After years of dormancy, there have been renewed rumblings in the DW/BI industry about the Kimball versus Inmon approaches. When I mentioned the resurrected debate to a former colleague who’s been away from our industry for nearly 15 years, she replied, “Again?!” I’m experiencing a serious case of déjà vu.
In fact, we’ve been contacted by several folks looking for us to jump into the fray in a public event. As far as we’re concerned, this supposed controversy is a non-topic. We’re not interested in further stirring the pot, especially when the motivation is commercial gain. We have written and presented our views consistently for nearly 20 years, focusing on understandable, fast dimensional delivery of data for the enterprise, integrated via conformed dimensions. There are no secrets about our approach. Everything’s been documented and available in the public domain; approximately 350,000 copies of our Toolkit books have been sold.
The recent criticism lobbed at the Kimball approach and dimensional modeling seems largely based on misunderstandings of our long-standing messages. We wrote a white paper in January 2008 entitled “Facts and Fables about Dimensional Modeling.” Unfortunately, some people making unfounded charges about the Kimball methodology and architecture apparently haven’t read it.
Nearly eight years ago, I wrote an article for Intelligent Enterprise magazine (since absorbed into InformationWeek) entitled “Differences of Opinion.” At the time, I tried to fairly contrast the Kimball bus architecture approach versus the Corporate Information Factory. I had several clients who were struggling to make a decision between the two dominant schools of thought, so I attempted to summarize the similarities and differences. As industry thinking has morphed over time, some of the differences have softened.
I’ll admit I have a biased opinion about the Kimball approach. We’ve seen our methods work time after time in client situations. We’ve received feedback from the thousands of students we’ve trained that our practical techniques work. However, I’m also willing to admit that organizations can be successful leveraging the approaches advocated by others. Whatever approach you choose, embrace it; read the books and go to training so you can fully adopt the methodology as it is intended.
Rather than devoting more consternation to the philosophical differences between the Kimball and Inmon approaches, the industry would be far better off devoting energy to ensure that whatever we deliver to the business from our DW/BI systems is broadly accepted by the business to make better, more informed business decisions. Allowing ourselves to be distracted by religious debates may interest some, but it doesn’t contribute to what I see as our industry’s true objectives.