Retention is all you need: what drives local news readers to return?
Jan 28, 2026 | 09:00 AM Eastern Time (US and Canada) | Zoom
Speaker: Teresa Mondría Terol | A.I., The New York TimesModerator: Dr. Ana Moya | Data Scientist | INFOMOTION
How do you turn a first visit into a second within seven days? In this session, Teresa Mondria Terol walks through the end-to-end methodology behind her research at the Brown Institute on predicting return behavior for local newsrooms. The research foundations are raw GA4 data from BigQuery, session-level feature engineering, and interpretable tree-based models (Decision Trees, Random Forests, XGBoost with SHAP visualizations). We’ll cover the variables that matter most (article interaction, page views, engagement time), why device and referrer context shape retention (desktop and search outperform mobile and social), and how to translate model thresholds into concrete UX and editorial moves—like nudging a second page view and optimizing content discovery over chasing longer first sessions. Expect practical patterns, pitfalls in imbalanced audience datasets, and plug-and-play ideas for newsroom teams aiming to grow loyal audiences fast.