Analysis of Teller Service Times in Retail Banks

Résumé

Abstract. Retail banks typically staff teller lines as if all tellers had the same service capabilities. Likewise,queuing models to recommend staffing levels often assume that all servers have the same service timedistribution and that this distribution is exponential. These assumptions are motivated more by operational andanalytical convenience than supported by data. Until recently there have been little available data to test theseassumptions and their impact on customer service. This article presents results from analyzing a large datasetof teller service times from a retail bank. The validity of common teller staffing assumptions are explored alongwith the impact that deviations from the assumptions have on staffing recommendations obtained from queuingmodels.