Time: 9:40 AM - 11:00 AM

Chair
Haruka Kozuka (Seikei University)

Mirai Igarashi (University of Osaka), Hiroshi Onishi (Keio University), Osamu Sakai (Hosei University), & Shohei Hasegawa (Hosei University)

Title

How Chatbot Anthropomorphism Shapes Customer Purchase Decisions: Evidence from a Field Experiment on a Cosmetic E-commerce Retailer

Abstract
Chatbots have become increasingly popular for interacting with consumers visiting websites. In particular, chatbots on e-commerce sites are useful tools for reducing companies' labor costs; however, consumers also seek personalized care and human-like interactions from chatbots. This study conducts a real-world field experiment on a Japanese cosmetic e-commerce website during a two-month campaign to examine the impact of chatbot anthropomorphism on customer behavior. It employs a 2x2 framework of form realism (human/cartoon appearance) and behavioral realism (warmth/competence response styles), enabling an evaluation on how chatbot anthropomorphic designs influence customer purchasing behavior across two purchasing types: a one-time order and a subscription order. While none of the previous studies has simultaneously examined both dimensions of realism, our results reveal that low form realism (cartoon appearance) when combined with high behavioral realism (warmth response) significantly enhances customers' subscription purchases, especially among less loyal customers, aligning with the expectancy-disconfirmation theory. In contrast, chatbots with high behavioral realism (competence response) are more effective for one-time order customers, reflecting central-route decision-making as described in the elaboration likelihood model (ELM). Over the study period, while the novelty effect of chatbots diminished for one-time purchases, it remained stable for subscription purchases, highlighting differences in customer engagement across purchasing order types. This study contributes to the literature by providing empirical insights from a field experiment, providing practical implications for optimizing the anthropomorphism of chatbot designs based on customer loyalty and order types.

Keywords

Chatbot, Anthropomorphism, Field experiment, One-time orders, Subscriptions, Elaboration likelihood model

Akira Shimizu (Keio University)

Title

Inside Head:  To find the decision-making process using eye-tracking and EEG

Abstract
Research on the decision-making process leading to consumer purchases has traditionally been conducted primarily through experiments using clickstream data or eye tracking. The former is actual online shopping data, making it more accurate than experiments; however, it has the disadvantage of not being able to identify where on the page the consumer was looking. The latter can determine where the consumer was looking, but most studies use experimental materials that are often disconnected from actual shopping experiences. In this study, we combined clickstream data, eye-tracking, and EEG to investigate the actual decision-making process of consumers in an online shopping environment. The results revealed that there are differences in brain waves between the stages of considering potential items and making a purchase, that brain wave patterns differ significantly just before adding an item to the cart compared to other situations, and that among the considered items, those viewed for more extended periods are more likely to be purchased, with this tendency mediated by the brain wave patterns during the viewing period.

Keywords
Eye-tracking, Decision-Making Process, EEG, Online Retailing

Fumiko Kano Glückstad (Copenhagen Business School)

Title

Testing a model of destination image formation - revisited

Abstract
In this presentation, I will revisit our work that was awarded as an excellent paper at “2018 Global Marketing Conference at Tokyo”. In this work, we have suggested a dynamic model of destination image formation (DDIF) explained as an iterative process of concept learning. To demonstrate the underlying mechanism of this model, we employed a machine learning algorithm called nonparametric Bayesian relational modeling that can simultaneously group people and destination attributes (so-called biclustering). This algorithm can not only do biclustering of one dataset but also can do biclustering of three datasets in parallel. We collected data (n = 521) and asked them to select what they associate with France, Germany, and Turkey from 70 generic destination attributes, and what information sources they access to learn about these destinations. By applying the nonparametric Bayesian relational modeling to analyze the collected data, we investigated three research questions: (i) How patterns of associations with the three specific destinations held by the individuals are shared among a group of individuals (segment) and distinguished across segments; (ii) How patterns of destination images held by the segments and their willingness to visit are related for the three respective destinations; and (iii) How groups of individuals who have developed distinctive destination images of the three destinations have accessed information sources to learn about the respective destinations. DDIF implies that individuals who have been exposed to similar stimuli (information sources) in the iterative image formation process may have a higher likelihood of associating a destination with similar groups of attributes. Thus, they develop a similar destination image. DDIF also assumes that types of attributes that the members of the respective segments associate with a destination are related with their positive or negative willingness to visit that destination. These assumptions were tested and demonstrated in our work.

Keywords
Tourism, Dynamic Destination Image Formation, Consumer Segmentation, Nonparametric Bayesian Relational Modeling, Willingness to Visit 

URL(s) / Related Work(s)

Glückstad, F. K., Schmidt, M. N., and Mørup, M. (2020). Testing a model of destination image formation: Application of Bayesian relational modelling and fsQCA, Journal of Business Research, 120, 351-363. https://doi.org/10.1016/j.jbusres.2019.10.014 

Mengzhou Zhuang (U of Hong Kong), Beibei Dong (Lehigh University), Eric Er Fang (Lehigh University) & Maggie Dong (U of New South Wales)

Title

Name for a Name, Name for a Gain: The Mitigating Effects of Sender and Content Personalization on Name Personalization in Promotional Campaigns

Abstract
Advances in technology have spurred the widespread use of personalization in marketing, with personalized promotions often deemed to deliver superior value to customers. However, name personalization, in which recipients’ names appear in marketing messages, can lead to contradictory consequences. This research documents the negative effects of name personalization and explores how coupling name personalization with sender personalization and content personalization can counteract or overturn the negative impact of name personalization on purchase behavior by increasing message credibility. Furthermore, the authors investigate the heterogeneous effects of personalization strategies across customer relationship stages (new vs. repeat customer). Empirically, they conduct two field experiments on WeChat, followed by three lab experiments. The findings offer managerial implications for the optimal personalization designs of promotional messages catered to different customer stages.

Keywords
Personalization, Name Personalization, Field experiments, Lab experiments

Time: 11:10 AM - 12:30

Chair
Takashi Teramoto (Chuo University)

Justin Huang (U of Michigan), Jangwon Choi (Santa Clara University) & Yuqin Wan (Northwestern University) 

Title

Politically Biased Moderation Drives Echo Chamber Formation: An Analysis of User-Driven Content Removals on Reddit

Abstract
Echo chambers -- online spaces where individuals are met with reinforcing viewpoints and insulation from opposing viewpoints -- are of increasing interest to social media platforms and policymakers amidst rising political polarization. Selective participation and algorithmic recommendations are commonly cited as drivers of echo chamber formation. In this work, we demonstrate a previously-undocumented mechanism behind the formation of echo chambers on social media sites: the politically biased content removal decisions of user moderators. Applying a combination of natural language processing and network analysis to characterize political leanings on a dataset of removed and unremoved comments from 1.2 million users on 110 location-based subreddits (forums or sub-communities dedicated to a specific topic) of Reddit, we document political bias in content removal: moderators are significantly more likely to remove content that differs from their political orientation. Further, using a matching approach, we show that content removals have an indirect chilling effect on censored users' subsequent political speech on the same channel. Finally, we conduct a counterfactual simulation and demonstrate that politically biased content removal increases echo chamber intensity. These findings are of broad interest to users, platforms and regulators who weigh the costs and benefits of freedom of expression with the logistics of moderating massive online spaces. 

Keywords

Echo Chambers, Political Polarization, Social Media, Content Moderation, Platform Governance

Paper

Politically Biased Moderation Drives Echo Chamber Formation: An Analysis of User-Driven Content Removals on Reddit

Xiang Wan (Santa Clara University), Xiaojing Dong (Santa Clara University), Yuchi Zhang (Santa Clara University) & Xiaosong Dong (Shanghai University of Engineering Science)

Title

Recommendation Effects in the Presence of Customer Learning and Privacy Concerns

Abstract
Product recommendation systems are widely used in mobile retailing. While these systems improve through gathering data from user interactions and delivering more precise suggestions, users themselves also learn through repeated app usage and become more self-reliant, potentially reducing their dependence on algorithmic suggestions. This observation raises an important yet underexplored question: how does the effectiveness of recommendations vary with user experience? We address this question using a quasi-experiment conducted on a major mobile shopping platform in Asia. Our findings reveal that the effectiveness of the recommendation engine declines as users gain more experience with the app. To explain this pattern, we propose and validate a dual learning process hypothesis, in which both the user and the platform learn concurrently but at different rates. We show that when user interactions involve product browsing or purchasing, users learn faster than the platform, leading to reduced dependence on recommendations. In contrast, when interactions are limited to session initiation, the platform retains a learning advantage. We also rule out privacy concerns as an alternative explanation. This study underscores the evolving nature of user-platform interactions and offers insights for developing adaptive recommendation strategies that align with users’ experience levels.

Keywords
Recommendation Systems, User Experience, Dual Learning, Privacy Concerns, Quasi-Experiment, Staggered Adoption, Counterfactual Estimators

Chaehyeon Lee (Yonsei University) & Sue Ryung Chang (Yonsei University)

Title

Consumer Reactions to AI-Generated Visual Content

Abstract
Amid the rapid development and widespread accessibility of generative AI tools, the emergence of AI-generated images is gaining remarkable traction. In this study, we examine whether consumers like AI-generated images and how image characteristics and creator characteristics influence the consumer reactions to AI-generated images. We especially focus on two important image characteristics in evaluating AI-generated images (i.e., naturalness and complexity) and two creator characteristics (i.e., professional and influencer status). 

For empirical analysis, we collect digital images, whether generated by human or AI, along with related information about the creators from a global showcase platform for artists. We employ automatic image analyses to operationalize image characteristics. We find that people tend to like images less when they recognize the images are generated using AI tools. Regarding image characteristics, however, the naturalness of image weakens the negative effect of using AI in generating images, but complexity enhances it. As for creator characteristics, the influencer status of the creator mitigates the negative reactions to AI-generated images, while art professional status of the creator strengthens them. Our research contributes to the literature of AI usage in marketing by providing valuable insights into consumer reactions to AI-generated image and effective company strategies for utilizing AI-generated image without eliciting repulsive emotions from consumers about the use of AI.

Keywords
Generative AI, AI-generated image, AI disclosure, User-generated content

Wenbo Wang (Hong Kong University of Science and Technology)

Title

Information Silo on Social Media

Abstract
An information silo on social media platforms like TikTok refers to a situation where users are primarily or exclusively exposed to a narrow range of content, perspectives, or information—often aligned with their existing interests, beliefs, or behaviors. This phenomenon can be reinforced by algorithmic personalization and confirmation bias. Through a field experiment on US TikTok, we examined the existence of information silos and consumer heterogeneity in its magnitude. We also explored whether low cost marketing nudge on the consumer side can reduce information silo. 

Keywords
Information Silo, Social Media, Echo Chambers, Field Experiment

Time: 14:00- 14:40

Chair
Satoshi Nakano (Chuo University)

Richard Tang (Loyola Marymount University) & Desmond Lo (Santa Clara University)

Title

Location Synergy and Its Implications on Establishment Survival

Abstract
The agglomeration and co-location of industries have long been recognized as drivers of economic advantage. Agglomeration yields both supply-side benefits, such as reduced transportation costs and access to specialized labor and knowledge spillovers, and demand-side benefits, including decreased search costs and increased customer traffic spillovers. While the co-location index of Billings and Johnson (2016) captures the tendency of industries to co-locate, we extend this metric to the establishment level by developing a location synergy index. Our new metric quantifies an establishment’s local co-location advantage based on: (i) its geographic coordinates, (ii) a predefined radius surrounding its location, and (iii) the pairwise industry co-location indices between the focal establishment’s industry and those of its neighbors within the radius. Using this index, we examine the survival of over 230,000 retail establishments (NAICS 44–45) in Los Angeles County over a 30-year period. We find that higher location synergy significantly reduces the hazard of exit. This effect is particularly pronounced for stores with lower consumer awareness, smaller retail chains, and establishments that sell non-durable goods. Our results are robust to company fixed effects and alternative identification strategies, including instrumental variable regressions. Our location synergy index offers a generalizable, establishment-level metric that highlights the demand-side benefits of firm clustering. It also serves as a practical tool for business owners making location decisions and for policymakers engaged in area development and urban planning.

Keywords
Retailing, Agglomeration, Co-location, Retail Stores, Location, Firm Survival 

Wouter Dessein (Columbia University), Desmond Lo (Santa Clara University), Ruo Shangguo (Jinan University) & Chieko Minami (Kobe University)

Title

Horizontal Communication, Vertical Communication, and Organization Design

Abstract
Advances in organizational economics highlight the importance of allocating decision-making rights, addressing coordination needs, and facilitating communication in organizational design and adaptation. For instance, theoretical work suggests that delegation and communication are jointly determined, and that an increase in the need for coordination enhances horizontal communication but impairs vertical communication. A recent empirical study shows that the relationship between delegation and uncertainty is contingent on coordination needs. Our study synthesizes this body of work by examining the complementarity between the allocation of decision rights and communication patterns, using micro-level and performance data from managers working in over 400 department stores of a major Japanese retailer. Using exogenous local weather shocks as an instrument for increased delegation, we find that task delegation to managers rises in response to local weather shocks, coordination needs, and managerial experience. Notably, we find that delegation enhances communication with peers and subordinates. However, the increase is less pronounced with superiors, suggesting a shift in the structure of internal information flow following the transfer of authority. Importantly, the complementarity between delegation and communication is particularly evident in departments with high coordination needs, as well as in high-performing departments. High-performing managers exhibit the strongest link between delegation and communication, aligning with the notion that decentralization is most effective when capability is high. Our data also show that managerial characteristics, such as leadership style, influence both the likelihood of receiving delegation and the intensity of communication, in line with prior work on managerial heterogeneity. Our results provide new empirical support for the view that firms structure authority and communication jointly, both of which are shaped by uncertainty, coordination needs, and managerial traits.

Keywords
Retail Management, Communication, Delegation, Coordination, Uncertainty

   Time: 14:50 - 15:40

Chair/Moderator
Masakazu Ishihara (New York University)

Tulin Erdem (New York University)

Title
Structural Modeling: Few Remarks

Slides
Structural Modeling: Few Remarks

Time: 15:50 - 16:50

Chair
Makoto Abe (Chuo University)

Shosuke Noguchi (Rice University / Waseda University)

Title

Strategic Buyers and Inventory Choice under Dynamic Price in the Flower Market

Abstract
Dynamic pricing is a prevalent strategy for firms selling perishable products with limited inventory. However, empirical analysis of its welfare effects is limited, particularly when buyers strategically delay purchases and sellers can endogenously adjust initial inventory size. This paper develops and estimates a structural model where strategic buyers choose purchase timings considering price dynamics and stockout risks, while sellers optimize initial inventory levels. I estimate the model using a neural-network technique well-suited for this structural framework, applied to granular data from a wholesale flower market. The records of individual purchase histories and fluctuating inventory size allow for a detailed examination of buyers’ incentives to wait and supply costs. Applying the estimated parameters to the simulations of a uniform pricing mechanism, I perform counterfactual analyses to assess welfare implications. The counterfactual results reveal that sellers would increase initial inventory under uniform pricing, highlighting that ignoring inventory endogeneity can lead to biased welfare conclusions. My quantitative analysis demonstrates that dynamic pricing might not dominate uniform pricing in enhancing total welfare. Furthermore, contrary to previous literature that overlooks strategic buyers and endogenous inventory decisions, my findings suggest that dynamic pricing can hurt sellers more than buyers.

Keywords

Strategic Buyers, Dynamic Pricing, Revenue Management, Inventory Management, Structural Estimation

Eunkyung An (Stevens Institute of Technology), Dmitry Mitrofanov (Boston College), Masakazu Ishihara (New York University)

Title

The Impact of In-Store Service Level on Consumer Behavior

Abstract
Consumers experience variations in in-store service levels due to differences in store inventory availability. In brick-and-mortar retail settings, it is common for consumers to encounter stockouts of desired products. Experiencing low service levels can negatively affect both short- and long-term consumer behavior. While retailers understand the importance of inventory management, maintaining consistently high service levels is not only operationally challenging but also financially burdensome. To investigate the impact of in-store service levels on consumer behavior, we collaborated with an American multinational retailer in the apparel industry. Using quasi-experimental methods on both consumer-level and store-level data, we estimate (1) the effect of service level improvements on consumer purchase behavior, (2) the immediate impact of service levels on store sales, and (3) the short- and long-term effects of experienced service level on consumer behavior. We find that service level improvements significantly increase in-store purchase behavior but have no spillover effects on online purchases. The immediate impact of service levels on store sales varies by store characteristics, promotional intensity, product categories, and product types. Our findings suggest diminishing returns to service level and offer actionable insights on how to tailor service levels across different store locations in the apparel retail sector.

Keywords

Stockout, Service Level, Customer Relationship Management, Customer Lifetime Value, Quasi-experimental Design

Jong Yeob Kim (Nanyang Technological University)

Title

Health Insurance and the Dynamics of Patient Decision Making

Abstract
Increasing life expectancies and an aging population globally have significantly burdened government sponsored health care programs for the elderly. This article exploits variations in eligibility age thresholds and coverage amounts in the South Korean dental market to examine health care utilization, strategic delays, and the dynamics of treatment choices by the elderly patients. We use two primary data sources: a repeated cross-sectional representative health survey, and a long panel data on individual-level prescription and treatment choices from a large university hospital in South Korea. Our empirical approach exploits temporal variation in eligibility thresholds, and out-of-pocket costs for different forms of treatments (dentures vs. implants) to compare age cohorts just meeting the insurance criterion vs. not. We find large increases in healthcare utilization post insurance expansion, and strong evidence of strategic delays—a particular form of moral hazard—in treatment choices for patients just below eligibility thresholds. Evidence from reduced-form analysis guides us to develop a dynamic structural “patient life-cycle” model to examine patients’ strategic delays of treatment and adoption of new technology (implants) under different dental insurance policies. From a methodological point of view, we show how the age threshold can be used as an exclusion restriction to estimate discount factors in dynamic models of health care choices. Structural parameters replicate many of the patterns observed in the data and allow us to conduct a variety of counterfactuals to provide welfare estimates for different health policies. For example, we find that lowering the eligibility threshold from age 65 to 60 increases treatment timing and improves patient welfare by 6.8% but also leads to a significant increase in government spending, resulting in a net negative overall change in social welfare. Implications of our findings for other contexts, such as the expansion of Medicare in the United States to include dental benefits and lowering age thresholds, are discussed.

Keywords

Public Health, Policy Intervention, Dynamic Structural Model, New Technology Adoption

Time: 17:00 - 18:00

Chair
Takuya Satomura (Keio University)

Minjung Kwon (Syracuse University)

Title

Price Transparency in Healthcare: Understanding the Impacts of the Price Disclosure Policy in Maine

Abstract
This research investigates the operational and competitive effects of Price Transparency Regulation (PTR) in healthcare through both theoretical and empirical lenses. We develop a framework analyzing how price information influences consumer choice on the demand side while affecting pricing decisions on the supply side. We leverage the launch of Maine’s price transparency website as a natural experiment and construct a Difference-in-Differences model to identify causal effects. Our analysis demonstrates how PTR restructures the dynamics of healthcare operations: (1) Demand-side, it increases demand elasticity, redirecting patient flow to lower-cost providers and reducing system-wide costs; (2) Supply-side, it alters revenue management strategies, with dominant insurers securing lower payments and high-demand hospitals commanding pricing power. This study contributes to healthcare operations management by empirically linking three core operational processes: patient flow patterns, revenue management systems, and payment negotiation dynamics. The results confirm that PTR achieves its policy objectives of reducing consumer costs through operational efficiency gains, validating its broad implementation. However, the uneven distribution of benefits—where larger insurers and hospitals capture disproportionate advantages—indicate that policymakers should consider measures to level the playing field, ensuring that smaller and newer entities are not disadvantaged.

Keywords

Price Transparency, Insurer Price Bargaining, Operations of Healthcare Services, Medical Decision Making

Peng Liu (Santa Clara University), Tat Chan (Washington University) & Hai Che (UC Riverside)

Title

Social Status-Seeking in Online Game Community and Its Effects on User Engagement and Purchases

Abstract
We study the motivation for pursuing higher social status among players in an online game community and how such motivation influences both players’ game engagement and in-game purchase decisions. We develop and estimate a dynamic model that captures how a player makes engagement and purchase decisions and how these decisions impact future promotions within the guild rank, which is a proxy for the social status of the players. Results show that attaining a higher guild rank significantly increases a player’s utility, pointing to the existence of social status-seeking motivation in the community. Because game item purchases and engagement help increase the chance of promotion, the player will be motivated to spend and play more. We then use counterfactuals, in which the focal game company offers price discounts,

to explore how it can leverage such social status-seeking motivation to boost its revenue. We find that price discounts for hedonic game items dominate price discounts for utilitarian items. At the optimum, however, offering a 20% discount for hedonic and utilitarian items simultaneously can boost the company’s profit by 18.86%. These findings provide valuable insights into how game companies can benefit from integrating social status features in the design of games.

Keywords
Structural models, Electronic commerce, Dynamic Programming, Choice modeling

Xuyang Xia (Duke University), Mantian (Mandy) Hu (Chinese University of Hong Kong), Jingwei Dai (Carnegie Mellon University) & Daniel Xu (Duke University)

Title

Exclusive Contracts and Prisoner’s Dilemma: Evidence from Hong Kong Food Delivery Industry

Abstract
This paper examines the competitive and welfare implications of exclusive contracts on a two-sided food delivery platform in Hong Kong. Policymakers have expressed concerns about these contracts, which require restaurants to provide their services exclusively on a single platform in exchange for additional benefits. In theory, dominant platforms could use exclusive contracts to hinder the growth of weaker rivals in markets if there were strong network externalities. However, the overall impact on consumers, restaurants, and market competition remains empirically ambiguous.

We use detailed consumer digital receipt data from Hong Kong to estimate consumer preference and find that exclusive status significantly boosts the restaurant’s chance of being chosen on the platform. The underlying mechanism seems to differ across platforms --- some are “diminishing”, where having more exclusive restaurants on the platform weakens its positive effect; others are “reinforcing”, where the benefit of being exclusive is strengthened by a larger exclusive restaurant share. We illustrate that, in our setting, they can have opposite welfare implications.

We then use an equilibrium model of restaurants’ exclusive status choices to evaluate the impact of exclusive contracts and conduct various counterfactual policies. First, the exclusive contract poses a prisoner’s dilemma for restaurants and platforms: when banning exclusive contracts, both platforms and restaurants receive more orders, via expanding the market demand withheld by the diminishing exclusive effect. Second, if exclusive contracts were only allowed for less popular platforms, based on a policy proposal by the Hong Kong authority, there would be substantial catching-up in market share without significantly damaging restaurant and consumer welfare, especially when the smaller platform exhibits a reinforcing exclusive effect. Lastly, banning exclusive contracts improves the overall consumer welfare as reflected by higher order volume. Taken together, banning exclusive contracts potentially improves consumer, restaurant, and platform welfare at least in the short run, but limiting exclusive contracts to startup platforms can also be a useful tool to help new entrants compete with big players in the market without significant social costs.

Keywords
Structural Model, Platform competition, Exclusive Contracts, Multihoming, Food Delivery Market