Top-ten decision lever best practices

“We need to win more work,” says the CEO.  “Can you think through how we could lower our prices to become more attractive to our customers?”

A good decision engineer can’t help but ask “the lever question” at this point: “Ma’am, are we only to consider pricing, or would you be open to other approaches to winning more work?”  Confirming the scope of levers allowed like this is the most fundamental decision lever best practice, but there are many more.

Levers are the parts of a decision model that you can change.  They represent the actions that you set in motion with your decisions.  A decision model simulation or diagram can help you to understand how lever settings lead to intermediate results and then to outcomes to understand the impact of those choices: whether it is a personal financial choice, a major strategic decision for a company, or a new policy for a government.

I introduced decision levers in a previous post; here I’ll go into some specific decision engineering best practices around creating great decision levers, with a specific focus about how I conduct the lever brainstorming session in a collaborative decision modeling workshop.

As with all decision modeling, you can use these ideas with pen and paper to draw a decision model on your own, you can use them, as I do, with a team to help align an understanding of your decision situation; and you can also go as far as creating an automated decision model using computer simulation to drive intuition as to the often difficult-to-discern emergent behavior of complex situations.

Here are the ten most important considerations when creating decision levers

  1. Do your lever elicitation after outcomes, to create a focus on the particular levers that the team believes can have an impact on those outcomes.
  2. As above, make sure you understand what levers are in- or out-of-scope for your particular decision.  As in the example above, be sure to ask if that pricing idea should be considered a helpful suggestion or a hard constraint.
  3. To the degree allowed as above, begin with a brainstorming session to consider all possible levers.  No idea is bad; write them all down.
  4. Your team members will almost always talk about the implications of levers: chains of events including externals, intermediates, dependencies, and outcomes.  Respectfully write down their ideas, then bring the conversation back to the levers, saying something like: “Your chain of events is so important I want to be sure to devote dedicated time to it, instead of trying to cover it as a side topic while we’re brainstorming levers.”
  5. Be careful to distinguish levers from externals. Ensure that candidate levers are only those items that can be changed by this team / within your recommendation scope.

    A sample decision model showing an investment lever

  6. Keep in mind that it’s commonplace for levers to become externals and vice versa over the course of using a decision model.
  7. As you start to think about how to visualize levers, keep the fact that this model is, first and foremost, a tool to aid understanding. So use pictures and interactive graphics (most read-only controls like pie charts are good candidates for levers: imagine dragging pie slices to change their size).  Don’t limit yourself to just sliders.  More on lever visualization: Borrow from reality: think about what is natural for your audience.  Avoid text whenever possible. And study Perceptual Learning for a deep dive into the topic of designing the right controls to maximize understanding and performance.
  8. For each lever, ask whether the decision will be made once or continuously.  One-time decision examples include investing in a company, merging, launching a particular new product, or investing in training.  In contrast, a continuous decision might be setting a product price, experimenting with new ad copy in an A/B test, or approving a bank application.  Clearly, there’s a grey area in between: the point is to keep the re-use of the lever in mind as you build.your model and choose how to use it.
  9. Keep in mind that, in many organizations, “that’s how we’ve always done it” prevails, and the levers have become disconnected from their cause-and-effect structure.  When the situation changes, resulting in the same levers produce different outcomes than they used to, we might call this the “ideology” pattern of unintended consequences.
  10. As always, put the data aside (really) so that the team can focus, for now, on the qualitative structure of the model.  Skeleton before meat, tree before ornaments…

In my experience, the lever brainstorming session is one of the most rewarding aspects of a decision modeling workshop.  I’ve found that the combination of the freedom of brainstorming, breaking old assumptions about what could be done to achieve outcomes, a visual model to keep everyone aligned, and a focus on just levers and not the rest of the model, can produce breakthrough results.

Lorien Pratt

Pratt has been delivering AI and DI solutions for her clients for over 30 years. These include the Human Genome Project, the Colorado Bureau of Investigation, the US Department of Energy, and the Administrative Office of the US Courts. Formerly a computer science professor, Pratt is a popular international speaker and has given two TEDx talks. Her Quantellia team offers AI, DI, and full-stack software solutions to clients worldwide. Previously a leading technology analyst, Pratt has also authored dozens of academic papers, co-edited the book: Learning to Learn, and co-authored the Decision Engineering Primer. Her next book: Link: How Decision Intelligence makes the Invisible Visible (Emerald Press), is in production. With media appearances such as on TechEmergence and GigaOm, Pratt is also listed on the Women Inventors and Innovator’s Mural. Pratt blogs at www.lorienpratt.com.

You may also like...