A leading regional telecommunications provider uses tetrel’s churn prediction to analyze and prevent customer churn.
Customer churn has a significant impact on overall profitability for the telco due to high acquisition costs. Industry benchmarks show a higher churn rate than comparable competitors.
Since broadband is a low-touch product, it is important to avoid waking customers who would otherwise not have churned. Before defining retention measures, root causes of churn must be clearly understood. Broad marketing initiatives to reduce churn had mixed success in the past.
Understanding root causes for churn at an individual customer level was paramount for the telco to succeed. Targeted retention measures had to be designed, implemented and tested against each other to identify the most effective way to increase customer lifetime value.
Root causes for churn were identified from historical data using causal inference modeling. Predictive models in the customer intelligence platform were trained to predict the churn propensity and the most probable causes of churn individually for each customer.
Based on these results, the telco and tetrel defined a suite of new retention measures to specifically target at-risk customers. The effectiveness of these measures was validated using causal inference and A/B tests.
By combining churn risk, root causes and validated retention measures, the churn prediction regularly suggests the best course of action for each customer. By weighing the probable impact of retention measures against their cost, customer lifetime value is maximized on an individual customer level.