The subject of the study is a Singapore-based third-party logistics provider (3PL) that manages a network of retail stores on behalf of a public sector client. The network currently stocks a large number of slow-moving items (SMI) at store-level, which leads to significant working capital tied-down on inventory. This paper studies into and quantifies the effectiveness of a proposed approach to centralize the retail network’s inventory of slow-moving items to a pooling point. Results from the analysis show that under the base scenario with just 5% of items (by unit sales volume) stocked centrally, the aggregation of demand across retail outlets can contribute to a net 27.3% reduction in fulfillment costs. Furthermore, a selective centralization strategy can also be designed to be quite robust (i.e. costs would be close to optimum) within a broad range of SMI definitions.