Recently, an executive at an online media firm had asked me to take a look at some data. His team had found some interesting results using some correlations of data points for his web activity. Unfortunately, he wasn’t convinced of what they were saying, because his intuition was telling him otherwise. However, he couldn’t refute the analysis, it was fairly sound. He decided to get another opinion.
This story is similar to a post I made a few months back on the Yin Yang of Analytics, where intuition is just as important as the mathematics. Knowing the business well gives you an advantage and provides great insight, in the same way a jockey keeps a horse on track.
When I looked at the data, it took me a few minutes to realize what might be going on. The correlations were correct, and the other analysis they conducted was numerically correct; however, there were some flaws in their assumptions. This led to an incorrect assessment, and the results of any actions, might not have produced the intended outcomes. We looked at the data and found some other interesting information using some multidimensional techniques. We then sat down with the team to review the findings.
A former NYPD detective once told me, as a police officer “we are taught to trust our gut feeling, if something isn’t right, its probably not.” The same holds true for analytics. Trust your gut, and ask some questions. Challenge the assumptions and validate the results.