The Evaluation of Potential New Product Alternatives

In evaluating new product opportunities through the use of a rating scheme, account must be taken of the fact that the individual factors on which the product is being assessed interact in producing an overall result. A score for a product may be thought of as a point in n dimensional space in which are located two density functions, one representing scores of successful products and the other of unsuccessful products. Linear discriminant analysis provides the means for utilizing the likelihood ratio of success to failure for any given potential product (point). The question then arises as to whether to accept the decision dictated by the discriminant analysis or to collect more information concerning the location of the point. Both a posterior and a preposterior distribution can be obtained, yielding the value of additional information and thus determining whether to make the decision now or to postpone it.

• 저널Management Science · 13(6) · B–279
• 토픽Innovation Diffusion and Forecasting · Management Science and Operations Research
• DOI10.1287/mnsc.13.6.b279
• 저자Marshall Freimer, Leonard S. Simon