Stuff Warehousing (step #1: Detecting the need)

My job in the Business Intelligence field often involves dealing with data warehouses, which are big databases organized to suit reporting and data analysis needs, it is an organised place to store data.
Just like a closet is an organized place to store stuff in the house. I would like to stress the parallel between building a data warehouse (which I know how to do) and building a closet (which I will soon know about as well). I actually need to build 2 of them for my new apartment.

First of all you detect there is a need for something:

  • We have all our data spread out there in excel spreadsheets, we have to do something about it.
  • we have all our stuff spread out there at our parent’s home and all over the apartment, we have to do something about it.

Then you go see your users and set up the general requirements:

  • I want all my data in there, available when I want and easily so I can still use my excel spreadsheets to do secret calculations that I do not want the others to know about.
  • I want all the stuff that wanders in the house in there. And I don’t want to use a stool. And I want to pick my dress without having to bend or to stand on my tiptoes.

Meanwhile you need to perform technological watch:

  • What skills do I need, what are the different approaches to building this thing?
  • What amount of effort, money and time do I have to put in it?
  • What are the solutions available on the market? What are their advantages, disadvantages according to my needs? What is their price? What’s their compared ROI? (nooo, come on! nobody ever does that in data warehousing :-) )

Then you have to start confronting the needs with reality, you have to show your users what is possible, you have to make them talk and drive them to a solution that you know feasible and affordable. The key here is to keep them involved by making the product real, by taking their needs into consideration:

  • ok the rod will by 1.4 m high but you’ll have some lost space beneath.
  • ok you’ll have your data daily but we can’t keep 10 years of daily data.
  • how about we have a geographical dimension to analyse how our clients are spread on the territory
  • how about we have a 2 columns closet, one with shelves and the other to hang clothes.

Coming soon: Step #2: General and detailed specifications.

This entry was posted in Business Intelligence, DIY, opinion, Warehousing. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong> <pre lang="" line="" escaped="">