Knowledge that Transforms

Marketing Science greatly benefited from the admirable and fastidious efforts of more than 200 different individuals who provided manuscript reviews last year. Beyond those individuals already recognized on the editorial board, the editor-in-chief and senior/guest editors of Marketing Science are indebted to the many guest editors-in-chief, guest associate editors, and ad hoc reviewers who provided expert counsel and guidance on a voluntary basis. The following list acknowledges the contribution of guest editors-in-chief, guest associate editors, and ad hoc reviewers who served from January 1, 2015 to December 31, 2015. Finally, let us not forget to thank the authors. Marketing Science requires and receives outstanding submissions from many leading researchers and prestigious organizations. K. Sudhir Yale University

Consumers’ preferences can often be represented using a multimodal continuous heterogeneity distribution. One explanation for such a preference distribution is that consumers belong to a few distinct segments, with preferences of consumers in each segment being heterogeneous and unimodal. We propose an innovative approach for modeling such multimodal distributions that builds on recent advances in sparse learning and optimization. We apply the model to conjoint analysis where consumer heterogeneity plays a critical role in determining optimal marketing decisions. Our approach uses a two-stage divide-and-conquer framework, where we first divide the consumer population into segments by recovering a set of candidate segmentations using sparsity modeling, and then use each candidate segmentation to develop a set of individual-level heterogeneity representations. We select the optimal individual-level heterogeneity representation using cross-validation. Using extensive simulation experiments and three field data sets, we show the superior performance of our sparse learning model compared to benchmark models including the finite mixture model and the Bayesian normal component mixture model. Data, as supplemental material, are available at http://dx.doi.org/10.1287/mksc.2016.0992 .

We study the effects of information content in 59,814 pharmaceutical sales calls on doctors’ prescription decisions for statins, in the face of entry of competing brands and generics, using a hierarchical Bayesian distributed lag model. We conclude that adding information content to the prescription response model improves the in- and out-of-sample performance of the model. In the first six months following generic entry, it is more effective for incumbent brands to detail on drug contraindications and indications, compared to other periods, to positively differentiate from generics. In the first six months following branded entry, it is less effective for incumbent brands to detail on drug indications and costs, given increased competitive clutter. We also document substantial heterogeneity among doctors in their response to information content. Our model is helpful for analysts to more accurately assess the effectiveness of detailing. Our empirical results are also informative for drug manufacturers as they set or change their messaging policies in response to entry and help firms to tailor their message content at the doctor level. Data, as supplemental material, are available at https://doi.org/10.1287/mksc.2015.0971 .

The marketing and operations disciplines have increasingly accounted for the presence of strategic consumer behavior. Theory suggests that such behavior exists when consumers are able to consider future distribution of prices, and that this behavior exposes firms to intertemporal competition that results with a downward pressure on prices. However, deriving future distribution of prices is not a trivial task. Online decision support tools that provide consumers with information about future distributions of prices can facilitate strategic consumer behavior. This paper studies whether the availability of such information affects transacted prices by conducting an empirical analysis in the context of the airline industry. Studying the effect at the route level, we find significant price reduction effects as such information becomes available for a route, both in fixed-effects and difference-in-differences estimation models. This effect is consistent across the different fare percentiles and amounts to a reduction of approximately 4%–6% in transactions’ prices. Our results lend ample support to the notion that price prediction decision tools make a statistically significant economic impact. Presumably, consumers are able to exploit the information available online and exhibit strategic behavior. Data, as supplemental material, are available at http://dx.doi.org/10.1287/mksc.2015.0965 .

We study the pass-through of wholesale price changes onto regular retail prices using an unusually detailed data set obtained from a major retailer. We model pass-through as a two-stage decision process that reflects both whether as well as how much to change the regular retail price. We show that pass-through is strongly asymmetric with respect to wholesale price increases versus decreases. Wholesale price increases are passed through to regular retail prices 70% of the time while wholesale price decreases are passed through only 9% of the time. Pass-through is also asymmetric with respect to the magnitude of the wholesale price change, with the magnitude affecting the response to wholesale price increases but not decreases. Finally, we show that covariates such as private label versus national brand, 99-cent price endings, and the time since the last wholesale price change have a much stronger impact on the first stage of the decision process (i.e., whether to change the regular retail price) than on the second stage (i.e., how much to change the regular retail price). Data, as supplemental material, are available at http://dx.doi.org/10.1287/mksc.2015.0947 .

The proliferation of free trials for high-tech services calls for a careful study of their effectiveness, and the drivers thereof. On one hand, free trials can generate new paying subscribers by allowing consumers to become acquainted with the service free of charge. On the other hand, a disappointing trial experience might alienate potential customers, when they decide not to adopt the system and are lost for good. This dilemma is particularly worrisome in early periods, when service quality has not been “tried and tested” in the field, and breakdowns occur. We accommodate these phenomena in a model of consumers’ free-trial and regular adoption decisions. Among other effects, it incorporates usage- and word-of-mouth-based learning about quality in a setting where quality itself is evolving. Consumers are forward-looking in that they account for changes in quality and anticipate uncertainty reduction due to trial usage. We estimate our model and run simulations on the basis of a rich and unique data set that incorporates customers’ trial subscription, adoption, and usage behavior for an interactive digital television service. The results underscore that free trials constitute a double-edged sword, and that timing and consumers’ usage intensity during the trial are key to the effectiveness of these promotions. Implications for managers are also discussed. Data, as supplemental material, are available at http://dx.doi.org/10.1287/mksc.2015.0973 .