Case Study: Mediapro improves TV audience share prediction with WhiteBox

MediaPro improves TV audience prediction with WhiteBox using Darts and gradient boosting

Mediapro, a leader in television production in Spain, faced the challenge of accurately predicting audience shares for its programs and those of its competitors, a key factor in determining advertising costs. To lead in technological innovation, the company collaborated with vendor WhiteBox to improve its audience prediction methodology.

WhiteBox implemented a solution using the Darts open-source library and a robust gradient boosting algorithm, moving beyond conventional time series analysis for a more efficient and scalable model. This provided Mediapro with a powerful and easy-to-use system for improving its audience predictions during both training and production use.

MediaPro