Design Optimization of a Laptop Computer Using Aggregate and Mixed Logit Demand Models With Consumer Survey Data

Abstract

Laptop computers are designed in a variety of shapes and sizes in order to satisfy diverse consumer preferences. Each design is optimized to attract consumers with a particular set of preferences for design tradeoffs. Gaining a better understanding of these tradeoffs and preferences is beneficial to both laptop designers and to consumers. This paper introduces an engineering model for laptop computer design and a demand model derived from a main-effects choice-based conjoint survey. Several demand model specifications are compared, including linear-in-parameters and discrete part-worth specifications for aggregate multinomial logit and mixed logit models. An integrated optimization scheme combines the engineering model with each demand model form for profit maximization. The solutions of different optimal laptop designs and market share predictions resulting from the unique characteristics of each demand model specification are examined and compared.