There seems to be a lot of discussion lately about Business Analytics best practices, specifically the best way in which to use analytics to manage your business. One popular opinion has been interpreted by many to mean eliminating entirely, the use of your “gut” or intuition and replacing it with finely honed analytics. There are plenty of examples cited of how relying on your intuition and instincts to make decisions can take you down the wrong road that a statistical analysis would have avoided: risky mergers and acquisitions, war waged in error, products over or under stocked. The premise is that simply looking at the facts, as brilliantly portrayed in a well-executed analytics environment, will spell out the correct path to follow.
In reality, as Run DMC so accurately put in many years ago: It’s tricky.
This trickiness came to light during a recent client engagement for a large, national restaurant company. We demonstrated the advanced analytic capabilities of a software package by running several years of point-of-sale data through a market basket analysis. The objective was to algorithmically determine the most likely pairing of menu items in an effort to guide the marketing department in deciding on the product mix of a value-pack type offering.
The initial pass through seemed to suggest a particular product grouping that was sure to be popular and raise sales. Certainly, the statistician at the helm of the data-mining tool made his determination that the answer had been discovered, and it was clear. That was until those with the right industry experience on the team spoke up, correctly realizing that the grouping would actually lead to cannibalization and a decrease in margin. Why discount a grouping of products that are already purchased together, was the question. Instinct and experience prevailed to modify the original query from the company’s marketing group to ask for the most likely pairing of items that would maximize profit margin.
One danger in developing analytical solutions is that when they are wrong, they don’t automatically sound off alarm bells or start blinking in bright red colors.
In this case the analytics were correct- they did answer the first question correctly. However it took instinct to realize that they were answering the wrong question. Use of one’s gut is certainly still valid and is an absolute requirement. I believe the correct interpretation of the original opinion isn’t that your gut should be eliminated- it’s very necessary in both designing the analytics and interpreting them. In fact, it seems that when it comes to developing an advanced analytics environment, the goal is to mechanize that gut instinct in a way that is provable, repeatable and reliable.
The key to managing your business analytically isn’t to replace your brain with a giant eight-ball that you shake for answers. Analytics is a tool, but at its best it is also a culture that asks for more information on which to make important decisions—to ask “Why?” when given facts, figures and opinions. Great instincts will make great decisions when supported by deep, accurate and insightful information on the topic at hand. Instincts without facts to support them are often betrayed by ego, desperation or even wishful thinking.
So, go ahead and use your gut and draw on all your experience. Just make sure you have the facts to explain yourself.
No comments:
Post a Comment