Rewarding Suppliers' Performance via Allocation of Business

摘要

biProblem definition : Buyers can encourage competition among multiple suppliers of goods or services by redistributing their shares of the business based on recent observable performances. This is usually needed because actual efforts or investments by suppliers are typically unobservable, a moral hazard issue. This poses the problem: How should such redistribution be linked to the suppliers’ recent performances to motivate the best supplier performance? biAcademic/practical relevance : Such performance-based allocations of business are a common practice in supply management when a buyer sources from multiple suppliers. But there is little research on the structural characteristics of the optimal allocation rule in the presence of moral hazard. biMethodology : We apply principal-agent theory to model the buyer’s incentive design problem. We analytically derive the optimal allocation rule and conduct numerical experiments to evaluate the effectiveness of several simple heuristics against the optimal rule. biResults : Our key result is that the buyer should divide the purchase among suppliers based on their relative performance deviations from the targets set by the buyer. A ratio rule is used for positive deviations, whereas a winner-take-all may be applied when some suppliers underperform. Our result generalizes to a range of environments with different cost structures, distributions of performance outcomes, correlated performance, etc. When the optimal rule is used to benchmark several more practical heuristics, the only really poorly performing heuristic is one that does not use the relative performance deviations. biManagerial implications : Buyers’ allocations should reflect suppliers’ relative performance deviations from the desired targets, not absolute performances. Performance deviations can be weighted in some way, but this is a second-order gain. A ratio rule is needed to reward good performance, whereas a winner-take-all rule with more extreme allocations is used for penalizing poor performance.