Search

Due to the sheer number of available choices, online retailers frequently use tiered assortment to present products. In this case, groups of products are arranged across multiple pages or stages, and a customer clicks on "next" or "load more" to access them sequentially. Despite the prevalence of such assortments in practice, this topic has not received much attention in the existing literature. In this work, we focus on a sequential choice model that captures customers' behavior when product recommendations are presented in tiers. We analyze multiple variants of tiered assortment optimization (TAO) problems by imposing "no-duplication" and capacity constraints. For the offline setting involving known customers' preferences, we characterize the properties of the optimal tiered assortment and propose algorithms that improve the computational efficiency over the existing benchmarks. To the best of our knowledge, we are the first to study the online setting of the TAO problem. A unique characteristic of our online setting, absent from the one-tier MNL bandit, is partial feedback. Products in the lower priority tiers are not shown when a customer has purchased a product or has chosen to exit at an earlier tier. Such partial feedback, along with product interdependence across tiers, increases the learning complexity. For both the noncontextual and contextual problems, we propose online algorithms and quantify their respective regrets. Moreover, we are able to construct tighter uncertainty sets for model parameters in the contextual case and thus improve the performance. We demonstrate the efficacy of our proposed algorithms through numerical experiments. This paper was accepted by J. George Shanthikumar, data science. Supplemental Material: The data files and e-companion are available at https://doi.org/10.1287/mnsc.2023.4940. [ABSTRACT FROM AUTHOR]

Random coefficient multinomial logit models are widely used to estimate customer preferences from sales data. However, these estimation models can only allow for products with positive sales; this selection leads to highly biased estimates in long tail markets, that is, markets where many products have zero or low sales. Such markets are increasingly common in areas such as online retail and other online marketplaces. In this paper, we propose a two-stage estimator that uses machine learning to correct for this bias. Our method first uses deep learning to predict the market shares of all products, where the neural network's structure mirrors the random coefficient multinomial logit model's data generating process. In the second stage, we use the predictions of the first stage to reweight the observed shares in a way that corrects for the induced bias and maintains the causal interpretation of the structural model. We show that the estimated parameters are consistent in the number of markets. Our method performs well on simulated and real long tail data, producing accurate estimates of customer behavior. These improved estimates can subsequently be used to provide prescriptive policy recommendations on important managerial decisions such as pricing, assortment, and so on. This paper was accepted by Gabriel Weintraub, revenue management and market analytics. Supplemental Material: The data files are available at https://doi.org/10.1287/mnsc.2023.4893. [ABSTRACT FROM AUTHOR]

Through repeated interactions, firms today refine their understanding of individual users' preferences adaptively for personalization. In this paper, we use a continuous-time bandit model to analyze firms that recommend content to multihoming consumers, a representative setting for strategic learning of consumer preferences to maximize lifetime value. In both monopoly and duopoly settings, we compare a forward-looking recommendation algorithm that balances exploration and exploitation to a myopic algorithm that only maximizes the quality of the next recommendation. Our analysis shows that, compared with a monopoly, firms competing for users' attention focus more on exploitation than exploration. When users are impatient, competition decreases the return from developing a forward-looking algorithm. In contrast, development of a forward-looking algorithm may hurt users under monopoly but always benefits users under competition. Competing firms' decisions to invest in a forward-looking algorithm can create a prisoner's dilemma. Our results have implications for artificial intelligence adoption and for policy makers on the effect of market power on innovation and consumer welfare. This paper was accepted by Dmitri Kuksov, marketing. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2023.4722. [ABSTRACT FROM AUTHOR]

Advancements in targeting technology have allowed firms to engage in more precise targeting based on several aspects of consumers' preferences. Exposed to more targeted ads, consumers are becoming increasingly aware of being targeted and respond accordingly. This paper provides a theoretical analysis of multidimensional targeting under which consumers can draw inferences about multiple components of their utility from the advertised product. We show that the firm can be worse off under multidimensional targeting than under single-dimensional targeting, in which the firm targets consumers based only on a single component of their utility. This is because, with multidimensional targeting, targeted consumers may face greater uncertainty about on which specific dimension(s) they can expect to enjoy the advertised product. Therefore, they may be less willing to exert a costly effort of clicking the ad and making a purchase decision. When this result holds, the firm may want to adopt a single-dimensional targeting strategy. However, we show that the firm cannot credibly commit to such a strategy once given access to multiple dimensions of customer data. Interestingly, a higher unit cost of advertising can mitigate the firm's commitment problem for using customer data and thus increase the firm's profit. Moreover, the firm can sometimes lower the price to recover some of, but not entirely offset, the drawbacks of multidimensional targeting. We discuss the implications of our results regarding the current practice of targeted advertising and data privacy protection policies. This paper was accepted by Dmitri Kuksov, marketing. Funding: This work was supported by the Korea Advanced Institute of Science and Technology [Grant G04210017]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2022.4604. [ABSTRACT FROM AUTHOR]

Research in psychology shows that consumption goals can help consumers avoid excessive consumption of vice goods and the associated long-term harm. In this paper, we propose a model of self-control with consumption goals and examine how goals moderate the behavior of consumers and the firm's strategy. We find that consumers' personal goals lead to a lower price for a less unhealthy product but a higher price for a more unhealthy product. Furthermore, even though personal goals reduce the sales of a product, the firm can be better off if consumers have goals rather than no goals. The improvement in the firm's profits need not be at consumers' expense. In fact, consumer welfare increases with personal goals. In some contexts, consumption is not driven by personal goals but shaped by social norms, such as the advice of experts or social groups. We find that, unlike personal goals, normative goals make consumers less sensitive to price and do not always improve consumer welfare. Furthermore, normative goals can hurt the firm's profits in contexts in which personal goals could improve profits. Finally, we show that our framework with dynamically inconsistent preferences yields results that are consistent with alternative formulations of consumer self-control problems, such as the dual-self model of Thaler and Shefrin and the costly self-control model of Gul and Pesendorfer. This paper was accepted by Duncan Semester, marketing. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2022.4567. [ABSTRACT FROM AUTHOR]

This paper exploits a major food safety crisis to estimate a full demand model for the unsafe product and its substitutes, recovering consumers' preference parameters for different product characteristics. Counterfactual exercises quantify the relevance of different mechanisms driving consumers' responses, such as changes in safety perceptions, idiosyncratic taste, product characteristics, and price. We find that consumers' reactions are limited by their preferences for the product's observable and unobservable characteristics. Because of the costs associated with switching from the affected product, the decline in demand following a product-harm crisis tends to understate the true weight of such events in consumers' utility. We find that unobservable taste is a crucial driver of consumers' responses. Our counterfactual exercises illustrate that the demand would have declined further if consumers had had access to a closer substitute. For an accurate assessment of product-harm crises, managerial strategies should therefore account for how different demand drivers bind consumers' substitution patterns. This paper was accepted by Matthew Shum, marketing. Funding: This work was supported by the Ministerio de Ciencia e Innovación [Grants ECO2008-01116 and ECO2010-15052], the Ministerio de Economía y Competitividad [Grant ECO2013-43011-P], and the Deutsche Forschungsgemeinschaft [Grant CRC TR 224]. R. Ferrer received support from [Grant ECO2016-76998-P], and the Severo Ochoa Programme for Centers of Excellence in R&D [SEV-2015-0563]. [ABSTRACT FROM AUTHOR]

Existing research on word-of-mouth considers various descriptive statistics of rating distributions, such as the mean, variance, skewness, kurtosis, and even entropy and the Herfindahl-Hirschman index. But real-world consumer decisions are often derived from visual assessment of displayed rating distributions in the form of histograms. In this study, we argue that such distribution charts may inadvertently lead to a consumer-choice bias that we call the histogram distortion bias (HDB). We propose that salient features of distributions in visual decision making may mislead consumers and result in inferior decision making. In an illustrative model, we derive a measure of the HDB. We show that with the HDB, consumers may make choices that violate well-accepted decision rules. In a series of experiments, subjects are observed to prefer products with a higher HDB despite a lower average rating. They could also violate widely accepted modeling assumptions, such as branch independence and first-order stochastic dominance. This paper was accepted by Chris Forman, information systems. Funding: This work was supported by the National Natural Science Foundation of China [Grants 72131001 and 92146003] and the Research Grants Council, University Grants Committee [GRF 14500521, GRF 14501320, GRF 14503818, and TRS:T31-604/18-N]. Supplemental Material: Data and the online appendices are available at https://doi.org/10.1287/mnsc.2022.4306. [ABSTRACT FROM AUTHOR]

This paper investigates duopoly competition when horizontally differentiated firms are able to make personalized product-price offers to returning customers, within a behavior-based discrimination model. In the second period, firms can profile old customers according to their preferences, selling them targeted products at personalized prices. Product-price personalization (PP) allows firms to retain all old customers, eliminating second-period customer poaching. The overall profit effects of PP are shown to be ambiguous. In the second period, PP improves the matching between customers' preferences and firms' offers, but firms do not make any revenues in the rival's turf. In the Bertrand outcome, second-period profits only increase for both firms if the size of their old turfs are not too different or initial products are not too differentiated. However, the additional second-period profits may be offset by lower first-period profits. PP is likely to increase firms' overall discounted profits when consumers' (firms') discount factor is low (high) and firms' initial products are exogenous and sufficiently different. When the location of initial products is endogenous, profits are hurt because of an additional location (strategic) effect aggravating head-to-head competition in the first period. Likewise, when a fraction of active consumers conceals their identity, PP increases second-period profits at the cost of aggressive first-period price competition. Finally, we show that the room for profitable PP enlarges considerably if firms rely on PP as an effective device to sustain tacit collusive outcomes, with firms credibly threatening to respond to first-period price deviations with second-period aggressive relocations of their standard products. This paper was accepted by Matthew Shum, marketing. Funding: D. Laussel gratefully acknowledges support by the French National Research Agency [Grant ANR-17-EURE-0020] and the Excellence Initiative of Aix-Marseille University [Grant A*MIDEX]. J. Resende was supported by Portuguese public funds [Project Reference NORTE-01-0145-FEDER-028540] and through FCT (Fundação para a Ciência e a Tecnologia) [Project Reference POCI-01-0145-FEDER-006890]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2022.4298. [ABSTRACT FROM AUTHOR]

In this paper, we model a firm that can reduce its carbon footprint in the presence of a segment of eco-conscious consumers, who consider the product's carbon footprint when making purchasing decisions. The firm can reduce its controllable emissions at increased fixed and variable costs. The firm can also buy carbon offsets, at the price set by a nongovernmental organization (NGO), for both its controllable and uncontrollable emissions in the supply chain. We find that the firm should not use these two emission reduction methods as simple substitutes. In particular, as the offset price decreases, the firm may spend more efforts reducing its controllable emissions. Also, a firm's decision to buy carbon offsets depends on the correlation between consumers' preferences for the product function and for its environmental attributes, and this has implications for NGOs selling offsets. Specifically, although NGOs should price offsets as low as possible in most cases, we find some instances where a premium pricing strategy may be more effective in promoting lower-carbon footprint products, especially when eco-conscious consumers have a significantly higher valuation for the product than those who do not care about the environment, on average. This paper was accepted by Jayashankar Swaminathan, operations management. Supplemental Material: The data files and online appendix are available at https://doi.org/10.1287/mnsc.2021.4293. [ABSTRACT FROM AUTHOR]

This paper estimates the effects of time-varying consumer preference bias on trade flows and welfare. We use a unique data set from the BBC World Service Poll, which surveys (annually during 2005–2017 with some gaps) the populations of a wide array of countries on their views of whether an evaluated country is having a mainly positive or negative influence in the world. We identify the effects on consumer preference parameters due to shifts in these country image perceptions and quantify their general equilibrium effects on bilateral exports and welfare (each time for an evaluated exporting country, holding the exporting country's own preference parameters constant). We consider five important shifts in country image: the George W. Bush effect, the Donald Trump effect, the Senkaku Islands Dispute effect, the Brexit effect, and the Good-Boy Canadian effect. We find that such changes in bilateral country image perceptions have quantitatively important trade and welfare effects. The negative impact of Donald Trump's "America First" campaign rhetoric on the U.S.'s country image might have cost the United States 4%–5% of its total exports and welfare gains from trade. In contrast, the consistent improvement of Canada's country image between 2010 and 2017 has amounted to more than 8% of its total welfare gains from trade. This paper was accepted by Matthew Shum, marketing. [ABSTRACT FROM AUTHOR]

A decision maker (DM) must choose an action in order to maximize a reward function that depends on the DM's action as well as on an unknown parameter Θ. The DM can delay taking the action in order to experiment and gather additional information on Θ. We model the problem using a Bayesian sequential experimentation framework and use dynamic programming and diffusion-asymptotic analysis to solve it. For that, we consider environments in which the average number of experiments that is conducted per unit of time is large and the informativeness of each individual experiment is low. Under such regimes, we derive a diffusion approximation for the sequential experimentation problem, which provides a number of important insights about the nature of the problem and its solution. First, it reveals that the problems of (i) selecting the optimal sequence of experiments to use and (ii) deciding the optimal time when to stop experimenting decouple and can be solved independently. Second, it shows that an optimal experimentation policy is one that chooses the experiment that maximizes the instantaneous volatility of the belief process. Third, the diffusion approximation provides a more mathematically malleable formulation that we can solve in closed form and suggests efficient heuristics for the nonasympototic regime. Our solution method also shows that the complexity of the problem grows only quadratically with the cardinality of the set of actions from which the decision maker can choose. We illustrate our methodology and results using a concrete application in the context of assortment selection and new product introduction. Specifically, we study the problem of a seller who wants to select an optimal assortment of products to launch into the marketplace and is uncertain about consumers' preferences. Motivated by emerging practices in e-commerce, we assume that the seller is able to use a crowd voting system to learn these preferences before a final assortment decision is made. In this context, we undertake an extensive numerical analysis to assess the value of learning and demonstrate the effectiveness and robustness of the heuristics derived from the diffusion approximation. This paper was accepted by Omar Besbes, revenue management and market analytics. [ABSTRACT FROM AUTHOR]

We develop a data-driven approach for the multiproduct pricing problem, using the theory of a representative consumer in discrete choice. We establish a set of mathematical relationships between product prices and demand for each product in the system, including that of the outside option. We provide identification conditions to recover the underlying representative consumer model and show that, with sufficient pricing experiments, the approach can identify the underlying demand model (more precisely, the associated perturbation function in the representative consumer model) accurately up to a constant shift and a given tolerance level. This holds even when the demand data obtained are a noisy realization of the theoretical demand. We use this approach to solve the multiproduct pricing problem using a (mixed integer) linear optimization method. Extensive tests using both synthetic and industry data clearly demonstrate the benefits of this approach, which addresses the issue of model misspecification in traditional pricing methods using discrete choice models and circumvents the computational issues associated with pricing methods that assume a known consumer valuation of each product. This paper was accepted by David Simchi-Levi, big data analytics. [ABSTRACT FROM AUTHOR]

Integrating inventory and assortment planning decisions is a challenging task that requires comparing the value of demand expansion through broader choice for consumers with the value of higher in-stock availability. We develop a stockout-based substitution model for trading off these values in a setting with inventory replenishment, a feature missing in the literature. Using the closed form solution for the single-product case, we develop an accurate approximation for the multiproduct case. This approximated formulation allows us to optimize inventory decisions by solving a fractional integer program with a fixed point equation constraint. When products have equal margins, we solve the integer program exactly by bisection over a one-dimensional parameter. In contrast, when products have different margins, we propose a fractional relaxation that we can also solve by bisection and that results in near-optimal solutions. Overall, our approach provides solutions within 0.1% of the optimal policy and finds the optimal solution in 80% of the random instances we generate. This paper was accepted by David Simchi-Levi, optimization. [ABSTRACT FROM AUTHOR]

Choice-based demand estimation is a fundamental task in retail operations and revenue management, providing necessary input data for inventory control, assortment, and price-optimization models. The task is particularly difficult in operational contexts where product availability varies over time and customers may substitute into the available options. In addition to the classical multinomial logit (MNL) model and extensions (e.g., nested logit, mixed logit, and latent-class MNL), new demand models have been proposed (e.g., the Markov chain model), and others have been recently revisited (e.g., the rank list-based and exponomial models). At the same time, new computational approaches were developed to ease the estimation function (e.g., column-generation and expectation-maximization (EM) algorithms). In this paper, we conduct a systematic, empirical study of different choice-based demand models and estimation algorithms, including both maximum-likelihood and least-squares criteria. Through an exhaustive set of numerical experiments on synthetic, semisynthetic, and real data, we provide comparative statistics of the predictive power and derived revenue performance of an ample collection of choice models and characterize operational environments suitable for different model/estimation implementations. We also provide a survey of all the discrete choice models evaluated and share all our estimation codes and data sets as part of the online appendix. This paper was accepted by Vishal Gaur, operations management. [ABSTRACT FROM AUTHOR]

On online platforms, consumers face an abundance of options that are displayed in the form of a position ranking. Only products placed in the first few positions are readily accessible to the consumer, and she needs to exert effort to access more options. For such platforms, we develop a two-stage sequential search model where, in the first stage, the consumer sequentially screens positions to observe the preference weight of the products placed in them and forms a consideration set. In the second stage, she observes the additional idiosyncratic utility that she can derive from each product and chooses the highest-utility product within her consideration set. For this model, we first characterize the optimal sequential search policy of a welfare-maximizing consumer. We then study how platforms with different objectives should rank products. We focus on two objectives: (i) maximizing the platform's market share and (ii) maximizing the consumer's welfare. Somewhat surprisingly, we show that ranking products in decreasing order of their preference weights does not necessarily maximize market share or consumer welfare. Such a ranking may shorten the consumer's consideration set due to the externality effect of high-positioned products on low-positioned ones, leading to insufficient screening. We then show that both problems—maximizing market share and maximizing consumer welfare—are NP-complete. We develop novel near-optimal polynomial-time ranking algorithms for each objective. Further, we show that, even though ranking products in decreasing order of their preference weights is suboptimal, such a ranking enjoys strong performance guarantees for both objectives. We complement our theoretical developments with numerical studies using synthetic data, in which we show (1) that heuristic versions of our algorithms that do not rely on model primitives perform well and (2) that our model can be effectively estimated using a maximum likelihood estimator. This paper was accepted by Gabriel Weintraub, revenue management and market analytics. [ABSTRACT FROM AUTHOR]

Personalized marketing in retail requires a model to predict how different marketing actions affect product choices by individual customers. Large retailers often handle millions of transactions daily, involving thousands of products in hundreds of categories. Product choice models thus need to scale to large product assortments and customer bases, without extensive product attribute information. To address these challenges, we propose a custom deep neural network model. The model incorporates bottleneck layers to encode cross-product relationships, calibrates time-series filters to capture purchase dynamics for products with different interpurchase times, and relies on weight sharing between the products to improve convergence and scale to large assortments. The model applies to loyalty card transaction data without predefined categories or product attributes to predict customer-specific purchase probabilities in response to marketing actions. In a simulation, the proposed product choice model predicts purchase decisions better than baseline methods by adjusting the predicted probabilities for the effects of recent purchases and price discounts. The improved predictions lead to substantially higher revenue gains in a simulated coupon personalization problem. We verify predictive performance using transaction data from a large retailer with experimental variation in price discounts. This paper was accepted by Gui Liberali, Management Science Special Section on Data-Driven Prescriptive Analytics. [ABSTRACT FROM AUTHOR]

We consider a logit model-based framework for modeling joint pricing and assortment decisions that take into account customer features. This model provides a significant advantage when one has insufficient data for any one customer and wishes to generalize learning about one customer's preferences to the population. Under this model, we study the statistical learning task of model fitting from a static store of precollected customer data. This setting, in contrast to the popular learning and earning paradigm, represents the situation many business teams encounter in which their data collection abilities have outstripped their data analysis capabilities. In this learning setting, we establish finite-sample convergence guarantees on the model parameters. The parameter convergence guarantees are then extended to out-of-sample performance guarantees in terms of revenue, in the form of a high-probability bound on the gap between the expected revenue of the best action taken under the estimated parameters and the revenue generated by a decision maker with full knowledge of the choice model. We further discuss practical implications of these bounds. We demonstrate the personalization approach using ticket purchase data from an airline carrier. This paper was accepted by J. George Shanthikumar, Management Science Special Section on Data-Driven Prescriptive Analytics. [ABSTRACT FROM AUTHOR]

We consider dynamic repositioning when competing firms try to follow the evolution of consumer preferences while taking into account the competitive interaction, both in terms of static market competition and the dynamic effects of different firm positionings. We fully characterize the dynamic market equilibrium, which includes the timing of the firms' repositionings depending on consumer preferences. As consumer preferences evolve away from where both firms are located, one firm first moves to follow consumer preferences, with the second firm only moving if the consumer preferences continue evolving away from that firm. The model predicts rich market dynamics, where firms stay for some period in different positionings if consumer preferences are in a relatively middle ground or where a firm repositions to follow consumer preferences but then repositions back to the original position if consumer preferences return. We find that, when the variability of the consumer preferences or the discount rate is greater or when the importance of the repositioning attribute is smaller, firms are less likely to follow consumer preferences. Firms are more heterogeneous in their responses, which leads to longer periods of differentiation when the variability of the consumer preferences, the discount rate, or the importance of the repositioning attribute increases. We also find that competing firms reposition less frequently than what is socially optimal and what collusion would imply, and we find more differentiation under collusion than under competition. This paper was accepted by Matthew Shum, marketing. [ABSTRACT FROM AUTHOR]

Reviews for products and services written by previous consumers have become an influential input to the purchase decision of customers. Many service businesses monitor the reviews closely for feedback as well as detecting service flaws, and they have become part of the performance review for service managers with rewards tied to improvement in the aggregate rating. Many empirical papers have documented a bias in the aggregate ratings, arising because of customers' inherent self-selection in their choices and bounded rationality in evaluating previous reviews. Although there is a vast empirical literature analyzing reviews, theoretical models that try to isolate and explain the bias in ratings are relatively few. Assuming consumers simply substitute the average rating that they see as a proxy for quality, we give a precise characterization of the self-selection bias on ratings of an assortment of products when consumers confound ex ante innate preferences for a product or service with ex post experience and service quality and do not separate the two. We develop a parsimonious choice model for consumer purchase decisions and show that the mechanism leads to an upward bias, which is more pronounced for niche products. Based on our theoretical characterization, we study the effect on pricing and assortment decisions of the firm when potential customers purchase based on the biased ratings. Our results give insights into how quality, prices, and customer feedback are intricately tied together for service firms. This paper was accepted by David Simchi-Levi, operations management. [ABSTRACT FROM AUTHOR]

We study how user-generated content (UGC) about new products impacts a firm's advertising and pricing decisions and the effect on profits and market dynamics. We construct a two-period model where consumers value quality and are heterogeneous in their taste for the new product's positioning and examine three information-transfer structures across generations: no information (benchmark), average rating (AR), and the joint distribution of ratings and taste locations (reviews). First, we show that in the AR case, the firm advertises to a smaller set of consumers and prices higher relative to the other cases. This occurs because the firm has an incentive to use advertising strategically to bump up the first-generation average rating in order to increase second-generation quality perceptions. The firm charges a higher first-period price as consumers who are relatively close to the product's location are advertised to. Second, we find that as more information is transferred across generations, there is a greater likelihood that average ratings will exhibit an increasing pattern over time. This happens because with reviews, second-generation consumers are able to base purchase decisions on the product's positioning in addition to its quality. Interestingly, because of the firm's narrower advertising strategy in the AR case and the richer UGC in the reviews case, average ratings will exhibit a reversal over time: higher in the AR case in the first period but higher in the reviews case in the second period. Third, a firm's expected profits can exhibit a nonmonotonic relationship with respect to the amount of UGC transferred: highest in the AR case if the marginal return on advertising is high (e.g., when advertising is cheap to execute, consumers are insensitive to product fit or consumers greatly value quality) but highest in the reviews case when the return on advertising is low. We then examine a setup whereby second-generation consumers become aware of the product through social interaction with first-generation buyers. We find that advertising and social contagion interact: the firm generally has a greater incentive to advertise as homophily increases, that is, as social interactions occur between consumers who more likely share similar preferences. However, this pattern can reverse with reviews. This paper discusses several robustness checks and model extensions and concludes by highlighting the managerial implications of the results. This paper was accepted by Matthew Shum, marketing. [ABSTRACT FROM AUTHOR]

We investigate the economic and environmental implications of jointly implementing leasing and modularity, two prominent circular economy strategies, and their interaction. We develop a durable goods model to obtain insights for firms considering the use of these strategies. We first analyze the effect of adopting leasing on a firm's product architecture choice. We find that a modular product architecture is more attractive under leasing only if off-lease products depreciate to a greater extent. We next analyze the effect of adopting a modular product architecture on the firm's business model choice. We find that adopting a modular architecture can cause a leasing firm to switch to selling but will not lead a selling firm to switch to leasing. Building on these results, we show that leasing a modular product is preferred by the firm only under limited conditions (i.e., when production costs are low, off-lease products are in relatively better condition, and when there is a large difference between depreciation levels of different modules). Finally, we show that jointly implementing leasing and modularity can also lead to higher environmental impact than only adopting leasing or modularity in isolation. This paper was accepted by Vishal Gaur, operations management. [ABSTRACT FROM AUTHOR]

When preferences are sensitive to context, firms may influence purchase decisions by designing the environment of consumption choices. This paper studies how competitive retailers optimally design their product line if preferences at the store depend on whether the choice set draws consumer attention to the quality or price of a product. Before making a purchase decision, a consumer chooses among retailers without (fully) anticipating that her preferences at the store are malleable. We show that this setup can align evidence on retailer marketing practices and the existence of loss leaders with the experimental literature on decoy effects: in equilibrium, retailers use loss leaders to attract the consumer to the store and decoy products to draw consumer attention at the store to more profitable alternatives featuring a higher price (inducing a profitable upsell) or lower production cost (inducing a profitable downsell). We embed and compare the predictions of three recent specifications of choice-set dependent attention—namely, Salience, Focusing, and Relative Thinking. All three specifications predict that firms construct choice environments that utilize decoy effects to up- or downsell the consumer. Because they differ in how attention can be directed using the product line, however, they differ in their predictions of how the decoy will be positioned in the price-quality space. This paper was accepted by Yan Chen, decision analysis. [ABSTRACT FROM AUTHOR]

We study customer preference for the bike-share system in the city of London. We estimate a structural demand model on the station network to learn the preference parameters and use the estimated model to provide insights on the design and expansion of the bike-share system. We highlight the importance of network effects in understanding customer demand and evaluating expansion strategies of transportation networks. In the particular example of the London bike-share system, we find that allocating resources to some areas of the station network can be 10 times more beneficial than others in terms of system usage and that the currently implemented station density rule is far from optimal. We develop a new method to deal with the endogeneity problem of the choice set in estimating demand for network products. Our method can be applied to other settings in which the available set of products or services depends on demand. This paper was accepted by Gabriel Weintraub, revenue management and market analytics. [ABSTRACT FROM AUTHOR]

We use detailed clickstream data on online word-of-mouth (WOM) to uncover mechanisms underlying its influence on consumer decision making. A feature launch on a major doctor appointment booking platform allows us to examine the effects of online WOM on three dimensions of a consumer's choice process: the consideration set size, the time taken to consider alternatives (web session duration), and the geographic dispersion of the choices considered. Results indicate that the effects of WOM on decision-making processes are not monotonic but rather are contingent on the abundance of WOM (number of rated doctors) in a market. When the abundance of WOM is high, the introduction of WOM makes patients consider fewer doctors, browse for a shorter duration, and focus on doctors that are geographically more proximate. In contrast, when the abundance of WOM is low, the introduction of WOM makes patients consider more doctors, browse for longer duration, and consider doctors that are geographically more dispersed. We also find that WOM can lead to a cannibalization effect: when ratings are published, the highly rated doctors reap the benefits (in the form of increased demand) at the expense of unrated doctors. Our study contributes to the extant literature on online WOM by providing new insights into how WOM influences consumer decision making and by examining this question at a more granular level than prior work. This paper was accepted by Anandhi Bharadwaj, information systems. [ABSTRACT FROM AUTHOR]

This paper investigates the impact of customer compatibility—the degree of fit between customers' needs and the capabilities of the operations serving them—on customer experiences and firm performance. We use variance decomposition analysis to quantify the relative importance of customer, employee, process, location, and market-level effects on customer satisfaction. In our models, which explain roughly a quarter of the aggregate variance, differences among customers account for 96%–97% of the explainable portion. Further analysis of interaction-level data from banking and quick-service restaurants reveals that customers report relatively consistent satisfaction across transactions with particular firms but that some customers are habitually more satisfied than others. A second set of empirical studies provides evidence that these customer-level differences are explained in part by customer compatibility. Customers whose needs, proxied by differences in demographics and product choices, diverge more starkly from those of their bank's average customers report significantly lower levels of satisfaction. Consistently, banks that serve customer bases with more dispersed needs receive lower satisfaction scores than banks serving customer bases with less dispersed needs. Finally, a longitudinal analysis of the deposit and loan growth of all federally insured banks in the United States from 2006 to 2017 reveals that customer compatibility affects a firm's financial performance. Branches with more divergent customers grow more slowly than branches with less divergent customers. Institutions serving customer bases with more dispersed needs have branches that exhibit slower growth than those of institutions serving customer bases with less dispersed needs. This paper was accepted by Vishal Gaur, operations management. [ABSTRACT FROM AUTHOR]

Amid growing calls for transparency and social and environmental responsibility, companies are employing different strategies to improve consumer perceptions of their brands. Some pursue internal initiatives that reduce their negative social or environmental impacts through responsible operations practices (such as paying a living wage to workers or engaging in environmentally sustainable manufacturing). Others pursue external responsibility initiatives (such as philanthropy or cause-related marketing). Through two experiments conducted in the field and complementary online experiments, we compare how transparency into these internal and external initiatives affects customer perceptions and sales. We find that transparency into both internal and external responsibility initiatives tends to dominate generic brand marketing in motivating consumer purchases, supporting the view that consumers take companies' responsibility efforts into account in their decision making. Furthermore, the results provide converging evidence that transparency into a company's internal responsibility practices can be at least as motivating of consumer sales as transparency into its external responsibility initiatives, incrementally increasing a consumer's probability of purchase by 6.40% and 45.85% across our two field experiments, conducted in social and environmental domains, respectively. Our results suggest that it may be in the interest of both business and society for managers to prioritize internal responsible operations initiatives to achieve both top- and bottom-line benefits while mitigating social and environmental harms. This paper was accepted by Charles Corbett, operations management. [ABSTRACT FROM AUTHOR]

We study the role of ad networks in the online advertising market. Our baseline model considers two publishers that can outsource the sale of their ad inventories to an ad network, in a market where consumers and advertisers multi-home. The ad network increases total advertising revenue by tracking consumers across outlets and reduces competition between publishers by centralizing the sale of ads. Consequently, outsourcing to the ad network benefits the publishers, but may penalize the advertisers. We show that the ad network's ability to track consumers may either expand or reduce the provision of ads, depending on consumers' preferences for the publishers and how advertisers use tracking information. Specifically, tracking is more likely to expand (respectively, reduce) the provision of ads when consumers' preferences for the publishers are positively (respectively, negatively) correlated. Tracking is also more likely to expand (respectively, reduce) the provision of ads when advertisers use tracking information to cap the frequency of impressions (respectively, target specific consumers). Furthermore, we study the implications of consumers' choice to block tracking. Generally, blocking negatively impacts the advertising industry by making ad allocation less effective. Blocking also entails an externality on consumers, which is negative when tracking reduces the provision of ads. Given these conditions, regulatory restrictions on tracking may reduce consumer surplus as well as advertising revenue. These findings contrast with the presumption that regulation should make it easier for consumers to avoid tracking. We propose further extensions, including competing ad networks, more than two publishers, and networks that do not sell ads, but only tracking information to the advertisers. This paper was accepted by Juanjuan Zhang, marketing. [ABSTRACT FROM AUTHOR]

We consider the use and value of social-network information in selectively selling goods and services whose value derives from exclusive ownership among network connections or friends. Our model accommodates customers who are heterogeneous in their number of friends (degree) and their proclivity for social comparisons (conspicuity). Firms with information on either (or both) of these characteristics can use it to make the product selectively available and to personalize prices. We find that firms' preferred customers are low degree and high conspicuity, with the conspicuity threshold nondecreasing in degree. Interestingly, although both degree and conspicuity levels are relevant to curating the desired customer base, we find that firms do not need conspicuity information to do so; its absence can be substituted by incentivizing customers to self-select. There is no such recourse for the absence of degree information. As a result, degree information is typically more valuable than conspicuity information. Our analysis suggests that there are two canonical categories of social information—less valuable "consonant" information on characteristics where firm and customer preferences are aligned and more valuable "competing" information where preferences are misaligned. Customers can be incentivized to act in a way that their actions are a perfect substitute for consonant information, making it less valuable. This paper was accepted by Gad Allon, operations management. [ABSTRACT FROM AUTHOR]

We develop a model in which reward-based crowdfunding enables firms to obtain a reliable proof of concept early in their production cycle: they learn about total demand from a limited sample of target consumers preordering a new product. Learning from the crowdfunding sample creates a valuable real option because firms invest only if updated expectations about total demand are sufficiently high. This is particularly valuable for firms facing a high degree of uncertainty about consumer preferences, such as developers of innovative consumer products. Learning also enables firms to overcome moral hazard. The higher the funds raised, the lower the firms' incentives to divert them, provided third-party platforms limit the sample size by restricting campaign length. Although the probability of campaign success decreases with sample size, the expected funds raised are maximized at an intermediate sample size. Our results are consistent with stylized facts and lead to new empirical implications. This paper was accepted by Gustavo Manso, finance. [ABSTRACT FROM AUTHOR]

We study the contagious switching behavior related to a consumer's choice of wireless carriers, that is, that a consumer is more likely to switch wireless carriers if more of their contacts from the same carrier have switched. Contagious switching (or a positive network effect) can be driven by information-based social learning, as well as other mechanisms related to network size. Although previous marketing literature has documented the social-learning effect, most of the applications studied involve products in which consumers usually do not enjoy any direct benefits from a large network other than from information-based social learning. We explore the importance of the social-learning effect relative to other mechanisms that may also lead to the network effect. We propose a dynamic structural model with interpersonal interactions. To model the social-learning effect, a consumer uses feedback from his or her contacts who have switched from a focal carrier to update his or her quality expectations of alternative carriers. Our model further accounts for two unique aspects of consumer strategic learning: (i) the individual's perception on the signal of alternative carriers from contacts who switch is systematically different according to whether the signal comes from a loyal contact; and (ii) that the perceived noisiness of the signal on alternative carriers from a contact who has switched depends on the strength of the relationship between the individual and the contact. The remaining network effect not captured through social learning is modeled as a function of the size of the network. We solve the model with a two-step dynamic programming algorithm, with the assumption that a consumer is forward-looking and decides whether to stay with the same service carrier in each period by maximizing the total utility received from that day onward. We apply the proposed model to the data set of a mobile network operator in a European country. We find that churning/switching behavior is contagious in the network context and that one-third of general network effects can be attributed to social learning. We also detect strategic learning by consumers from their contacts in two ways: the experience signal on alternative carriers from a more loyal contact who has switched from the focal carrier is perceived to be more positive than that from a less loyal contact; and the social-learning effect is stronger from an individual's closest contacts. The simulation analysis demonstrates the value of our model in helping a company prioritize its customer relationship management effort. This paper was accepted by Matthew Shum, marketing. [ABSTRACT FROM AUTHOR]

Motivated by the increasing use of online appointment booking platforms, we study how to offer appointment slots to customers to maximize the total number of slots booked. We develop two models, nonsequential offering and sequential offering, to capture different types of interactions between customers and the scheduling system. In these two models, the scheduler offers either a single set of appointment slots for the arriving customer to choose from or multiple sets in sequence, respectively. For the nonsequential model, we identify a static randomized policy, which is asymptotically optimal when the system demand and capacity increase simultaneously, and we further show that offering all available slots at all times has a constant factor of two performance guarantee. For the sequential model, we derive a closed form optimal policy for a large class of instances and develop a simple, effective heuristic for those instances without an explicit optimal policy. By comparing these two models, our study generates useful operational insights for improving the current appointment booking processes. In particular, our analysis reveals an interesting equivalence between the sequential offering model and the nonsequential offering model with perfect customer preference information. This equivalence allows us to apply sequential offering in a wide range of interactive scheduling contexts. Our extensive numerical study shows that sequential offering can significantly improve the slot fill rate (6%–8% on average and up to 18% in our testing cases) compared with nonsequential offering. Given the recent and ongoing growth of online and mobile appointment booking platforms, our research findings can be particularly useful to inform user interface design of these booking platforms. [ABSTRACT FROM AUTHOR]

Sometimes we desire change, a break from the same, or an opportunity to fulfill different aspects of our needs. Noting that consumers seek variety, several approaches have been developed to diversify items recommended by personalized recommender systems. However, current diversification strategies operate under a one-shot paradigm without considering the evolution of preferences resulting from recent consumption. Therefore, such methods often sacrifice accuracy. In the context of online media, we show that by recognizing that consumption in a session is the result of a sequence of utility-maximizing selections from various categories, one can increase recommendation accuracy by dynamically tailoring the diversity of suggested items to the diversity sought by the consumer. Our approach is based on a multicategory utility model that captures a consumer's preference for different categories of content, how quickly the consumer satiates with one category and wishes to substitute it with another, and how the consumer trades off costly search efforts with selecting from a recommended list to discover new content. Taken together, these three elements allow us to characterize how an individual selects a diverse set of items to consume over the course of a session and how likely the individual is to click on recommended content. We estimate the model using a clickstream data set from a large media outlet and apply it to determine the most relevant content to recommend at different stages of an online session. We find that our approach generates recommendations that are on average about 10% more accurate than optimized alternatives and about 25% more accurate than those diversified using existing diversification strategies. Moreover, the proposed method recommends content with diversity that more closely matches the diversity sought by readers, exhibiting lower concentration–diversification bias than other personalized recommender systems. Using a policy simulation, we estimate that recommending content using the proposed approach would result in visitors reading 23% additional articles at the studied website and deriving 35% higher utility. This could lead to immediate gains in revenue for the publisher and longer-term improvements in customer satisfaction and retention at the site. [ABSTRACT FROM AUTHOR]

The mere preannouncement of a new product can affect consumer choice, thus complicating preannouncement strategy. This is because a preannounced product that is unavailable immediately can still be one of the alternatives in a consumer's mind at the time of choice. Such unavailable products, also known as phantom products, influence the reference point that consumers compare alternatives to when making a choice, as has been widely demonstrated in experimental studies. Thus, in addition to encouraging consumers to postpone purchase in favor of a future product, preannouncement also changes their preference for the currently available products when consumers do not prefer to postpone. In this paper we explore preannouncement strategy by analyzing a model that incorporates the effect of new product preannouncement (NPP) on consumer preferences and compare the results with a benchmark case in which consumer preferences across the existing products are not influenced by preannouncement. We find that when we take into account the effect of NPP on consumer preferences across the existing products, although postponement of purchase by some consumers remains beneficial, the preference for the current product offering with a lower quality can suffer so much that the significant lowering of current profits is not offset by future gains. Thus, preannouncement may no longer be the optimal strategy for the firm with a lower-quality product, which in turn explains the "Osborne effect." Our results also challenge the conventional wisdom in new product preannouncement literature. [ABSTRACT FROM AUTHOR]

We study strategic communication between a customer and an advisor who is privately informed about the most suitable choice for the customer but whose preferences are misaligned with the customer's preferences. The advisor sends a message to the customer who, in turn, can secure herself from bad advice by acquiring costly information on her own. In our experiments, we find that making the customer's information acquisition less costly leads to less prosocial behavior of the advisor. This can be explained by a model of shared guilt, which predicts a shift in causal attribution of guilt from the advisor to the customer if the latter could have avoided her ex post disappointment. We conclude that providing better access to information through, for example, consumer protection regulation or digital information aggregation and dissemination, may have unintended negative consequences on peoples' willingness to take responsibility for each other. Data files and the online appendices are available at https://doi.org/10.1287/mnsc.2018.3101. [ABSTRACT FROM AUTHOR]

Healthy eating goals influence many consumer choices, such that evaluating the healthiness of food portions is important. Given that both the type and quantity of food jointly contribute to weight and overall health, evaluations of a food portion's healthiness ought to consider both type and quantity. However, existing literature tends to examine food type and food quantity separately. Across seven studies, we show that consumers treat type as a primary dimension and quantity as a secondary dimension, such that a change in type (versus quantity) has a greater impact on perceived healthiness or health goal impact, even when holding objective impact constant in terms of calories. We also examine whether one reason this effect occurs is because most consumers consider type (a categorical attribute) before quantity (a continuous attribute). We conclude by discussing extensions of these ideas to other perceptual assessments involving both type and quantity (e.g., price perceptions). The online appendix and data files are available at https://doi.org/10.1287/mnsc.2018.3098. [ABSTRACT FROM AUTHOR]

Product and content personalization is now ubiquitous in e-commerce. There are typically not enough available transactional data for this task. As such, companies today seek to use a variety of information on the interactions between a product and a customer to drive personalization decisions. We formalize this problem as one of recovering a large-scale matrix with side information in the form of additional matrices of conforming dimension. Viewing the matrix we seek to recover and the side information we have as slices of a tensor, we consider the problem of slicerecovery, which is to recover specific slices of "simple" tensors from noisy observations of the entire tensor. We propose a definition of simplicity that on the one hand elegantly generalizes a standard generative model for our motivating problem and on the other hand subsumes low-rank tensors for a variety of existing definitions of tensor rank. We provide an efficient algorithm for slice recovery that is practical for massive data sets and provides a significant performance improvement over state-of-the-art incumbent approaches to tensor recovery. Furthermore, we establish near-optimal recovery guarantees that, in an important regime, represent an order improvement over the best available results for this problem. Experiments on data from a music streaming service demonstrate the performance and scalability of our algorithm. The e-companion is available at https://doi.org/10.1287/mnsc.2018.3092. [ABSTRACT FROM AUTHOR]

Customer preferences may not be rational, so we focus on quantifying the limit of rationality (LoR) in choice modeling applications. We define LoR as the cost of approximating the observed choice fractions from a collection of offer sets with those from the best-fitting probability distribution over rankings. Computing LoR is intractable in the worst case. To deal with this challenge, we introduce two new concepts, rational separation and choice graph, through which we reduce the problem to solving a dynamic program on the choice graph and express the computational complexity in terms of the structural properties of the graph. By exploiting the graph structure, we provide practical methods to compute LoR efficiently for a large class of applications. We apply our methods to real-world grocery sales data and identify product categories for which going beyond rational choice models is necessary to obtain an acceptable performance. The online appendix is available at https://doi.org/10.1287/mnsc.2018.3030. This paper was accepted by Vishal Gaur, operations management. [ABSTRACT FROM AUTHOR]

We present a lab-field experiment designed to systematically assess the external validity of social preferences elicited in a variety of experimental games. We do this by comparing behavior in the different games with several behaviors elicited in the field and with self-reported behaviors exhibited in the past, using the same sample of participants. Our results show that the experimental social preference games do a poor job explaining both social behaviors in the field and social behaviors from the past. We also include a systematic review and meta-analysis of previous literature on the external validity of social preference games. Data are available at https://doi.org/10.1287/mnsc.2017.2908. This paper was accepted by John List, behavioral economics. [ABSTRACT FROM AUTHOR]

It is now well understood that social media plays an increasingly important role in consumers' decision making. However, an overload of social media content in product search engines can hinder consumers from efficiently seeking information. We propose a structural econometric model to understand consumers' preferences and costs on search engines to improve user experience under unstructured social media. Our model combines an optimal stopping framework with an individual-level random utility choice model and analyzes click behavior in conjunction with purchase choices. Our model accounts for three major constraints in a consumer's decision-making process: (1) interdependency in decision making for different alternatives, (2) sequential arrival of information revealed by click-throughs, and (3) nonnegligible search cost. Our approach allows us to jointly estimate consumers' heterogeneous preferences and search costs under the interplay of social media and search engines, and to predict search and purchase behavior for each consumer. We validate the model using an individual session-level data set of approximately seven million observations resulting in room bookings in 2,117 U.S. hotels. Interestingly, our analysis allows us to quantify the trade-off between consumers' benefits and cognitive costs from using large-scale unstructured social media information during decision making. Our policy experiments show that providing a carefully curated digest of social media content during the earlier stages of consumer search (i.e., on the search results summary page) can lead to a 12.01% increase in the overall search engine revenue. The online appendix is available at https://doi.org/10.1287/mnsc.2017.2991. This paper was accepted by Anandhi Bharadwaj, information systems. [ABSTRACT FROM AUTHOR]

Online product ratings offer information on product quality. Scholars have recently proposed the potential of designing multidimensional rating systems to better convey information on multiple dimensions of products. This study investigates whether and how multidimensional rating systems affect consumer satisfaction (measured by product ratings), based on both observational data and two randomized experiments. Our identification strategy of the observational study hinges on a natural experiment on TripAdvisor when the website started to allow consumers to rate multiple dimensions of the restaurants, as opposed to only providing an overall rating, in January 2009. We further obtain rating data on the same set of restaurants from Yelp, which controls for the unobserved restaurant quality over time and allows us to identify the causal effect using a difference-in-differences approach. Results from the econometric analyses show that ratings in a single-dimensional rating system have a downward trend and a higher dispersion, whereas ratings in a multidimensional rating system are significantly higher and convergent. Findings from two randomized experiments suggest that the multidimensional rating system helps consumers find products that better fit their preferences and increases the confidence of their choices. We also show that the observed results cannot be explained by the priming effect due to rating system interface or a list of other alternative explanations. The combined evidence from the natural experiment and randomized experiments support the view that the multidimensional rating system enhances rating informativeness and provide implications for designing online rating systems that help consumers match their preferences with product attributes. Data and the online appendix are available at https://doi.org/10.1287/mnsc.2017.2852. This paper was accepted by Anandhi Bharadwaj, information systems. [ABSTRACT FROM AUTHOR]

Firms employ various techniques to obtain information about consumer taste/location and valuation prior to making product design decisions. User-generated content has become an important information source. The vast variety and volume of user-generated content makes firms better informed about consumers (precision-improving effect), and the common and public nature of user-generated content makes firms' information more correlated (correlation-increasing effect). We examine the impact of user-generated content in a setting in which two competing firms that are uncertain about consumer location or valuation design and sell horizontally differentiated products. We find that user-generated content has very different implications for competing firms' location decisions and quality decisions. When firms are uncertain about consumer taste and choose their product locations, whether firms and/or consumers benefit from the user-generated content depends on which of the two effects dominates. We find that only when the correlation-increasing effect is moderate, a win–win scenario for both firms and consumers occurs, but the society always benefits from user-generated content. Stronger consumer preference strengthens the overall impact of user-generated content. In sharp contrast, when firms face uncertain consumer valuation of quality and choose product quality, they do not benefit from user-generated content, but consumers may benefit or lose from it. When the correlation-increasing effect is significant, both firms and consumers, and therefore the society, are hurt by user-generated content. Stronger consumer preference mitigates the negative impact but amplifies the positive impact of user-generated content in this case. The online appendix is available at https://doi.org/10.1287/mnsc.2017.2839. This paper was accepted by Chris Forman, information systems. [ABSTRACT FROM AUTHOR]

Products such as Nike running shoes, Gillette razors, and Gatorade sports drink serve as the standard against which consumers evaluate other members of the category. Empirical evidence suggests that consumers care about not only the consumption utility derived from a product, but also the gain–loss utility in comparison to the reference product of the category. This paper examines how reference-dependent utility affects price competition in a horizontally differentiated market where consumers’ tastes are diverse. When consumer valuations are low, the reference product is priced lower than a nonreference product. In contrast, when consumer valuations are high, the reference product is priced higher than a nonreference product. Moreover, loss aversion on the price dimension always intensifies competition among low-valuation goods, whereas loss aversion on the taste dimension softens competition among high-valuation goods only if consumer sensitivity to the difference in match quality is above a threshold. We also find that an increase in the diversity of consumers’ tastes reduces equilibrium profits if consumer valuation is low but has the opposite effect if consumer valuation is high. We further explore how an increase in the popularity of the reference product affects price competition. Finally, we assess the robustness of the findings by extending the model in different directions. The online appendix is available at https://doi.org/10.1287/mnsc.2017.2834. This paper was accepted by Juanjuan Zhang, marketing. [ABSTRACT FROM AUTHOR]

Despite criticisms regarding their effectiveness, reward programs today represent a prevalent and apparently successful form of marketing investment for several industries such as airlines, hotels, gas stations, and credit cards. To understand what might contribute to their success, this paper investigates the purchase behavior of consumers who participate in a reward program from the travel industry. We estimate a dynamic demand model of gasoline purchase, and compare the value that consumers attach to rewards with the value they attach to the money spent for gasoline. We find that there exists a significant portion of frequent travelers who extract more value from one dollar’s worth of rewards than from one dollar spent for gasoline. Most of the effectiveness of the reward program is due to the behavior of these consumers. Their insensitivity to price induces firms to increase fuel prices by 1 euro cent per liter, which corresponds to about 10% increase in their margins. The program also contributes to softening competition, increasing margins by 4%. The web appendix is available at https://doi.org/10.1287/mnsc.2017.2821. This paper was accepted by Matthew Shum, marketing. [ABSTRACT FROM AUTHOR]

Consumers search for product information to resolve valuation uncertainties before purchase. We incorporate search cost into consumer choice models and study the two-stage consider-then-choose policy. In the first stage, a consumer forms her consideration set by balancing utility uncertainty and search cost. In the second stage, she evaluates all products in her consideration set and chooses the one with the highest net utility. We show that the revenue-ordered assortment (i.e., the offer set that includes products in the revenue-decreasing order) fails to be optimal, although it can obtain at least half the optimal revenue. We propose a k-quasi-attractiveness-ordered assortment and show that it can be arbitrarily near optimal for the market share maximization problem. The assortment problems with search cost are generally NP-hard, so we develop efficient approximation or relatively fast exact algorithms for a variety of assortment problems under the consider-then-choose models with search cost. For the joint assortment planning and pricing problem with homogeneous consumers, we show that the intrinsic-utility-ordered assortment and the quasi-same-price policy, which charges the same price for all products except at most one, are optimal. The online appendix is available at https://doi.org/10.1287/mnsc.2017.2790. This paper was accepted by Gad Allon, operations management. [ABSTRACT FROM AUTHOR]

In retail operations, customer choices may be affected by stockout and promotion events. Given panel data with the transaction history of customers, and product availability and promotion data, our goal is to predict future individual purchases. We use a general nonparametric framework in which we represent customers by partial orders of preferences. In each store visit, each customer samples a full preference list of the products consistent with her partial order, forms a consideration set, and then chooses to purchase the most preferred product among the considered ones. Our approach involves: (a) defining behavioral models to build consideration sets as subsets of the products on offer, (b) proposing a clustering algorithm for determining customer segments, and (c) deriving marginal distributions for partial preferences under the multinomial logit model. Numerical experiments on real-world panel data show that our approach allows more accurate, fine-grained predictions for individual purchase behavior compared to state-of-the-art alternative methods. [ABSTRACT FROM AUTHOR]

We develop a behavioral model of durable good usage with mental accounting for sunk costs. It predicts higher-than-rational usage that attenuates at a rate that increases with sunk costs. Singapore government policy varied the sunk cost of buying a new car. Using Singapore data, we estimate the elasticity of driving with respect to sunk costs to be 0.048, which implies that government policy between 2009 and 2013 was associated with 86 kilometers per month, or 5.6%, more driving. The results are robust to specifying sunk costs as relative to buyer income and estimation with Hong Kong data. We believe this to be the first field evidence of the sunk cost fallacy in usage of a major durable good. [ABSTRACT FROM AUTHOR]

This paper proposes a novel account of impatience: People pay more attention to the opportunity costs of choosing larger, later rewards than to the opportunity costs of choosing smaller, sooner ones. Eight studies show that when the opportunity costs of choosing smaller, sooner rewards are subtly highlighted, people become more patient, whereas when the opportunity costs of choosing larger, later rewards are highlighted this has no effect. This pattern is robust to variations in the choice task, to the participant population, and to whether the choices are incentivized or hypothetical. We argue that people are naturally aware of the opportunity costs of delayed rewards but pay less attention to those associated with smaller, sooner ones. We conclude by discussing implications for theory and policy. Data, as supplemental material, are available at . This paper was accepted by Yuval Rottenstreich, judgment and decision making. [ABSTRACT FROM AUTHOR]

Network externality arises when the utility of a product depends not only on its attributes but also on the number of consumers who purchase the same product. In this paper, we study consumer choice models that endogenize such network externality. We first characterize the choice probabilities under such models and conduct studies on comparative statics. Then we investigate the assortment optimization problem under such choice models. Although the problem is generally NP-hard, we show that a new class of assortments, called quasi-revenue-ordered assortments, which consist of a revenue-ordered assortment plus at most one additional item, are optimal under mild conditions. We also propose an iterative estimation method to calibrate such choice models, for both uncensored and censored data cases. An empirical study on a mobile game data set shows that our proposed model can provide better fits for the data, increase the prediction accuracy for consumer choices, and potentially increase revenue. This paper was accepted by Noah Gans, stochastic models and simulation. [ABSTRACT FROM AUTHOR]

We study behavior-based pricing (BBP) in a vertically differentiated model. Vertical differentiation is innately asymmetric because all customers prefer the higher-quality product when prices are equal. This asymmetry causes BBP to have different properties than symmetric horizontally differentiated models. We highlight two dimensions that affect the analysis: the role of quality-adjusted cost differences between the firms and the role of consumer discounting relative to firm discounting. In the second period, consistent with the prior literature, we find that there are conditions based on market shares and quality-adjusted costs under which either the low-quality firm or the high-quality firm-but not both-will reward its current customers with lower prices than it charges to new customers. We then consider whether these conditions can arise in a two-period equilibrium. We find that if consumers sufficiently discount the future periods, then firms at enough of a competitive disadvantage will reward their customers: i.e., the low-quality firm will reward its current customers if the quality-adjusted cost differential between the two firms is small, while the high-quality firm will reward its current customers if this cost differential is large. Conversely, we find that if consumers do not discount the future very much, then the firm at a competitive disadvantage (i.e., a low-quality firm competing against a low-cost high-quality firm, or a high-quality firm that has very high costs) can earn greater profits with BBP than without BBP, although there are cases where both firms may benefit from BBP. This paper was accepted by J. Miguel Villas-Boas, marketing. [ABSTRACT FROM AUTHOR]

In the last decade, interest in the 'wisdom of crowds' effect has gained momentum in both organizational research and corporate practice. Crowd wisdom relies on the aggregation of independent judgments. The accuracy of a group's aggregate prediction rises with the number, ability, and diversity of its members. We investigate these variables' relative importance for collective prediction using agent-based simulation. We replicate the 'diversity trumps ability' proposition for large groups, showing that samples of heterogeneous agents outperform same-sized homogeneous teams of high ability. In groups smaller than approximately 16 members, however, the effects of group composition depend on the social decision function employed: diversity is key only in continuous estimation tasks (averaging) and much less important in discrete choice tasks (voting), in which agents' individual abilities remain crucial. Thus, strategies to improve collective decision making must adapt to the predictive situation at hand. This paper was accepted by Yuval Rottenstreich, judgment and decision making. [ABSTRACT FROM AUTHOR]