Look Closely at the Outliers in Your Data

Kellogg Insight has written an article about the research conducted by Joel Shapiro, Clinical Associate Professor and Executive Director for the Program on Data Analytics.   Shapiro has some terrific advice for data scientists attempting to discover opportunities for improving the customer experience.  He recommends avoiding the usual practice of "scrubbing the data" of outliers.  Instead, he advocates mining those outliers for interesting insights about consumer needs and pain points.  Here's an excerpt from the article: 

“The mere presence of outliers in customer experience data means that really good or bad things can happen to customers,” says Shapiro. “Maybe you can move that [experience] toward something that either increases the number of positive experiences or doesn’t detract from them.”  When data scientists come across an outlier, their first inclination may be to discard it in favor of “cleaning” or “smoothing out” the data. After all, the data might have been entered incorrectly or appear as the result of a modeling error. Or it may represent a freak accident—a set of circumstances unlikely to replicate itself. Why waste time accounting for the easily discountable?  Resist that urge, Shapiro says. It is always worth examining why the outlier occurred.

Shapiro's point actually connects quite well with a technique employed by design thinking experts as they conduct qualitative/ethnographic research.   Design thinkers do not simply interview and observe "average/typical users."  They look for "extreme users" - people who out of the mainstream.  Perhaps, if you were studying a project on grocery stores, you might study someone who buys fresh food daily at the store, and prepares home-cooked meals for his or her large family each day.   At the same time, you might study a few individuals who never cook for themselves, and how rarely buy food at the grocery store because they eat out on a regular basis.   In the end, you are not designing a new product or service for these extreme users.  Instead, you are using these extreme users/outliers to gain insight and inspiration.