Continuing from Part 1.

So now we have our transactional model and a basic user story:

Our first request from the business for our data mart is that they want to be able to query all of the orders by date, by customer and/or by region (state, city or country). They want to be able to aggregate (sum and average) across those items.

The first thing we need to do is talk to the business and find out exactly what that request means. Do the users want to see information about daily orders in general?

As I mentioned a while back (a loooong while back), I have been thinking about writing up how I design data marts. The problem with that is that it is a huge topic. Even converting an existing schema (which doesn’t always exist) to a data mart (star schema style), still takes plenty of behind the scenes data analysis and prepwork. Still, I am not going to take a shot at it.

I could start with a laundry list of requirements but I don’t think that would be interesting to very many people.