← back to writing

Churn models that explain themselves

A churn score is useless if your CS team doesn’t know what to do with it. We use SHAP values to give per-user explanations.

The problem with black-box churn

Most churn models output a probability. “User 4821 has a 73% chance of churning.” So what? Your CS team needs to know why — so they can act.

SHAP explanations per user

We build every churn model with SHAP (SHapley Additive exPlanations) baked in. Each prediction comes with the top 3 factors driving that user’s risk score. Your CS team gets actionable intelligence, not just a number.