SPEAKER

Dr. Ping Xiao
Associate Professor of Marketing
Melbourne Business School
University of Melbourne

ABSTRACT

With the rise of the internet and recent advancements in digital technology, firms across various industries such as fast food, coffee, and retail, are increasingly embracing digital transformation. We analyze the case of Starbucks and Luckin to explore the impact of a digital entrant (Luckin) on the digital transformation of the incumbent (Starbucks) and to quantify the value generated by this transformation. We categorize the coffee stores into digital stores which facilitate services like “order online, pick up in store,” and traditional stores, which lack such features. Specifically, we address the following research questions: (1) How has a digital entrant (Luckin) affected the incumbent’s (Starbucks) development path, especially with regards to their digital transformation journey? (2) What benefits has Starbucks obtained from the digital transformation of their distribution channels? and (3) To what extent has either firm benefited from the strategic impact they’ve exerted on their competitors? We developed a dynamic model framework that considers both firms’ distribution adjustment decisions for traditional stores and digital stores as well as their strategic interaction. Our model incorporates unobserved market preference for coffee consumption, which affects firm profitability and therefore distribution adjustment decisions, into the model. We collected the entire development path of the major players in the coffee chain category across various markets in China to estimate the model. To the best of our knowledge, our study is among the first to structurally quantify the impact of digital transformation. Our research will also offer insights into the relationship between traditional stores and digital stores.

SPEAKER

Prof. Bernd Schmitt
Robert D. Calkins Professor of International Business
Marketing Division
Columbia Business School

ABSTRACT

Humanoid robots look and act more and more like humans. Consumers can already interact with humanoid robots as salespeople, social companions, and even intimate partners. Soon, these robots may be indistinguishable from human beings. In this presentation, I explore how consumers might respond to these “perfect” robots and how they are likely to shape consumer behavior and consumer society in the coming decades. I introduce the construct of anti-robot speciesism—a tendency to treat robots as fundamentally different from humans, even when humans and robots are indistinguishable from each other in appearance and behavior in consumer relevant settings. Several studies test hypotheses related to anti-robot speciesism. Together, these studies illustrate how speciesism is deeply entrenched in human psychology and likely to shape consumer behavior in the time of perfect humanoid robots.

SPEAKER

Dr. Eddie Ning
Assistant Professor of Marketing
Behavioural Science Division
University of British Columbia

ABSTRACT

When gathering information to make decisions, individuals often have to delay making a decision because the process of gathering information is interrupted, and the individual is not yet ready to make a decision. The paper considers a model of choice deferral based on time-varying search costs, potentially based on search fatigue, in which individuals have to strategically decide whether to defer choice when information gathering is interrupted, taking into account the current available information, and when they will be able to resume gathering information. We find that individuals are more likely to defer choice when information gathering is interrupted less frequently, when individuals can resume gathering information sooner, and when they discount less the future. We also consider the case in which individuals incur costs of re-starting a process of information gathering, and cases in which the individual has greater or less information about the extent of search fatigue. The paper also considers optimal pricing and shows how pricing should respond to the length of consumer browsing sessions, and gaps between browsing sessions

SPEAKER

Dr. Yan Xu
Assistant Professor of Marketing
The Pamplin College of Business
Virginia Tech

ABSTRACT

Workers often have unobserved characteristics, e.g., soft skills, that are important for teamwork. In this paper, we formulate and estimate a model of teamwork that imposes no functional form restrictions on the complementarity between different unobserved types of workers. Our modeling approach builds on the stochastic blockmodels in statistics (e.g., Bickel et al., 2013) and the econometric framework developed by Bonhomme (2021), that quantifies the complementarity between different types of individual contributions when only teams’ outputs are observed. We apply our model to a data set from a leading Chinese real estate company; the data contain the complete history of team assignments, team performances, and detailed property characteristics. We find evidence that complementarities between different agent types are heterogeneous and cannot be captured by commonly used production functions. More specifically, workers of intermediate solo performance complement all other types of workers the most; the best solo-performing workers, however, are not the best team players. Our findings suggest that firms can improve productivity by redesigning teams without incurring additional hiring costs. Leveraging our complementarity estimates, we use counterfactual experiments to show that restructuring teams can improve overall team output by up to 28.4%.

SPEAKER

Prof. Qiaowei Shen
Professor of Marketing
Guanghua School of Management
Peking University

ABSTRACT

Online question-and-answer (Q&A) platforms, as an important type of user generated contents (UGC), allow users to learn and share different perspectives of information and knowledge. Such platforms’ success critically depends on the quantity and diversity of the knowledge contents. This paper utilizes a novel dataset from one of the largest Q&A platforms and studies how the amount of information and the quality of the early-stage knowledge content influence the growth of future knowledge content. We measure and characterize knowledge content’s growth in quantity and diversity using an unsupervised learning method, which allows us to account for similarity across documents based on their overall meaning. Our empirical results suggest the amount of information in the early knowledge content has a negative effect on the quantity of future knowledge content but a positive effect on diversity. We also find high-quality early knowledge content drives more future knowledge quantity but has no influence on diversity. Our analysis provides important managerial implications for platform strategies. 

SPEAKER

Prof. Carlos Torelli
Professor of Marketing
Head of the Department of Business Administration
Anthony J. Petullo Professor of Business Administration
University of Illinois

ABSTRACT

Religion has been identified by international charities as an important factor to promote generosity. However, extant research on the relationship between religion and donations toward distant beneficiaries yields inconsistent findings. Whereas the universal prosociality hypothesis suggests a positive effect of religion, the minimal prosociality hypothesis argues that religious individuals donate less toward distant beneficiaries. This research identifies political ideology as a novel factor that may help to reconcile these inconsistent findings. Specifically, we theorize that religious conservatives (vs. liberals) believe more in a punitive God and are more likely to be driven by fear-based prosociality, which reduce their donations toward distant recipients. A multi-method approach provides evidence for these predictions. An analysis of average household donations from 3,080 U.S. counties reveals that in more religious counties, a pervasive conservative ideology decreases donation toward distant recipients (i.e., national charities) but not toward local churches. Three lab studies further demonstrate that religious conservatives (vs. liberals) donate less to distant beneficiaries. Furthermore, this effect is mediated by religious conservatives’ greater beliefs in a punitive God.

Introduction

The Conference on Behavioural Science in the Age of Smart Technology 2023 will be held on July 3, 2023, at The University of Hong Kong. The conference is sponsored by the HKU Business School’s Institute of Behavioural and Decision Science (IBDS) and Institute of Digital Economy and Innovation (IDEI).

The objective of the conference is to create a forum for researchers from multiple disciplinaries to discuss the behavioural impact of the smart technology. We welcome papers in all areas of business research with a behavioural focus.

Keynote Speaker

Professor Andrew Burton-Jones

MIS Quarterly Editor-in-Chief

School of Business, The University of Queensland

Submissions

Authors are invited to submit empirical papers on the above-mentioned topics. An extended abstract should be submitted in PDF format to the submission page.

Important Dates

Submission Deadline:    May 30, 2023

Notice of Acceptance:      June 10, 2023

Organising Committee

Professor Echo WAN

Director, Institute of Behavioural and Decision Science

HKU Business School

Professor Yulin FANG

Director, Institute of Digital Economy and Innovation

HKU Business School

Professor Zhenghui Jack JIANG

Academic Area Head, Innovation and Information Management

HKU Business School

Dr Michael JIA

Associate Director, Institute of Behavioural and Decision Science

HKU Business School

Dr Yiwen ZHANG

Associate Director, Institute of Behavioural and Decision Science

HKU Business School

Contact

Please contact the Institute of Behavioural and Decision Science at ibds@hku.hk for enquiries.

SPEAKER

Prof. Jiwoong Shin
Professor of Marketing
School of Management
Yale University

ABSTRACT

We study an advertiser’s optimal targeting strategy and its implications for the consumer’s data privacy choices, both of which determine the advertiser’s targeting accuracy. When consumers are uncertain about their preferences, an ad targeted to a consumer carries an implicit message: an algorithm predicts that the product matches her preferences. This implicit recommendation influences the consumer’s beliefs and purchase decision but also introduces misaligned incentives; the advertiser may want to exploit the consumer’s beliefs by sending ads even to the wrong consumers predicted to have a bad match with the product. As the prediction accuracy improves, the consumer makes stronger inferences from targeted ads but so does the firm’s incentives to engage in mistargeting. Thus, under exogenous price, as the advertiser’s prediction becomes more accurate, the advertiser adopts a less precise targeted advertising strategy. Even if the prediction is perfect, the advertiser intentionally targets the wrong consumers, some of whom unknowingly purchase the product of a bad match. Despite the negative consequences, the consumer surplus can remain positive because the advertiser can better identify consumers with a good fit for the product, and thus consumers do not withhold information from the firm. In contrast, under endogenous price, a better prediction leads to a more targeted advertising strategy, although mistargeting persists. To better exploit the recommendation effect of advertising, the advertiser raises its price instead of diluting its recommendation power. The higher price leads to lower consumer welfare, which sometimes induces consumers to opt-out of data collection.

SPEAKER

Dr. Lan Luo
Associate Professor of Marketing
Marshall School of Business
University of Southern California

ABSTRACT

Profile pictures are a key component of many freelancing platforms, a design choice that can impact hiring and matching outcomes. In this paper, we examine how appearance-based perceptions of a freelancer’s fit for the job (i.e., whether a freelancer “looks the part” for the job), as inferred from profile pictures, can impact hiring outcomes on such platforms. Leveraging computer vision techniques and choice models, we analyze six-month data from Freelancer.com (63,014 completed jobs that received 2,042,198 applications from 160,014 freelancers) and find that, above and beyond demographics and beauty, freelancers who “look the part” are more likely to be hired. Interestingly, we do not find a strong correlation between “looking the part” and job performance. Supplementing our large-scale observational study with two choice experiments, we find that (i) the effect of perceived job fit is stronger when reputation systems are not sufficiently diagnostic to differentiate candidates and (ii) that by considering perceptions of job fit, participants are more likely to choose freelancers with fewer reviews, lower ratings, and/or without certifications. Last, we find that “platform recommendations” can only partially mitigate the unintended consequences of profile pictures, and recommending multiple freelancers can further increase the role of “looking the part.”

SPEAKER

Dr. Tong Guo
Assistant Professor
The Fuqua School of Business
Duke University

ABSTRACT

We propose a novel strategy to causally identify the impact of news coverage on product entries in local markets via intermediaries. Our identification relies on interacting the common time-series of the global social media discussion obtained from a semi-supervised topic model with the local shares of media consumption irrespective of the products being studied. We demonstrate our identification strategy in the case of the early-stage launching of impossible meat, a novel food technology that synthesizes meat substitutes by closely simulating the texture, flavor, and appearance of real meat. To study the impact of social media news on restaurant adoption of impossible meat products, we construct a novel location-specific adoption metric that accurately measures the decisions of local standalone restaurants and stores using their social media announcements. We further construct the exogenous measure of county-quarter-level intensity of topic-specific news coverage as the interactions between the global time series of social media discussion about various aspects of impossible meat products (e.g., financials of the key manufacturer, Beyond Meat) during 2015-2019 and local share of genre-specific media consumption in 2014 (e.g., percentage of financial content in social media news among food industry). Arguably, the constructed measures are exogenous to local demand shocks given the local share of media consumption is pre-determined thus irrespective of the new product being studied. We further control for county and quarter fixed effects, local-dynamic confounders, and cross-regional information spillovers. Our results suggest that local news coverage on financing of the new technology is the most impactful topic among all news topics in increasing the regional launching of impossible meat products.