Oops! The input is malformed!
Originally published 16 January 2006
In terms of analytics, the greatest challenge is knowing the right question to ask. Once you know the right question to ask, the knot unravels itself. But until you know this, the question is as imponderable as the infamous Gordian knot.
So how do you know the right question to ask? There are a million ways to know the right question to ask, but ironically there is not one particular formula that leads to the right path. One way to proceed is to start with the lifeblood of the company, which usually is related to cash and cash flow. From there, you can proceed to sales, production and other areas that are relevant to the business of the corporation. Oftentimes, however, these paths are circuitous and lead to blind alleys. Or worse, they lead to blind alleys that you do not recognize as such.
An even worse problem with these paths is that they only focus on one aspect of the question that you are examining. You end up looking at only a fraction of what is important, while convincing yourself that you are looking at the whole picture.
So you are back where you started. What is the right question to ask?
Knowing the right question often requires both intuition and experience. There are just no formulas or procedures to directly lead you to asking the right question.
Eventually, however, you can begin “fishing in the right pond.” Even if you don’t know the right question to ask, you can search for relevant information. Such information can frequently be helpful. By making enough observations, or observations of the right kind, you can get a feel for what information is there and what questions can be answered. Now you are on the path to determining the right question to ask.
Nevertheless, this path must be tempered by intuition and experience. And sometimes even past experience isn’t enough. As valuable as experience can be, it often closes the door to new ways of thinking.
As an example of past experience becoming a liability, I struggled greatly with the question of justifying the costs the data warehouse. Although I tried to apply all the models of cost justification that I knew, nothing actually worked. Then I began thinking in a completely different way. Instead of trying to understand the problem in a big picture context, I examined the problem at the micro level. And it worked! Suddenly I was able to understand and explain how a data warehouse could easily be cost justified. Only after ignoring my experience and intuition, could I go back and examine the problem in a new way. By doing this, I was able to ask the right question and see the problem in the proper light.
So asking the right question is the key to success in analytics. But this path is as undefined and murky as the road to the river Styx.
Recent articles by Bill Inmon