Marketers routinely use timing as a segmentation device through sequential product releases. While there has been much theoretical research on the optimal introduction strategy of sequential releases, there is little empirical research on this problem. This paper develops an econometric model to empirically solve the inter-release timing problem: it involves (1) developing and estimating a structural model of consumers’ choice for sequentially released products and (2) using the estimates of the structural model to solve for the optimal inter-release time. The empirical application focuses on the movie industry, where we specifically address the issue of the inter-release time between a theatrical movie and its DVD version. We find that consumers are indeed forward looking; a shrinking movie-DVD release window does negatively impact box office revenues, but there is a tradeoff in that there is greater residual buzz from the movie marketing that supports the sales of DVD due to the shorter time window. This leads to an inverted U shaped relationship between movie-DVD release window and revenues, and the theater-DVD window that maximizes industry revenue for the average movie during the data period is 2.5 months.
The paper develops a modeling framework to study how sustainability interventions impact consumer adoption of durable goods innovation, firm profit and environmental outcomes in equilibrium. Our two period model with forward looking consumers and a monopoly firm introducing an innovation in the second period accommodates three key features: (1) it builds on the psychology literature linking reactive and anticipatory guilt to consumers’ environmental sensitivity on initial purchase and upgrade decisions; (2) it disentangles environmental harm over the product life into that arising from product use and dumping at replacement; and (3) it clarifies how a taxonomy of innovations (function, fashion and use-efficiency) differ in how they provide value and cause environmental harm during use and dumping. Given how guilt impacts environmental sensitivity, the model allows for owners upgrading a product to be more environmentally sensitive than first time buyers; this makes dumping harm and in-use harm from products not fungible. We find that with fashion and function innovations, increasing consumer sensitivity to environmental harm can surprisingly result in increased environmental harm. Further, when consumers are very sensitive to environmental harm, firms will not inform (pre-announce to) consumers about the impending arrival of use-efficiency innovation; to minimize environmental harm, a sustainability advocate needs to inform consumers. Thus, contrary to conventional wisdom, consumer environmental sensitivity does not always substitute for the role of sustainability advocates. Our results clarify how to design win-win policies for firms and the environment; and when advocates have complementary/adversarial roles relative to firms to achieve sustainability goals.
This paper introduces the problem of coresets for regression problems to panel data settings. We first define coresets for several variants of regression problems with panel data and then present efficient algorithms to construct coresets of size that depend polynomially on 1/ε (where ε is the error parameter) and the number of regression parameters – independent of the number of individuals in the panel data or the time units each individual is observed for. Our approach is based on the Feldman-Langberg framework in which a key step is to upper bound the “total sensitivity” that is roughly the sum of maximum influences of all individual-time pairs taken over all possible choices of regression parameters. Empirically, we assess our approach with synthetic and real-world datasets; the coreset sizes constructed using our approach are much smaller than the full dataset and coresets indeed accelerate the running time of computing the regression objective.
The paper develops the first structural model of organizational buying to study innovation diffusion in a B2B market. Our model is particularly applicable for routinized exchange relationships, whereby centralized buyers periodically evaluate and choose contracts, then downstream users or- der items on contracted terms. The model captures different utility tradeoffs for users and buyers while accounting for how buyer and user choices interact to impact user adoption/usage and buyer contracting. Further, the paper considers the dynamics induced by share of wallet (SOW) pricing contracts, commonly used in B2B markets to reward customer loyalty with discounts for buying more than a threshold share from a supplier. We assemble novel panel data on surgeon usage, SOW contracts, contract switching, and hospital characteristics. We find two segments of hospitals in terms of the relative power of surgeons and buyers: a buyer-centric and a surgeon-centric segment. Further, innovations diffuse faster in teaching hospitals and when surgeries are concentrated among a few surgeons. Finally, we answer such questions as: Should the marketer focus on push (buyer-focused) or pull (user-focused) strategies? Do SOW contracts hurt the innovations of smaller firms? Surprisingly, we find that the contracts can help speed the diffusion of major innovations from smaller players.
We study the problem of constructing coresets for clustering problems with time series data. This problem has gained importance across many fields including biology, medicine, and economics due to the proliferation of sensors for real-time measurement and rapid drop in storage costs. In particular, we consider the setting where the time series data on N entities is generated from a Gaussian mixture model with autocorrelations over k clusters in Rd. Our main contribution is an algorithm to construct coresets for the maximum likelihood objective for this mixture model. Our algorithm is efficient, and, under a mild assumption on the covariance matrices of the Gaussians, the size of the coreset is independent of the number of entities N and the number of observations for each entity, and depends only polynomially on k, d and 1/ε, where ε is the error parameter. We empirically assess the performance of our coresets with synthetic data.
Lookalike targeting is a widely used model-based ad targeting approach that uses a seed database of individuals to identify matching “lookalikes” for targeted customer acquisition. An advertiser has to make two key choices: (1) who to seed on and (2) seed-match rank range. First, we find that seeding on others’ journey stage can be effective in new customer acquisition; despite the cold start nature of customer acquisition using Lookalike audiences, third parties can indeed identify factors unobserved to the advertiser that move individuals along the journey and can be correlated with the lookalikes. Further, while journey-based seeding adds no incremental value for brand marketing (click-through), seeding on more downstream stages improves performance marketing (donation) outcomes. Second, we evaluate audience expansion strategies by lowering match ranks between the seed and lookalikes to increase acquisition reach. The drop in effectiveness with lower match rank range is much greater for performance marketing than for brand marketing. Performance marketers can alleviate the problem by making the ad targeting explicit, and thus increase perceived relevance; however, it has no incremental impact for higher match lookalikes. Increasing perceived targeting relevance makes acquisition cost comparable for both high and low match ranks.
Lookalike Targeting is a widely used model-based ad targeting approach that uses a seed database of individuals to identify matching “lookalikes” for targeted customer acquisition. An advertiser has to make two key choices: (1) who to seed on and (2) seed-match rank range. First, we assess if and how seeding by others’ journey stages impact clickthrough (upstream behavior desirable for brand marketing) and donation (downstream behavior desirable in performance marketing). Overall, we find that lookalike targeting using other’s journeys can be effective-third parties can indeed identify factors unobserved to the advertiser merely from others’ journey stage to improve targeting. Further, while it is sufficient to seed on upstream journey stages for brand marketing, seeding on more downstream stages improves performance marketing outcomes. Second, we assess the effectiveness of expanding the target audience with lower match ranks between seed and lookalikes. The drop in effectiveness with lower match rank range is much greater for performance marketing (donation) than for brand marketing (click-through). However, performance marketers can alleviate the reduction in ad effectiveness for low match ranks by making targeting more salient; but increasing salience has little impact for high match rank. Overall, by increasing salience, performance marketers can make acquisition cost comparable for high and low match ranks.
The paper broadens the focus of empirical research on salesforce management to include multitasking settings with multidimensional incentives, where salespeople have private information about customers. This allows us to ask novel substantive questions around multidimensional incentive design and job design while managing the costs and benefits of private information. To this end, the paper introduces the first structural model of a multitasking salesforce in response to multidimensional incentives. The model also accommodates (i) dynamic intertemporal tradeoffs in effort choice across the tasks and (ii) salesperson’s private information about customers. We apply our model in a rich empirical setting in microfinance and illustrate how to address various identification and estimation challenges. We extend two-step estimation methods used for unidimensional compensation plans by embedding a flexible machine learning (random forest) model in the first-stage multitasking policy function estimation within an iterative procedure that accounts for salesperson heterogeneity and private information. Estimates reveal two latent segments of salespeople- a “hunter” segment that is more efficient in loan acquisition and a “farmer” segment that is more efficient in loan collection. Counterfactuals reveal heterogeneous effects: hunters’ private information hurts the firm as they engage in adverse selection; farmers’ private information helps the firm as they use it to better collect loans. The payoff complementarity induced by multiplicative incentive aggregation softens adverse specialization by hunters relative to additive aggregation, but hurts performance among farmers. Overall, task specialization in job design for hunters (acquisition) and farmers (collection) hurts the firm as adverse selection harm overwhelms efficiency gain.
We develop the first structural model of a multitasking salesforce to address questions of job design and incentive compensation design. The model incorporates three novel features: (i) multitasking effort choice given a multidimensional incentive plan; (ii) salesperson’s private information about customers and (iii) dynamic intertemporal tradeoffs in effort choice across the tasks. The empirical application uses data from a micro nance bank where loan officers are jointly responsible and incentivized for both loan acquisition repayment but has broad relevance for salesforce management in CRM settings involving customer acquisition and retention. We extend two-step estimation methods used for unidimensional compensation plans for the multitasking model with private information and intertemporal incentives by combining flexible machine learning (random forest) for the inference of private information and the first-stage multitasking policy function estimation. Estimates reveal two latent segments of salespeople-a “hunter” segment that is more efficient in loan acquisition and a “farmer” segment that is more efficient in loan collection. We use counterfactuals to assess how (1) multi-tasking versus specialization in job design; (ii) performance combination across tasks (multiplicative versus additive); and (iii) job transfers that impact private information impact firm profits and specific segment behaviors.
Charities often send annual “thank you letters” to express gratitude to donors, but seek to defray these costs by inviting additional donations or engagement. However, the additional asks may backfire if potential donors see the thank you message as “insincere” or “manipulative.” We test this trade-off by conducting a field experiment in cooperation with a leading charity in India. We find that an explicit ask for additional donations or even a request to follow the organization on Facebook reduces giving. However, these effects are not only heterogeneous, but asymmetric by past giving behavior. Recent, frequent, and higher monetary value donors react negatively to additional asks by reducing giving, but lapsed, infrequent, and lower monetary value donors react positively by giving more. Our results highlight that findings based on purely cross-sectional experiments may offer incomplete insight. We estimate that differentially targeted ask messages based on past donation behavior, data readily available to charities, can increase donations overall by 6-11%.
Charities routinely send “thank you letters” and small gifts to express gratitude to donors but seek to defray these costs by making additional asks for donations and/or engagement. But the “ask for more” can backfire if potential donors perceive persuasive intent in the expression of gratitude, inducing reactance. We hypothesize that such reactance and its impact on giving will vary by donor loyalty. Loyal donors are more likely to experience reactance to additional asks, muting the feeling of reciprocity aroused by the expression of gratitude to suppress giving. In contrast, non-loyal donors are less likely to experience reactance, and therefore more likely to channel the feeling of reciprocity toward giving. We test our hypothesis using a large-scale natural field experiment involving nearly 180,000 past donors to a leading charity in India. We find evidence in support of our hypothesis. We therefore recommend that additional asks only be made to nonloyal donors. Such differentially targeted ask messages based on past donation behavior, using data readily available to charities, can increase overall donation amounts by 12.8-17.5%. Our findings highlight that purely cross-sectional experiments that do not account for past donor/customer history may offer incomplete insight and lead to erroneous managerial implications.
The authors address two novel and significant challenges in using online text reviews to obtain attribute level ratings. First, they introduce the problem of inferring attribute level sentiment from text data to the marketing literature and develop a deep learning model to address it. While extant bag of words based topic models are fairly good at attribute discovery based on frequency of word or phrase occurrences, associating sentiments to attributes requires exploiting the spatial and sequential structure of language. Second, they illustrate how to correct for attribute self-selection—reviewers choose the subset of attributes to write about—in metrics of attribute level restaurant performance. Using Yelp.com reviews for empirical illustration, they find that a hybrid deep learning (CNN-LSTM) model, where CNN and LSTM exploit the spatial and sequential structure of language respectively provide the best performance in accuracy, training speed and training data size requirements. The model does particularly well on the “hard” sentiment classification problems. Further, accounting for attribute self-selection significantly impacts sentiment scores, especially on attributes that are frequently missing.
A critical element of word of mouth (WOM) or buzz marketing is to identify seeds, often central actors with high degree in the social network. Seed identification typically requires data on the full network structure, which is often unavailable. We therefore examine the impact of WOM seeding strategies motivated by the friendship paradox to obtain more central nodes without knowing network structure. But higher-degree nodes may communicate less with neighbors; therefore whether friendship paradox motivated seeding strategies increase or reduce WOM and adoption remains an empirical question. We develop and estimate a model of WOM and adoption using data on microfinance adoption across 43 villages in India for which we have data on social networks. Counterfactuals show that the proposed seeding strategies are about 15-20% more effective than random seeding in increasing adoption. Remarkably, they are also about 5-11% more effective than opinion leader seeding, and are relative more effective when we have fewer seeds.
The authors address two significant challenges in using online text reviews to obtain fine-grained attribute level sentiment ratings. First, they develop a deep learning convolutional-LSTM hybrid model to account for language structure, in contrast to methods that rely on word frequency. The convolutional layer accounts for the spatial structure (adjacent word groups or phrases) and LSTM accounts for the sequential structure of language (sentiment distributed and modified across non-adjacent phrases). Second, they address the problem of missing attributes in text in construct-ing attribute sentiment scores—as reviewers write only about a subset of attributes and remain silent on others. They develop a model-based imputation strategy using a structural model of heterogeneous rating behavior. Using Yelp restaurant review data, they show superior accuracy in converting text to numerical attribute sentiment scores with their model. The structural model finds three reviewer segments with different motivations: status seeking, altruism/want voice, and need to vent/praise. Interestingly, our results show that reviewers write to inform and vent/praise, but not based on attribute importance. Our heterogeneous model-based imputation performs better than other common imputations; and importantly leads to managerially significant corrections in restaurant attribute ratings.
A critical element of word of mouth (WOM) or buzz marketing is to identify seeds, often central actors with high degree in the social network. Seed identification typically requires data on the full network structure, which is often unavailable. We therefore examine the impact of WOM seeding strategies motivated by the friendship paradox to obtain more central nodes without knowing network structure on adoption. Higher-degree nodes may be less effective as seeds if these nodes communicate less with neighbors or are less persuasive when they communicate; therefore whether friendship paradox motivated seeding strategies increase or reduce WOM and adoption remains an empirical question. We develop and estimate a model of WOM and adoption using data on microfinance adoption across 43 villages in India for which we have data on social networks. Counterfactuals show that the proposed seeding strategies are about 15-24% more effective in increasing adoption relative to random seeding. These strategies are also about 5-13% more effective than the firm’s leader seeding strategy, and are relatively more effective when we have fewer seeds.