Communication service providers (CSPs) operate in highly competitive, saturated markets. Subscriber growth relies on two factors: the retention of existing subscribers (subs) and the capture of new subs from competitors. It costs less for CSPs to retain their existing subs than to acquire new ones. And because churn—the number of customers who stop using the provider—is a problem across the industry, every CSP has a methodology to try to address it. CSPs seek a solution that helps them confidently identify the key drivers behind churn and subs at high risk to churn. Many CSPs are turning to machine learning to build churn prediction models, which identify high-risk subs and use subs’ experiences and profiles to personalize offerings in an attempt to retain them.
Prescriptive architectural diagrams, sample code, and technical content