Kahan Data Solutions on YouTube · Open on YouTube
Professional practice: how analytics teams actually work · Competent · Lesson 33 of 99 · 17 min video, then about 5 hours of practice
Dimensional modelling: facts, dimensions and grain
After this you can design a simple star schema, state the grain of a fact table, and explain why analysts query star schemas rather than raw application tables.
Why it matters
Almost every company warehouse and BI model is built on Kimball-style dimensional modelling, and Power BI in particular is designed around star schemas. Learn the vocabulary (fact, dimension, grain, conformed dimension, slowly changing dimension) and practise declaring the grain first: 'one row per order line'. The trap is mixing grains in one table, which makes every aggregate wrong in a slightly different way.
How to use this resource
A practical star-schema tutorial that builds fact and dimension tables from a normalised source.