EURO Working Group on Stochastic Modelling

FIRST MEETING: 19-21 APRIL 2006, AMSTERDAM

meeting 1
     program
     practical information
     register
     participants
meeting 2
meeting 3

Abstracts

Robert Shumsky, Capacity Estimation and Optimal Staffing for an email Contact Center

We develop a method to estimate the capacity of agents that answer email in an outsourced contact center (a 'vendor'), given historical data that has been truncated at an upper limit. In our system, customers of an outsourcer send emails to the outsourcer with requests for sales or service. These emails are collected in a universal queue, and the outsourcer hires vendors that can access the universal queue and send replies to the emails. For our vendor, the outsourcer sets a target number of emails that the vendor should serve for each day. The Service Level Agreement (SLA) in the contract between the outsourcer and the vendor specify that the vendor should serve at least 90% and not exceed 110% of the daily target. Therefore, the number of emails actually answered on any given day is truncated at 110% of each day's target, and capacity estimates based on the observed number of emails answered are extremely biased. We use the data to find maximum likelihood estimates of the actual capacity and use these estimates to calculate optimal staffing levels to meet future demand. The implementation results demonstrate that the method provides accurate staffing recommendations.