Effect of delivery timing interaction and disasters on co-supplier evaluation

Most supplier evaluation and selection models found in the literature evaluate candidate suppliers individually on various dimensions such as price, quality and delivery timing. Combinations of suppliers used together are modelled as linear combinations of the individual suppliers' characteristics on each relevant dimension. The existing models are contextual in that they incorporate factors of importance to the buyer. However, they do not consider the interaction between co-suppliers and the interaction between the suppliers' characteristics and the ordering policy used by the buyer. The present paper proposes a set of stochastic models that explicitly incorporate these interactions. Computational experiments are performed and the results reported. The results show that taking into account the ordering policy and the interaction effects it can induce can lead to co-supplier evaluations that can differ significantly from those obtained from existing models that do not consider these effects.