Analytics and Truth
Guest blog post by Martin Jetton, Senior Quantitative Research Scientist at Kronos, Inc.
A definition found in Thomas Davenport’s book “Competing on Analytics” for analytics: “Analytics is defined as the extensive use of data, statistical and quantitative analysis, exploratory and predictive models, and fact based management to drive decisions and actions.” As a practitioner of analytics I’m always asked to define what I do. I like the above definition but it’s a little wordy. I’m leaning towards a simpler but more controversial definition based upon the above: “Analytics is the extensive use of data to explore truths and then change business processes to improve future decisions and actions.” I might even simplify this to “Analytics is the exploration of truths to improve future decisions.” When managing the workforce, these truths provide companies with rich data needed to increase productivity and stay within budget. Now I can hear you saying, “Well Martin you’ve driven that truck off the road. Please elaborate on ‘truths.’”
Truth comes in several forms; formal, factual and value truths. Formal truths are those logic statements of truth such as “If A is true then B is true.” These are truths proven by proofs and never ever really related to business issues. Factual truths start adding real world nouns to the situation. Something like: “all swans are white.” It might be that all known swans are white, but somewhere out there unmeasured is a black swan. So facts can be disputed with observation. The final truth crosses closer into the realm of business decisions and that is “value” truths. These are the soft observational truths carried around by human decision makers. Something like: “Bob likes to make money.” Truths are important because they allow us to “justify” our decision making processes. They are a means to an end, our survival in business.
Why is analytics and truth important to your business? With computers comes inexpensive data collection through the natural process of business automation. We have data coming out of our ears with computer automated processes. Value truths on the other hand are ingrained in our humanness of decision making. Decision makers have developed them to survive in the depth of data today. The struggle for analytic practitioners is defining and exploring the relationships between factual truths and value truths in the context of business decisions. I think that is what Davenport is getting at with his definition of analytics, but the world of factual testing of values is tough for those not ready for it.
As we analytical sorts swim comfortably in the morass of data these days, we need to be cognizant of facts relative to the values in place. While factual truths are easy to explore, value truths do not lend themselves easily to arguments of validation (or verification) and are essential to providing decision makers the necessary visibility to effectively manage the workforce.