An Approach to Modeling Customer Preference Uncertainty by Applying Bootstrap to Choice-Based Conjoint Analysis Data
Analysis of customer preferences is among the most important tasks in a new product development. How customers come to appreciate and decide to purchase a new product affects the products market share and therefore its success or failure. Unfortunately, when designers select a product concept early in the product development process, customer preference response to the new product is unknown. Conjoint analysis is a statistical marketing tool that has been used to estimate market shares of new product concepts by analyzing data on the product ratings, rankings or concept choices of customers. This paper proposes an alternative to traditional conjoint analysis methods that provide point estimates of market shares. It proposes two approaches to model market share uncertainty; bootstrap and binomial inference applied to choice-based conjoint analysis data. The proposed approaches are demonstrated and compared using an illustrative example.