Johnson Marketing Professor Nominated for Top Research Honor
Sachin Gupta’s paper one of four finalists for Journal of Marketing Research’s top honor
Sachin Gupta, Henrietta Johnson Louis Professor of Management and professor of marketing at the Samuel Curtis Johnson Graduate School of Management, and Sungho Park, PhD ’10, have been named finalists for one of the marketing disciplines most coveted research awards. Gupta and Park, now assistant professor of marketing at Arizona State University, are finalists for the 2014 O’Dell Award that honors the Journal of Marketing Research article published in 2009 that has made the most significant long-term contribution to marketing theory, methodology and/or practice.
Their paper, “Simulated Maximum Likelihood Estimator for the Random Coefficient Logit Model Using Aggregate Data,” appeared in the August 2009 issue of the journal, which is published by the American Marketing Association. In the paper, the authors present a new method for using aggregate sales data to model the dependence of demand on marketing mix activities. The paper is based on Chapter 1 of Sungho Park’s Ph.D dissertation, completed while he was at Cornell University.
Until the development and publication of the method presented by Park and Gupta in the paper, researchers relied upon a method commonly known as “BLP,” which was created by three economists in 1995. One drawback to the BLP approach is that it assumes that observed market shares of brands have no sampling error. This assumption was easily met in the economists’ study, as they were analyzing automobile shares in the entire U.S. market, greatly reducing the likelihood of sampling error in the observed market shares.
Yet in many other situations, the sample sizes underlying observed market shares are not as large as in the case of the U.S. auto industry, which can result in significant sampling errors. The method developed by Park and Gupta assumes there is sampling error in observed market shares, and is therefore more applicable—and realistic—when market shares arise from smaller sample sizes. Examples of such situations occur frequently in consumer packaged goods industries, where point-of-sale measurement systems report market shares at the retail store or chain level.
The Park and Gupta method also addresses another weakness of the BLP method. The latter does not provide good estimates of the distribution of consumer preferences, which are of critical importance to marketers, who use these data for segmentation and targeting. The Park and Gupta method provides efficient estimates of the distribution of consumer preferences.
The Park and Gupta method, while relatively young compared with the BLP approach, already has been cited in numerous academic research papers, as well as in working papers and books. Marketing researchers expect the method to have broad appeal in multiple disciplines outside of marketing, such as economics, ecology, and operations management, due to its strengths in modeling dependence of demand on a mix of marketing activities.
The winner of the O’Dell Award will be announced from four finalists’ papers in June 2014. Gupta previously won the prize, in 2008, for the paper “Is 75% of the Sales Promotion Bump Due to Brand Switching? No, Only 33% Is” with co-authors Harald J. van Heerde and Dick R. Wittink. Two of his previous papers have also been finalists for this award in 2001 and 2002.