Smarter Segmentation with Audiences and Predictions
summary
Shows how Firebase combines dynamic audiences (new dynamic user‑property evaluation, exclude groups, time‑windows, 30‑day membership) with Firebase Predictions to auto‑segment users and trigger actions. Example builds an audience of “frustrated novices” (20 level starts without a level end, >10 sessions in 3 days, level == 1) then pushes a Remote Config difficulty change and a long‑running daily push; Predictions reports ~90% accuracy on a spend model and supports A/B tests and targeted ad toggles.
AI-written summary · about
chapters (5)
related talks
13mCreate better user experiences with Remote Config for advanced app personalization
11mPersonalizing your app in real-time: Powered by Firebase and ML
15mAdding on-device recommendations to your app using TensorFlow and Firebase
19mHow to Leverage Predictive Analytics to Enhance iOS Marketing Performance
21mDemystifying Machine Learning for Programmatic UA & Retargeting
19mSession: Re-Engagement In 2026: How To Design & Scale Winning Retargeting Campaigns
28mAutomating SEA Retargeting for Smarter Audience Engagement and Higher Conversions
23mThe unreasonable effectiveness of data for better monetisation