The guide
Guide to automatic stock replenishment
Replenishing at the right time means keeping the delivery dates promised to customers without tying up too much cash in the warehouse. This guide explains how to calculate a reorder point and safety stock, choose quantities and prepare purchases automatically without losing control.
How does automatic stock replenishment work?
Automatic replenishment rests on a simple idea: watch each item's stock continuously and trigger an order as soon as it falls below a threshold, the reorder point. That threshold is calculated so that new goods arrive before the remaining stock runs out.
For this to work, you need accurate real-time stock, which perpetual inventory provides, and reliable parameters per item: average consumption, supplier lead time, safety stock, pack size and minimum order. In Numezis, the stock taken into account is available stock, meaning physical stock minus customer reservations, plus stock on order from purchase orders already placed. The agent prepares the purchase order, but it is only sent after approval by an authorised person. You keep the decision, without having to watch every item yourself.
How do you calculate a reorder point?
The basic formula is: reorder point = average daily consumption × lead time in days + safety stock. If an SME uses 20 boxes of gloves a day, its supplier delivers in 6 working days and it keeps a safety stock of 120 boxes, the reorder point is 20 × 6 + 120 = 240 boxes.
The difficulty is keeping these parameters up to date. Consumption changes with seasons and customers, and supplier lead times sometimes lengthen without warning. Numezis recalculates average consumption over a rolling period, for example 90 days, and measures each supplier's real lead time at every receipt. Reorder points adjust automatically or are proposed for approval, as you prefer. Seasonal items can have different parameters depending on the time of year.
How much safety stock should you keep to avoid stock-outs?
Safety stock covers the unexpected: higher consumption than planned, a late supplier, an incomplete delivery. Too low, it exposes you to stock-outs; too high, it ties up cash and warehouse space, and raises the risk of a write-down at the close.
A simple approach is to set it in days of consumption, for example five or ten days depending on how important the item is. A more precise approach calculates it from the variability of consumption and lead times, and the target service level: 95 % for a standard item, 99 % for a critical one. Numezis offers both. It can also distinguish items by ABC class: A items, which weigh heavily in stock value, keep tight safety stock and are watched closely; inexpensive C items can afford a wider margin.
What quantity should you order each time?
Once the reorder point is reached, you still have to decide how much to order. Ordering little and often lowers average stock but multiplies ordering and transport costs. Ordering a lot gets better prices but ties up cash. The economic order quantity, or Wilson formula, looks for the balance between these two costs.
In practice, other constraints come in: the supplier's minimum order, pack multiples, full pallets, discount tiers, free-delivery thresholds. Numezis takes these constraints into account to propose a realistic quantity, and groups needs for the same supplier into one order to reach free delivery. The approver sees the proposed quantity, the days of cover it provides and the amount, and can adjust it before sending. The category's purchasing budget is checked along the way.
Can you automate purchasing without losing control?
Yes, as long as preparation and decision are clearly separated. Automation is very useful for preparation: watching hundreds of items, calculating quantities, choosing the usual supplier, grouping orders. The decision to commit money must stay human and framed by rules.
In Numezis, the agent prepares each proposal with its calculation and places it in approvals. Your rules set who can approve up to what amount, and a second signature is required above a threshold. The requester cannot approve their own order. You can approve from the web, Teams, Slack or WhatsApp in one click, or change the quantity before sending. Every decision stays traced with its author, date and the calculation behind it. The result: more regular purchasing, fewer stock-outs, and no order sent without a person wanting it.