Why we created a Data Warehouse (for the Web Analysts and everyone else)

On Thursday we launched our Data Warehouse. It was a special moment since it took us about a year of blood, sweat and tears to get all data in the same place. Much credit goes to Jon, Andrew, Harb and all the other guys involved in creating this masterpiece.

Why we created a data warehouse?

Our business are 100% online so we use mainly Omnitures various products to collect and analyze the data but it was not enough. The main problem is the lack of flexibility in the Omniture product suite hindering us to truly customize according to our need. We have about 60 websites that we keep track of and on top of that we have no control of the revenue data since it is tracked by a third party making our revenue based decisions more or less a guessing game. With the magic of the Data Warehouse we are able to eliminate this blindspot and make better business decisions.

What is the value of a data warehouse for a web analyst?

I must admit that in the beginning I had a hard time understanding the value of the data warehouse, although my attitude towards data is the more the better. I thought that the Omniture products were enough, and in some sense that is correct. The problem was that we have so many websites, and so many third party cooperation’s that when something went wrong no one could really make a good analysis. Now we can. Now I can see all the way from a visit to a purchase. Priceless.

So, how valuable is a data warehouse for a web analyst and can one get by with just a web analytics package? Well, I think it depends from case to case. As a general rule the more complex your business model is the bigger the need for a data warehouse.  At my previous employer we turned our tool into a data warehouse by customizing it to the extreme (thank you Yahoo Analytics for letting us do it). I know for example Dustin, the wholesaler of electronics are harvesting the potential of a data warehouse. For example by breaking down the figures they could micro segment their users sending out special offers depending on their interests and location.

How are we getting smarter by the data warehouse?

So far we have only had the thing for a few days but already I used it once. In future I see the data warehouse doing a lot of cool stuff such as correlation analysis of our top keywords and brands, projection of future and past trends and many more other things that does not come into my mind right now.

Of course it is like many others have pointed out, the tool itself is a dumb box, and it’s the people using it that will make the difference. Hopefully we will be able to get to the next level but I think road will be bumpy since it involves a mindset change for many people within the organization. If it is done the right way I think this will save a lot of money and energy by letting us to take the right decisions.

I think the web analytics market is going towards a more data warehouse approach

The field of web analytics is a fairly new one but fast evolving. I see more and more signals that the tools are becoming more sophisticated as their users, which is pushing the development towards a more data warehouse driven approach. For example the other day we had a representative of Web Trends that showed us their products and it was clear that they are going towards a data warehouse approach (and also copying Omniture).

I will most likely get back to the data warehouse when we have used it a bit more. Until then feel free to comment below with your own experience on your thoughts or experiences with a data warehouse.

2 thoughts on “Why we created a Data Warehouse (for the Web Analysts and everyone else)”

  1. Great post, Mr. Conversion. Our company is slowly reaching the child-teenager phase (according to the Gartner Web Analytics Maturity model) In North America, we call them “tweeners”. lol

    Yes, the road ahead will be bumpy – no one likes change! 😉 So long as the momentum continues, this is a good step in the right direction for the company. I am passing the torch off to you, though…

    Cheers to becoming a true data-driven organization!

    Rock on dude.

Leave a Reply

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