In my last post I introduced the term “behavior architecture,” and this time I would like to explore what that concept means. One approach is to start with the basics: given a business process with a set of decision points and a number of participants, the behavior architecture is the model of the operational processes, the identified decision points, the type of insight that can be presented to the participant at the decision point, all coupled with the desired expected actions each participant performs at each decision point.

We might refer to this as a behavior model that represents the expectation for potential improvements related to the delivered results of data analyses. For example, in my last post I provided the example of analyzing and scoring sales prospects in relation to the expectation of a positive response based on a number of different factors.

In the behavior model that best benefits from this analysis, the sales person is presented with a list of prospects and scores that is sorted by score from highest to lowest. The expectation is that the sales person will review the list and contact the top 50 prospects on that list, despite the sales person’s perception of how promising the lead is.

The goal of the behavior model is to define the best case scenario and then compare existing behaviors against the best-case model to determine where the participants are not taking advantage of the analytics process. That will highlight who, where and when the objectives of the analytics process are not being met, and will lead to the next step: behavior modification.