All posts
postJune 5, 2026

Star schema vs snowflake schema?

#data-modeling#star-schema#warehouse
Star schema vs snowflake schema — which should I use?

Star for almost everything. One fact table, denormalized dimension tables — simple joins, fast queries. Snowflake schema normalizes dimensions further to save storage, but storage is cheap now and the extra joins slow queries down. Default to star unless you have a specific reason not to.

A star schema has one fact table (transactions, events) surrounded by denormalized dimension tables (users, products, dates). Each dimension is a single table — names, categories, attributes all in one place. Joins are simple and queries are fast.

A snowflake schema normalizes those dimensions into multiple tables. Instead of one product table, you have product, category, and subcategory tables linked by keys. This saves storage but requires more joins. Most modern warehouses prefer star — storage is cheap and query simplicity wins.