CHAPTER 1
What Is Decision Analysis?
And Why Should I Care?
The premise of this book is simple: there is a set of tools and mental frameworks referred to as decision analysis (DA) that can help you, your family, and the teams and groups that you work with improve your decision making. You donât need to hire expensive consultants to use the toolsâthe concepts are not that difficult and the tools are very useful.
Here is the context Iâd like you to envision: you and I have just sat down in first class seats for a three-hour plane trip. You know that I have a background in both management and DA consultation and you have asked me how DA could help you in your job as a manager (or team leader or senior professional). We have a pad of paper to draw on, and we will talk about concepts, examples, and stories rather than the mathematics underpinning DA.
Many of the examples weâll discuss involve personal decisionsâmy friend David Skinner (DA practitioner, author, and entrepreneur) and I discovered while teaching DA at Conoco that people absorb the concepts more quickly from personal examples than from business examples.
So get comfortable, and letâs talk about management and DA tools! Since you have an inquisitive mind and have several questions, weâll begin with FAQs (frequently asked questions) about DA:
Then weâll talk about the tools and how you, as a manager, a team member, or a family member can use them.
Why Is Decision Analysis Important?
Decision analysis is important because your personal and business success (getting what you want) depends on:
understanding what you want,
luck (i.e., uncertainty), and
good decision making.
Thinking through what is important to youâwhat you wantâsounds obvious, but many people simply donât take the time to think through their objectives, let alone those of others. Sometimes removing ambiguity concerning your objectives (and those of the people you are working with) is enough to add clarity and reach a consensus on what to do going forward.
Luck and Uncertainty
We cannot control our luck, but we can use DA to improve the quality of our decisions and subsequently increase the chances of getting what we want (see Figure 1.1).
An important distinction DA makes is that good outcomes and good decisions are correlated, but there is no guarantee of a good outcome as a result of making a good decision. We cannot control outcomes; all we can do is control the decisions we make. Improving the quality of our decision making through time improves our odds, but we are still at the mercy of luck. My friend Patrick Leach noted that it is ironic that in professional poker, knowing the odds is just the start of being able to play, whereas in business, âthe players have the skills but do not understand the odds.â1 DA helps you understand the odds.
Figure 1.1 Luck versus decision quality
When I introduce this concept while teaching DA, I stop and ask the students for examples of poor decisions that can have good outcomes. Invariably a student will bring up winning the lottery, as the expected value (probability of winning times the amount you win) is considerably less than what you have to pay for the ticket. Lifestyle choices we make fit into this part of Figure 1.1: some of us can smoke cigarettes all our lives and have no adverse health effects; others will get cancer or heart disease from making the decision to smoke.
When I ask for an example of a disappointing outcome from a good decision, usually a wildcat oil well comes up as an example. You can carefully analyze the seismic data and pick the best spot to drill a well and still come up with a dry hole. In terms of personal decisions, you can do careful research on a stock or mutual fund and invest carefully, only to have some unforeseen factor drive down the value of your investment. Health is also an example of this part of Figure 1.1: we can eat healthy foods, exercise appropriately, and still get cancer or heart disease. All we can do is make the best decisions possible, thereby improving our odds of getting what we want.
The other point I make when discussing Figure 1.1 is that making poor decisions, especially with regard to lifestyle choices, will eventually catch up with you. Weâll talk about the implications of poor decision making relative to ethics and lifestyle choices at the end of the book. For example, Ralph Keeney concluded in a recent study that about half of the deaths in the United States are a direct result of poor lifestyle decisions (weâll talk more about this in Chapter 14).
One of the biggest problems we have in business is that managers are usually rewarded (or penalized) for outcomes rather than the quality of their decisions. The problem with rewarding luck is that sooner or later luck runs out. If you are in a position where you are evaluating managersâ performances, long-term success of your company depends on making this distinction (talent versus luck) as fairly as you can.
Companies Use DA
Another reason that DA is important is that many companies are using DA as part of their normal business management. If you understand the tools and go to work for a company using DA, youâll be able to contribute quickly. There is strong evidence (as weâll discuss later in this chapter) that companies that use DA are more likely to succeed than those that donât, so if your company is unwilling to use DA in an industry where the competitors do use it, you might consider changing jobs if the opportunity arises.
In fact, one of the most interesting articles Iâve ever read was by a professor named Kathleen Eisenhardt.2 She studied the decision-making processes of eight âdot.comâ companies in detail and correlated the processes with financial and performance results. Decision quality varied from negative to neutral to very high, and financial performance results tracked with decision quality. Therefore, if your company has poor decision-making practices (group think, who shouts the loudest, highly political, etc.), you need to be aware that your competitors are likely to win in the long run.
Also, please note that DA is not a fad like so many other ideas that have been advanced by consulting firms and academia. If youâve been a manager for very long, youâve experienced some of these fads, which at best were a distraction and at worst damaged your companyâs strategy and morale. DA has been successfully used in oil and gas, chemicals, pharmaceuticals, transportation, and manufacturing for decades and is part of the normal business practice at many companies.
Consensus and Alignment
Another reason these tools are important is that they are designed to help groups and teams of people reach consensus and engage in the decision-making process. It is sometimes critical for a group to buy in to a course of action and align together to reach a goal. Caution, though: there are other times when the technical accuracy of the decision is what ultimately drives success.
Very recently, a team of which I am a member was divided over the role of a potential new piece of equipment being developed in the laboratory. A detailed technical program had been laid out and agreed to earlier, but the role of the new equipment was ambiguous. I drew a simple tree (weâll discuss decision trees in a later chapter) on the white board, noting that there are only three possible outcomes of the technical program now in place:
Current technology is adequate to address the needs of the program, in which case the new equipment is interesting but not necessary to complete the mission.
Current technology wonât adequately address the needs of the program, in which case the current program should go on hold until the new equipment is available.
Current technology addresses part but not all the needs of the program, in which case a cost/benefit analysis should be done to determine whether to proceed or wait. We labeled this outcome the âsort-ofâ case; the team deemed this the most likely (and least desirable) outcome.
Once the team understood the uncertainty (adequacy of current technology) and the decisions that would logically flow from resolving the uncertainty, the team aligned and agreed on how to proceed.
What Is Decision Analysis?
Weâve been talking about how important DA is, so it is time to define what it is that we are talking about. David Skinner3 developed the following definitions:
âA decision is a conscious, irrevocable allocation of resources with the purpose of achieving a desired outcome.â
âDecision analysis is a methodology and set of probabilistic frameworks for facilitating high quality, logical discussions, illuminating difficult decisions, and leading to clear and compelling action by the decision maker.â
Note that in real life, thereâs no âcontrol zâ undo on the commitment of our resources. You may subsequently have to reverse a decision, but only with the loss of money, time, and resources. Davidâs point about clear and compelling action is importantâDA should always add clarity. Once a course of action is clear and agreed upon, your analysis is complete, and it is time to focus on planning and implementation.
Note the adjective probabilisticâthereâs a significant component of statistics incorporated into the analysis tools of DA. Somewhat tongue-in-cheek, David and I used to use a chart we called the âFour Dsâ that noted the common elements of statistics in statistical process control (defects), health risk assessment (deaths), fault tree analysis (destruction), and DA (decisions).
Here is my definition of decision analysis:
Decision analysis is a set of tools (frameworks) that can help people or groups of people:
clarify and reach alignment on their goals and objectives,
develop and examine alternatives,
systematically examine the effect of uncertainty, and
maximize the probability of achieving their goals and objectives.
My definition is more of a working definition of DAâyou first have to figure out (and get alignment on) what you want, then figure out what alternatives you really have, and then use probabilistic analysis to handle uncertainty. The toolset is a bit broader than any of the definitions imply. I think that clear and compelling action is important but not as important as maximizing the probability of achieving your goals (you need to have goals to maximize your chances of achieving them!).
The term decision analysis was originally developed by Professor Ronald Howard (not the movie director) of Stanford in the 1960s.4 Since that time, many books and academic papers have expanded the set of tools referred to by the term decision analysis.
You may now be curious about the tools. I split the tools into framing and analysis tools. Framing tools help remove ambiguity concerning the nature of the problem, the objectives, what is known and unknown, and what alternatives or sets of alternatives are available. Sometimes just framing a problem correctly can achieve enough clarity that a team can align on the decisions that need to be made and proceed with planning and implementation.
Analysis includes financial modeling, assessment, and examining how uncertainty affects the potential outcomes for the decision(s) at hand. There are also several special-purpose tools that weâll consider near the end o...