Jeff Stonebraker: Get Decision Makers Involved in Your Analytic Models!

Jeff Stonebraker, a leading professor of decision analysis at NC State’s Poole College of Management supply chain program, discussed some of the important issues in decision analysis in his presentation at the SCRC meeting today.

Jeff emphasized that if you only have one alternative, the decision is easy….what do you need decision analysis for?  But in most cases, you have multiple alternatives, and in this case,  you also need to think about how to develop meaningful and reliable information, employ a clear set of values and tradeoffs, and most importantly, apply logically correct reasoning.  Upfront, there also needs to be a commitment to action on the part of leadership that they will follow-through.  A sound decision-making process promotes decision quality by allowing the organization to minimize bias, as we are often over-confident, and also tend to think we know more than what we know.  Expert bias is based on thinking we are experts, and is also influenced by politics and “who has the best PPT”!  Being compelling is important, but shouldn’t replace logic and correct reasoning.  But the importance of all of these components is superseded by having the right decision-makers at the table, and that it is relevant, and it is provided by an expert.  Framing involves identifying mental structures….and in this case, decision-makers may be able to do it better than you do as an analyst.  These are mental structures and different ways of looking at the problem, and that is why it is important to have many different decision-makers on the team.  This may also mean pushing to have someone who is not initially involved on the team be a part of it.  As an analyst, it is important to make sure you get all of the right people on the team.

Jeff pointed out that there may also be disagreement on the result and outcome of a decision analysis process. This should be viewed as a good thing, as it shows that decision-makers are engaged and part of the decision.  We don’t know if there is a guarantee of good outcomes, but generating the right approaches is a function of a good decision-making process.  Having the appropriate frame is the number one issue ….namely, what is the problem we are seeking to answer?  A proper framing exercise will answer the following questions:

Why are we doing this?

What are we going to do?

How will we know if we are successful?

Framing the mission statement may be 60% of the problem (especially for SCRC projects) is often one of the biggest challenges our students encounter in supply chain projects.  Decision analysis provides a rational and consistent way to make decisions, and provides concepts and language to help the decision-maker.

After framing the issue, the next step involves obtaining decision-relevant information which often is a-process that involves different stages of dialogue.  The team frames the situation, recognizes the situation, and create the frame for the decision including the parameters and outcomes required.  This leads to approval from the decisionmaker.   A key element involves buy-in with the decision-makers early on and this includes having a decision analysis who leads the process, and decision-makers lead the process.

Some of the tools Jeff discussed in decision analysis include strategy tables, tornado diagrams, risk-return plots, rainbow diagrams, decision pyramids, influence diagrams, and others.  In all cases, he advises, “uncertainty is your friend”, (otherwise the problem would be obvious.)  Uncertainty can be expressed meaninfully and clearly through probability estimates, but there is a limit on how much to spend to resolve uncertainty.  Sometimes, killing a project is the best way to add to the portfolio if the uncertainty cannot be effectively measured and identified.

We look forward to more decision analysis projects and insights with our SCRC partners.