Even though many business decisions can be made intuitively, based on hunches and experience, there are more and more situations in which quantitative models can assist. Some managers, however, fear that the use of a formal analysis process will reduce their decision-making power. Others fear that it may expose some previous intuitive decisions as inadequate. Still others just feel uncomfortable about having to reverse their thinking patterns with formal decision making. These managers often argue against the use of quantitative methods.

Many action-oriented managers do not like the lengthy formal decision-making process and prefer to get things done quickly. They prefer “quick and dirty” techniques that can yield immediate results. Once managers see some QA results that have a substantial payoff, the stage is set for convincing them that quantitative analysis is a beneficial tool.

We have known for some time that management support and user involvement are critical to the successful implementation of quantitative analysis projects. A Swedish study found that only 40% of projects suggested by quantitative analysts were ever implemented. But 70% of the quantitative projects initiated by users, and fully 98% of the projects suggested by top managers, were implemented.

Just as managers’ attitudes are to blame for some implementation problems, analysts’ attitudes are to blame for others. When the quantitative analyst is not an integral part of the department facing the problem, he or she sometimes tends to treat the modeling activity as an end in itself. That is, the analyst accepts the problem as stated by the manager and builds a model to solve only that problem. When the results are computed, he or she hands them back to the manager and considers the job done. The analyst who does not care whether these results help make the final decision is not concerned with implementation.

Successful implementation requires that the analyst not tell the users what to do but rather work with them and take their feelings into account. An article in Operations Research describes an inventory control system that calculated reorder points and order quantities. But instead of insisting that computer-calculated quantities be ordered, a manual override feature was installed. This allowed users to disregard the calculated figures and substitute their own. The override was used quite often when the system was first installed. Gradually, however, as users came to realize that the calculated figures were right more often than not, they allowed the system’s figures to stand. Eventually, the override feature was used only in special circumstances. This is a good example of how good relationships can aid in model implementation.