New Subscriber Predictor
Predict churn probability for newly registered users using all 18 variables sorted by impact
Critical & High Priority Inputs
These parameters have the largest mathematical influence on predictions. Adjusting these will yield the largest shifts in churn risk.
The single largest predictor of customer loyalty; churn drops exponentially as tenure grows.
Average session duration in minutes. Longer sessions correlate with deep content attachment.
Weekly hours. Low engagement levels represent a major cancellation risk warning.
Offline downloads indicate a committed watchlist use and high content consumption.
Higher price points raise customer billing sensitivity and drop loyalty.
Premium subscription packages show statistically lower churn.
Filing multiple complaints indicates active product bugs and user friction.
Manual checks increase monthly payment failures and cancellation triggers.
Medium Priority Inputs
Secondary behavioral preferences that contribute to the customer's risk profile.
Users who utilize subtitles display slightly higher retention rates.
Accounts with parental controls enabled have a slightly lower churn risk.
Lower Priority Inputs
Variables that have a minimal or marginal effect on predictions, but help polish model precision.