A Bayesian approach to service level performance monitoring in supplier,provider relationships

Techniques are presented for monitoring the performance of a delivery service. The techniques were developed from the practical need to support two parties: a virtual retailer that out-sources the home delivery of its goods and the logistics business that provides its courier service. The techniques are relevant to contractual negotiations and performance monitoring. A graphical tool for hypothesis testing is presented as an aid to contract negotiation and a dynamic model is developed to track performance over time for operational management. The approach is based upon the beta-binomial distribution, within a Bayesian decision context.