The Value of Oddball Data

Key Thought:

Solving complex, messy business problems by NOT ignoring data that does not fit.

Background:

Harvard Business Review published an excerpt of discussion between Roger Martin – author of “The Design of Business”, and Stephen Scherer, world-renowned autism spectrum disorder and genomics researcher where it illustrate the 3 stages strategy described in the book.  1) staring into the mystery, 2) developing a heuristic, or rough rule of thumb; and 3) creating an algorithm, or step-by-step formula for addressing the problem.

Summary:

Martin

- most problem solvers would start with a hypothesis, test it, and then look for data that confirm or disprove it

-  identify the mystery, then attacked the mystery by asking what is true, rather than what was true

- scientific method focus on creating reliable outcome, but major leap in knowledge comes from achieving a valid outcome

- using well-horned template saves time, but we can forget to look at the mysteries entirely

Scherer

– my belief is that answers to really difficult problems can often be found in the data points that don’t seem to fit existing frameworks

- quantitative measures that produce reliability often strip away nuance and context and thus sometimes even prevent the discovery of a valid outcome

My Comments:

A startling thought consider how well-drilled we are in the reliability reproduction of outcome as the bedrock for the scientific method.  In complex, messy problems, it is often better to look at the mysteries first.  In conclusion, read the book first, and try apply it to a problem we have – like how to improve profitability!

Reference: Conversation: two leading researchers discuss the value of oddball data, Harvard Business Review, November 2009, pg 26.