Data Modeling

This chapter explores how data modeling underpins growth engineering. It looks at why growth engineers must think in two systems: online transaction processing (OLTP) for capturing interactions cleanly and online analytical processing (OLAP) for analyzing them effectively. The chapter also explores the characteristics of each system, the types of data they handle, and the different modeling priorities they require. For OLTP modeling, it discusses how relational databases, entity-relationship diagrams, normalization, and constraints help maintain integrity and prevent silent data corruption in transactional systems. The chapter shows how modeling with these principles in mind supports reliable telemetry and experimentation infrastructure. For OLAP modeling, it focuses on speed and analytical clarity. The chapter introduces facts and dimensions as the foundation for analytical models, and examines two common design patterns–the star schema and the snowflake schema—along with their trade-offs and when to use each.