Essay 1. An economic approach to the moderating effects of sales promotion on category structure (with A. Montgomery). Sales promotions are widely recognized as a powerful and ubiquitous tool to spur demand. They are reputed to have that extra special, almost magical quality of generating more sales than a mere price cut of the same magnitude would entail. The traditional wisdom maintains that when a brand is placed on promotion, its sales will increase, and the increase will partly come at the expense of the non-promoted rival brands, thus enhancing the inherent substitution effects in the category. Two factors are usually pointed out as conducive to this impact: the promotion will attract and focus consumers' attention on the promoted brand (informational role), as well as present a compelling reason for the choice of that brand as consumers recognize its temporarily enhanced value (incentive role).;As the ability of promotions to raise sales above and beyond the level implied by commensurate price cuts has been widely acknowledged, many demand models have incorporated them as main effects, as well as factors moderating the impact of the price variable. In many cases, empirical models have been developed in a relatively ad hoc fashion without explicit reference to the underlying economic theory. Drawing on prior research in economics to motivate the proposed methodology, we develop a model consistent with economic fundamentals for a more comprehensive, theoretically sound analysis of demand. The model enables us to capture the immediate propagating impact of promotion that affects the substitution relationships even within a single category, thus altering the category structure.;Central to our approach is the premise that promotions may affect customer perceptions of prices, expenditure, and products, and temporarily alter their decision-making processes. We employ the theoretical framework proposed by Lewbel (1985) to demonstrate how promotion can be incorporated to reflect the posited changes in the consumer decision space, and construct a relatively parsimonious model that affords some interesting new insights into the changing structure of the category. Specifically, we find that promotions of particular brands can foster supplemental purchases of non-promoted brands in the same category. In contrast, promotions of other brands that are characterized with sufficient in-brand variety can simply enhance the underlying substitution patterns. This dichotomy of impact can be attributed to the mediating role of variety-seeking in the category. We also find that promotion may change the substitution patterns between non-promoted brands, although these effects can be rather subtle.;We compare the performance of our model with a series of popular ad-hoc models of varying degrees of economic consistency and parsimony, and demonstrate that in addition to its inferential advantages, our model can outperform or at least match the performance of the benchmark models. We illustrate how the inferences obtained by the model can inform the design of optimal pricing and promotion policies and estimate the profit implications of the proposed changes. We believe that our methodology has important implications for managers and researchers interested in understanding the role of promotions in setting an optimal pricing strategy.;Essay 2. Consumer learning and brand valuation: An application on over-the-counter drugs (with M. T. Akeura and F. F. Gonul, Marketing Science, 2004, Winter, pp. 156-169). We develop a brand choice model with consumer learning based on the Kalman filter methodology. The model enables us to separate the contemporaneous effects of marketing promotions from the impact of the perceived quality valuation accrued through product usage over time. We estimate the model on scanner data of over-the-counter drugs (leg-and-back pain relievers) purchased by a panel of consumers, which enables us to examine the evolution of idiosyncratic consumer learning and preferences over time. The insights from the model estimation afford inferences regarding brand diagnostics based on consumer response to marketing mix variables, brand memorability, and quality valuation updating.;The results point to the presence of a considerable amount of heterogeneity in the valuation carryover coefficients across consumers and brands, but even more importantly, highlight potential deficiencies in the marketing strategies for several of the studied brands. In contrast to our expectations, a higher price seems to be of no concern for a large percentage of consumers in the sample. We suggest that price promotions can be potentially ineffective as a tool to substantially increase market share in the studied product category. Other promotional vehicles such as prominent retail space allocation and more informative advertising can be of greater significance. The inferences drawn from our model, given availability of relevant data, can enable companies to develop educational and product awareness programs that are targeted at segments most likely to benefit from extra information about the application of a particular OTC brand, thus improving its positioning in the market.;This paper applies a relatively new methodology in a specific marketing context and illustrates its analytical potential by demonstrating its inferential power. The model is particularly useful for product categories where consumers are sufficiently motivated to learn their preferences through experience. Our substantive contributions of potential benefit to marketing decision makers are in the development of a diagnostic tool for brand performance in a competitive market setting. Brand managers can employ the model to pinpoint marketing problems in the case of dissatisfactory market performance of their brand. The insights obtained from the analysis can be further used for sales simulations and forecasting under different price and promotion scenarios.;Essay 3. Promotion of prescription drugs and its impact on physicians' choice behavior (with F. F. Gonul, F. Carter, and K. Srinivasan, Journal of Marketing, 2001, July, pp. 79-90). We investigate whether and how the pricing strategies of pharmaceutical companies and their personal selling activities such as detailing and free samples influence the prescription decisions made by physicians. The data used in this study are compiled from nationally representative samples of physicians' diagnosis and prescription data, personal selling data, and retail price data. The prescriptions are written for a specific therapeutic state that is chronic and relatively more common among the elderly population. The end result is a unique data set for this field, combining patient's insurance coverage data, retail price data, detailing and samples data, and the physicians' prescription choice data.;The model we employ is a latent class multinomial logit model with main effects, quadratic terms, and interactions among the key variables of interest, as well as time discounting terms acting on the promotional variables to reflect the notion of memory decay and fading impact of past detailing and samples over time.;We find that the effectiveness of personal selling follows an inverse U-pattern, so that there are optimal values of both detailing minutes and free drug samples. Exceeding those values has dissipative economic impact on the company and potentially, on society. Hence, we find some reasons for concern related to potential waste of resources if this type of marketing activity is not administered systematically and is not monitored more stringently by the pharmaceutical companies. We find that physicians are characterized by fairly limited price sensitivity; that detailing and samples have a mostly informative effect on physicians; and that physicians with a relatively large number of Medicare or HMO patients are less influenced by personal selling efforts than other physicians are.;These insights underscore the need for enhancing the effectiveness of direct promotional efforts to physicians by better segmentation, targeting, and positioning strategies tailored to reflect the intrinsic brand preferences demonstrated by these healthcare professionals and the prevalent type of insurance held by their patients. The amount and scheduling of detailing and free samples can be additionally optimized for maximizing the return on direct selling activities employed by the pharmaceutical companies.
展开▼
