Data Warehouses

Why would you want all your data in one place?

Otherwise, data is scattered and inconsistent, and you can't access it all at once easily.

However, not all data is the same, so to bring it together sometimes you need some work.

This is called ETL: Extract, Transform, and Load.

Data warehouses can be relational and multidimensional. We use the term OLAP cube to describe the structure.

Remember users want Graphs, Grids, and Reports. Mostly.

These things all have in common: they are measuring something (e.g. money), and they have dimensions (e.g. date).

So having an OLAP cube and a way to interact with it (e.g. Tableau) makes it simpler to access all our data and analyze it.