Embedded Decision Analysis

Last year at a conference for decision professionals (DAAG 2013), David Matheson articulated a vision for Embedded Decision Analysis, where Decision Analysis (DA) is embedded as part of the natural work process. People doing their routine work would get the benefit of the wisdom behind DA, without necessarily doing a formal DA consulting project.

Back in December, 2013, David was requested to elaborate on his vision by the Society of Decision Professionals. In this talk, he outlined key shifts that the DA community will need to make if DA is to serve a much broader community than it currently does.

Instead of focusing on an intervention, we will need to focus on the business process that defines and drives the work of an organization. Perhaps a bigger shift is to move from delivering decisions to delivering learning. This implies that the nature of our work has to change from one-time one-off engagements to be more iterative and ongoing. Instead of relying exclusively on expert judgment, we’d have to think about consuming and producing data that the organization needs. Careful assessments would give way to good-enough assessments. Decision Analysis would not be front and center, taking all the attention. Rather, it would play a specific role in driving evaluations that are embedded in the business process. We’d have to shift from providing power tools for specialists to scalable tools for non-DA users. More fundamentally, we’d need to shift our rhetoric from “producing good decisions” to “producing valuable results.”

This also requires a personal identity shift. From relying on a highly specialized DA consultant who intervenes in a situation to facilitate a group to a great decision as measured by decision quality, we’d empower ordinary people doing routine work to drive their organization to great decisions as measured by great results.

The usage of our tools would be refocused to deliver learning. For instance, Tornados would not be used as a precursor to a more careful uncertainty analysis. Instead, by simplifying the framing of the problem, the Tornado would be used to directly communicate the impact of uncertainty and focus conversations on learning gaps so that action can be taken.

If you are a decision analyst, or have engaged with decision analysis in some shape or form, either as a practitioner or a consumer, we’d love to hear from you. Please feel free to peruse David’s slides and watch his webcast, and share your reflections/questions in the comments section below this post.