How to win at monetizing games as a service
summary
Practical roadmap for live-game monetization: unify telemetry with marketing and CRM into BigQuery via Pub/Sub and Dataflow, govern it with Dataplex and BigLake, share via Analytics Hub, then run BQML or Vertex AI models. Specifics: churn model to catch 25% day‑1 loss, predictive LTV to find the ~20% of players that drive ~80% of revenue, combine churn×LTV as “expected value at risk,” and use Vertex AI Pipelines to automate retrain/production (87% of models otherwise never ship).
AI-written summary · about
chapters (17)
- 0:00Introduction and goals
- 1:02Data-driven culture
- 3:23Unifying and governing data
- 6:25Data governance strategy
- 13:02BigQuery architecture
- 17:38Operational telemetry
- 20:39Dataplex and BigLake
- 23:27Analytics Hub and use cases
- 27:52Enhancing monetization with ML
- 30:13Churn and LTV modeling
- 33:15Model activation scenarios
- 36:43Personalization and recommendations
- 40:59Predictive acquisition
- 43:43Vertex AI for ML operations
- 50:44ML pipelines and monitoring
- 52:23Quantifying business impact
- 54:11Q&A and closing
related talks
23mCustom-made games with machine learning and big data
20mUsers acquired - Now what?: Maximising revenue from existing players
27mBig Data in Mobile Games
23mThe unreasonable effectiveness of data for better monetisation
20mBuilding Intelligent Retention Engines: Real-Time Churn Detection
55mHow to Improve Your Game's Retention & Monetization Through Data-Driven Experimentation
32mMaking Models Work: How an optimized ML lifecycle can drive better ROI
18mSession: The Science Of Success – Driving Growth With Prediction, Incrementality, And Scalable....