Modeling for analysis
Dimensional modeling is a design method that shapes data around how the business asks questions, rather than how the source systems store it. It makes analytics intuitive and fast.
The two building blocks
- Facts are the measurable events, like an order with quantity and revenue. They live in a fact table at a defined grain, the level of detail of one row.
- Dimensions are the descriptive context, like who, what, when, and where, that you filter and group by.
The design steps
A common approach follows four steps:
- Pick the business process to model, such as sales.
- Declare the grain, for example one row per order line.
- Identify the dimensions that describe each event.
- Identify the facts, the numeric measures.
Why grain matters
Choosing a clear, consistent grain prevents double counting and lets dimensions attach cleanly. Mixing grains in one table is a frequent source of wrong totals.
Key idea
Dimensional modeling organizes data into facts and dimensions at a clear grain so business questions are easy and fast to answer.