Dimension Table

A dimension table stores the descriptive context of a business process — the who, what, where, when that measurements are sliced by: customer, product, store, date. Where a fact table is long and narrow, a dimension is wide and comparatively short: many descriptive attributes, far fewer rows.

Dimensions do the work in analytics: nearly every filter, group-by, and report label comes from a dimension attribute. Their two design questions are how to key them (usually a surrogate key) and how to handle attribute change over time — the slowly changing dimension problem. Shared across fact tables, they become conformed dimensions, the mechanism that makes metrics comparable across an organisation.

Go deeper: Fact Table vs Dimension Table: The Core Distinction