Decision support systems (DSSs) combine models and data in an attempt to analyze semistructured problems and some unstructured problems that involve extensive user involvement. Models are simplified representations, or abstractions, of reality. DSSs enable business managers and analysts to access data interactively, to manipulate these data, and to conduct appropriate analyses.

Decision support systems can enhance learning and contribute to all levels of decision making. DSSs also employ mathematical models. Finally, they have the related capabilities of sensitivity analysis, what-if analysis, and goal-seeking analysis, which you will learn about next. You should keep in mind that these three types of analysis are useful for any type of decisionsupport application. Excel, for example, supports all three.

Sensitivity Analysis

Sensitivity analysis is the study of the impact that changes in one or more parts of a decision-making model have on other parts. Most sensitivity analyses examine the impact that changes in input variables have on output variables.

Most models include two types of input variables: decision variables and environmental variables. “What is our reorder point for these raw materials?” is a decision variable (internal to the organization). “What will the rate of inflation be?” is an environmental variable (external to the organization). The output in this example is the total cost of raw materials. Companies generally perform a sensitivity analysis to determine the impact of environmental variables on the result of the analysis.

Sensitivity analysis is extremely valuable because it enables the system to adapt to changing conditions and to the varying requirements of different decision-making situations. It provides a better understanding of the model as well as of the problem that the model purports to describe.

What-if Analysis

A model builder must make predictions and assumptions regarding the input data, many of which are based on the assessment of uncertain futures.

The results depend on the accuracy of these assumptions, which can be highly subjective. What-if analysis attempts to predict the impact of a change in the assumptions (input data) on the proposed solution.

For example, what will happen to the total inventory cost if the originally assumed cost of carrying inventories is 12 percent rather than 10 percent? In a well-designed BI system, managers themselves can interactively ask the computer these types of questions as often as they need to.

Goal-Seeking Analysis

Goal-seeking analysis represents a “backward” solution approach. It attempts to calculate the value of the inputs necessary to achieve a desired level of output.

For example, let’s say that an initial BI analysis predicted a profit of dollar 2 million. Management might want to know what sales volume would be necessary to generate a profit of dollar 3 million. To find out, they would perform a goal-seeking analysis.

The managers, however, cannot simply press a button labeled “increase sales.” Instead, the company will need to take certain actions to bring about the sales increase. Options include lowering prices, increasing funding for research and development, paying the sales force a higher commission rate, enhancing the advertising program, and, of course, implementing some combination of these actions. Whatever the action is, it will cost money, and the goalseeking analysis must take this into account.