Informs Annual Meeting 2017

TB48

INFORMS Houston – 2017

2 - Level Up: Leveraging Skill and Engagement to Maximize Player Retention in Online Video Games Yan Huang, University of Michigan, 701 Tappan Street, Ann Arbor, MI, 48109, United States, yphuang@umich.edu, Stefanus Jasin, Puneet Manchanda We propose a novel two-stage data analytic modeling approach, combining statistical methods with optimization techniques to maximize customer retention via matching in the online video game industry. In Stage 1, we build a Hidden Markov Model (HMM) to capture the evolution of gamers’ latent engagement state and calibrate the HMM using a longitudinal dataset obtained from a major video game company. We find high-, medium- and low-engagement-state gamers respond differentially to motivations such as achievement and challenge. In Stage 2, we develop a matching algorithm that learns a gamer’s current engagement state and exploits that learning to match the gamer to a round to maximize retention. 3 - The Impact of an Augmented Reality Game on Local Businesses: A Study of Pokemon Go on Restaurants Pokemon Go, an augmented reality technology based game incentivizes it’s user to move in the physical world. Such incentives can have potential externalities to businesses associated with them. Using online reviews of restaurants as a proxy for sales and quality, we study the impact of Pokemon Go on local restaurants. We treat the release of Pokemon Go as a natural experiment and study the post- release impact for the associated restaurants. We find that restaurants that are located near PokeStops indeed observe a higher number of customers compared to restaurants that do not have PokeStops nearby. However, the online quality perception isnt affected due to this effect. 4 - Platform Control and Complementor Performance: Evidence from Mobile Apps Anparasan Mahalingam, Purdue University, 403 W. State Street, Krannert Building, West Lafayette, IN, 47906, United States, anparasan@purdue.edu We study how changes in platform control affect the performance of complementors. Leveraging the jailbreak of iOS 10 as a natural experiment, we examine how an exogenous deciency in iOS’s gatekeeping policy affects the performance of mobile apps on iOS compared with apps on Android. We also show several boundary conditions shaping the effect of the jailbreak on apps’ performance Vandith Pamuru Subramanya Rama, Purdue University, West Lafayette, IN, United States, vandith@purdue.edu, Warut Khern-am-nuai, Karthik Kannan 361A BOM Working Paper Competition Finalists Sponsored: Behavioral Operations Management Sponsored Session Chair: Michael Becker-Peth, University of Cologne, Cologne, 50923, Michael Becker-Peth, University of Cologne, Albertus-Magnus- Platz, Supply Chain Management & Man. Sci., Cologne, 50923, Germany, michael.becker-peth@uni-koeln.de Finalists in the BOM Best working paper competition will present. 2 - Discrimination with Incomplete Information in the Sharing Economy: Evidence from Field Experiments on Airbnb Ruomeng Cui, Emory University, 1300 Clifton Rd, Atlanta, GA 30322, Atlanta, GA, 30322, United States, ruomeng.cui@gmail.com, Jun Li,, Dennis J Zhang Recent research has found widespread discrimination by hosts against guests of certain races in online marketplaces, which endangers the very basis of a sharing economy â € ” building trust in the communities. In this paper, we explore the root cause of discrimination and how to reduce discrimination. We conducted two randomized field experiments among 1,256 hosts on Airbnb by creating fictitious guest accounts and sending accommodation requests to them. We find that requests from guests with distinctively African American names are 19 percentage points less likely to be accepted than those with distinctively White names. However, a public review posted on a guestâ € ™s page mitigates discrimination: when guest accounts receive a positive review, the acceptance rates of guest accounts with distinctively White and African American names are statistically indistinguishable. We further demonstrate that a negative review also significantly reduces discrimination. Our finding is consistent with statistical discrimination: when lacking perfect information, hosts infer the quality of a guest by race and make rental decisions based on the average predicted quality of each racial group when enough information is shared, hosts do not need to infer guestsâ € ™ quality from their race, and discrimination is eliminated. Our results Germany, michael.becker-peth@uni-koeln.de 1 - BOM Working Paper Competition Finalists TB48

offer direct and clear guidance for sharing-economy platforms on how to reduce discrimination. Platform owners should motivate users to write reviews of one another and design a better mechanism to facilitate information sharing â € ” especially information that signals guest quality. 3 - Modeling Newsvendor Behavior: A Prospect Theory Approach Bhavani Shanker Uppari, INSEAD, 1 Ayer Rajah Avenue, Singapore, 138676, Singapore, BhavaniShanker.Uppari@insead.edu, Sameer Hasija Problem definition: Studies have shown that the behavior of subjects in newsvendor experiments is not consistent with expected profit maximization as an assumption that is often made in inventory and supply chain management literature. Although prospect theory has been established as a popular model of behavioral decision making under uncertainty, it was considered to be inconsistent with observed newsvendor behavior (in particular, the pull-to-center effect) until a recent study proposed a prospect theory model that is consistent with the pull-to-center effect however, this model’s ability in representing newsvendor behavior compared to other plausible prospect theory models is unexplored in the literature. This paper takes a more comprehensive approach in building several prospect theory-based newsvendor models, and evaluates their competence in representing the observed newsvendor behavior. An important feature of these models is that they are not only consistent with the pull-to-center effect, but they also can accommodate individual level heterogeneity in order quantities. This is in accordance with the findings from recent research that there is remarkable heterogeneity in the order quantities in newsvendor experiments, and that the behavior of a significant proportion of individuals is inconsistent with the pull-to-center effect. 361B Practice/eBusiness II Contributed Session Chair: Xinyi Ren, University of Maryland-College Park, College Park, MD, United States, xinyi.ren@rhsmith.umd.edu 1 - Measuring Consumer Surplus in the On-demand Economy: The Case of Ride Sharing Meng Liu, Massachusetts Institute of Technology, 100 Main Street, Sloan Room 412, Cambridge, MA, 02142, United States, mengliu@mit.edu To understand the consumer welfare impact of ride-sharing platforms, we estimate a demand model for rides. Leveraging on the unusually rich taxi, Uber, and Lyft trip records from NYC as well as Uber and Lyft dynamic pricing and wait time, we find substitution patterns that vary by location, time, and demographics. We further provide strong evidence that consumers value time and dislike waiting. The consumer welfare gain mainly comes from two channels: first, the tech-aided ride-sharing platforms represent a more efficient “meeting” technology, causing an overall reduction in wait time; second, the platforms “fill in” the previously under-served communities by taxi. 2 - Neighborhood Selection Methods to Improve Recommendation Diversity Pelin Atahan, Assistant Professor, Ozyegin University, Nisantepe Mah Orman Sok Cekmekoy, Istanbul, 34794, Turkey, pelin.atahan@ozyegin.edu.tr Recommendation systems help match users with products in order to ease information overload of users. Recommendation systems are being utilized in many domains such as e-commerce, social networks, entertainment, and apps. While majority of the literature has focused on improving recommendation accuracy, recent studies have focused on other aspects of recommendation quality, such as diversity. However, there is typically a trade-off between recommendation accuracy and diversity. We propose a method that carefully selects neighbors in a way that will help learn the unique tastes of the users in order to improve recommendation diversity, while not compromising recommendation accuracy. 3 - Customer Differentiation with Shipping as an Ancillary Service: Free Service, Prioritization, and Strategic Delay Arvind Sainathan, Nanyang Business School, S3-B2a-03, Nanyang Business School, 50 Nanyang Avenue, Ntu, Singapore, 639798, Singapore, Asainathan@ntu.edu.sg A service provider/retailer offers ancillary service (e.g. shipping by an online retailer) to impatient and patient customers, who may be heterogeneous both in their delay sensitivities and service valuations. She can use prioritization and/or strategic delay to differentiate them. We characterize the optimal solutions under both exogenous and endogenous capacities, and find conditions in which these key service delivery features are present: (i) free service, (ii) single/differentiated service, (iii) split of customers, and (iv) strategic delay. . TB49

320