Compare methods

	Latent-Class Choice Segmentation	Hierarchical Bayes Choice Model
Field	Marketing Science	Marketing
Family	Latent structure	Regression model
Year of origin	1989	2005
Originator	Wagner A. Kamakura & Gary J. Russell	Peter E. Rossi, Greg M. Allenby & Robert McCulloch
Type	Finite-mixture choice model for simultaneous segmentation and response estimation	Hierarchical Bayesian random-coefficients discrete-choice model
Seminal source	Kamakura, W. A., & Russell, G. J. (1989). A Probabilistic Choice Model for Market Segmentation and Elasticity Structure. Journal of Marketing Research, 26(4), 379-390.	Rossi, P. E., Allenby, G. M., & McCulloch, R. (2005). Bayesian Statistics and Marketing. John Wiley & Sons. ISBN: 9780470863671
Aliases	Finite-Mixture Logit Segmentation, Latent-Class MNL, Mixture Choice Model, Concomitant-Variable Latent-Class Choice Model	HB Choice Model, Bayesian Random-Coefficients Logit, Hierarchical Bayesian Conjoint, Individual-Level Partworth Model
Related	3	3
Summary	Latent-class choice segmentation estimates consumer market segments and their preferences at the same time, by fitting a finite mixture of discrete-choice models to individual purchase or choice data. Wagner Kamakura and Gary Russell introduced the approach in their 1989 Journal of Marketing Research paper, which fit a probabilistic choice model whose latent segments differ in both brand preference and price sensitivity, yielding a unified picture of market structure and elasticities. Rather than clustering consumers first and modeling choice afterward, the method treats segment membership as an unobserved (latent) variable and recovers it jointly with the segment-level choice parameters by maximum likelihood. Each segment is a multinomial logit model with its own coefficient vector, and the mixing proportions describe how large each segment is. Michel Wedel and Wagner Kamakura's authoritative monograph later codified the finite-mixture framework as the methodological backbone of model-based market segmentation. The result links the pattern of brand switching to the magnitudes of own- and cross-price elasticities, giving managers a behaviorally grounded segmentation tied directly to demand response.	Hierarchical Bayes (HB) choice models estimate a separate set of preference weights — partworths — for every individual respondent, while borrowing strength across respondents through a shared population distribution. The model has two levels: at the lower level each person's choices follow a logit driven by their own coefficients, and at the upper level those individual coefficients are treated as draws from a common multivariate distribution whose mean and covariance are themselves estimated. Inference is Bayesian and proceeds by Markov chain Monte Carlo — typically Gibbs sampling with Metropolis steps — which yields a full posterior for each respondent's partworths rather than a single point estimate. The approach, codified by Rossi, Allenby, and McCulloch, solved a long-standing problem in choice modeling: how to recover genuine individual-level heterogeneity from the sparse data each person provides. Sparse individual estimates are stabilized by shrinkage toward the population mean, giving reliable person-level coefficients usable for segmentation, targeting, and realistic market simulation. HB is now the default estimator for conjoint and scanner-based choice analysis.