Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

WB15

3 - Research on Supplier Selection in Logistics Service Supply Chain of China Railway Express Based on Improved DEA Method Yinying Tang, Southwest Jiaotong University, Chengdu, China, Qinglin Li, Qiyuan Peng, Yang Ge, Jingru Ren In this paper, a new framework and a new method are proposed to do the supplier selection in logistics service supply chain of CHINA RAILWAY Express (CRE). Firstly, considering CRE’s actual needs, we analyze the basic process of CRE selecting logistics supplier. secondly, considering the demand & supply characteristics of the freight market between China and Europe, a new selection framework containing service quality, service price, service capacity and cooperation risk as second-level indicators is given. Thirdly, considering the uncertainty of CRE’s international operation, the improved DEA (data envelopment analysis) Model is proposed and solved to evaluate and select suppliers. 4 - Analysis of a Manufacturer Issuing Free Gift Cards Yuefeng Li, University of Electronic Science and Technology of China, School of Management of Economics, West Hi-Tech Zone, Chengdu, 611731, China, Jingming Pan, Xiaowo Tang Manufacturers offer “free gift cards to consumers who buy the particular product. We structure a model of such gift card promotion in a two-product supply chain. It includes the manufacturer who plans to sponsor gift card promotion and takes the gift card cost, another manufacturer who is passively involved in the promotion and a retailer who sells the two products to consumers. The result shows that the two products are complementary with respect to consumers’ demand in the promotion. Moreover, the gift cards have a cannibalization effect on the retailer and a spillover effect on the passive manufacturer. Furthermore, we offer a straightforward strategy which could lead to a win-win-win situation. 5 - Supply Chain Coordination with Sales Effort under Cap-and-Trade Regulation Ting Ji, University of Science and Technology of China, Hefei, China This paper explores the production decision and the sales effort level in a make- to-order supply chain consisting of a manufacturer and a retailer under cap-and-trade regulation. We explore the supply chain coordination with wholesale price and cost sharing contracts. First, as the marginal sales effort cost increases, the retailer’s profit firstly increases and then decreases, while the monotonicity of the manufacturer’s profit depends on the cap. Second, both wholesale price and cost sharing contracts can coordinate the supply chain. Third, cap-and-trade regulation has a positive effect on supply chain coordination with the two contracts. n WB14 North Bldg 126C Choice Modeling in Online Platforms Sponsored: Manufacturing & Service Oper Mgmt/Service Operations Sponsored Session Chair: Heng Zhang, Marshall School of Business, Bridge Memorial Hall - BRI 401, 701 W. Exposition Boulevard, Los Angeles, CA, 90007, United States Co-Chair: Negin Golrezaei, Yale School of Management, Massachusetts Institute of Technology, 30 Memorial Dr, Cambridge, MA, 02142, United States 1 - Dynamic Learning in Assortment Selection: A Thompson Sampling Approach Vashist Avadhanula, Columbia University, 3022 Broadway, 4th Floor West, New York, NY, 10027, United States, Assaf Zeevi, Shipra Agrawal, Vineet Goyal We consider a dynamic assortment selection problem, where in every round the retailer offers a subset (assortment) of N substitutable products to a consumer, who selects one of these products according to a multinomial logit (MNL) choice model. The retailer observes this choice and the objective is to dynamically learn the model parameters, while optimizing cumulative revenues over a selling horizon of length T. We present an approach to adapt Thompson Sampling to this problem and show that it achieves near-optimal regret as well as attractive numerical performance on real world data. 2 - An Optimal Assortment for Complements Xi Shan, University of Texas at Dallas, Richardson, TX, 75080, United States, Dorothee Honhon, Suresh P. Sethi, Chenglin Zhang Using complements is common practice in assortment planning: Restaurants offer meals with drinks or a music environment, and even allow consumers to bring their own wine. Department stores use one category as complement to the other categories to stimulate purchases. We consider complements in the canonical assortment model (van Ryzin and Mahajan 1999) wherein a newsboy stimulates purchases by offering complements. We show that with complements the retailer uses lower variety and achieves higher profit under certain conditions. We obtain

the Nash equilibrium for category managers in the category management problem and demonstrate the coordination mechanism. 3 - Assortment Optimization for an Omnichannel Retailer with Features-based Value Boosts and Discounts Venus Lo, Cornell University, 206 Rhodes Hall, Cornell University, Ithaca, NY, 14853, United States, Huseyin Topaloglu We consider a retailer who sells online and in a physical store offline. Customers have initial product valuations, but boost/discount values if they observe under/over-hyped product features in-store. The retailer selects an assortment to offer in-store to maximize expected revenue while managing customers’ valuation. Our model organizes products on a tree so that a leaf is a product and a non-leaf vertex is a feature common to leaves in its subtree. Online and offline customers have different consideration sets but both adjust valuations based on the offline assortment. This problem is NP-hard and we present a FPTAS, which performs much better than its theoretical guarantee under computation. n WB15 North Bldg 127A Queueing Games Sponsored: Applied Probability Sponsored Session Chair: Dongyuan Zhan, University College London, London, United Kingdom 1 - Invite Your Friend and You’ll Move Up in Line: Optimal Design of Referral Priority Programs Luyi Yang, Johns Hopkins University, 100 International Drive, Baltimore, MD, 21202, United States In a referral priority program, existing customers on a waitlist for products or services (e.g., mobile apps) can invite their friends to also sign up so that they can jump up in line and get early access. Firms vary in the amount of priority incentives they give, which begs the question of what the optimal program should be. We take a mechanism design approach combined with the achievable region method to this question. Our unifying framework subsumes different schemes used in practice (namely, FIFO, full priority, partial priority, and strategic delay) as special cases and characterizes the conditions under which each scheme is optimal. 2 - An Equilibrium Analysis of a Multiclass Queue with Endogenous Abandonments in Heavy Traffic Xiaoshan Peng, Indiana University, Kelley School, Bloomington, IN, United States, Baris Ata We study a multiclass queueing system with endogenous abandonments where congestion affects customers’ abandonment behavior and vice versa. In the abandonment model, customers take the virtual waiting time distribution as given. Customers are forward looking and make wait or abandon decisions dynamically to maximize their expected discounted utilities. The queueing model takes the customers’ abandonment time distribution as an input and studies the resulting virtual waiting time distribution. We show that there exists a unique equilibrium in which the customers’ abandonment time and the virtual waiting time for the various classes are consistent in the two models. 3 - Service Pricing when Customers Collude Chenguang (Allen) Wu, Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong, Chen Jin, Senthil Veeraraghavan Customers’ strategic behavior of service procurement is typically modeled by assuming that customers are rational individuals. However, the understanding of customers’ behavioral heterogeneity and the resulting implication on the service provider’s operational problems is lacking. In this paper, we model the behavioral heterogeneity of customers by considering two customer segments, individual and group customers, who are endowed with different decision structures. Our results explain the operational incentive of a common practice in service systems: group customers receive special rates for service and enjoy a high priority in the admission line. 4 - Harnessing the Double-edged Sword: Information Disclosure on Ride-hailing Platforms Dongyuan Zhan, University College London, School of Management, Gower Street, London, WC1E 6BT, United Kingdom, Leon Yang Chu, Zhixi Wan We consider a ride-hailing platform that provides free information service to taxi drivers. Upon receiving a rider’s request, the platform broadcasts the rider information to the idle taxi drivers, who may accept or decline the request based on the customer’s profitability. The first driver who accepts the request gets the ride. If no driver accepts it then the rider leaves. As a result, information disclosure is a double-edged sword for drivers’ profits: they may take more profitable rides via inefficient idling. We study the profit implication of this information disclosure, and propose a simple broadcast policy that can achieve close to the first best when the system is large.

445