A reorder point is the stock level at which a replenishment order should be raised — classically, the demand expected during the resupply lead time plus a safety buffer — so that new stock arrives before the existing stock runs out.
Why the reorder point matters
The reorder point encodes the single most useful fact in inventory control: ordering is not instant. Between deciding to order and receiving goods, demand continues. A business that reorders when stock hits zero is guaranteed to be out of stock for the whole lead time.
It also converts a continuous judgement into a trigger. Rather than asking "should we order this?" for every item every week, the question becomes "which items have crossed their line?" — which is answerable by a system rather than by a person with good memory.
How the reorder point is calculated
The standard form is:
reorder point = (average demand per period × lead time in periods) + safety stock
- Average demand — consumption over a trailing window long enough to smooth noise.
- Lead time — time from raising the order to the stock being available to sell, which includes receiving and putaway, not just transit.
- Safety stock — cover for variability in both demand and lead time. Without it, the reorder point protects only against average conditions, and average conditions are not when stock-outs happen.
Safety stock is where judgement enters. Statistically it scales with the variability of demand and lead time and with the service level targeted — the higher the confidence of never running out, the more capital is parked. Chasing the last few percent of availability is disproportionately expensive.
An example
An item sells 20 units a week. Lead time is two weeks. Lead-time demand is 40 units. Demand varies by about 5 units a week, and the supplier is occasionally a week late, so a safety stock of 25 units is set. The reorder point is 65. When stock falls to 65, an order is raised; if everything runs to average, roughly 25 units remain when it arrives. If the supplier is a week late, the buffer absorbs it.
Common variations
- Min/max. The minimum is the reorder point; the maximum sets how much to order — order up to the maximum whenever the minimum is breached.
- Fixed-quantity reorder. Always order the same quantity, typically an economic or lot size, whenever the point is crossed.
- Periodic review. Check at a fixed interval rather than continuously, in which case the reorder point must also cover demand until the next review.
- Dynamic reorder point. Recomputed as demand and lead time are observed, rather than set once.
Limitations worth stating
The formula assumes demand is reasonably stable and lead time reasonably known. Promotions, seasonality and new launches all break the first assumption, and they break it exactly when volumes are largest. A reorder point calculated from parameters entered once and never revisited will be confidently wrong — and, because it is arithmetic, it will look authoritative while being so. Reorder points also treat each item independently, so they do not naturally account for constraints shared across the catalogue, such as a budget or a vehicle.
How xMatix supports reorder points
In xMatix Inventory, stocking parameters are held per item: minimum and maximum stock level, reorder level and reorder quantity, minimum and maximum order quantity, lot size, lead time and average consumption. These feed the replenishment engines in Procurement — minimum stock level, replenish-to-maximum, and consumption-based forecasting — which can run on a schedule to produce dated suggestion runs.
Crucially, the calculation nets what is already committed before comparing to the reorder point: stock on hand, in-transit shipments, pending purchase orders, open requisitions and unfulfilled sales orders. A line that looks below its reorder point on hand alone may be fully covered once the pipeline is counted, and this is where most naive implementations over-order.
The same netting powers a check in the other direction: when a purchase order line would bring in stock the pipeline already covers, Sense Assist flags it with the arithmetic shown and offers the corrected quantity. Every suggested line also carries its own inputs — projected quantity, available quantity, each pipeline component, lead time and average consumption — so a buyer can see which assumption drove the number.
Note that the rules-based engines read average consumption and lead time as item attributes, so the quality of the trigger depends on those values being kept current. Where demand is worth predicting rather than assuming, statistical and machine-learning forecasting model it from historical sales, movement patterns and seasonal variation instead.
Related: suggested order quantity · ABC and FSN analysis · three-way match