Conjoint analysis, a research technique that enables optimal new product development by identifying the levels of product attributes desired by consumers for each important attribute considered during product purchase, has been developed and refined in various forms to improve the predictive power of the model. However, traditional conjoint analysis methods using full-profile or trade-off approaches have not provided satisfactory results in predicting consumer preference for each profile or consumer choice probabilities for brands when the number of product attributes used in profile generation increases. This paper proposes a Hierarchical Bayes hybrid conjoint model that combines the Hierarchical Bayes analytical technique—designed to reflect heterogeneity in part-worths across individual consumers—with a hybrid conjoint model that incorporates both the self-explicated importance evaluation method and the full-profile evaluation method, as a new approach to overcome the limitations of traditional conjoint analysis methods. Through empirical analysis, this paper demonstrates that this model is superior in predictive power compared to other types of conjoint analysis models.