In 1982, my master’s degree thesis had the weighty title of, “User-Developed Decision Support Systems: Concept and Control”. The paper examined the current literature of the time regarding individual users (not IT) developing their own systems to support decision making.
What is remarkable about re-reading this nearly 30 year-old paper was not my academic brilliance (far from it.) Rather, it is striking how the concepts leading minds of the late ‘70’s and early ‘80’s put forth on the subject of decision support that still apply today.
At the time PCs (then more commonly called “microcomputers”) were just beginning to appear in the business setting. Academic authors predicted their growth would soar in the coming years bringing both pros and cons to decision support activities.
Up to that time, users had been restricted to computing resources over which the Information Technology department (IT) had more control. The PC put increased, uncontrolled computing power in the hands of non-IT people. This was predicted to inevitably result in more user-developed decision support applications.
The two major advantages of such “home brew” efforts were thought to be:
- Increased productivity (quicker development time, less IT resources needed).
- User-developers already knew their requirements (no need for business analysts).
Despite the advantages, the potential dangers to such a trend were troubling:
- Eliminating the Business Analyst role could lead to sloppy requirements management.
- Lack of user-developer knowledge in applying quality assurance techniques.
- Data integrity and report accuracy concerns.
Of course, all of this occurred as predicted. PCs are ubiquitous in the workplace and software like EXCEL has allowed non-IT departments to create their own customized decision support tools. At the same time, lack of rigorous processes to create and maintain these tools has led to many of the predicted problems.
While leading authors had many suggestions for maximizing the positives and minimizing the negatives, the most agreed on three basic principles:
- Data is a centrally controlled corporate asset. All users need to draw from and reconcile to one source of truth.
- IT provides and supports a standard toolset that is analytically powerful, flexible, and easy to use.
- IT provides ongoing user training and support.
Today, the means to apply these principles are readily available.
- Many organizations have developed rigorous data management practices that have greatly improved the quality, accuracy, and security of their data.
- There is a dizzying array of high quality analysis tools on the market today. The biggest challenge is often choosing the best fit for an organization’s needs.
- Enlightened Information Technology departments have adopted a customer service attitude toward training and supporting users of centralized data and standard toolsets. The benefit is a more effective enterprise decision support platform while maintaining the necessary rigor brought about by IT oversight.
Organizations need to examine their decision support processes to determine how well they are applying these principles. Many enterprises are still struggling with the disadvantages of unmanaged decision support activities. (See my article from June 15 Escape From Excel Hell.) Like a weedy farm field, these activities are hard to rein in. However, making an enterprise effort to apply these three sound principles from the age of disco will help your organization stop signing “I Will Survive” and move to a more effective and efficient decision support position singing “We Are Family”.