Game theoretic derivations of competitive strategies in conjoint analysis

While conjoint analysis has been applied in a wide variety of different contexts in Marketing, most applications fail to explicitly consider retaliatory reactions from competitors. In this paper, a methodological extension is developed for conjoint analysis by explicitly modeling competition in a game theoretic context. The Nash equilibrium concept is employed to model competitive reactions to produce design, and its implications for reactive product strategies are discussed. The optimal product design problem for each firm is formulated as a nonlinear integer programming problem, which is solved via a specialized branch and bound method combined with a heuristic. In order to compute a Nash equilibrium, a sequential iterative procedure is proposed. The proposed procedure is illustrated under several scenarios of competition using previously published conjoint data.This research has been supported by the Henry Rutgers Research Fellowship, Rutgers University.     

A spatial interaction model for deriving joint space maps of bundle compositions and market segments from pick-any/J data: An application to new product options

We propose an approach for deriving joint space maps of bundle compositions and market segments from three-way (e.g., consumers x product options/benefits/features x usage situations/scenarios/time periods) pick-any/J data. The proposed latent structure multidimensional scaling procedure simultaneously extracts market segment and product option positions in a joint space map such that the closer a product option is to a particlar segment, the higher the likelihood of its being chosen by that segment. A segment-level threshold parameter is estimated that spatially delineates the bundle of product options that are predicted to be chosen by each segment. Estimates of the probability of each consumer belonging to the derived segments are simultaneously obtained. Explicit treatment of product and consumer characteristics are allowed via optional model reparameterizations of the product option locations and segment memberships. We illustrate the use of the proposed approach using an actual commercial application involving pick-any/J data gathered by a major hi-tech firm for some 23 advanced technological options for new automobiles.     

A new clustering methodology for the analysis of sorted or categorized stimuli

This paper introduces a new stochastic clustering methodology devised for the analysis of categorized or sorted data. The methodology reveals consumers' common category knowledge as well as individual differences in using this knowledge for classifying brands in a designated product class. A small study involving the categorization of 28 brands of U.S. automobiles is presented where the results of the proposed methodology are compared with those obtained from KMEANS clustering. Finally, directions for future research are discussed. Wayne S. DeSarbo is the S. S. Kresge Distinguished Professor of Marketing and Statistics, and Michael D. Johnson is Associate Professor of Marketing, both at the University of Michigan's School of Business Administration. Kamel Jedidi is Assistant Professor of Marketing at Columbia University's Graduate School of Business. The authors gratefully acknowledge DuPont Incorporated for providing financial support for this research.     

Simultaneous multidimensional unfolding and cluster analysis: An investigation of strategic groups

This paper develops a maximum likelihood based methodology for simultaneously performing multidimensional unfolding and cluster analysis on two-way dominance or profile data. This new procedure utilizes mixtures of multivariate conditional normal distributions to estimate a joint space of stimulus coordinates and K ideal points, one for each cluster or group, in a T-dimensional space. The conditional mixture, maximum likelihood methodology is introduced together with an E-M algorithm utilized for parameter estimation. A marketing strategy application is provided with an analysis of PIMS data for a set of firms drawn from the same competitive industry to determine strategic groups, while simultaneously depicting strategy-performance relationships.INSEAD     

Latent class metric conjoint analysis

A latent class methodology for conjoint analysis is proposed, which simultaneously estimates market segment membership and part-worth utilities for each derived market segment using mixtures of multivariate conditional normal distributions. An E-M algorithm to estimate the parameters of these mixtures is briefly discussed. Finally, an application of the methodology to a commercial study (pretest) examining the design of a remote automobile entry device is presented.     

Market segmentation with choice-based conjoint analysis

Choice-based conjoint analysis has increased in popularity in recent years among marketing practitioners. The typical practice is to estimate choice-based conjoint models at the aggregate level, given insufficient data for individual-level estimation of part-worths. We discuss a method for market segmentation with choice-based conjoint models. This method determines the number of market segments, the size of each market segment, and the values of segment-level conjoint part-worths using commonly collected conjoint choice data. A major advantage of the proposed method is that current (incomplete) data collection approaches for choice-based conjoint analysis can still be used for market segmentation without having to collect additional data. We illustrate the proposed method using commercial conjoint choice data gathered in a new concept test for a major consumer packaged goods company. We also compare the proposed method with ana priori segmentation approach based on individual choice frequencies.     

Representing Heterogeneity in Consumer Response Models 1996 Choice Conference Participants

We define sources of heterogeneity in consumer utility functions relatedto individual differences in response tendencies, drivers of utility, formof the consumer utility function, perceptions of attributes, statedependencies, and stochasticity. A variety of alternative modelingapproaches are reviewed that accommodate subsets of these various sourcesincluding clusterwise regression, latent structure models, compounddistributions, random coefficients models, etc. We conclude by defining anumber of promising research areas in this field.     

A latent class poisson regression model for heterogeneous count data

In this paper an approach is developed that accommodates heterogeneity in Poisson regression models for count data. The model developed assumes that heterogeneity arises from a distribution of both the intercept and the coefficients of the explanatory variables. We assume that the mixing distribution is discrete, resulting in a finite mixture model formulation. An EM algorithm for estimation is described, and the algorithm is applied to data on customer purchases of books offered through direct mail. Our model is compared empirically to a number of other approaches that deal with heterogeneity in Poisson regression models.