When We Don’t Believe Our Data

Submitted on: Mon, 03.28.2016 06:01pm - Annie Eissler |


Note: The following blog post was written by guest blogger Andy Weith.

“Beauty is in the eye of the beholder.” This phrase is used to describe a subjective approach to art, and whether or not it is “good” or “bad” it cannot apply to data, where an objective view is necessary. This may seem straightforward and to many folks, myself included, begs the question “Why does this matter, facts don’t lie?”

Confirmation bias can be understood as the selective collection of evidence, or in layman’s terms, favoring information that confirms one’s preexisting beliefs. As people use the data in powerful software tools like MITS Distributor Analytics, they unknowingly bring an art approach to science.

University of Virginia psychologist Jonathan Haidt says “We may think we're being scientists, but we're actually being lawyers. Our ‘reasoning’ is a means to a predetermined end—winning our ‘case’—and is shot through with biases. They include ‘confirmation bias,’ in which we give greater heed to evidence and arguments that bolster our beliefs, and ‘disconfirmation bias,’ in which we expend disproportionate energy trying to debunk or refute views and arguments that we find uncongenial.”

Relating this to sales data, or in my recent experience, tying Cost to Serve information to a set of sales accounts, I found the human element to be fraught with confirmation and disconfirmation bias. Even when agreeing to the calculations that define our Cost to Serve, including but not limited to the below, we find that all too often users argue with the findings:

  • Operational expenses to support the business
  • Average delivery expense based on type of truck delivery
  • Percent of lines returned
  • Average Days to Pay

When presented with the fact that some of their biggest (and sometimes favorite) customers are actually some of their least profitable, or at times even not profitable at all, this bias rears it’s ugly head.

“Well they have a lot of small orders but they usually pick up a bunch at the same time so it doesn’t really cost as much as we say it does!”

When shown a customer with 1/10th the sales and greater net profit, the argument quickly takes a 180 and now this person attempts to explain how these numbers are lying and this customer is not as golden as the facts are presented.

While we are always careful that our data doesn’t trick people by excluding certain details or by using poorly designed data visualization, this is a battle we must also face and overcome. The best solution is perhaps to maintain an objective perspective and consider that our findings may not actually be as we expect them! An outrageous statement, I know, but it seems we are far too likely to attempt to reason with fact rather than accept the possibility that we may be wrong about something!
Andy Weith is the Founder of ASW Global Consulting, a consulting firm that helps businesses better utilize the technology they have and improve the processes they employ.