Recently, I received an e-mail from a colleague I worked with on a project for a division of GE here in the U.S. He mentioned a system we built more than ten years ago. It was one of the first real business intelligence systems for that organization.

I was reminded of the process of moving because when you're moving, you are consumed with all the work of the task. You're carrying boxes, you're directing movers where to put boxes, and all the while, in the back of your mind, you're working on a seemingly endless checklist of new phone lines, address changes, finding a new doctor, locating your favorite mayonnaise at the new grocery store, etc.

In this mad commotion and confusion of moving a household, there is inevitably the placement of the moving box. Either you, or someone you are directing, will put down a box. You will have every intention of moving it to a final resting place later. Perhaps it will be "temporarily" placed along a garage wall or in a basement corner. Wherever it ends up is fine right now, because there are more important things going on (such as placing a box of china) and you have every intention of returning to it later to put it into its proper, final spot.

Possibly a few years later, but certainly the next time you move, you will find the moving box still in the exact place you put it down the day of the previous move. It will be a little dusty and may have something else piled on top of it, but there it will be. Abandoned.

This is not to say that you didn't have the best of intentions. You really did plan to come back and put that box away, perhaps empty it of its contents, sort it out and make it right. You just got caught up in the rest of the moving activities that day, and then there was signing the kids up for school, soccer, Little League and then parent teacher conferences. You know the story.

All the while, the moving box sat, alone and abandoned, waiting for the well-intentioned owner who never returned.

In a business intelligence (BI) project, the moving box is filled with all the tasks on our list of things to do that would have been next in line had life not gotten so busy. These tasks remain unfinished.

I thought of the process of moving because my former GE colleague told me that next year, if things go as planned and the new OLTP system is completed and implemented, they are going to finally turn off the BI system we built more than 10 years ago. For more than a decade, it's been chugging along, providing value to the users, answering questions that can't be answered anywhere else in the organization.

While that BI system has been running, the division has more than tripled in revenues. New lines of business have emerged. New types of customers have been engaged. Mergers and acquisitions have taken place. New measures and metrics of performance have been implemented and, of course, countless people have cycled through the organization, both as users and supporters of the BI application.

All the while, it kept pumping out answers; and all the while, more than one moving box sat quietly against the back walls and closets of that BI application.

When we were building the system, changes, additions, alterations and enhancements were happening at a fast and furious pace. We made countless on-the-spot, arbitrary decisions regarding the design of the databases, user interfaces, semantics, etc. We would often be forced to leave something "not quite done" while we put out a different fire among the beta testers or early-adopter user community. We had to leave many things unfinished because there were higher priorities established by the users and the management team, especially as we suffered the "curse of success" ­– everyone wanted to use it, and all the users wanted more data in it (the sooner the better).

In all these cases, we had the best intentions of going back to complete the incomplete, correct the incorrect and enhance the ungainly. In almost every case, these items became moving boxes.

The lesson here is that when you are building a BI system, it is critically important to realize that once the system enters production, there will be zero time and resources to go back and fix or complete anything that is not entirely mission critical. This is especially true of meta data. There is never time to go back and retroactively populate or correct meta data.

During design and construction of a BI system, things are still malleable. The shape of the system is not yet in its final form; changes, additions and alterations can be made somewhat quickly and with comparatively little effort. However, once the system is being utilized and the change requests, bug reports and enhancement demands overwhelm the team, the die is cast. The moving boxes, so easily rearranged only minutes or hours before, become permanent fixtures.

The trick is to ensure that the most important features, capabilities and fixes get placed properly before the clock runs out. Make sure you are planning ahead and work to minimize the number of moving boxes cluttering the hallways, closets and garages of your BI system.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access