Data Warehousing: ETL--Part III

Enter the data warehousing team. What they do is magically make the information match and put it where you can get to it, thus solving the two major problems outlined in parts I and II. This process is known as Extraction/Transformation/Loading or ETL. Sometimes this is done with straight up coding and sometimes a tool is used, such as Informatica, but either way, it's a lot of work.

The first step is extraction, which is much like what your analyst does when they pull data from an individual system.

The second step is transformation. This is where much of the value is added by the data warehousing team. When they get the data, it is one holy mess: in one system female is denoted as F, in the next, it's female, in the third, it's 02. There are different data types, calculations, holes in the data, and other horrors we don't even want to hear about.

And it's your DW team's job to get all that data squeaky clean, integrated and moved, every night, without fail, so you can get your reports when you walk in the door in the morning. This final step is loading, which organizes the data so that you can get to it easily and quickly.

Okay, so that's just the beginning. It doesn't sound like much, but once you get to this point in your data warehouse project, you are well on your way to intelligent results!

Thanks for reading!