Big Data

Bayesian Applications In Marketing

Download Now Date Added: May 2010
Format: PDF

This paper reviews applications of Bayesian methods to marketing problems. Key aspects of marketing applications include the discreteness of response or outcome data and relatively large numbers of cross-sectional units, each with possibly low information content. Discrete response data require the development of non-standard likelihoods and low information content requires careful use of informative priors. One particularly important form of informative prior is embodied in hierarchical models. Given the importance of the prior, it is important to assure flexibility in the prior specification. Non-standard likelihoods and flexible priors make marketing a very challenging area for Bayesian applications.