Guest Post: Making Levers Work: How Behavior Enablers Connect Decision Intelligence and Human Action (Part 1)

When we build decision models, there is often a particularly satisfying moment. After mapping a complex situation with a causal decision diagram (CDD), the system finally reveals itself, and the decision mapping participants can breathe a sigh of relief that it’s no longer invisibly in our heads but captured online.  

And as we step back to view the CDD, we often observe that a few powerful levers stand out – the actions that, if we took them, should move us most effectively toward our goals. 

Or, at least, that’s how it appears.  In the diagram, everything makes sense. But when it’s time to actually take the actions shown by the CDD, things can be messier.  Sometimes, teams don’t pull the levers as expected. Or they do, but the impact is weaker than the CDD suggested. Even worse: when faced with a compelling impact of the results of behaving in the “old way”, people keep making the same bad choices.  

When actions aren’t taken as expected in these and other ways, decision modelers are left asking, “Did we misunderstand the system, or are people just resistant to change?” 

As it turns out, there can be powerful psychological forces at play that prevent taking the best actions.  What we’ve found is that something else is sometimes going on: the model quietly assumes that people can and will behave in a certain way – while the real conditions for the behaviours it shows are not yet in place.  

Visualizing behavior enablers

The good news is that there’s a way to audit your decision diagram to discover these problems as you’re modeling.  You’re looking to discover missing Behaviour Enablers.

Before showing how that’s done, you’ll need to understand a bit about how Decision Intelligence (DI) and Causal Decision Diagrams (CDDs) help us understand where and how to intervene in a complex system. DI helps you to understand how the actions available to a particular role or person influence outcomes via a causal chain of intermediates.  

DI and CDDs represent a huge step forward compared to gut-feel decision-making alone, because they invite collaborators and careful scrutiny. 

But DI alone does not necessarily tell us whether the decision maker can realistically do what the diagram expects them to. Somewhere between “we draw this lever on the CDD” and “the outcome improves” there are usually a number of behavioural assumptions which, if violated, stop the outcome from occurring.  What are the reasons that, despite a good model, the person or role for whom the decision model was built won’t actually follow its recommendation, e.g.  will take a different action than is modelled? 

Behavioural science gives us a systematic way to identify these assumptions. One of the most practical models is called COM-B, which stands for Capability, Opportunity, and Motivation.  COM-B says that any given behaviour (action) will only be likely to occur if three conditions are present: the decision maker must have the Capability (physical/psychological) to act, the Opportunity (physical/social) provided by their environment, and the Motivation (reflective/automatic to follow through.  

We can assess each CDD intermediate through the COM-B lens. Behaviour Enablers are the capabilities, opportunities, and motivations that must be in place for an intermediate in the CDD to actually occur.. They are designable conditions inside the organization – training, routines, forums, psychological safety, expectations – that we can influence, but often forget to model because we treat people like technical components that will simply “execute” the plan. 

Click here for Part Two of this two-part series, where we’ll show how to identify behavior enablers within a CDD, and how to use this analysis to unblock actions that might otherwise be blocked for psychological reasons.