Safety stock and reorder point calculator

Enter your average daily sales and how long a delivery takes. The calculator returns the safety stock, the level at which you have to place the next order, and — if you enter what is on the shelf — how many days you have left before that moment.

The busiest day you would still want to cover.
STDEV over daily sales in a spreadsheet.
Share of cycles that end without a stockout.
Optional — tells you how long you have left.
Safety stock
Reorder point
Days of cover

Sold during the lead time
Safety stock covers, days
StockSafety stockReorder pointLead timeTime
Order when stock touches the reorder point — what is left covers the lead time.

The formulas

reorder point = average daily sales × lead time + safety stock

safety stock, worst case   = (peak daily × worst lead) − (average daily × average lead)
safety stock, service level = Z × standard deviation × √lead time

The reorder point is not a target stock level. It is a trigger: the moment stock touches it, the order has to go out, because what remains is exactly what you will sell while waiting.

Which method to use

Worst case needs two numbers you already know: your busiest day and your slowest supplier. It is easy to justify and it errs towards more stock, because it assumes both bad things happen at once — a peak week during a delayed shipment. In reality they rarely coincide, so the result is generous.

Service level asks instead: how often are you willing to run out? At 95% you accept a stockout in one replenishment cycle in twenty. It needs the standard deviation of daily sales — one STDEV over a column of daily figures in a spreadsheet — and gives a smaller, better-argued number.

Service levelZOne stockout in
90%1.2810 cycles
95%1.6420 cycles
97.5%1.9640 cycles
99%2.33100 cycles
99.9%3.091000 cycles

Notice how the last few points cost: going from 95% to 99.9% nearly doubles the safety stock to remove one stockout in fifty. On a slow-moving item that is money sitting still for a year.

Why the square root

The service-level formula multiplies by √lead time, not by lead time. A supplier who takes four times longer needs only twice the safety stock.

The reason is that daily variation partly cancels out. Over sixteen days some are above average and some below, and the swings offset each other; the uncertainty grows with the square root of the period, not in proportion to it. Multiplying by the full lead time — a common shortcut — buys a large amount of stock that protects against nothing.

This is also the strongest argument for shortening lead times: cutting a delivery from sixteen weeks to four halves the frozen capital before you negotiate a single price.

Worked example

Twenty units a day, fourteen days of lead time, peaks of thirty-two units and a worst case of twenty-one days:

  • sold during the lead time = 20 × 14 = 280
  • safety stock = 32 × 21 − 280 = 392
  • reorder point = 280 + 392 = 672

The same item by service level, with a standard deviation of 6 units a day and a 95% target: 1.64 × 6 × √14 = 37 units of safety stock and a reorder point of 317.

Ten times less cushion for one stockout in twenty. The gap is the price of assuming the worst day and the worst delivery arrive together — which is the honest answer when you have no sales history, and an expensive one once you do.

What this calculator does not cover

Seasonality. If December sells four times October, one average is meaningless. Recalculate per season, or per month for anything with a strong cycle.

How much to order. The reorder point says when; the order quantity is a separate decision driven by minimum order quantities, freight breakpoints and how much capital you are willing to freeze — what that freezing costs per year is worked out in what holding stock really costs.

Marketplace fulfilment. Selling through FBA adds a second lead time — the inbound leg and check-in at the warehouse — which has to be added to the supplier lead time, not treated separately.

Frequently asked questions

Put daily unit sales for the last two or three months in a column and apply STDEV. Exclude promotion days and stockout days: a day when you sold nothing because there was nothing to sell is not low demand, and it drags the average down as well.

Between 90 and 95% for most goods. Push higher only where a stockout is genuinely expensive — a hero product, a marketplace listing whose ranking depends on availability, or a spare part a customer cannot wait for.

In the worst-case method that means the peak you entered is no worse than the average — check the numbers. Genuine zero variability does not exist in retail.

Yes. Measure from the moment you decide to order to the moment the stock is sellable: supplier production, transport, customs, inbound and check-in. The internal steps are often a third of the total and are the easiest to forget.

Monthly for fast movers, quarterly for the rest, and immediately after any change of supplier or shipping mode. A reorder point set for sea freight is wildly wrong once you switch to air.

Related tools