טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
M.Sc Thesis
M.Sc StudentNakibly Efrat
SubjectEstimating Waiting Times in Telephone Service Systems
DepartmentDepartment of Industrial Engineering and Management
Supervisor Professor Emeritus Avishai Mandelbaum


Abstract

Customer satisfaction is today the major concern for almost all companies. Businesses compete on both quality of products and service level.  One of the key attributes of service level is the time customers are waiting to be served, which has a direct effect on operational costs.

Surveys indicate that customer satisfaction with waiting can be improved by managing their expectations and perception of the waiting. One way of managing expectation is by providing information about the anticipated waiting.


With information about the anticipated waiting, the uncertainty level involved in waiting decreases, and abandonment after waiting is often replaced with balking upon arrival. The results are a higher customer satisfaction, as well as some objective improvements to the performance of the service system. Information regarding the anticipated waiting is of a particular importance in cases of invisible queues, where the uncertainty involved in waiting is high, and does not decrease over time.

The most common examples of service systems with invisible queues are call centers. Call centers have become the main service channel, and are estimated to handle about 70% of all customers’ interactions. Modern call centers employ many agents with different skills, and provide a wide range of services to various customers.


The goal of this work is to develop methods for estimating the expected waiting time of customers in different schemes of call centers. Estimations are made based on the system state at the time of estimation. We model the call center as a queueing system, and use accurate analytical models, approximations and simulations in order to estimate the waiting times.

Though we focus on call centers, the work can be applied more broadly to service systems in general.