SPEAKER

Prof. Huachao Gao
Associate Professor of Marketing and International Business
Peter B. Gustavson School of Business
University of Victoria

ABSTRACT

Consumer food waste, with its vast social, economic, and environmental implications, has been spotlighted during the COVID-19 pandemic, which disrupted food supply chains and heightened food insecurity. This study delineates the contrasting effects of infectious salience and restriction salience on food waste behavior. We find that infectious salience tends to increase food waste due to a heightened safety mindset, whereas restriction salience reduces waste by inducing a scarcity mindset. To address these dynamics, we propose and evaluate interventions that leverage the notion of resource scarcity and the financial consequences of wasted resources to decrease food waste during pandemic conditions. Additionally, we introduce a safety-focused intervention designed to neutralize the excessive safety mindset driven by infectious salience. Our empirical investigation includes a comprehensive field study, analysis of a secondary dataset, a laboratory experiment focused on actual food waste behavior, and three auxiliary experiments, all of which substantiate our conceptual model. These varied methodologies highlight the effectiveness of safety interventions implemented through different mediums, such as table tents, napkins, and to-go containers. This research harmonizes conflicting views on pandemic-induced changes in food waste, deepens the theoretical understanding of pandemic-related food waste phenomena, and suggests practical approaches for marketers and policymakers to curb consumer food waste in the context of pandemics.

SPEAKER

Prof. Tong Wang
Assistant Professor of Marketing
Yale School of Management
Yale University

ABSTRACT

The rapid development of large language models (LLMs) presents both opportunities and challenges in deploying them in goal-oriented dialogues for complex human interactions, such as customer support and persuasion. Advanced LLMs like GPT-4 excel in these domains but are too large and cost-prohibitive, while smaller and more economical models like LLaMa 2 offer limited performance. This paper proposes a novel approach to enhance the capabilities of smaller LLMs by leveraging the strategic prowess of their more advanced counterparts. Unlike traditional methods that focus on direct response learning, we introduce a strategy-centric imitation learning framework. Here, the advanced LLM acts as a teacher, imparting strategic thinking to the prompts of a lesser LLM and refining it iteratively until the student mimics the teacher effectively. We design an iterative process which alternates between scenario generation and strategy learning and returns a customized library of various scenarios and the optimized strategies. Crucially, our approach requires only black-box access to the models, facilitating easier integration across different platforms without the need for direct parameter manipulation. This strategy not only improves the functional capacity of smaller LLMs but also contributes to broader AI safety and interpretability by enabling the scrutiny of learned strategies by domain experts. The results indicate significant potential for using strategic knowledge transfer in real-world applications, enhancing the utility of LLM deployments in cost-sensitive environments.

SPEAKER

Prof. Yupeng Chen
Assistant Professor of Marketing
Nanyang Business School
Nanyang Technological University

ABSTRACT

Firms routinely use choice-based conjoint (CBC) data to estimate consumers’ heterogeneous preferences. Since the amount of information elicited from each respondent is often limited, effective information pooling across respondents is critical for accurate CBC estimation. In this paper, we propose a novel affine-subspace shrinkage approach to pooling information in CBC estimation. Our approach, formulated as a simple and efficient convex optimization problem, models preference heterogeneity by shrinking the individual-level partworth estimates toward an affine subspace of the partworth space, which itself is selected as part of the estimation. Using an extensive CBC simulation experiment and two field CBC data sets, we show that our model outperforms a strong multitask learning model, and it performs comparably to a hierarchical Bayes model with a Dirichlet process prior which requires a considerably more sophisticated solution algorithm.

SPEAKER

Prof. Jessica Fong
Assistant Professor of Marketing
Ross School of Business
University of Michigan

ABSTRACT

Delay in negotiations is common in many settings, but the effects of delay have rarely been studied empirically in the field. We measure the causal effect of delay using data from millions of negotiations on eBay. We find that for both buyers and sellers, the longer the bargaining party delays, the less likely the opponent is to continue the negotiation by countering. However, the downstream consequences vary. The more the seller delays, the more likely the negotiation will fail, but the more the buyer delays, the more likely the seller will accept the buyer’s offer. The effects of delay are robust; they exist even under short amounts of delay (under 6 hours) and for negotiations for low-priced goods. We find that these effects are consistent with models of strategic delay, in which delay acts as a signal of bargaining power.

SPEAKER

Prof. Xiaolin Li
Assistant Professor of Marketing
The London School of Economics and Political Science
The University of London

ABSTRACT

This paper examines the effects of incentive plans on the quantity and quality of creative production. We study a serial publishing platform which switched from a uniform commission plan to a quantity-based incentive plan offering higher commission rates if a writer’s production meets higher quantity brackets. Our analysis shows that, for a given book, the chapters published in the time periods when writers reached higher brackets of quantity (hence higher commission rates) had higher quality measured by chapter-to-chapter customer retention rates. Such a positive correlation is not significant in books published when the platform offered a uniform commission plan. We theorize that the quantity-based commission plan can enhance the quantity-quality complementarity in a writer’s payoff function. With the enhanced complementarity, the creators who reach a higher bracket of quantity will produce a higher quality under a quantity-based plan than under a uniform commission plan. Further empirical analysis indicates that the degree of enhanced complementarity is weaker for the writers who earn commissions from multiple books. Overall, our results underscore the importance of proper incentive design in improving the platform’s performance in managing mass creativity.

SPEAKER

Prof. Thomas Otter
Professor of Quantitative Marketing
Faculty of Economics and Business
Goethe University Frankfurt

ABSTRACT

Observations of product and service reviews suggest that textual product reviews may contain statements that talk about the overall experience (“We had a great time”) or, similarly, whether to recommend a particular product. We argue that such statements encapsulate an overall assessment and hence are not independently informative about, but rather reflect overall ratings. We propose a model that allows for the distinction between topics that contribute to and topics that merely reflect an overall evaluation and apply it to a data set consisting of luxury hotel reviews. Compared to a standard supervised LDA, we find our model to better fit the data and to improve customer insights by resulting in more semantically coherent topics which point at specific attributes with positive and negative relationships to customers’ evaluation of their experience.

SPEAKER

Professor Adelle Xue Yang
Assistant Professor of Marketing
NUS Business School
National University of Singapore

ABSTRACT

The reduction of decision speed is a central goal in the development of AI (artificial intelligence) applications. However, do people always prefer a faster AI decision-maker? Across twelve pre-registered experiments (N = 6,971), we find that people have greater appreciation for an AI decision-maker when it is slower at resolving moral tradeoffs than structurally identical non-moral tradeoffs. The effect was replicated across a variety of classic moral dilemmas (e.g., trolley problem) as well as generic resource-allocation problems. The effect shows insensitivity to decision consequences, and holds even when slower moral decision-making produces non-adaptive outcomes. Critically, the effect no longer emerges when either the “moral” or “tradeoff” component is removed from the slower decision. Results from moderation studies and measured process variables consistently suggest that the effect is attributable primarily to overgeneralized moral intuitions about “good” moral decision-makers, rather than specific inferences about the AI decision-makers’ “mind” or decision quality per se. Additional analyses of text responses using a large language model (LLM) corroborate these mechanism insights. Theoretical and practical implications are discussed.

SPEAKER

Professor Andre Bonfrer
Professor of Marketing
Deakin Business School
Deakin University

ABSTRACT

Increasingly, consumers turn to mobile applications (apps) to achieve personal development and wellness goals, such as improving their physical, mental, social, professional or financial well-being. To better assist consumers in their pursuit of these goals, service providers often introduce goal-enabling technologies (GETs) within their mobile apps that allow customers to define and monitor their service-related goals. However, empirical evidence regarding the presence and magnitude of the impact of GET adoption on goal-congruent behavior remains scarce. We use rich panel data from an investment app that introduced a GET to examine how customers who set a savings goal via the GET within the app changed their real-world savings behavior over time. Controlling for the potential issues related to self-selection into adopting GET and endogeneity, our results indicate that, on average, GET adoption increases customers’ goal-congruent behaviors. However, 25% of customers experienced zero or a negative impact of GET adoption on their savings behavior, and this impact depends on how customers customize the three goal difficulty dimensions. We identify an inverted Ushape relationship for end-goal, sub-goal and distance-to-goal difficulty dimensions on goal-congruent behavior. A field experiment highlights how this knowledge can be used to better assist customers in achieving their savings goal.

SPEAKER

Professor Karsten Theil Hansen
Professor of Marketing
UCSD Rady School of Management
University of California San Diego

ABSTRACT

In recent years, the emergence of smart technologies—including wearable devices like smartwatches, smartphones, smart scales, and e-bikes—has dramatically transformed how consumers interact with products. Embedded with sensors and tracking technology, these devices provide companies with unparalleled real-time insights into consumer engagement with their products. This shift enables a transition from solely analyzing purchase behaviors to understanding nuanced consumption patterns. Our research underscores the value of studying consumer engagement for businesses.

This paper presents our findings within the realm of streaming media consumption, drawing on a comprehensive dataset from a large panel of households. We introduce several metrics for quantifying media engagement and develop empirical models to identify distinct engagement segments. By training these models on an extensive dataset, we elucidate how engagement varies by show and viewer characteristics, offering predictions on future consumer behavior. Our work highlights the importance of analyzing consumption patterns, demonstrating their potential to inform content creation, marketing strategies, and enhance user experience.

SPEAKER

Professor Ganesh Iyer
Edgar F. Kaiser Professor in Business Administration
Senior Editor for Marketing Science
Haas School of Business
University of California, Berkeley

ABSTRACT

We study how market competition influences the algorithmic design choices of firms in the context of targeting. Firms face a general bias-variance tradeoff when choosing the design of a supervised learning algorithm in terms of model complexity or the number of predictors to accommodate. Each firm has a data analyst who uses the chosen algorithm to estimate demand for multiple consumer segments, based on which, it devises a targeting policy to maximize estimated profits. We show that competition induces firms to strategically choose simpler algorithms which involve more bias but lower variance. Therefore, more complex/flexible algorithms may have higher value for firms with greater monopoly power.