When a decision is required, a manager typically begins by making certain assumptions about the environment in which the decision will be carried out. Such assumptions might involve, for example, the level of demand over the next few months, the cost of capital, achievable production rates, or the price of raw materials.
Once these assumptions are made, the manager then considers the consequences of adopting alternative policies. Two critical questions must be considered: What if one policy is adopted, rather than another? What if the original assumptions are inaccurate?
Many credit the electronic spreadsheet as the invention that changed the business community's view of personal computers (see John Walkenbach's "spreadsheet history" page), showing that, far beyond being mere game machines, they were tools which could be used to substantially improve managerial decision-making. By building a spreadsheet model of a business situation, with some cells holding assumptions, others containing tentative values for the decision variables, and the rest holding formulas which predict the effect of the assumptions and decision variables on critical objectives, managers could play the "What-If?" game: The decision variables could be varied, and predicted outcomes compared. For any given policy, the assumptions could be varied, and the consequence of errors in those assumptions measured.
Many of the features of modern spreadsheet programs were specifically developed to ease the task of evaluating alternative policies. First came "data tables," which quickly summarize the impact of varying a few decision variables over a wide range of values, and which facilitate sensitivity analysis by showing how changes in critical parameters of a business model might affect the outcome of a decision. Later came "goal-seeking" capabilities, which automate the process of seeking out the policy for which the predicted outcome is optimal. Recently, personal computers have become fast enough that simulation models, built within spreadsheets, can be used both to predict the expected outcome of a decision in an uncertain environment, and to measure the amount of risk associated with the decision.
In today’s business world, a manager who knows how to use spreadsheets for tabulation, optimization, and simulation in support of decision analysis has a decided advantage over those who lack these crucial skills.