2016 INFORMS Annual Meeting Program

TD31

INFORMS Nashville – 2016

TD31 202C-MCC Information Economics in Operations Sponsored: Manufacturing & Service Oper Mgmt, Service Operations Sponsored Session Chair: Senthil Veeraraghavan, University of Pennsylvania, The Wharton School, Philadelphia, PA, 19104, United States, senthilv@wharton.upenn.edu Co-Chair: Shiliang Cui, Georgetown University, 548 Rafik B. Hariri Building, 37th and O Streets NW, Washington, DC, 20057, United States, shiliang.cui@georgetown.edu 1 - Listen To The Crowd: Network Effects And Online Reviews In Restaurant SalesForecasting Shawn Mankad, Cornell University, Ithaca, NY, United States, smankad@cornell.edu, Qiuping Yu, Masha Shunko Using a comprehensive dataset from a major restaurant franchise, we forecast weekly store sales using classical measures of service quality from internal surveys at focal and neighboring stores, in addition to online ratings. Our results show that higher quality at the neighboring stores leads to higher sales at the focal store. We also find that the accumulation of online reviews reduces the importance of internal quality surveys at focal and neighboring stores as predictors. 2 - Information Sale And Competition Kostas Bimpikis, Stanford University, kostasb@stanford.edu, Davide Crapis, Alireza Tahbaz-Salehi This paper studies the strategic interaction between a seller of information and a set of firms competing in a downstream market. We show that the nature and intensity of competition play a first-order role in determining the seller’s optimal strategy. When firms’s actions are strategic complements (Bertrand competition), it is optimal for the seller to trade with all her customers. In contrast, when actions are strategic substitutes (Cournot competition), the seller maximizes her profits by restricting the supply of information and/or distorting its content. Furthermore we establish that her incentives to restrict the supply of information grow stronger in the presence of information leakage. 3 - Efficient Information Heterogeneity In A Queue Yang Li, CUHK Business School, liyang@baf.cuhk.edu.hk, Ming Hu, Jianfu Wang How would the growing prevalence of real-time delay information affect a service system? We consider an M/M/1 queueing system in which only a fraction of customers are informed about real-time delay. Perhaps surprisingly, we find that throughput and social welfare can be unimodal in the fraction of informed customers. In other words, some amount of information heterogeneity in the population can lead to strictly more efficient outcomes, in terms of the system throughput or social welfare, than information homogeneity. Moreover, we show that the impacts of growing information prevalence on system performance measures are determined by the equilibrium joining behavior of uninformed customers. 4 - Multi-stage Intermediation In Display Advertising Ozan Candogan, University of Chicago, ozan.candogan@chicagobooth.edu, Santiago Balseiro, Huseyin Gurkan We consider a setting where an advertiser seeks to acquire impressions from an advertising exchange through a network of intermediaries, and characterize mechanisms offered by strategic intermediaries when the advertiser’s value is private. Our results indicate that the position in the intermediation process has a significant impact on the profits of the intermediaries and the most profitable position depends on the underlying value distribution. Intuitively, when the private value distribution is heavy tailed, downstream intermediation positions are more profitable, and otherwise upstream positions are more profitable. We also analyze merger decisions of intermediaries. TD32 203A-MCC Revenue Mgt, Pricing IV Contributed Session Chair: Yanqiao Wang, UC Berkeley, Berkeley, CA, United States, yanqiao@berkeley.edu 1 - Application Of Optimization Techniques And Survival Analysis On Pricing And Revenue Management In Semiconductor Industry Amir Meimand, Pricing Scientist, Zilliant Inc, 1781 Spyglass Drive, Apt 359, Austin, TX, 78746, United States, amir.meimand@zilliant.com, Steve Tao, Lee Rehwinkel Prices in some markets such as the semiconductor industries tend to decrease

over time due to market pressure, product life-cycle, etc. Hence, it is desirable to reduce discounting behavior while maximizing sale/profit simultaneously. To meet this goal, we present a novel two-phase method. The first phase is based on an optimization model relies on elasticity estimation emerging from historical transactions. The second phase is modification of optimization solution based on price survival analysis to minimize the discount rate considering transaction date. We also present the numerical result of model applied to a real world problem with +2,000 products and +5,000 customers over a year. 2 - A Customized Dynamic Pricing Based Optimization Model For In- house Electricity Consumption Scheduling With Energy Storage Renewable Option Goutam Dutta, Professor, Indian Institute of Management, Wing 3, Room No 3H, Production Quantitive Methods Area, Ahmedabad, Gujarat, 380015, India, goutam@iima.ac.in, Krishnendra Mitra In this paper we propose a scheduling model for electrical appliances in a dynamic pricing environment. Initially we provide a vector of price points for the next twenty four hours. Then we develop an optimization model that minimizes cost to customer subject to different operating constraints of the appliances. We consider five different cases of price variation. We also study the effects of including energy storage and renewable energy generation at the consumer level. In this case we propose a linear price function that helps in automatically generating a price value for a time slot. 3 - Pricing Consumable Products To Maximize Profit Randy Robinson, Assistant Professor, Bemidji State University, 1500 Birchmont Dr. #30, Bemidji, MN, 56601, United States, rrobinson@bemidjistate.edu An introduction of a new consumable product to market is expected to have a sales growth rate that follows a sigmoid growth curve. If changes in price will affect the growth rate of this curve, what price should the producer charge to maximize profit over the time period in which they believe the product will remain popular? This talk will explore an explicit solution for the optimal price and the associated sensitivity analysis. 4 - Joint Optimization Of Capacitated Assortment And Pricing Problem Under The Tree Logit Model Yanqiao Wang, UC Berkeley, Berkeley, CA, United States, yanqiao@berkeley.edu, Zuo-Jun Max Shen Assortment and pricing decisions are of significant importance to firms and have huge influences on profit. How to jointly optimize over both assortment and prices draws increasing attention recently. However, in the existing literature, there is no flexible and comprehensive way to deal with the joint effects of assortment and price since the tangle between them makes the joint optimization problem less tractable. In this paper, we study the joint assortment and price optimization problem under the d-level nested logit model. Assume there are k lowest-level nests and each has n products, we develop an efficient algorithm that runs in O(kn^2) time to locate the joint optimal assortment and prices. TD33 203B-MCC Queueing Models III Contributed Session Chair: Petra Vis, VU Amsterdam, De Boelelaan 1105, Amsterdam, 1081 HV, Netherlands, petra.vis@vu.nl 1 - Meeting Service-level Constraints In Multi-class Service Systems Rene Bekker, VU Amsterdam, De Boelelaan 1081a, Amsterdam, 1081 HV, Netherlands, r.bekker@vu.nl, Ger M Koole, Petra Vis In many service systems, the service level (SL) is defined in terms of the tail probability of the waiting time. Different customer classes typically have different SL constraints. We first study a call blending system with an urgent (inbound) and best effort (outbound, email, call back) class, where the former has a severely more stringent SL. For threshold control, we show that the waiting time distribution is a mixture of exponentials. Second, we identify how to optimally assign agents to customers by exploiting the waiting time process of the first customer in line; we derive the value function for an isolated customer class and then apply one-step policy improvement. 2 - A General Workload Conservation Law With Applications To Queueing Systems Muhammad A El-Taha, Professor, University of Southern Maine, Department of Mathematics and Statistics, 96 Falmouth Street, Portland, ME, 04104-9300, United States, el-taha@maine.edu In the spirit of Little’s law L=\lambda W and its extension H=\lambda G we use sample-path analysis to give a general conservation law. For queueing models the law relates the asymptotic average workload in the system to the conditional asymptotic average sojourn time and service times distribution function. This law generalizes previously obtained conservation laws for both single and multi-server systems, and anticipating and non-anticipating scheduling disciplines. Applications to single and multi-class queueing and other systems that illustrate the versatility of this law are given.

344