Marketing - HKU Business School

17 January

The Impact of Recommender Systems on Content Consumption and Production: Evidence from Field Experiments and Structural Modeling

SPEAKER

Prof. Tat Chan
Philip L. Siteman Professor of Marketing
Olin Business School
Washington University in St. Louis

ABSTRACT

Online content-sharing platforms such as TikTok and Facebook have become integral to daily life, leveraging complex algorithms to recommend user-generated content (UGC) to other users. While prior research and industry efforts have primarily focused on designing recommender systems to enhance users’ content consumption, the impact of recommender systems on content production remains understudied. To address this gap, we conducted a randomized field experiment on one of the world’s largest video-sharing platforms. We manipulated the algorithm’s recommendation of creators based on their popularity, excluding a subset of highly popular creators’ content from being recommended to the treatment group. Our experimental results indicate that recommending content from less popular creators led to a significant 1.34% decrease in video-watching time but a significant 2.71% increase in the number of videos uploaded by treated users. This highlights a critical trade-off in designing recommender systems: popular creator recommendations boost consumption but reduce production. To optimize recommendations, we developed a structural model wherein users’ choices between content consumption and production are inversely affected by recommended creators’ popularity. Counterfactual analyses based on our structural estimation reveal that the optimal strategy often involves recommending less popular content to enhance production, challenging current industry practices.
Thus, a balanced approach in designing recommender systems is essential to simultaneously foster content consumption and production.

07 January

Effects of Government-to-Contractor Revolving Door Directors on Customer Relationship Performance

SPEAKER

Prof. Ju-Yeon Lee
John and Connie Stafford Professor in Business
Associate Professor of Marketing
Ivy College of Business
Iowa State University

ABSTRACT

Firms in business-to-government (B2G) marketplaces often invite former government officials to join their boards of directors, in search of their critical knowledge and access. However, the actual impacts of these revolving door directors are unclear for marketing outcomes. By analyzing multisource, secondary panel data of 1,677 publicly traded U.S. firms in the B2G market between 2005 and 2021, the authors find that revolving door directors significantly improve three dimensions of customer relationship performance: customer acquisition, customer retention, and cross-selling performance. The beneficial effects of revolving door directors are contingent on demand volatility and the competitive-bidding preferences of the government customers the firm serves. That is, revolving door directors are more effective for improving customer relationship performance when firms face higher demand volatility from customers but less effective if customers prefer competitive-bidding processes. The customer relationship performance also mediates the relationship between revolving door directors and financial outcomes (contracting performance). These results provide unique contributions to marketing theories and implications for practitioners.

01 November

Income Tax and Salesforce Performance: A Micro Perspective

SPEAKER

Prof. Yi Xiang
Professor of Marketing
China Europe International Business School

ABSTRACT

How does changes in income tax affect sales performance? Our paper explores the link between economic policy and salesforce management at the transactional level, using data from a large apparel retailer in China. The analysis indicates that a nationwide personal income tax cut in October 2018 incentivized those salespersons who benefited from the policy to increase their sales productivity and improved their performance. The performance gain was particularly significant in low-income regions, and was mainly attributed to higher sales in each transaction, rather than more reliance on discounts to boost the number of transactions. In addition, the study shows that the tax cut resulted in a net increase in the government revenue because of higher sales and the subsequent rise in corporate tax payments.

01 November

Identities without Products: When the Preference for Self-Linked Products Weakens

SPEAKER

Prof. Liad Weiss
Associate Professor of Marketing
Warwick Business School
The University of Warwick

ABSTRACT

Extant literature and common marketing practices converge around the idea that stronger self-links to a brand and its products lead to increased brand loyalty. In this article, we challenge this conventional notion by revealing situations where the preference for self-linked brands diminishes, despite the self-links remaining unchanged. We introduce a key distinction between two types of consumer identities based on whether identity expression relies on specific products: product-dependent (e.g., chef) and product-independent (e.g., foodie). Our theory posits that self-links to products exert less influence on preference when a product-independent identity is prominent. Across five studies examining consumer leisure identities, we find that priming a product-independent (vs. product-dependent) identity reduces preference for self-linked products/brands. Interestingly, it can also enhance preference for negatively self-linked (dissociative) products/brands among materialistic consumers. In a sixth experiment and a real-world Facebook study, we illustrate that the extent to which consumers’ identity is chronically product-independent can be assessed either directly or indirectly from social media interests, allowing for more effective targeting of brand-switching appeals. Adding to the literature on the symbolic role of products in identity expression, our research uniquely investigates the functional role of products in identity expression and its profound impact on product/brand preference.

01 November

Over-Tailored Support: Support-Giving Consumers’ (Misguided) Inclination Toward Domain-Specific Gifts and the Role of Attentiveness Signaling

SPEAKER

Prof. Soo Kim
Assistant Professor of Marketing
Nanyang Business School
Nanyang Technological University

ABSTRACT

Witnessing the struggles of close others is an inevitable part of life. During these delicate moments, consumers often turn to material gifts to express their support and encouragement. Accordingly, the marketplace offers a wealth of options that assist consumers in conveying their support, such as products featuring affirming messages. Across six preregistered studies—including an analysis of customized gift designs by support-giving participants and an incentive-compatible dyadic choice study—this work demonstrates that support-giving consumers favor gifts that offer expressions of support tailored to struggling close other’s challenges:
Support-givers tend to choose gifts that express support specifically in the domain of the disclosed struggle (vs. in a supportive but non-domain-specific manner), perceive these gifts as more effective in alleviating close others’ struggles, and expect higher recipient demand for them. Studies find that this inclination is driven by support-giving consumers’ belief that domain-specific support-expressive gifts are more indicative of attentiveness. However, findings reveal that this inclination does not resonate well with receivers’ preferences. Whether domain-specific support-expressive gifts appear prescriptive (thus no longer simply attentive, but overbearing) serves as a boundary condition. Implications for consumer relational welfare and content and product design strategies for businesses that cater to support-giving consumers are discussed.

14 June

Designing experiments on social, healthcare, and information networks

SPEAKER

Prof. Edoardo Airoldi
Millard E. Gladfelter Professor of Statistics and Data Science
Professor of Finance
Fox School of Business
Temple University

ABSTRACT

Designing experiments that can estimate causal effects of an intervention when the units of analysis are connected through a network is the primary interest, and a major challenge, in many modern endeavors at the nexus of science, technology and society. Examples include HIV testing and awareness campaigns on mobile phones, improving healthcare in rural populations using social interventions, promoting standard of care practices among US oncologists on dedicated social media platforms, gaining a mechanistic understanding of cellular and regulation dynamics in the cell, and evaluating the impact of tech innovations that enable multi-sided market platforms. A salient technical feature of all these problems is that the response(s) measured on any one unit likely also depends on the intervention given to other units, a situation referred to as “interference” in the parlance of statistics and machine learning. Importantly, the causal effect of interference itself is often among the inferential targets of interest. On the other hand, classical approaches to causal inference largely rely on the assumption of “lack of interference”, and/or on designing experiments that limit the role interference as a nuisance. Classical approaches also rely on additional simplifying assumptions, including the absence of strategic behavior, that are untenable in many modern endeavors. In the technical portion of this talk, we will formalize issues that arise in estimating causal effects when interference can be attributed to a network among the units of analysis, within the potential outcomes framework. We will introduce and discuss several strategies for experimental design in this context centered around a useful role for statistics and machine learning models. In particular, we wish for certain finite-sample properties of the estimator to hold even if the model catastrophically fails, while we would like to gain efficiency if certain aspects of the model are correct. We will then contrast design-based, model-based and model-assisted approaches to experimental design from a decision theoretic perspective.

04 June

Goodbye My Friends and Goodbye My Career: Evidence from the Movie Industry

SPEAKER

Prof. Andrew Ching
Professor at Johns Hopkins Carey Business School
Johns Hopkins University

ABSTRACT

By analyzing a dataset with over 180,000 actors, directors and producers in the motion picture industry over a span of 20 years, this research provides causal evidence that network connections can significantly influence an actor’s career development. Specifically, we focus on quantifying the impact of losing a connection with a director/producer due to their death on an actor’s career trajectory. Our identification strategy leverages the insight that the timing of one’s death is exogenous to the network evolution, and hence it creates an exogenous shock to the network. Following the passing of a director/producer, our study reveals a decline of 6.76% in acting opportunities for actors who had previously collaborated with them. Our findings further reveal that more experienced actors are less impacted by the connection loss resulting from the passing of colleagues. Moreover, we find evidence that the adverse effect intensifies with time. Our results also show that the impact of such a connection loss on an actor’s career trajectory hinges on the initial impact of the shock to their network centrality.

22 May

Conference on Behavioural Science in the Age of AI and Social Media

20 May

Does Human-algorithm Feedback Loop Lead To Error Propagation? Evidence from Zillow’s Zestimate

SPEAKER

Prof. Meng Liu
Assistant Professor of Marketing
OLIN Business School
Washington University in St.Louis

ABSTRACT

We study how home sellers and buyers interact with Zillow’s Zestimate algorithm throughout the sales cycle of residential properties, with an emphasis on the implications of such interactions. In particular, leveraging Zestimate’s algorithm updates as exogenous shocks, we find evidence for a human-algorithm feedback loop: listing and selling outcomes respond significantly to Zestimate, and Zestimate is quickly updated for the focal and comparable homes after a property is listed or sold. This raises a concern that housing market disturbances may propagate and persist because of the feedback loop. However, simulation suggests that disturbances are short-lived and diminish eventually, mainly because all marginal effects across stages of the selling process—though sizable and significant—are less than one. To further validate this insight in the real data, we leverage the COVID-19 pandemic as a natural experiment. We find consistent evidence that the initial disturbances created by the March-2020 declaration of national emergency faded away in a few months. Overall, our results identify the human-algorithm feedback loop in an important real-world setting, but dismiss the concern that such a feedback loop generates persistent error propagation.

14 May

Learning from Online Ratings

SPEAKER

Prof. Xiang Hui
Assistant Professor of Marketing
OLIN Business School
Washington University in St.Louis

ABSTRACT

Online ratings play an important role in many markets. However, how fast they can reveal seller types remains unclear. We propose a simple model of rating behavior where learning about the seller type influences the rating decision. We calibrate the model to eBay data and find that ratings can be very informative. After 25 transactions, the likelihood of correctly predicting the seller type is above 95 percent.