Retail Sales Snowflake Schema

Retail sales snowflake schema

This snowflake schema records retail sales in FactSale and keeps the descriptive dimensions in related tables. Each sale identifies a date, product and store, then records quantity and revenue. Product rolls up through subcategory to category. Store rolls up through city to region. Those normalized chains are the difference between a snowflake schema and a simpler star schema: the fact table remains central, while repeated hierarchy attributes move into their own tables. Use this pattern when reporting needs consistent category and geography rollups across several facts or when those hierarchies change independently. It also gives data stewards a distinct place to maintain the product and location hierarchies used in reports.

類型 ERD標準 Chen 1976 / Everest 1976 (crow's foot)引擎 schematex-erd更新時間 2026/9/24

需求

“A retail sales snowflake schema: a FactSale table linked to Date, Product, and Store dimensions; normalize Product through Subcategory to Category and Store through City to Region.”

接著試試Add a customer dimension to FactSaleAdd a brand table below Product

這張圖裡有什麼

看懂背後的決策。

01Fact grain

Store one row for each sale line.

02Product hierarchy

Keep category data outside the product dimension.

03Geography hierarchy

Normalize stores through city and region.

Use this when product and store hierarchies are shared and need separate maintenance.

瀏覽所有 ERD 範本 →

常見問題

What makes this a snowflake schema?

Its dimensions are normalized into hierarchy tables, such as Product to Subcategory to Category.

What is the fact table grain?

Each FactSale row represents one recorded sale line with its measures.

Why normalize store geography?

City and region can be maintained once and reused by every store.