Marketing Papers
Document Type
Working Paper
Date of this Version
7-3-2016
DOI
10.2139/ssrn.2282570
Abstract
Conjoint is one of the most popular methods in marketing research, widely used to understand how customers trade-off features of a product. Since product images have a strong influence on customer choice, it is natural to want to include images in conjoint studies, yet this has proven to be difficult, since images are difficult to parsimoniously characterize in the utility function. This paper proposes a novel approach to account for the effect of images on respondents’ choices, in which consumer heterogeneity in the appeal of the images is modeled through the covariance structure in a probit model. The covariance structure is informed by a separate task where respondents rate the images included in the study. In our application to midsize crossover vehicles, we show that our approach readily scales to a large number of images, fits better than several alternatives commonly used in practice, and makes more reasonable predictions about product substitution when a new product enters the market. We discuss how this approach could be used predict the effect of other difficult-to-characterize product attribute such as sound quality or taste on product choice.
Keywords
choice models, conjoint, images, visual design, styling, Bayesian methods, automotive industry, multinomial probit model
Recommended Citation
Dotson, J. P., Beltramo, M. A., Feit, E. M., & Smith, R. C. (2016). Modeling the Effect of Images on Product Choices. http://dx.doi.org/10.2139/ssrn.2282570
Included in
Advertising and Promotion Management Commons, Behavioral Economics Commons, Business Administration, Management, and Operations Commons, Business Intelligence Commons, Marketing Commons, Sales and Merchandising Commons
Date Posted: 15 June 2018
Comments
This is an unpublished manuscript.