You are analyzing customer satisfaction for a subscription service and want to separate real drivers from variables that only move with satisfaction. You have customer-level data on satisfaction, service interactions, outages, billing issues, tenure, and price changes.
How would you use regression to understand what drives customer satisfaction?