Data MODELING MASTERY

Data Modeling Mastery: Techniques, Estimation & T-Shirt Sizing Best Practices

How do you estimate a project without pretending to know the future? And how do you design data systems that stay fast, consistent, and maintainable as they grow?

Data Modeling Mastery answers both questions with a practical, no-fluff approach for data engineers, analysts, and technical leads.

Part estimation guide, part architecture handbook, this book teaches you to think clearly about two of the hardest problems in data work: predicting effort and structuring information. You'll learn why humans are bad at hour-based estimates but surprisingly good at relative comparison—and how T-shirt sizing turns that insight into a fast, honest planning process. You'll also explore Planning Poker, Affinity Estimation, the Bucket System, and Dot Voting, so you can pick the right technique for any team or backlog.

On the modeling side, the book walks through dimensional (star schema), normalized, denormalized, and nested data models—with real SQL examples optimized for BigQuery and Google Cloud Platform. A complete e-commerce warehouse use case ties it all together, from requirements analysis to fact and dimension table design to ETL pipelines, performance tuning, and cost control. You'll also go beyond the warehouse into OLTP systems, NoSQL document databases, data lakes, and real-time streaming.

Whether you're building your first pipeline or an enterprise analytics platform, this guide gives you the frameworks, the trade-offs, and the confidence to choose well—and the judgment to know when to break the rules.

What you'll learn:

• The S/M/L/XL estimation framework and how to run consensus-driven sizing meetings
• Five estimation techniques compared across team size, speed, and accuracy
• When to use star schemas, normalization, denormalization, or nested fields
• A full e-commerce warehouse build: facts, dimensions, ETL, materialized views, partitioning, and clustering
• Decision frameworks for warehouses, OLTP, NoSQL, data lakes, and real-time analytics