Track 1 Session 6

9:10 to 10:10 a.m. Wednesday, June 16, 2010

Augmenting Standard Forecasting Methods by Leveraging Warranty Data to Inform Call Center Staffing Models

The management of staffing levels by qualified personnel is critical to the success of any customer relationship management (CRM) program. One type of customer service operation found in many industries is call center operations. In organizations where a formalized CRM exists, call center operations will usually fall under this domain. This presentation demonstrates the importance of call center staffing and how warranty analysis can be a valuable tool in informing the call center operations. Working under a proposed baseline scenario, a plausible set of modeling conditions are identified and subsequently varied within the boundaries of the design space.

Factors of interest were as follows: 1) Time series analysis. 2) Warranty analysis that imposed differing hazard functions following the two-parameter Weibull distribution (infant mortality, random failures and wearout) and was used in the context of how warranty data are typically collected. 3) Inclusion of inventory lag values. 4) Use of Erlang_c as a basis for staffing levels for given constraints.

Key Words: Warranty, Call Center Operations, Product Reliability, Staffing Models, Time Series, Regression, Erlang, Weibull, Poisson Processes

Jason Overstreet and Sankaran Mahadevan

Vanderbilt University

Nashville, Tennessee