Inventory is often discussed as a number of days on hand. That can hide the real engineering question: which uncertainty is the buffer meant to absorb, and where should it sit? For African operations with variable import, transport and supplier lead times, a blanket increase in stock ties up cash while leaving critical items exposed.
Separate demand from replenishment risk
Start with one product family. Record weekly demand, actual order-to-receipt lead times and the causes of delay. A nominal supplier lead time is not the same as the distribution of real arrivals. If a port hand-off, payment release or internal approval adds unpredictable days, the buffer calculation must reflect that variation.
Set a service target by business consequence. A low-value component that stops an entire production line deserves different treatment from a slow-moving substitute. Use a reorder point built from expected demand during lead time plus a safety allowance for observed variability. Test the result against past stockouts and excess stock, rather than treating a formula as proof.
Place the buffer where it protects flow
More inventory at every site can multiply working capital without improving customer service. Ask whether pooled stock at a regional hub, supplier-held inventory, a second source or a shorter approval path would give better protection. MIT supply-chain research found that supply lead-time variation drove high safety-stock values in two studied products; reducing the underlying variation was a practical lever for lowering buffers.
This is a systems decision. A local purchasing team rewarded only for unit price may choose large, infrequent orders, while operations absorbs the resulting shortages and storage burden. Make the trade-off visible across procurement, finance, warehousing and customer service. Our earlier supply-chain control-tower guide shows how to give those teams one view of exceptions and ownership.
Run a measured six-week test
Choose 20 to 30 items with a mix of criticality and demand patterns. Establish baseline fill rate, stockout days, expediting cost and inventory value. Change one policy at a time, assign an owner for exceptions and review actual arrivals weekly. Compare service and cash together before expanding the policy.
When the same shortage returns despite higher stock, investigate the feedback loop rather than ordering more. The systems-thinking guide on recurring fixes offers a way to trace those unintended consequences. The aim is reliable flow with the smallest justified buffer, supported by better supplier and internal decisions.
Research context: MIT Center for Transportation and Logistics, 2020.