Observations on using data and predictive models in the context of knowledge management, and the possibility of using data and predictive models as a mechanism for creating explicit knowledge from tacit knowledge:

In the business of knowledge management, knowledge is commonly classified into two main types: “tacit” and “explicit.”

Tacit knowledge is typically associated with deep expertise derived from repetition of key information, experiences, and repeated use of skills, etc., over long periods of time. The experts in key areas of any organization, the sales leader, marketing genius, IT wizard, visionary team leader, etc., “just know” some things that the rest of us don’t. They often seem to have a natural gift for taking what they’ve learned from the past and applying it to future cases, and this kind of knowledge has great potential value for the business that can leverage it. Tacit knowledge, however, is often hard to articulate by the person who possesses it, and therefore challenging to capture and share. As such, valuable tacit knowledge is normally difficult to explicitly codify in a way that others in the organization can easily use.

Explicit knowledge, on the other hand, is that which typically lends itself well to codification:

  • As document content
  • As documented business processes
  • As processes automated by means of software

Explicit knowledge can also be generated from your business data, if the data is processed into information that conveys knowledge to its consumers within your organization.

Furthermore, with that data comes the possibility of gaining even greater benefit… by creating advanced analytical models (e.g., predictive models) where the model is the knowledge. The (predictive) model is derived from your data using mathematical algorithms, thus providing your business with the new-found ability to detect complex patterns in data undetectable by the naked eye; once identified, these patterns can illuminate potent success factors and suggest key strategies for growing your business and increasing your bottom line. Once built, the predictive model can be deployed and shared across your organization.

If we see that predictive models are effectively the output of a “learning” process, created from observations recorded in historical data, isn’t it possible to then frame tacit knowledge capture in a new way, as a learning process by which a predictive model is constructed from “experience data” and applied to future cases by anyone using that model? Parallels can be drawn to the same process by which an expert learns from their past experience and applies mental models to new cases. Specifically, if we could capture the experience data of experts over time, couldn’t we apply mathematical algorithms to that experience data, and induce the general from the specific, building a predictive model that mimics the mental model of the expert in knowing what to do when a new experience presents itself? Couldn’t we construct mobile apps, tracker apps and cloud-based solutions to capture experience data from the day-to-day activities of our experts, as well as capturing the outcomes of the decisions they made about how to proceed given a new problem to solve?

By expanding their methods of capturing key business data and employing predictive analytics to evaluate this data in more powerful ways, progressive companies are beginning to extract, codify and leverage the largely untapped wealth of tacit knowledge from the minds of their organization’s experts. In this way any organization might advance the effectiveness of their knowledge management strategy around the capture of tacit knowledge by converting it to a more explicit form, making it more easily accessible for use by the larger organization, and with the proper incentives in place to do so.

If you’d like to explore the intersection of tacit knowledge management and predictive analytics, drop me a line at gene@connollyconsultants.com.