Multi-attribute utility (MAU) models are mathematical tools for evaluating and comparing alternatives to assist in choosing among them. They are designed to answer the question, "Given the factors we care about, what's the best choice?" MAU models are based on the assumption that the desirability of a particular alternative depends on how well its attributes measure up against key evaluation factors. For example, if you are shopping for a new car, you will prefer one over another based on how well each one scores on the factors you think are important, such as price, reliability, safety ratings, fuel economy, and style. These models can be applied in all kinds of decision situations and are often used in the technical and programmatic parts of procurement evaluations.

Methods to evaluate alternatives. MAU models offer a structured way to weight, evaluate, and compare possible alternatives. They offer a quantifiable method for choosing among options.

Ways to conduct sensitivity analysis. A MAU model can also be used to explore the consequences of changing the attributes, their weights, or the scores they receive. Since the model usually is embodied in a simple spreadsheet, it is possible to make any number of changes and review the results. For example, if it appears that some attribute is too important in determining the results, the weights can be adjusted to produce different overall scores and to see if those differences really matter to the final decision.

Clarifying the nature and importance of evaluation criteria. One of the most useful benefits of using a MAU model is that it makes clear to all involved the basis on which the alternatives are being evaluated. This is particularly important in group decision-making situations in which many different points of view and decision alternatives have to be reviewed and taken into account.

Managing complex comparisons. Some choices need to reflect evaluation of many criteria. A MAU model helps manage that complexity by converting the evaluation to a numerical score while still presenting the logic behind the score.

Requires group consensus. MAU models are typically used in a group situation. To be effective, the group must be able to come to consensus on the attributes in the model and on the weights to be used to indicate their relative importance. It may be very difficult and time consuming, or even impossible to achieve consensus on very controversial decisions.

Conflicts often arise. The level of detail and specification necessary in the discussion of attributes, their overall importance, and the extent to which each alternative meets them can result in considerable conflict and contention, rather than the move toward consensus. However, conflict often indicates that some important criterion has not be made explicit and this possibility should be explored through candid group discussion.

Edwards, W. (1982). Multiattribute Evaluation. Beverly Hills, CA: Sage Publications.