Using Scheduled Ordering to Improve the Performance of Distribution Supply Chains

We study a supply chain with one supplier and many retailers that face exogenous end-customer demands. The supplier and the retailers all try to minimize their own inventory-related costs. In contrast to the retailers' newsvendor-type ordering behavior (under which retailers may place orders freely in every period), we propose two scheduled ordering policies: the scheduled balanced ordering policy (SBOP) and the scheduled synchronized ordering policy (SSOP). Under both the SBOP and SSOP, retailers are allowed to order freely only in one period of an ordering cycle, and receive fixed shipments in other periods. Retailers take turns to order freely under the SBOP, while under the SSOP all retailers order freely in the same period. With the average supply chain cost per period as the performance measure, we identify mathematical conditions under which scheduled ordering policies outperform the newsvendor-type ordering. Through a large-scale numerical study, we find that scheduled ordering policies are most effective when (i) the supplier's holding and expediting costs are high and the retailer's backorder cost is small, (ii) the end-customer demand variance and correlation are high, and (iii) the supplier's capacity is high. In addition, we observe that the behavior of the SSOP often complements that of the SBOP. Whereas the SBOP is better than SSOP when the supplier's capacity is low and when the end-customer demand correlation level is high, the SSOP is better when the opposite conditions prevail.