Probabilistic Inventory Planning for a Hardware Retailer: Safety Stock and Reorder Point Assessment

Abstract

This study examines the information required to apply a probabilistic inventory policy to 9 mm plain square bars at Toko Besi Sadarman, Gunungsitoli. The analysis uses the annual demand totals reported for 2022 and 2023, checks the internal consistency of the proposed inventory thresholds, and compares their assumptions with the cited case study. Annual demand increased from 4,319 to 4,621 units, equivalent to 6.99%. This change describes annual volume but does not establish daily demand variability or a probability distribution. The proposed safety stock of 79 units and reorder point of 206 units imply mean demand during replenishment lead time of 127 units. However, the daily demand series, lead-time observations, service target, and parameter estimates needed to reproduce these thresholds are unavailable in the supplied evidence. A continuous-review framework is therefore specified to separate order quantity, expected lead-time demand, and safety stock. The analysis identifies the records and performance measures required before implementation. The resulting framework can support inventory planning and staff learning, but neither an optimal stochastic policy nor realized cost savings can be established from the available data. Further validation requires consistent operating-day definitions, reliable transaction records, and out-of-sample policy evaluation.