2015 Informs Annual Meeting

WB53

INFORMS Philadelphia – 2015

2 - Efficient Workforce Size and its Schedule in the Retail Store Peeyush Pandey, Doctoral Student, IIM INDORE, FPM block. room no.-315, IIM Indore, Prabandh Shikhar,, Indore, MP, 453331, India, f12peeyushp@iimidr.ac.in, Bhavin Shah, Ashish Sadh, Hasmukh Gajjar A Stochastic model is proposed to determine optimal workforce size at the different point of time in the retail store. Further this optimal size is used as an input to workforce-scheduling model considering uncertain and uneven customer traffic, union contracts, labor laws, company policies etc. Proposed optimization model and solution methodology guarantees to provide efficient workforce schedule. 3 - Beyond the Forecast – Risk Based Promotion Management Applied in a Grocery Retail Environment Ted Matwijec, Managing Director, ACT Operations Research, 1345 Legendary Lane, Morrisville, NC, 28202, United States of America, ted.matwijec@act-operationsrearch.com, Raffaele Maccioni Promotions are one of the biggest challenges for management for businesses focused on consumer retailing. Each product promotion campaign must contribute to the businesses to attract new customers and still retain existing customers. In this paper we analyze the modeling, which is applied to promotions used at a national grocery retailer. The solution minimizes the risk of running promotions by using optimization techniques which ultimately benefits a retailer profitability. 4 - The Softer Side of Assortment Planning Nazrul Shaikh, Assistant Professor, University of Miami, 268, Our research focuses on the sensitivity of the optimal assortment plans to soft costs, such as backorder costs, and proposes a methodology for generating robust solutions to the assortment planning problem. This adds a useful element to extant research that only focuses on improving future demand estimates and substitution probabilities. WB53 53-Room 107B, CC Frontiers of Behavioral Operations Research Sponsor: Behavioral Operations Management Sponsored Session Chair: Karen Zheng, MIT, 77 Massachusetts Avenue, Cambridge, MA, 02139, United States of America, yanchong@mit.edu 1 - A Cognitive Strategy for Reducing Managerial Bias under Censored Demand Jordan Tong, Assistant Professor, University of Wisconsin at Madison, WI, United States of America, jordan.tong@wisc.edu, Daniel Feiler, Richard Larrick Existing evidence suggests managers exhibit a censorship bias: demand beliefs and order decisions are biased low when demand is censored by the inventory level. We propose a new cognitive strategy, which is easily implementable in practice, for reducing this bias. Two experiments show that the strategy improves managerial performance across multiple profit margins and outperforms other plausible debiasing techniques. The results also illuminate the underlying causes of the censorship bias. 2 - The Bright and Dark Sides of Perception Biases in Inventory Decisions Yaozhong Wu, National University of Singapore, NUS Business School, Singapore, Singapore, yaozhong.wu@nus.edu.sg We study the impact of perception biases in competing inventory decisions. We analyze how a manager’s perception bias affects each other’s inventory decisions and performances in strategic interactions, and more importantly who benefits from these biases in the short and long runs. We show that a perception bias can serve as a competitive advantage in the sense that a biased manager can achieve a higher profit than an unbiased competitor. 3 - The Exploration – Execution Transition in Product Development: An Experimental Analysis Stephen Leider, University of Michigan, 701 Tappan Ave R4486, We examine experimentally the effect of exogenous and endogenous transition times on performance in a creativity task. Subjects who choose their own transition perform worse, even when the transition time is similar. We find that early design prototyping , testing and even failure improves performance, while the number of ideas does not. Idea selection and execution accounts for more of the performance difference than idea generation. Ann Arbor, MI, 48104, United States of America, leider@umich.edu, Evgeny Kagan, William Lovejoy McArthur Engineering Building, University of Miami, Coral Gables, FL, 33146, United States of America, n.shaikh@miami.edu, Shelby Koos

4 - Pricing When Customers have Limited Attention Tamer Boyaci, McGill University, Montreal, Canada, tamer.boyaci@mcgill.ca, Yalcin Akcay

We study optimal pricing when customers ha e limited attention and capability to process information about the value (quality) of the offered products. We model customer choice based on the theory of rational inattention in the economics literature, and capture not only the impact of true qualities and prices, but also the intricate effects of customer’s prior beliefs and cost of information acquisition and processing. We consider both monopolistic and competitive settings. WB55 55-Room 108B, CC Decision Analysis I Contributed Session Chair: Xiaoya Xu, PhD, University of Macau, S9-7025, Macau, Macau, xlwxxy@gmail.com 1 - Managing Rental Products with Breakdown Mohammad Firouz, PhD Candidate, The University of Alabama, 610 13th St, Apt. 19, Tuscaloosa, AL, 35401, United States of America, mfirouz@crimson.ua.edu, Burcu Keskin, Linda Li We investigate capacity planning problem of a rental system. Products owned by the system have a life time distribution and may breakdown for an uncertain duration of time. We solve the proposed Quasi-Birth and Death (QBD) model via matrix analytic methods. We also compare the result of our proposed approximation to the problem with the QBD. In our analysis, we prove the convexity of the approximate cost objective function and show conditions under which it may or may not be accurate. 2 - Repeat Purchase Prediction of Loyalty Customers using Survival Analysis The prediction of purchase behavior of loyalty customers in terms of who will repeat the purchase and when the repeat purchase of a particular category will be done. We use Survival Model combined with Logistic Model and Decision Tree to find when and who will repeat the purchase while retaining the robustness, interpretability of the models. 3 - Selling Probability Service: Profiting from Market Segment and Discrimination Xiaoya Xu, PhD, University of Macau, S9-7025, Macau, Macau, xlwxxy@gmail.com, Zhaotong Lian, Xin Li, Pengfei Guo We consider a setting where goods A and B are offerred to customers of three types: buyers who desire for A, buyers who desire for B, and the third type of buyers who are flexible.A probability selling service is created by the seller to offer the option of getting an unknown item either A or B, targetting at the third type of customers. This paper investigates the role of probability selling service provider in such a setting as Priceline and Hotwire in market segmentation. WB57 57-Room 109B, CC Advances in Sustainable Energy and Economic Systems Analysis Sponsor: ENRE – Energy II – Other (e.g., Policy, Natural Gas, Climate Change) Sponsored Session Chair: Soheil Shayegh, Postdoctoral Research Scientist, Carnegie Institution for Science, 260 Panama St., Stanford, Ca, 94305, United States of America, sshayegh@carnegiescience.edu 1 - Adapting to Rates of Climate Change Soheil Shayegh, Postdoctoral Research Scientist, Carnegie Institution for Science, 260 Panama St., Stanford, Ca, 94305, United States of America, sshayegh@carnegiescience.edu, Ken Caldeira, Juan Moreno-Cruz Most of the discussion around adaptation in IPCC AR5 and other sources has focused on amounts of climate change. However, it is becoming increasingly clear that, as climate continues to change, people and ecosystems will need to continuously adapt to a moving target. We have developed a model that convincingly makes this point and illustrates it with a quantitative example involving coastal development in the face of ongoing sea level rise. U Dinesh Kumar, Dr., Professor in Quantitative Methods & Information Systems, Indian Institute of Management, Bannerghatta Road, Bangalore, 560076, India, dineshk@iimb.ernet.in

418