Business Intelligence is supposed to deliver the information for making the right decisions. Well, how can executives, managers and professionals make the right decisions when, with today’s Business Intelligence approach, vital information for making decisions becomes invisible?
Where are the Lessons Learned?
Right now, Business Intelligence (BI), with buzzwords such as big data, analytics and data warehouses is one of the big trends. No large organization can afford to miss this trend, it seems. But, Business Intelligence is not new. It evolved from Decision Support Systems. Yet, something is missing.
When I started to ask questions about lessons learned, and especially about how to prevent high-impact traps, there were no answers. That is more than worrying, as the future of companies depends on making the right decisions! Let’s have a closer look at this.
As a mathematician, I liked to do BI during the ‘80s and ‘90s. SAS was already a powerful BI tool. It delivered excellent results, up to a world-wide lead in IT storage efficiency. That’s to say, when the following requirements were met:
- The data is readily available or can be created with a modest effort.
- The data is clearly definable.
- The data can be processed through mathematical formulas.
- The maintenance aspects are manageable and their costs acceptable.
After many recent BI discussions, several presentations and exceptions granted, I am afraid that I have to conclude that today’s BI is limited to requirements 1 to 3 above. That’s indeed where it will work. But watch out: There are important lessons learned with requirement 2. In addition, it appears requirement 4 has been lost and that the associated lesson will be re-learned the hard way.
These days, there is something new: mathematics-based Artificial Intelligence (AI) for BI. It has demonstrated impressive results. Though, more computer power and more formulas will not make above requirements go away. There is a tipping point to be aware of (see also generally applicable Guiding Question 4). As long as AI-based solutions stay prior to the tipping, they are bound to continue delivering impressive results. Beyond the tipping point, however, it is a different matter.
Let us call the business intelligence based on above requirements BI-data. In today’s complex world, however, these four requirements are quite often not given. This implies that all business intelligence coming with data, information, experiences and know-how not meeting above requirements is out of scope for Business Intelligence! More so, no trend or accepted best practice can be seen that delivers the missing BI part. With that, it can be predicted that the missing part will not get priority or budget. Add cost savings pressures and elements that may still be in production are bound to be removed. This brings us to the hidden root cause dilemma introduced in the previous post.
The Hidden Root Cause Dilemma of BI
Information not fitting the BI-data requirements above becomes invisible, as its root cause is located outside of BI and outside the parts receiving priority and budget. The tragedy is this: Sooner or later, it will become visible that the wrong decisions have been made while the decision makers thought to have made the right decisions. People will be made responsible and will lose their jobs. It doesn’t end there. Their successors will fall into the same trap. In addition, it will be no surprise when productive employees lose their jobs. During the ‘80s and ‘90s, this was not the case where I worked.
At my employer, it was common practice to drop data, definitions and mathematics when the four requirements above were not given. In such situations, we looked for lessons learned, knowledge and experience. We reduced an overload of patterns down to those showing the way to the best possible solutions. Let us call this BI-knowledge. Quite often, we used both BI-data and BI-knowledge. This delivered a stream of solutions and services that were ahead of their time. Clients were more than happy. While others were still trying to convince colleagues to follow boring definitions and bureaucratic standards, our solutions were already in production.
With this, here is my collection of Guiding Question for this field.
Guiding Questions for Business Intelligence
Solving this trap starts with a general question:
Where is the solution approach that delivers what executives, managers and professionals need to make the best possible decisions? To be more precise:
- Where is the training and coaching of BI-professionals in the lessons learned and in the associated solutions of the past 25 years?
- Where are the BI-professionals who ensure that BI remains prior to the tipping point, beyond which the quality of data drops and costs grow exponentially?
- Where is the priority for both BI-data and BI-knowledge in decision making processes?
- Where is the solution for delivering BI-knowledge?