Selecting between alternatives without Research

Occasionally it may be possible for managers to select the course of action to be pursued without first carrying out some kind of research or experiment. The manager’s experience may tell him or her that a certain alternative seems preferable over the other possible courses of action being considered. Sometimes managers use their judgment or intuition to select the best course of action.

Although there are no decision models or formulas that managers can use to help them choose between alternatives, there is a method of analysis that might be useful to them in such situations. This method is called decision tree analysis, and it can be useful in two ways: it may help a manager choose the best alternative from those under consideration without having to do research; or it may help a manager identify more precisely the information he or she needs to obtain through research in order to select the best alternative.

Decision Tree Analysis:

Decision trees are tree like diagrams consisting of four major components: (1) a decision fork, (2) outcome forks, (3) probabilities associated with each outcome, and (4) rewards or penalties associated with each outcome.

To illustrate, assume that the Acme Bakery is planning a new promotion program in the city in which it is located. The program will consist of either newspaper advertisements or direct mail coupons. News paper advertisements are less costly and will reach a larger audience than direct mail coupons but also are more easily observed by the competing bakery. Consequently, the use of newspaper advertisements is more likely to trigger a counter promotion by the competing bakery. After careful consideration, the marketing manager constructs a decision tree of this situation.

1. Decision Fork: The left side of figure shows the decision fork. The decision fork represents all of the alternative courses of action being considered. If three alternatives were being considered instead of only two, the decision fork would consist of three branches than only the two branches.
2. Outcome Forks: Moving to the right along one of the branches of the decision fork, one encounters an outcome fork. An outcome fork represents all possible outcomes if the alternative associated with that branch is selected. In this case, the possible outcomes of interest are those concerning competitive response. For example, if the decision is made to use newspaper advertisements. Figure shows the possible occurrence of two, outcomes – the competing bakery reacts to the Acme Bakery promotions, or it does not.
3. Outcome Reward or Penalty: Finally, using the best information available, the Acme Bakery marketing manager estimates that Acme will either realize a profit of $500,000 or a loss of $100,000 if it uses newspaper advertisements, depending upon whether or not the competing bakery reacts to the promotion. These are the rewards and penalties associated with the different possible outcomes if newspaper advertisements are used.

The same reasoning applies to the branch associated with the decision alternative of using direct mil coupons. The latter are more costly to use, buy they are less likely to be noticed by the competitor. These facts are reflected in the figure, which shows a larger probability of no competitive response and a smaller profit for the direct mail coupon branch. Thus, figure is a tree diagram for the Acme Bakery promotion decision, and it contains a decision fork, outcome forks, outcome probabilities, and outcome rewards or penalties.

4. Calculating Expected Values: When the decision tree is completed, the Acme bakery marketing manager can use it to evaluate the two alternatives being considered. This is done by calculating an expected value for ach alternative. An expected value consists of two components – the reward or penalty associated with an event and the probability that the event will occur. For example, if one buys a ticket in a $500 lottery in which 100 tickets are sold, the expected value of buying the ticket is equal to the probability of winning the lottery (1 in 100, or 0.01) multiplied by the reward associated with winning ($500). This expected value of $5 (= $500 x 0.01) is a measure of the value of one lottery ticket.