Date: December 12, 2014

Time: 10:00 am

Venue: Room 1004, Administration Building, Zijingang Campus

Speaker: ZHANG Xiaowei, Assistant Professor, the Hong Kong University of Science and Technology

Topic: Are Call Center Arrivals Poisson?

Host: Prof. ZHOU Weihua, Associate Dean of ZJUSOM

Abstract: Poisson arrival process is one of the most fundamental assumptions in call center management. There have been mixed opinions on whether the Poisson model is appropriate, from both theoretical and statistical perspectives. In the first part of the talk, we will argue that timescale is the key that underlies the different conclusions in the literature regarding the Poisson-ness of the arrivals via various hypothesis tests on two well-known call center datasets. We show that the arrival process exhibits Poisson-like behaviors only on the microscopic timescale (e.g. seconds). However, non-Poisson features, such as excessive stochastic variability and autocorrelation, dominate on the timescale that really matters to call center management (e.g. 15 minutes and longer). In the second part of the talk, we propose a doubly stochastic Poisson process model that is parsimonious, analytically tractable, and capable of incorporating the critical elements on different timescales. We will further discuss the implication of this observation on call center staffing. In particular, the well-known “square-root” safety staffing rule may fail and a new staffing rule will be proposed.

About the Speaker: Xiaowei Zhang is an Assistant Professor in the Department of Industrial Engineering and Logistics Management at Hong Kong University of Science and Technology. He received his Ph.D. in Operations Research from Stanford University in 2011. His research interests include data-driven service engineering and Monte Carlo simulation.