earlier at the SMB level.

The 5 Agentic Events

# Event Trigger Benchmark
1 First Activation Signed up, zero work units 98% who never activate churn in 2 weeks (Amplitude 2025)
2 Gone Quiet Zero activity for 7 consecutive days Strongest predictor. Cancellation within 30 days.
3 Portfolio Milestone 100 units/week threshold Retention roughly doubles past this point
4 Upsell Signal 30 units/week for 2+ weeks Top SaaS: NRR above 120%
5 Health Score Active/total SMB ratio 34% more at-risk found, 41% fewer false positives

Key Numbers

  • 256 partners analysed
  • 54.5% under-monetized (SMB revenue below subscription cost)
  • 31% churn reduction benchmark (AI prediction vs control)
  • 5 leading indicators designed into the detection model

Northstar Validation — Platform AI Penetration

  • Platform-wide AI penetration: 4.5% (7,586 of 169,288 paying SMBs have AI deployed)
  • Kaseya proof case: 103 AI deployed, 100% Gross Revenue Retention, zero churn — AI-deployed SMBs do not churn
  • Neighborly benchmark: 3,011 AI deployed across 3,975 paying SMBs (75.7% AI penetration) — highest penetration in the partner portfolio
  • The retention signal: Partners with high AI penetration show materially lower churn. Kaseya's 100% GRR across 103 AI-deployed SMBs validates the framework's core thesis — AI engagement is the strongest churn prevention lever

What the PM Did

  1. Analysis -- Analysed 256 partner portfolios to identify the two-level churn chain
  2. Framework Design -- Designed 5-event detection model with market-grounded thresholds
  3. AI Configuration -- Wrote 150+ line system prompt for the AI data analyst agent
  4. Prototype -- Built and deployed a live Streamlit demo
  5. Stakeholder Pivot -- Moved from standalone dashboard to platform-native contribution model
  6. Market Validation -- Benchmarked against Amplitude 2025, Recurly 2025, and cohort studies

Live Demo

  • Portfolio page: Hosted via GitHub Pages from the docs/ folder
  • AI Data Analyst: Streamlit App

Related Projects

Author

Sangeetha K -- Product Manager

About

Churn Detection — 5-event framework, 256 partners analysed. Platform: 4.5% AI penetration (7,586/169,288). Kaseya 100% GRR proof, Neighborly 75.7% benchmark.

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors