Author: Noe Gutierrez
Throughout the years, I have been witness to many Data Warehousing projects that started as an effort to help the business users make business decisions using information and along the way transformed into data centralization exercises. Somewhere during the project execution, the team lost focus on what the organization needed to be more efficient and productive, and became tangled in the complexity of the data. The Data Warehouse became an information bank, storing all valuable data elements into an electronic vault.
The expectations from the project teams were very similar, all making assumptions under the principle “If we build it, they will come”. The teams believed that if they could put enough valuable information in a single location, the business users would come knocking on the door, begging to be given access to the information. Reality prove them wrong, as the expectations from the business users grew up disproportionately given the amount time it was taking IT to complete the project . The business users expected a silver bullet, kind of “click here and your information problems will be solved button”, but instead they got a mountain of disorganized data. As a business user colorfully put it: “(finding information in the Data Warehouse) is like going through a huge bin of laundry trying to find your favorite pair of socks. You know they are in there somewhere, but when you can not find, you really start getting frustrated.”
If anything, the early Data Warehousing efforts did make the business users realize the value that information could bring to the operation of the business. The business community also concluded that waiting for IT was a long and tedious process and at the end they were not getting what they needed. The solution was simple, many business teams started hiring their own “shadow” IT organizations, bringing people who had the skills and knowledge of the systems into functional areas bypassing IT controls and governance. The irony came as the Data Warehousing effort was supposed to bring the data into a single location, but these spawned initiatives looked to solve specific business problems by creating small functional data marts across the enterprise. Some other organizations went as far as to look for outside help, asking third party companies, and some times their own vendors to provide insight into the operation of the business by looking and analyzing the data. Although “more responsive than IT”, these functional silos became very expensive to maintain and operate as they did not encourage knowledge sharing, resource pooling or reusing components across departments. Furthermore, as people were using different methods and sources to retrieve and calculate the data, it was a rare occurrence for the numbers to match.