Question

1. What is the first step in the process of developing new products at McDonald’s?

2. What type of statistical methodology would be appropriate to analyze consumer panel taste tests?

3. What sorts of conditions and assumptions might need to be checked?

4. What types of data does McDonald’s collect in the final steps before a new product is launched nationally?

Answer 1

1. The first step in McDonald’s new product development process involves gathering data to track customer trends across the country, determine gaps that exist relative to these trends in the McDonald’s menu, and determine what is on McDonald’s competitors’ menus that works for them.

2. Various types of experimental designs could be used depending on the number of different factors being considered and whether or not blocking is used. For instance, a randomized block design could be used in which each consumer tastes and rates all of the different new products. Similarly, a paired t-test could be used in the scenario described in which McDonald’s is trying to decide if a product tastes better if made on one piece of equipment versus another (each consumer tastes both products made on each piece of equipment).

3. The conditions and assumptions required for ANOVA would have to be checked. If blocking is used, the order in which consumers taste the products should be randomized. The ratings data should have similar variances for the different products. The ratings should also be normally distributed to allow statistical inference.

4. Based on their carefully designed experiments, McDonald’s continuously changes the product until the desired taste test ratings are achieved. If the new product does well in a few of their restaurants, then McDonald’s begins an advertised sales test. They promote the new product in a few markets across the country and, based on the data collected, determine whether it is ready to be launched nationally.