THE SEVEN STEP DECISION METHOD AND COMPUTER-BASED EXECUTIVE DECISION SUPPORT TOOLS A Management Technique Briefing from Dalton Dialogic, Inc. There are decisions, and then there are decisions. Some are of trivial importance. Some have implications beyond the obvious. On one thing, however, the experts on decision making all agree: good decisions stem from good decision making techniques. The art of decision making is one that is learned and practiced. Regardless of what we may believe regarding "born" leaders, the decisions they make are not, usually, seat-of-the- pants guesses based solely on past experience. In fact, those types of "gut level" decisions are very often wrong. The best decision makers, both on the personal and professional levels, use a decision support system to organize and direct their decision making process. This support system may be unconscious, born of years of practice, or it may quite formalized. In either case, it is structured and it imposes discipline on the decision making process. Decision making is integral to problem solving as well. Often, you can use the decision making process to decide between more than one potential solution to a problem. In other words, every problem has multiple possible solutions. For example, we might need to deal with a problem of, simply, being hungry. There are many solutions to dealing with our hunger. Now we must decide which is best. The need to make a decision often proceeds from a menu of possible solutions to a dilemma. Many problems are trivial. Their solutions are, almost, automatic responses to a question. Should we name our new daughter Jane or Mahitable? While some may argue that Mahitable is a unique name, most would, without thinking (given those two choices), select Jane. However, people in all facets of business and commerce make volumes of decisions on an almost hourly basis in the course of their business affairs. Finally, in today's business world, management experts agree that most organizations are "flattening", that is, becoming less vertical and hierarchical in their management structure. That means that ever more managers, supervisors, administrators, and, just, plain workers, are being empowered to make decisions about the ways they do their jobs. In fact, the concept of employee empowerment is one that is gaining far reaching support throughout business. These decisions may or may not be trivial. They may or may not have far reaching impact on the company's future. But regardless of their weightiness, they take time to consider and they almost always have some sort of impact on the business or the people involved. So, there are decisions to be made. And they are important and the process time consuming. HOW DO WE MAKE DECISIONS? Paul Moody in his book Decision Making... Proven Methods for Better Decisions (McGraw-Hill, Inc. 1983) describes a five step problem solving loop. This process describes how most people solve problems. The model is applicable to decision making. Moody tells us that the first step is to become conscious of the problem. Then, once an awareness of the problem exists, we attempt to recognize it and define it. The third step is to analyze potential alternatives and consequences. That leads to the fourth step, selecting a solution. Next you will implement the decision and, finally, provide feedback as to the success or failure of the solution. A negative outcome would require that you start the process over again. Thus, the description of this process as a problem solving loop. There are, however, other, similar, processes for arriving at a decision. Moody describes several in his book, ranging from simple brainstorming to complex mathematical statistical analysis. In reality, for the bulk of all business and personal decisions, people require a simple, yet elegant, set of decision support rules. Decision support is an extremely important piece of phraseology, by the way. No system, from the simple Ben Franklin approach (write down all the negatives on one side of a sheet of paper and all of the positives on the other. The longest list wins) to complex mathematical analysis can make a decision for you. They can, at best, help you guide your thinking and provide tools for exploiting your experience and the experience of others. At the end of the process, it is still left to you, the decider, to accept or reject the decision. People tend to make decisions based upon a finite set of criteria whose content and relevance are suggested by personal experience. Although there are a great number of decision types supported specifically by a detailed analysis of numeric data (such as surveys or sales curves), the bulk of all day to day decisions are supported only by the weight of experience or "gut feel". EVERYBODY HAS A THEORY There are many theories of decision making. Moody's book defines the process. However, there are many authors who have provided us with a variety of decision making techniques. Some of these authors, recognized experts in the field of quality decision making, have identified two important areas of the decision making process. The first area is improving the quality of your decisions. The second is identifying and breaking down barriers to good decision making. Unlike theories advanced by self-proclaimed "experts" to advance a "system", this terrain in the vast tract of the decision making process includes the basic considerations with which you, as a decider, must deal. Simply, for a "system" or "theory" to be of any practical use, it must accommodate these basic precepts. 1. Making Quality Decisions Author and educator Irving L. Janis, professor of psychology at Yale University and his colleague Leon Mann, professor of psychology in the School of Social Sciences at Flinders University (South Australia) and former professor at Harvard outline seven major criteria for high quality decision making in their book Decision Making - A Psychological Analysis of Conflict, Choice and Commitment (Free Press - 1979). According to Janis and Mann, these seven criteria include: - Canvassing a wide range of alternatives and choices - Surveying the full range of objectives to be fulfilled and the values implicated by the choice - Carefully weighing costs and risks of negative consequences, as well as positive consequences of each alternative - Intensively searching for new information relevant to further evaluation of the possibilities - Taking account of any new information or expert judgement, even if it does not support his or her preferred course of action - Reexamining both positive and negative consequences of all known alternatives before making a final decision - Making detailed provisions for implementing the final decision, including contingency plans for the materialization of known risks Considering these qualities of superior decision makers should be paramount in the design of any decision support system, regardless of whether it is mental or mechanical. Deciders should, in other words, frame their decision making processes around these guidelines for making quality decisions. Developers of decision support software programs should, likewise, fully consider and, where practical, support the quality decision making process. 2. Avoiding Barriers to Good Decisions J. Edward Russo, professor of Marketing and Behavioral Science at the Johnson Graduate School of Management, and faculty member at Cornell University in the Field of Cognitive Studies, and Paul J. H. Schoemaker, associate professor of decision sciences and policy at the Graduate School of Business, University of Chicago, approach the decision making process from the other direction. In their book, Decision Traps - The Ten Barriers to Brilliant Decision-Making and How to Overcome Them (Simon and Schuster, October 1990), the authors help deciders break down barriers that get in the way of the making of quality decisions. Here are Russo and Schoemaker's 10 traps: = Plunging in - beginning to gather information and reach conclusions without thinking through the issues = Frame blindness - solving the wrong problem because have created a mental framework for the decision with little thought. This causes you to overlook the best options or lose sight of important objectives = Lack of frame control - failing to fully define the problem in multiple ways, or being unduly influenced by others = Overconfidence in your own judgement = Shortsighted shortcuts - using rules of thumb when solid research is indicated = Shooting from the hip or "winging it" = Group failure - assuming that just because good people are involved in the decision making process quality decisions will follow automatically without the need to manage the process = Fooling yourself about feedback - failing to interpret the evidence of past experience because you are protecting your ego = Not keeping track - failure to analyze past results because you did not systematically track outcomes of previous decisions = Failure to audit your decision making process - failing to create an organized approach to collecting the results of decisions As with the need to frame any decision support system around the criteria for making quality decisions,decision support systems must have the means to break down these 10 barriers. DECISION SUPPORT TOOLS - WHAT DO THEY DO? Typically, a decider will, consciously or unconsciously, apply the weight of experience, either personal or otherwise, to defining, refining and selecting the best option in a decision set. If a decision support system is to be useful in this venue, then, it must provide the decider with three things. First, it must allow for quantizing empirical information. That means that the experiential factors which will effect the decision's final outcome must be able to be stated in a manner that yields a finite response. Often, this response is binary - that is, allowing just one of two and only two outcomes. That response could be a yes or a no (as in a go-no go situation). Or, it could be selecting one of a pair of choices as superior to the other. Second, the system must allow the decider to impose relative weights on the various aspects of the decision. The system must not assume that all factors are of equal importance. Nor must it assume that, simply because one factor of a pair outweighs the other, such an outcome is consistent among all pairs of factors. In other words A being more important than B doesn't imply that it is also more important than C, even though B may be more important than C. The decision support system must allow for relative weighting based upon the merits of the individual factors alone. Finally, the system must allow the decider to manipulate factor weightings and the factors themselves in the easiest possible manner in order to remain consistent with the experiential sources of the factors. A simple analysis of decision pairs, consistent with Moody's Precedence system depends for its accuracy upon the decider's ability to apply experience to the decision and weight it properly. Such a support system implies that all relevant factors have been considered, each in relation to the others and that the decider's (or his or her advisor's) specific experience has been applied in a quantified fashion. So, we have three requirements for a coherent decision support system. Is that all we need? In a practical sense, such a system must also allow the user to apply consistency in the way he or she makes decisions. Many decisions are, in part, based upon politics: = Will I ruffle my boss' feathers if I decide this way, even though it's the right decision? Often deciders allow personalities to enter the decision making process: = I know he's best for the job, but I don't want that turkey in the next office. Irrelevant factors often enter the decision making process: = He wears the ugliest ties. We can't give him a bigger sales territory (even though he consistently performs ahead of quota). Finally personal opinions pollute the process: = I don't like him. I'm not going to promote him. The point is, a good decision support system will help guide the decider to consider and weigh all relevant factors in a decision, while ignoring the irrelevant ones altogether. It will also do one other thing. It will validate and document the final decision and the steps taken to arrive at it. Decision support systems are just that: support systems. They don't replace human experience. They don't make decisions. They simply guide the decider through the web of possibilities and then support and document the final decision. It is still the decider's experience that determines the final outcome. Also, these systems, by documenting the decision making process, allow two additional benefits. First the decider can, in the event of an unexpected outcome, review the factors and their weightings that lead to the decision, changing them if appropriate. Second, deciders can perform "what if" analyses by manipulating factors and weightings to suit other circumstances. Finally, when evaluating decision support systems, the intrusiveness of the system itself must be a consideration. People tend to think randomly when making decisions. Factors affecting outcome are rarely properly ordered and paired in the decider's mind. A good decision support system must be simple to use, non-technical in nature and supportive of change during the decision process. It must also offer the broadest possibilities for weighting. A simple "good-better-best" analysis may be excessively restrictive to allow fine tuning complex decisions. SEVEN STEPS TO MAKING GOOD DECISIONS One of the premier benefits of a good decision support system is that it promotes ordered thinking. Such a system, by virtue of its own superior organization, guides the decider into an ordered approach to assembling and considering the factors which will affect the outcome of the process. Dalton Dialogic, after examining several different decision making methods, has determined that most can be supported by a single seven step process. This process, though considered as a simplification of a complex activity, does not trivialize the organization, data or experience that define the decision making activity. Likewise, it is consistant with Moody's 5 step problem solving loop and the Moody Precedence Charts as well as supporting quality decisions and breaking down barriers to good decisions. Here are Dalton's seven steps. 1. List Your Goals and Mission Objectives No decision can be made, regardless of the sophistication of the support tools, if the decision itself hasn't been defined. The premier companies in the business world set their corporate objectives by formally listing goals and a mission objective. The military, always the most directed and focussed of organizations, accomplishes its tasks by defining missions. Once a mission has been completed, it can be analyzed for successes and failures along with the reasons for the outcome. Deciders should learn to make such goal and mission statements at the start of the decision process for two reasons. First, the activity itself forces the decider to state clearly the decision to be made and the objectives which will be met by the outcome. Second, it forces deciders to formulate a mission that can have a measurable result. Mission statements should never be fuzzy. In other words, the results should never be expected to proceed from a broad and conflicting set of goals. The goals and the mission statement should clearly define the expected result of the process so that it is clear what has been decided. Mission statements are lofty. They are global in nature. Why? Simply because individual goals and objectives must be capable of being aligned with the grander objectives articulated in the mission statement. As deciders set goals and determine the extent to which their decisions support those goals, they must always be able to refer back to the mission statement and judge the degree to which they are deciding in support of their overall mission. Finally, mission statements should not be trivial. Decision support systems are not for all decisions. They are only for the important ones . 2. Identify your Choices and Alternatives This step requires that you determine what the possible alternatives to the decision are. For example, if I was interested in deciding upon a particular type of local area network, my possible choices might be: Token Ring, Ethernet and Arcnet. At this point you are not interested in the relative merits of your possibilities, only the possibilities themselves. Usually it is best to include all logical choices at this point, but a good decision support system should permit you to go back later and add additional alternatives. 3. Set Decision Factors and Weights Now we get to the part where we decide what the factors affecting the outcome of the decision are and the individual importances of those factors. It is very important to understand that we are, initially, anyway, weighting these factors in a vacuum. In other words, we don't care how important one factor is against another at this point. All we care about is the factor itself. Once you have made the first pass at determining factors and weighting them, a good decision support system will allow you to perform a "sanity check" on your determinations. Now you can view the factors in general relationship to one another and, if necessary, make subtle adjustments. At this point you have completed a model for making a specific decision. You have established a set of goals. You have determined a finite set of potential outcomes. And you have determined the factors that affect the final outcome and placed them in relative importance to each other. 4. Gather and Input Data Now it's time to apply specific data to the factors and weights for each potential choice. It is now that the model becomes the specific analysis. The factors and weights that affect the selection of possible outcomes are applied against specific data for those possibilities. The result is an order of preference for the possible choices. 5. Evaluate Scored Choices Once the decision support system has applied its scoring methodology to the choices, factors and data, it is time for your personal experience to come back into play. You used your experience (or someone else's) to arrive at the factors and their weightings as well as the universe of possible outcomes. Now that the system has applied your knowledge to the decision at hand, it's time to impose the human factor once again. This time we want to know if the results are consistent with our own experience. We should be able to see choices that are clearly not acceptable, either because they don't fully apply to the stated objectives or because they are of such low scoring that they may safely be eliminated. In a capable decision support system, you should also get an explanation for why one choice was not selected over another. With that additional tool, you can decide what might be required to shift the outcome to one more desirable from your personal perspective. The best choices should be equally identifiable, and the ability to make subtle "what if" adjustments should also be present. This step is the first of two sanity checks on the outcome and the processes that lead to it. 6. Check Choice Conformance to Goals This is the final sanity check. Did the final selection(s) conform to the stated goals and objectives established in step 1? If not, perhaps you need to reevaluate either the factors or weightings to remain consistent with your mission. This is where you will notice and be able to remove irrelevant factors such as personal prejudice, opinion or politics. Occasionally, you will be faced with another possibility which a competent decision support system should help identify and deal with. There will be times when, based upon your existing data or weighting, there is no good choice, or, at least, not one which is in conformance with your objectives. 7. Act on Your Decision You have now arrived at a decision that is consistent with your stated objectives as well as your (or another expert's) experience. You can act on the decision. However, many management specialists believe that even the most experienced and competent decision makers are correct but 50% of the time. While modern information systems may have improved that somewhat, it is still clear that nobody is 100% correct 100% of the time. However, a good decision support system will provide the tools which insure that you will not only make better decisions, but you will also have more confidence in them. Thus, you can expect, even with the best decision support systems, some percentage of error. That percentage of possible error is called uncertainty. A competent decision support system should identify areas of uncertainty and allow adjustments in factors and weightings to fine tune the system and limit the uncertainty as much as possible. It should also permit a very important part of step 7: feedback. As Moody identified in his 5 step problem solving model, decision making, in order to approach perfection, must be fined tuned over time. That means, as decisions are made, the quality of those decisions, based upon the proverbial 20/20 hindsight, must be used in a feedback mechanism to continually improve the decision making model. APPLYING THE 7 STEPS TO THE QUALITY OF DECISIONS The Seven Step Decision Method (SSDM) is not a viable decision support system unless it enhances the quality of the decision making process. However, interestingly, the steps themselves have real correlation with Janis and Mann's approach to quality. For example, surveying the full range of objectives (Janis and Mann) is supported by listing goals and mission objectives (SSDM). By recognizing your objectives you are lead to exploring a wide rage of alternatives. Canvassing a wide range of alternatives and choices (Janis and Mann) is supported by step 2, identifying choices and alternatives (SSDM). Intensively searching for new information (Janis and Mann) is supported by step 4, gather and input data (SSDM). Virtually all of the Janis and Mann quality decision guidelines have corresponding steps in the Seven Step Decision Method. That, of course, is no accident, since the SSDM was designed to support the most consistent and well conceived decision making guidelines today. USING THE 7 STEP SYSTEM TO BREAK DOWN BARRIERS TO GOOD DECISIONS If the 7 steps support the quality decision making process, they shine when it comes to breaking down barriers to good decisions. Again, deferring to the experts, Russo and Schoemaker declare that the first major barrier is Plunging In. Step 1 of the SSDM, listing goals and mission objectives forces that extra measure of attentiveness to the question at hand. When you must list on paper (or on a computer screen) exactly what you wish to accomplish, you tend to stop and think a bit about it. You have, then, avoided the first of Russo and Schoemaker's traps. In the same way that the 7 steps support quality decision making, they also continue to break down barriers. Setting decision factors breaks down lack of frame control. Gathering and inputting data overcomes short sighted shortcuts and shooting from the hip. Checking choice conformance to goals avoids fooling yourself about feedback. And the whole SSDM process avoids the failure to keep track and the failure to audit your decision making process. In short, the SSDM promotes quality decisions and breaks down barriers because it was designed to do so. APPLYING THE 7 STEPS TO A DECISION SUPPORT TOOL A decision support tool, in order to be most useful, must follow a decision making model closely in both its analysis process and its user functionality. Users must be able, without a stringent technical background, to apply the model functionally within the tool. For example, if you are using the 7 step model to arrive at a decision, then the tools you use should functionally support that model. The SSDM is a proven model for arriving at complex decisions. It is applicable to most analytic methods (often as a simplification of an extremely complex process) and it provides results that are consistent, both with empirical testing and experiential analysis. Dalton Dialogic has developed, as a tool for applying SSDM in practical applications, a PC software package called Decisions?/!Decisions. Decisions?/!Decisions uses a special "light tunnel" user interface to guide the decider through each of the seven steps in the model. In addition it provides for fine tuning the factors and weightings during the decision making process using a graphical display for easy visualization of relative weighting of factors. Areas of uncertainty and non applicable results are displayed clearly and you can apply feedback directly to the model for additional fine tuning or "what if" analysis. Finally, Decisions?/!Decisions provides a dialog at the completion of each analysis which contains an explanation of the rationale for not selecting a particular choice along with the requirements for its future selection. The SSDM provides decision makers with the means to model most complex decisions as well as many that are of lesser complexity. It provides a means for establishing a company-wide platform for consistent decision making and documentation. It supports the principles of empowerment by providing consistent tools both for making decisions and honing the individual worker's decision making skills. Decisions?/!Decisions provides an easy to use PC software decision support tool for implementing SSDM. Dalton Dialogic, Inc 102 Runnymead Road, Toronto, Ontario, Canada M6S 2Y3 (416) 767-1291