/case-studies/the-times
The Times: Programmatic Experimentation
How we used statistical modeling and AI-driven personalization to increase subscription retention by 42%.
42%
Retention Lift
12M+
Data Points
6
Months
The Challenge
The Times needed a way to predict user churn before it happened. Traditional heuristics were failing to capture the complex, non-linear patterns of user engagement degradation.
The Mathematical Solution
We deployed a machine learning model using a random forest classifier to identify behavioral signals indicating imminent churn. By running multi-armed bandit A/B tests on personalized intervention strategies, we optimized the save-rate dynamically.