Abstract:

 

Consumer products are often viewed as combinations of components drawn from categories such as features, benefits, and imageries. Successful marketing of brands depends on skill in choosing these combinations. The staggering number of combinations that can result from even a relatively small number of components is not always appreciated. For example, with only 60 components in these combinations, there are over a quintillion (10   ) of them - if there was one penny for every combination, the pennies would cover the surface of the earth twice if laid out side by side! Nevertheless, there are techniques available to consider these astronomically large numbers of alternative combinations to find the single best combination or a small group of actionable alternatives.

 

Although regression-based techniques such as conjoint analysis are useful in situations where the numbers of components are relatively small, such tools are not well-suited to handling problems of the larger type described above. Fortunately, these problems can be analyzed using techniques from the mathematical field of graph theory. Graph theory has grown substantially because of its applications to internet search, social networking, and national security. The purpose of this report is to demonstrate the use of graph theory to optimize products built around sparkling fruit juice beverages.

​

​

​