Thursday, 4 July 2013

The Fragmentation (and Crisis) of Business Analytics

The current crisis has many concurrent causes and we don't want to analyze them all. However, we do wish to focus on Business Analytics (BA), as one of the indisputable contributors to the meltdown. Since BA is very popular, one cannot avoid thinking that, to a certain non negligible degree, these tools not only didn't help prevent the crisis, they are also incapable of pulling us out of the crisis. And isn't this what they are for?

From the Wikipedia: "
Business analytics (BA) refers to the skills, technologies, applications and practices for continuous iterative exploration and investigation of past business performance to gain insight and drive business planning. Business analytics focuses on developing new insights and understanding of business performance based on data andstatistical methods. In contrast, business intelligence traditionally focuses on using a consistent set of metrics to both measure past performance and guide business planning, which is also based on data and statistical methods.

Business analytics makes extensive use of data, statistical and quantitative analysis, explanatory andpredictive modeling, and fact-based management to drive decision making. Analytics may be used as input for human decisions or may drive fully automated decisions. Business intelligence is queryingreportingOLAP, and "alerts".

This is precisely the problem. The number of methods and techniques is immense. Selecting the appropriate method when facing a specific problem is becoming increasingly difficult as the number of techniques grows. The (subjective) choice of a technique impacts the result and this adds uncertainty to an already uncertain problem.

What does all this point to? Fragmentation, fractalization, lack of new ideas, a state of crisis. When, in a particular field, you witness fast multiplication of techniques, which are all similar to each other, all being incremental improvements of the same old concepts, we're talking of crisis. The fragmentation we see in BA is an eloquent reflection of its deep state of crisis. The crisis is so deep that there are companies which offer specific tools to match a BA problem with a BA tool. Just think of the number of combinations. More uncertainty on top of uncertainty. What is the point? One thing is sure, tools which help to select tools will only make people more ignorant. Method selection must remain in the hands of the analyst if he doesn't want to become a button pusher.

Thomas S. Kuhn has pointed this out in his "Structure of Scientific Revolutions", in which he claims that fragmentation indicates a state of crisis, and crisis precedes a revolution. What a revolution brings is a change of paradigm. In other words, changes in paradigms come as a result of revolutions, not as an outcome of incremental improvements. And the current situation calls for a revolution.

What we need is new ideas, not fractalization. We don't need more techniques or fancy tools that help identify the "right" tools. We need new ideas. Innovation is when you come up with something radically different, not when you cover up cracks on an unstable wall.