.png)
Set one trigger, not a stockout promise
For a continuously reviewed imported SKU, start with:
Reorder point = forecast demand during end-to-end replenishment lead time + safety stock
The result is the inventory position at which a replenishment decision should be triggered. It is not a guarantee against stockouts, and it does not tell you how many units to order.
Keep three terms separate:
- Lead-time demand is the quantity expected to be consumed while replacement stock moves from purchase-order release to saleable receipt.
- Safety stock is an additional buffer chosen for uncertainty and the business's service-versus-inventory trade-off.
- Order quantity is the amount purchased after considering the supplier's MOQ, order multiples, cash, capacity, storage and obsolescence risk.
That separation matters for importers because replenishment can span production, export handling, international freight, Australian clearance, local delivery and receiving. A simple supplier lead-time field may represent only one part of that chain.
Measure the lead time your inventory actually experiences
The lead time in the formula must use the same start and finish for every receipt. For an Australian importer, a practical end-to-end definition is the elapsed calendar time from authorised purchase-order release to the moment the stock is available for its intended use or sale.
Start and stop the clock consistently
Record the planned and actual dates for:
- purchase-order approval and deposit clearance;
- material readiness and production;
- final factory release and origin handover;
- export handling and international transport;
- Australian customs and biosecurity release where applicable;
- final delivery; and
- incoming inspection and release to available inventory.
The supplier production lead-time record and sea-freight transit plan can supply components of this timeline. Do not add independently sourced averages that double-count overlapping stages.
Use actual completed receipts to challenge the planning value. A 45-day supplier promise is not a 45-day replenishment lead time if production starts late, the cargo misses a cut-off or stock remains quarantined after delivery.
Forecast demand over the same time unit
Match the demand rate and lead-time unit. If demand is expressed in units per calendar day, lead time must also be in calendar days. If the planning system uses weeks, convert both inputs before calculating.
Build the demand input at SKU and stocking-location level. Start with clean sales or consumption history, then identify known events that make the future unlike the average:
- confirmed customer commitments;
- promotions or launch activity;
- seasonality;
- a listing, price or channel change;
- a substitute SKU entering or leaving the range; and
- periods when sales were constrained because stock was already unavailable.
A raw sales average can understate demand when stockouts suppressed recorded sales. Conversely, a one-off promotion can overstate the ordinary run rate. Record the adjustment and its owner rather than silently changing the number.
Choose safety stock separately
Safety stock sits above expected lead-time demand. Its purpose is to absorb some difference between plan and reality, such as demand arriving earlier than forecast or replenishment taking longer than expected.
More buffer is not automatically better. It uses cash, storage space and working capital, and it can increase markdown, expiry or obsolescence exposure. Less buffer accepts more service risk. The right trade-off depends on the SKU's demand pattern, replenishment variability, margin, criticality, substitution options and recovery choices.
Do not copy a universal number of safety-stock days from another importer. A statistical setting also needs fit-for-purpose data and a defined service objective. Intermittent demand, short history, lost-sales data, seasonality or linked demand-and-lead-time shocks can make a simple model misleading. Escalate those cases to a qualified inventory planner.
Calculate a transparent first trigger
Consider a fictional SKU with these planning inputs:
| Input | Example | Treatment |
|---|---|---|
| Forecast average demand | 18 units/day | Current SKU/location forecast |
| End-to-end lead time | 92 days | PO release to saleable receipt |
| Forecast lead-time demand | 1,656 units | 18 x 92 |
| Safety-stock policy | 420 units | Illustrative management input, not a benchmark |
| Reorder point | 2,076 units | 1,656 + 420 |
When the relevant inventory position reaches or falls below 2,076 units, the business reviews or releases replenishment under its approved ordering policy. The calculation exposes the assumptions; it does not prove that 420 units is the correct buffer, eliminate stockout risk or guarantee that the next receipt will arrive in 92 days.
The worked figure is deliberately fictional. Replace every input with the SKU's controlled data and retain the calculation date.
Trigger on inventory position, not a warehouse glance
Physical on-hand stock alone can give the wrong signal. Some units may already be committed to orders or unavailable for quality reasons, while credible purchase orders may already be in the replenishment pipeline.
Microsoft documents one planning concept as net flow position: on-hand plus on-order minus qualified demand. Oracle likewise describes reorder-point planning against available quantity that includes planned receipts. Using those concepts, OPL's practical control is:
Planning inventory position = usable on-hand + firm incoming supply due within the defined horizon - separately qualified demand not already reflected in the forecast or supply record
Apply the organisation's own planning horizon and inventory-state definitions consistently. Count incoming supply only when it is firm, due inside that horizon, available to the same SKU and location, and not already allocated. Do not count quarantined, damaged or allocated units as freely available. Most importantly, do not subtract a customer commitment again if it has already been included in the demand forecast or netted elsewhere in the planning record.
In the fictional example, assume:
- usable on-hand: 1,800 units;
- firm incoming supply due within the defined horizon, available to this SKU and location and not otherwise allocated: 500 units; and
- separately qualified customer demand not already included in the 18-unit/day forecast: 250 units.
The planning inventory position is 2,050 units. That is below the 2,076-unit reorder point, so the replenishment review is triggered. The 26-unit gap does not mean the new order should be for 26 units.
Keep reorder timing separate from order quantity
The reorder point answers when to act. The purchase quantity answers how much to buy. Combining them can hide an MOQ-driven overstock decision inside what appears to be a demand calculation.
Set order quantity only after checking:
- supplier MOQ and order multiples;
- unit economics and the cash-flow timeline;
- storage, pallet and container constraints;
- product life, packaging changes and obsolescence risk;
- production and freight economics; and
- alternative recovery options, including a split shipment or different mode.
If the MOQ is larger than the defensible replenishment need, treat that as a commercial exception. Options may include MOQ negotiation, a staged call-off, a revised assortment or a conscious excess-stock decision. Do not inflate the reorder point merely to disguise the order-size constraint.
Back-test and govern the setting
A reorder point should change when its inputs change. Review each important SKU on a fixed cadence and after material exceptions.
| Control field | Evidence | Escalation signal |
|---|---|---|
| Demand forecast | Version, period and owner | Repeated forecast error or stockout-suppressed sales |
| End-to-end lead time | Planned and actual milestone dates | Late stages recur or the route changes |
| Safety-stock policy | Method, assumptions and approval | Buffer is repeatedly consumed or never approached |
| Inventory position | On-hand, unavailable, incoming and commitments | System states do not reconcile to physical stock |
| Order outcome | Trigger date, PO date and saleable receipt | Manual delay after the trigger or receipt after need date |
Back-test the rule against completed cycles. Ask whether the trigger fired early enough, whether incoming orders were counted accurately, which lead-time stage moved, and whether the buffer covered ordinary variability or merely hid a persistent supplier problem.
Recheck immediately after a supplier, route, freight mode, warehouse, product, channel or promotion change. An air-versus-sea decision, for example, changes both cost and replenishment time; it should not be patched into the model as an unexplained manual override.
Put the next review on the calendar
For each material SKU, store the demand version, lead-time history, safety-stock method, reorder point, order-quantity rule, inventory-state definitions, approver and next review date. Give one owner responsibility for investigating exceptions rather than merely updating the number.
Begin with the transparent formula, operate it against a defined inventory position, and learn from every completed receipt. The value is not a permanently perfect trigger. It is an auditable decision rule that changes when the evidence changes.