Case Study: Vodafone improves churn prediction and retention targeting with WhiteBox's causal inference model

Vodafone evaluates churn risk for 5 million monthly customers with WhiteBox

Vodafone faced the challenge of customer churn in Spain's competitive telecommunications market. They needed to go beyond traditional prediction models to not only identify which customers were likely to leave but also to deeply understand the underlying reasons for their departure. The vendor WhiteBox implemented a solution using a causal inference framework built with tools like Causal ML and CausalNex.

WhiteBox's solution provided Vodafone with a model that predicts churn probability and deciphers the causes behind it, while also forecasting how specific retention actions would influence customer decisions. This allowed for a more effective and targeted application of retention strategies. The measurable impact included the development of models evaluating over 5,000 customer characteristics to assess the churn risk for more than 5 million customers each month.

Vodafone