Pradeep Chintagunta Research Articles

The Probit Choice Model Under Sequential Search with an Application to Online Retailing

We develop a probit choice model under optimal sequential search and apply it to the study of aggregate demand of consumer durable goods. In our joint model of search and choice, we derive an expression for the probability of choice that obeys the full set of restrictions imposed by optimal sequential search. Estimation of our partially analytic model avoids the computation of high-dimensional integrations in the evaluation of choice probabilities, which is of particular benefit when search sets are large. We demonstrate the advantages of our approach in data experiments and apply the model to aggregate search and choice data from the camcorder product category at Amazon.com. We show that the joint use of search and choice data provides better performance in terms of inferences and predictions than using search data alone and leads to realistic estimates of consumer substitution patterns. Data, as supplemental material, are available at https://doi.org/10.1287/mnsc.2016.2545 . This paper was accepted by Pradeep Chintagunta, marketing.

Measuring the Efficiency of Category-Level Sales Response to Promotions

In this study, we focus on measuring the efficiency of category-level sales response to promotions across various categories and stores. Our heterogeneous stochastic frontier model allows us to attribute portions of this efficiency to specific characteristics of the stores and categories. Using our full PEM (promotional efficiency frontier) model, we analyze the efficiency of 20 frequently bought categories of a supermarket retailer and apply it to store-category-level data. We find that the average efficiency of category and store sales response across all categories and stores is 84.34%, with low values in categories such as spreads and fresh seafood and high values in categories such as frozen entrées and meat. We find that the variation in efficiency of this sales response can be attributed to specific store and category characteristics such as selling area of store, distance to competition, number of stock-keeping units in the category, and average interpurchase time. Unobserved heterogeneity is captured by the latent class approach that provides support for the existence of three segments. An understanding of the roles played by these characteristics in the efficiency of sales response can aid managers in devising a strategy that maximizes sales.This paper was accepted by Pradeep Chintagunta, marketing.

Measuring Multichannel Advertising Response

Advances in data collection have made it increasingly easy to collect information on advertising exposures. However, translating this seemingly rich data into measures of advertising response has proven difficult, largely because of concerns that advertisers target customers with a higher propensity to buy or increase advertising during periods of peak demand. We show how this problem can be addressed by studying a setting where a firm randomly held out customers from each campaign, creating a sequence of randomized field experiments that mitigates (many) potential endogeneity problems. Exploratory analysis of individual holdout experiments shows positive effects for both email and catalog; however, the estimated effect for any individual campaign is imprecise, because of the small size of the holdout. To pool data across campaigns, we develop a hierarchical Bayesian model for advertising response that allows us to account for individual differences in purchase propensity and marketing response. Building on the traditional ad-stock framework, we are able to estimate separate decay rates for each advertising medium, allowing us to predict channel-specific short- and long-term effects of advertising and use these predictions to inform marketing strategy. We find that catalogs have substantially longer-lasting impact on customer purchase than emails. We show how the model can be used to score and target individual customers based on their advertising responsiveness, and we find that targeting the most responsive customers increases the predicted returns on advertising by approximately 70% versus traditional recency, frequency, and monetary value–based targeting. This paper was accepted by Pradeep Chintagunta, marketing.

Personal and Social Usage: The Origins of Active Customers and Ways to Keep Them Engaged

We study how digital service firms can develop an active customer base, focusing on two questions. First, how does the way that customers use the service postadoption to meet their own needs (personal usage) and to interact with one another (social usage) vary across customer acquisition methods? Second, how do firm-to-customer and customer-to-customer communications promote different types of usage? We study these questions using two data sets and by developing a multivariate hierarchical Poisson hidden Markov model (HMM), which fits the data significantly better than univariate and latent class approaches. We indeed find that postadoption behavior varies depending on customer acquisition method and dynamic states. At the total usage level, in one context (an annotation and note-taking service), customers who heard about the service through search and mass-invite exhibited significantly higher usage compared to those who joined through word of mouth (WOM), whereas in another context (a cloud-based file storage service), customers who joined through WOM referrals tended to exhibit higher usage. Yet, examining how routes of adoption relate to specific types of behavior, personal versus social usages, reveals a more nuanced picture. Furthermore, in both contexts, communications postadoption influenced engagement, albeit in different ways. Firm-to-customer communications, through company posts to Twitter and blog entries, had varying effects on customer behavior and in some cases led to lower personal and/or social usage; however, customer-to-customer communications tended to increase personal-use engagement across latent states and in both data sets. The findings suggest that firms offering digital services should pay attention to how the mode of customer acquisition is related to subsequent usage intensity, accounting for both personal and social activity, and encourage customers to interact with each other postadoption. The online technical appendix is available at https://doi.org/10.1287/mnsc.2017.2754 . This paper was accepted by Pradeep Chintagunta, marketing.

How Targeting Affects Customer Search: A Field Experiment

It has become common practice for retailers to personalize direct marketing efforts based on customer transaction histories as a tactic to increase sales. Targeted email offers featuring products in the same category as a customer’s previous purchases generate higher purchase rates. However, a targeted offer emphasizing familiar products could result in curtailed search for unadvertised products, as a closely matched offer weakens a customer’s incentives to search beyond the targeted items. In a field experiment using email offers sent by an online wine retailer, targeted offers resulted in decreased search activity on the retailer’s website. This effect is driven by a lower rate of search by customers who visit the site, rather than a lower incidence of search. There are several ways this could potentially hurt retailers and consumers, such as reduced cross-selling and fewer opportunities for customers to explore new products. This paper was accepted by Pradeep Chintagunta, marketing.

Segmentation of Spatially Dependent Geographical Units: Model and Application

We develop and test a new methodology that assigns geographical units (such as ZIP codes) to market segments by simultaneously considering bases of segmentation, such as customer attitudes and needs, such that the resulting segments display a high level of spatial concentration. Such concentrated segments are managerially desirable because of the logistical and administrative efficiencies created by implementing spatially concentrated segment-level strategies. The suggested approach captures spatial dependence flexibly while allowing us to identify managerially actionable segments that effectively balance homogeneity and spatial concentration. An illustrative empirical application demonstrates how the results can be utilized to derive consumer insights for actionable segmentation strategies. A Monte Carlo simulation and an assessment of the quality of the segments obtained from the empirical application support the value of the proposed spatially dependent segmentation method. This paper was accepted by Pradeep Chintagunta, marketing.

Signaling Low Margin Through Assortment

Oftentimes, close competitors carry partially overlapping assortments in seeming contradiction to the principle of maximum differentiation. One of the justifications of such practice is that an overlapping assortment with competitive prices on the common products may prevent further consumer search and therefore could be useful even when profits from the overlapping products do not justify the costs of carrying them. In this paper, we examine the validity of this intuition and show that such strategy may indeed be optimal when consumers are uncertain about prices they might find elsewhere and face shopping costs for discovery of all prices. Specifically, we show that the (larger) assortment with product overlap may signal a “competitive” price of the relatively unique product and prevent further consumer search for a lower price on it. An implication of this finding is that a consumer may rationally behave as if she likes a larger assortment even if the assortment is enlarged by adding products the consumer has no interest in. Furthermore, we show that the optimal pricing strategy may include pricing of common products or products with known costs at a loss, which provides a novel explanation of loss-leader pricing. Data, as supplemental material, are available at http://dx.doi.org/10.1287/mnsc.2015.2384 . This paper was accepted by Pradeep Chintagunta, marketing.

Two New Features in Discrete Choice Experiments to Improve Willingness-to-Pay Estimation That Result in SDR and SADR: Separated (Adaptive) Dual Response

Discrete choice experiments such as choice-based conjoint and dual response are popular preference elicitation methods, yet, they can suffer from context effects, extreme response behavior, and problems with estimating consumers’ willingness to pay accurately when the purchase probability is high. This study proposes two new features to avoid these limits: (a) strictly separating all forced and all free choice questions and (b) an adaptive mechanism to select fewer, but more informative, free choice questions. The use of these two features invokes two new discrete choice experiment methods: separated dual response (SDR), with just the first feature, and separated adaptive dual response (SADR), with both features. A conceptual comparison, simulation study, tests to address endogeneity concerns, and three empirical studies demonstrate the appeal of these two new features, relative to the benefits of existing discrete choice experiments, especially when estimating willingness to pay. Data, as supplemental material, are available at http://dx.doi.org/10.1287/mnsc.2015.2367 . This paper was accepted by Pradeep Chintagunta, marketing.

How Much Is a Win Worth? An Application to Intercollegiate Athletics

Intercollegiate athletics in the United States have become a multibillion-dollar industry over the past several decades. In this study, we investigate the short- and long-term direct monetary effects of operating a winning athletics program for an academic institution of higher education. We construct a unique panel data set from multiple sources and utilize the latest dynamic panel data estimation methods to account for heterogeneity while also addressing endogeneity concerns. We find that success in men’s football and basketball has a significant impact on a school’s respective football and basketball revenues; however, the effect is different based on the type of school. We find that regular season wins in football account for most of the increase in revenue for established schools, whereas invitations to prestigious postseason bowl games play a big part for less-established schools. Furthermore, we find that student population and education quality dissipate the effect of athletic success on monetary payoffs. We find that success in basketball carries over more from the past than in football with additional contemporaneous marginal effects for established schools. We do find, however, that past athletic success carries over significantly to the present in both football and basketball, suggesting the significance of the long-term monetary effect of athletic success to many academic institutions in the United States. This paper was accepted by Pradeep Chintagunta, marketing.

Deadlines in Product Development

Deadlines are common in product development and are often felt to be too harsh—many development efforts are still worth continuing at the time of mandated termination. We examine the value of deadlines from the agency-theoretic perspective. We consider a firm that pays an agent to lead product development activities. The chance of success depends on the viability of the project and the effort of the agent. As the project proceeds without success, doubts grow as to whether the project is viable. To motivate continued effort, the firm must promise the agent a generous reward if success is achieved during the late stage of development. However, rewarding late success undermines effort incentives in the early stage. The firm may find it more profitable to impose a hard, early deadline to eliminate the agent’s dynamic incentive to procrastinate. We derive conditions under which the firm should impose such deadlines. This paper was accepted by Pradeep Chintagunta, marketing.

Everything We Do, You Do: The Licensing Effect of Prosocial Marketing Messages on Consumer Behavior

Do prosocial corporate marketing messages promote consumers’ altruistic behaviors, or do they advance self-interested and self-indulgent actions? To answer this question, the current research investigates the impact of different framings of prosocial marketing messages on consumers’ behaviors and choices more generally. Results from six laboratory studies and a field experiment demonstrate that exposure to messages that praise customers for good deeds can increase subsequent self-interested and self-indulgent behaviors more than messages that publicize a company’s good deeds or thank consumers for their patronage. Our findings demonstrate the possibility that a temporary boost in one’s self-concept drives this observed effect. In addition, the recipient’s level of support for the issue praised for moderates the effect of customer-praise messages on the recipient’s less altruistic behaviors. This paper concludes with a discussion of the theoretical and managerial implications.This paper was accepted by Pradeep Chintagunta, marketing.

Choosing to Choose: The Effects of Decoys and Prior Choice on Deferral

Sellers are often more interested in inducing buyers to make a choice—any choice—than in influencing which option they choose. For example, many retailers are more concerned about maximizing customer purchases within categories than influencing which brand is purchased in any particular category. Although this issue is important to researchers and practitioners, research on how contextual factors, such as decoys and prior decisions, affect whether consumers will “choose to choose” is scarce. We propose a reference-dependent model to fill this gap. This model unites several streams of research and leads to novel predictions that are tested and supported in four experiments. The paper concludes with a discussion of the theoretical and practical implications of the findings. Data, as supplemental material, are available at http://dx.doi.org/10.1287/mnsc.2015.2289 . This paper was accepted by Pradeep Chintagunta, marketing.

Competing for Attention in Social Communication Markets

We investigate the incentives for social communication in the new social media technologies. Three features of online social communication are represented in the model. First, new social media platforms allow for increased connectivity; i.e., they enable sending messages to many more receivers, for the same fixed cost, compared to traditional word of mouth. Second, users contribute content because they derive status- or image-based utility from being listened to by their peers. Third, we capture the role of social differentiation, or how social distance between people affects their preferences for messages. In the model, agents endogenously decide whether to be a sender of information and then compete for the attention of receivers. An important point of this paper is that social communication incentives diminish even as the reach or the span of communication increases. As the span of communication increases, competition between senders for receiver attention becomes more intense, resulting in senders competing with greater equilibrium messaging effort. This in turn leads to lower equilibrium payoffs and the entry of fewer senders. This result provides a strategic rationale for the so-called participation inequality phenomenon, which is a characteristic of many social media platforms. We also show that social differentiation may enhance or deter sender entry depending on whether it can be endogenously influenced by senders. Finally, we examine how the underlying network structure (in terms of its density and its degree distribution) affects communication and uncover a nonmonotonic pattern in that increased connectivity first increases and then reduces the entry of senders. This paper was accepted by Pradeep Chintagunta, marketing.

The Effects of Age and Expertise on Product Evaluations: Does the Type of Information Matter?

To better understand two large consumer segments, the boomers and the echo-boomers, we examine whether and why experts and novices from these two segments rely on attribute versus benefit information in product evaluations. In four studies, we show that expertise affects younger consumers, such that younger novices evaluate products more favorably when the descriptions feature benefit information, whereas younger experts evaluate products more favorably when the descriptions feature attribute information. However, both older novices and experts evaluate products described with benefit information more favorably than those described with attribute information. We further show that differences in the perceived diagnosticity of different types of information mediate the effects of expertise and age on product evaluations. We theorize that age differences in perceived diagnosticity occur because older and younger consumers spontaneously construe information in different ways. Therefore, age differences in the effect of expertise on evaluations should disappear when construal levels are controlled. Consistent with our hypotheses, we demonstrate that when primed to construe information at concrete levels, older consumers behave just like younger consumers—older experts formed more favorable evaluations toward products described with attribute information, but older novices formed more favorable evaluations toward products described with benefit information. When younger consumers are primed to construe information at abstract levels, they prefer products described with benefit information regardless of expertise, just like older consumers. We discuss the implications of our results for both researchers and practitioners. This paper was accepted by Pradeep Chintagunta, marketing.

Cumulative Growth in User-Generated Content Production: Evidence from Wikipedia

Open content production platforms typically allow users to gradually create content and react to previous contributions. Using detailed edit-level data across a large number of Wikipedia articles, we investigate how past edits shape current editing activity. We find that cumulative past contributions, embodied by the current article length, lead to significantly more editing activity, while controlling for a host of factors such as popularity of the topic and platform-level growth trends. The magnitude of the effect is large; content growth over an eight-year period would have been 45% lower in its absence. Our findings suggest that other open content production environments are likely to also benefit from similar cumulative growth effects. In the presence of such effects, managerial interventions that increase content are amplified because they trigger further contributions. Data, as supplemental material, are available at http://dx.doi.org/10.1287/mnsc.2015.2253 . This paper was accepted by Pradeep Chintagunta, marketing.

Editorial—Marketing Science and Big Data

This article was downloaded by: [128.97.27.20] On: 25 May 2016, At: 09:44 Publisher: Institute for Operations Research and the Management Sciences (INFORMS) INFORMS is located in Maryland, USA Marketing Science Publication details, including instructions for authors and subscription information: http://pubsonline.informs.org Editorial—Marketing Science and Big Data Pradeep Chintagunta, Dominique M. Hanssens, John R. Hauser To cite this article: Pradeep Chintagunta, Dominique M. Hanssens, John R. Hauser (2016) Editorial—Marketing Science and Big Data. Marketing Science 35(3):341-342. http://dx.doi.org/10.1287/mksc.2016.0996 Full terms and conditions of use: http://pubsonline.informs.org/page/terms-and-conditions This article may be used only for the purposes of research, teaching, and/or private study. Commercial use or systematic downloading (by robots or other automatic processes) is prohibited without explicit Publisher approval, unless otherwise noted. For more information, contact permissions@informs.org. The Publisher does not warrant or guarantee the article’s accuracy, completeness, merchantability, fitness for a particular purpose, or non-infringement. Descriptions of, or references to, products or publications, or inclusion of an advertisement in this article, neither constitutes nor implies a guarantee, endorsement, or support of claims made of that product, publication, or service. Copyright © 2016, INFORMS Please scroll down for article—it is on subsequent pages INFORMS is the largest professional society in the world for professionals in the fields of operations research, management science, and analytics. For more information on INFORMS, its publications, membership, or meetings visit http://www.informs.org

The Impact of Walmart Supercenter Conversion on Consumer Shopping Behavior

This paper presents an empirical study of the impact of Walmart supercenter conversion on consumer shopping behavior. By using a difference-in-difference estimator, we find that Walmart gains 41% in weekly revenue from the conversion. Decomposing the revenue gains into components attributable to store visits and per-visit expenditures, we find that the majority of these gains were due to larger expenditures, with a much smaller impact from store visits. By contrast, among competing retailers, grocery stores experience the most significant loss (20% weekly revenue) mostly from fewer store visits, with a much smaller impact attributable to per-visit expenditure. Taken together, these findings show that consumers may benefit from reduced shopping costs by making fewer overall trips and increasing their Walmart basket sizes. In addition, we find that overall revenue gains for Walmart from conversion outweigh the small cannibalization loss at the existing Walmart supercenters located farther away. Finally, from category-level analyses, we find evidence of increases in category-level spending in preexisting categories in the converted supercenter. However, we also find that positive demand externality is more pronounced in food categories, mainly as a result of increased purchase incidence. We discuss the implications of our findings for academics and retail managers. This paper was accepted by Pradeep Chintagunta, marketing.

Correcting for Misspecification in Parameter Dynamics to Improve Forecast Accuracy with Adaptively Estimated Models

Adaptive estimation methods have become a popular tool for capturing and forecasting changing conditions in dynamic environments. Although adaptive models can provide superior one-step-ahead forecasts, their application to multiperiod forecasting is challenging when the underlying parameter variation process is not correctly specified. The authors propose a methodology based on the Chebyshev approximation method (CAM), which provides a parsimonious substitute for the measurement updating process in the forecasting period, to help forecasters improve multiperiod accuracy in the case of parameter variation misspecification. In two empirical applications concerning the sales growth of new brands, CAM exhibits superior forecasting performance compared to a variety of benchmarks. CAM’s properties are further explored through extensive simulations, which suggest that the proposed method is more likely to increase forecast accuracy when parameter variation is more systematic but misspecified because of uncertainty regarding its exact functional form.This paper was accepted by Pradeep Chintagunta, marketing.

Threshold Effects in Online Group Buying

This paper studies two types of threshold-induced effects: a surge of new sign-ups around the time when the thresholds of group-buying deals are reached, and a stronger positive relation between the number of new sign-ups and the cumulative number of sign-ups before the thresholds are reached than afterward. This empirical study uses a data set that records the intertemporal cumulative number of sign-ups for group-buying deals in 86 city markets covered by Groupon, during a period of 71 days when Groupon predominantly used “a deal a day” format for each local market and posted the number of sign-ups in real time. We find that the first type of threshold effect is significant in all product categories and in all markets. The second type of threshold effect varies across product categories and markets. Our results underscore the importance of considering product and market characteristics in threshold design decisions for online group buying. Data, as supplemental material, are available at http://dx.doi.org/10.1287/mnsc.2014.2015 . This paper was accepted by Pradeep Chintagunta, marketing.

Probabilistic Selling in Quality-Differentiated Markets

Probabilistic selling—the sale of synthetic products consisting of a lottery between two distinct goods—has been extensively analyzed in horizontal markets. In this research, we investigate probabilistic selling in quality-differentiated markets. This is an important new dimension of inquiry because of the widespread prevalence of quality-differentiated markets as well as significant differences in the preference structure across these markets. In fact, this latter consideration casts doubt as to whether probabilistic selling will even emerge in quality-differentiated markets. We find that probabilistic selling emerges in quality-differentiated markets as a way to profitably dispose excess capacity; moreover, probabilistic selling remains viable even under endogenous quality choice. In addition, in markets where sellers employ “strong” quality differentiation, the introduction of an intermediate probabilistic good actually causes closer quality levels in a product line and enhances consumer welfare. In contrast, in markets where sellers employ “weak” quality differentiation, the introduction of an intermediate probabilistic good increases quality separation and degrades consumer welfare. Overall, we view our contribution as one of characterizing the optimality, implementation, and policy implications of probabilistic selling in quality-differentiated markets. This paper was accepted by Pradeep Chintagunta, marketing.