A Brand Reputation Tracker Using Social Media
We present a brand reputation tracker for the world's leading global brands, based on an analysis of all Twitter comments made by the public. We measure volume and sentiment on drivers and sub-drivers of brand reputation, on a weekly, monthly, and quarterly basis, using the Rust-Zeithaml-Lemon customer equity driver framework. The resulting measures are housed in an online longitudinal database and may be accessed by brand reputation researchers. We illustrate some of analyses that are possible using the database. For example, we show how the tracker reflects important brand events, and show that drivers and sub-drivers of brand reputation vary in terms of the degree to which they reflect those events. We explore the stationarity of the drivers and sub-drivers and analyze the interdependence pattern between the drivers and sub-drivers. We further link the drivers to firm abnormal stock returns to demonstrate the nomological validity of the drivers and their actionability for brand management. Finally, we provide some important research questions that may be answered using the tracker.
Roland T. Rust, William Rand, Ming-Hui Huang, Andrew Stephen, Gillian Brooks and Timur Chabuk
Closing the Gender Pay Gap
The gender pay gap is a topic of increasing discussion in the boardroom, in the media and among policymakers. Firms are now more likely than ever to seek outside consulting to measure the gap and obtain "equal pay verification" from accreditation bodies. However, it is unclear when faced with a gender pay gap - or any demographic gap - how best to close it. In this research we derive optimization formulations that balance fairness and efficiency. We explore the role of fairness and seek a balanced approach to correct demographic pay gaps. We introduce both between efficiency and fairness, both during equal pay campaigns and during the annual salary review cycle. We also seek to answer some of the practical considerations when implementing data driven approaches for salary decisions in organizations, for example we answer the question "When is the gap closed" and practical bounds to answer that question.
Descriptive and Predictive Modeling for Patient Empowerment
This research focuses on the role and use of data and prediction models in difficult and often emotional decision-making scenarios. Research shows that doctors are often reluctant to discuss prognosis with their cancer patients. Studies further show that when doctors do have these conversations, they tend to be overly optimistic and biased towards good outcomes. As a result, patients often may not have the clear understanding of their scenario that they need in order to make the best possible choices for themselves and their families. In this work we developed state-of-the-art survival models and translate the results into decision-support tools aimed at supporting patient/provider discussions. In a new collaboration with the Maryland Medical Center, Dr. Bjarnadóttir and collaborators will focus on end-of-life care and on the optimal use of data and modeling to support end-of-life decisions.
The Digital Revolution and the Global Economy
Anil Gupta is working on a new book (working title: "Global Goes Digital"). It will focus on how the digital revolution - in particular, big data and AI - will transform the global economy and the nature of globalization itself. Since emerging markets are close to accounting for half of the world's GDP, one part of the analysis will focus on the impact of digitization on economic growth and employment in emerging markets. The second part will focus on how digitization is shifting the dynamics of global integration. Historically, cross-border trade in goods has been the primary driver of global integration. Digitization is causing a shift - from integration via trade in goods to integration by flows of capital, data, and knowhow. The second part of the book will drill down on the nature of this shift and its implications. Some of the ideas related to this book were published in HBR in August 2018 (The Dangers of digital protectionism – coauthored with Ziyang Fan, Head of Digital Trade Policies, World Economic Forum).
Emerging Urban Mobility Services and Business Models
This work addresses a broad set of issues related to existing and emerging services for transporting people in an urban setting. These include both for-profit and non-profit, shared and individual and ride-hailing and reservation-based services. For for-profit services issues addressed include both pricing strategies and driver compensation policies. Also, of interest are the nature of competition between ride-hailing services and reservation-based services and the impact of various forms of government regulation. On the non-profit side, strategies and algorithms are developed for maximizing social welfare through efficient matching of multiple passengers with drivers. The techniques employed include empirical analysis industry data, closed-form solution of analytic models and optimization algorithms.
Engaged to a Robot? The Role of AI in Service
This paper develops a strategic framework for using artificial intelligence (AI) to engage customers at the different touchpoints of the service process. For service delivery, mechanical AI can deliver consistent, reliable service repetitively, such as McDonald's using robots to deliver ordered foods to customers (automated service delivery). For service creation, thinking AI can identify new markets and create new service, such as Netflix using recommendation agents to suggest movies to viewers (automated service choice). For service interaction, feeling AI can communicate with and engage customers, such as Macy's On Call using natural language to interact with customers' in-store shopping (automated customer service). We illustrate various AI applications for the three service stages that provide managerial guidelines for service providers to leverage the advantages of AI.
Ming-Hui Huang and Roland T. Rust
The Feeling Economy: Managing in the Next Generation of AI
The capability of AI is currently expanding beyond mechanical and repetitive to analytical and thinking. We show, based on both theory and empirical evidence, that a "Feeling Economy" is emerging, in which AI performs many of the thinking tasks, and human workers gravitate more toward interpersonal and empathetic tasks. Although these "people skills" have always been important, our analysis shows that they are now becoming important to an unprecedent degree. To adapt effectively, workers must place increasing emphasis on feeling skills, and managers must change the nature of jobs to compensate for the fact that many of the thinking tasks are now performed by AI. The next generation of AI will be one in which humans and AI work together, with AI performing most of the mechanical and thinking tasks, and humans specializing in the interpersonal, empathetic, feeling tasks.
Forgoing Screening in Online Sharing Markets: An Empirical Investigation of User Behaviors in Airbnb
Screening, a mechanism for alleviating information asymmetry, is considered a necessity for online peer to peer market platforms, but has also raised concerns of increased discriminatory or biased behaviors in the sharing economy. However, left unexamined is the recent phenomenon where providers of goods and services may voluntarily forgo screening, even though it increases the risks and costs associated with "lemons." In this research study, we examine when and who may choose to forgo screening, and the impact this may have on their performance outcomes. We answer these research questions on the Airbnb platform, which recently instituted an "Instant Book" feature that enables hosts to forgo the screening of guests seeking lodging. Utilizing a unique panel dataset of all the listings in New York City between August 2015 and February 2017, we employ propensity score matching combined with difference-in-difference analysis to examine switching and re-switching behaviors of hosts. Our study provides evidence of the economic benefits of forgoing screening from increased occupancy even as reviewer ratings decline, and shows these effects to be stronger for African American and female hosts. Our study also provides insights about the strategic and social welfare implications of these findings, in the context of the current conversation regarding discrimination and bias in the sharing economy.
The Future of Marketing
The main thesis of this article is that several long-term trends are reshaping marketing and forcing marketing managers to change radically to keep up. These long-term trends are technological, socioeconomic and geopolitical. Advances in technology, in particular, are having a profound impact on marketing, resulting in the deepening of customer relationships and the continuous expansion of the service economy. Artificial intelligence, big data, the Internet, and the expansion of networks are creating a revolution in marketing that makes the 1960s-style 4 P's increasingly obsolete. Compounding the problem for marketers are the socioeconomic factors of diversity and inclusion, as well as major geopolitical threats. We explore the nature of change, extrapolate marketing practice into the future, and examine the implications for marketing managers, marketing education and academic research in marketing.
Motivating Effective Mobile App Adoptions: Evidence from a Large-Scale Randomized Field Experiment
Prior research has established a positive association between mobile app adoption and customers' purchase behaviors. However, it is not clear whether firms can actively influence customers' mobile app adoptions and increase their purchases through these induced adoptions. Using a randomized field experiment involving over 230,000 customers, we investigate: i) whether and how a firm can motivate customers to adopt mobile apps using external interventions and, ii) the causal effect of induced mobile app adoptions on customers' purchase behaviors. We find that: i) both providing monetary incentives and information can lead to a significant increase in customers' mobile app adoption; ii) the effect of mobile app adoptions varies greatly depending on how customers are induced. Although providing monetary incentives may lead to a larger increase in mobile app adoptions, such induced adoptions do not result in more purchases in the long run. In contrast, providing information leads to effective mobile adoptions that sustainably increase customers' purchases, and overall profits for the firm. We further explore the underlying drivers of such differences in the effect of induced app adoptions and find that information, as compared to monetary incentives, may serve as a better sorting device and can attract customers who would need the app more and use it more effectively. Firms cannot predict such 'customer types' from observable characteristics, and thus need to use appropriate interventions to induce sorting. Finally, we examine customers' multi-channel purchase behaviors and find evidence for how induced app adoptions affect customers' purchase behaviors across mobile and desktop channels. Specifically, there is a complementary effect between mobile app and the desktop channel for information-induced app adopters, but a substitution effect between the mobile app and mobile web channel for incentive-induced app adopters. For information-induced app adopters, the mobile app serves as a discovery tool and helps them find a greater variety of deals.
In summary, by leveraging a randomized field experiment, our study provides actionable insights for firms designing interventions to motivate effective mobile adoptions.
Tianshu Sun, Lanfei Shi, Siva Viswanathan and Elena Zheleva
Precision Decision Support for Opioid Prescription
Chronic opioid therapy (COT) has been associated with serious adverse outcomes and the social and economic impact of continuing opioid treatment is sizeable. Given the risks and adverse outcomes associated with COT, many consider COT a care choice of last resort. Thus, to the extent that clinicians need to make decisions regarding whether opioid use will begin or continue, decision support on the individual-patient probability of COT appears critical. Personalized guidelines, built on decision support systems (DSSs) utilizing state of the art machine learning algorithms, have the potential to influence care at the point of service. Building models that can serve as the foundation of such systems can therefore contribute to changes in physician prescribing.
Our ongoing studies of opioid prescribing focus on individual risk scoring, optimal interventions and economic impact of policy changes. This work is supported by the National Institute for Health Care Management (NIHCM).
Retail Firms' Use of Social Media – Insights from Analysis of Large-Scale Twitter Data
While social media platforms have been used by retailers for a variety of purposes, there is limited research on how traditional retailers compete on social media platforms and what the effects of such competition are on related outcomes. Our paper seeks to fill this gap by examining whether retailers that are close competitors in the traditional context adopt similar content strategies on Twitter. We find that dissimilar firms have higher online engagement and acquire new followers faster. In examining the underlying mechanism, we find that this is attributable to their ability to leverage higher-level affordances of Twitter (i.e., relationship formation, meta-voicing1, interactivity, collaboration, and competition). We find that it is the use of these higher-level affordances that leads to greater online engagement compared to the use of lower-level affordances, such as self-presentation, presence-signaling and communication. Our findings have important implications for firms' competitive strategies in online social media platforms.