Reorder point and safety stock: the formulas, a worked example and when to recalculate

The reorder point is the stock level at which you should place a new purchase order, so that the delivery arrives before the shelf is empty. The standard formula is short: reorder point = average daily demand × lead time in days + safety stock. Safety stock is the buffer that covers demand above average while you wait for the delivery.

The formula does not need software or statistics training. It needs three numbers you probably already have: how many units you sell on an ordinary day, how many days the supplier takes to deliver, and how much daily sales tend to swing. This guide works through each one, shows a full example, and explains how to choose the buffer without either running out or tying up cash in stock that does not move.

The two formulas

In the safety-stock formula, σ_d is the standard deviation of daily demand: a measure of how much a typical day differs from the average. Z is the service-level factor taken from the normal distribution. It answers the question "in what share of replenishment cycles am I willing to run out?"

The square root of the lead time appears because daily swings partly cancel out: a busy Monday is often followed by a quiet Tuesday. Over ten days, the variation grows with the square root of ten, about 3.2, rather than with ten itself.

This version of the formula assumes the lead time is reasonably steady and only demand varies. When a supplier is often late, the buffer must also cover that; a simple way is to calculate with the lead time the supplier actually achieves in a bad month rather than the one on their price list.

A worked example

An illustrative product: a consumable you sell to business customers.

• Average daily demand: 20 units, from the last 90 days of sales.
• Standard deviation of daily demand: 6 units (a spreadsheet's STDEV over the same 90 daily totals).
• Supplier lead time: 10 days from order to goods on the shelf.
• Target service level: 95%, so Z = 1.65.

1. Lead-time demand = 20 × 10 = 200 units.
2. Safety stock = 1.65 × 6 × √10 = 1.65 × 6 × 3.16 ≈ 32 units.
3. Reorder point = 200 + 32 = 232 units.

When available stock reaches 232, you place an order. On an average day you will have used 200 of those units by the time the delivery arrives, and the 32-unit buffer covers busier periods in about 19 cycles out of 20.

Notice what the buffer costs. Raising the service level to 99% gives a safety stock of 2.33 × 6 × 3.16 ≈ 44 units: twelve more units held all year to avoid a stock-out roughly once in a hundred cycles instead of once in twenty. Whether that is worth it depends on the item.

The order quantity is a separate decision. The reorder point says when to buy; how much to buy depends on the supplier's minimum order, price breaks, shelf space and how often you are willing to place orders. A common starting point is to order enough to cover two to four weeks of demand and adjust from there.

Getting the inputs right

Demand: use real sales, and the right window

Use daily shipped or sold quantities, including days with zero sales; leaving out the empty days inflates the average and understates the variation. Ninety days is a reasonable window for steady items. For seasonal products, calculate from the same season last year, or from the most recent weeks, not from a yearly average that mixes the peak with the trough.

Lead time: from decision to shelf

Lead time is not the supplier's shipping time. It runs from the moment the stock level is reached to the moment the goods are counted and available: your own delay in placing the order, the supplier's processing and transport, and the time to receive and put away. If you only review stock on Fridays, add up to a week to the lead time, because a level reached on Saturday is not acted on until the following Friday.

Variation: look at the spread, not just the average

Two products with the same average of 20 a day can need very different buffers. One sells 18 to 22 every day; the other sells nothing for days and then 100 in one order. The second has a much larger standard deviation and needs a larger buffer, or a different approach entirely, such as ordering against confirmed customer orders.

When a simple minimum is good enough

The formula earns its keep on items you sell often and cannot afford to lack. For the long tail, a rougher rule is fine: set the minimum to what you sell in one lead time plus a few days, and review it twice a year. A useful split:

• Fast movers that stop sales when missing: full formula, 95% to 99% service level, reviewed quarterly.
• Steady, cheap items: full formula or a generous rough minimum; extra stock costs little.
• Slow, expensive items: low or no safety stock; order against customer demand when possible.
• Items with lumpy demand from a few large customers: ask those customers for forecasts rather than trusting the formula.

Mistakes that make the numbers lie

• Counting orders, not units. Demand is the quantity that left the shelf. Ten orders of one unit and one order of ten are the same demand for the reorder point, and very different variation.
• Forgetting returns and write-offs. Units returned to stock reduce net demand; damaged units written off increase it. Leave either out and the average drifts.
• Using the price-list lead time. A supplier who promises five days and delivers in nine needs nine in the formula. The last five deliveries tell you more than the brochure.
• Setting one minimum for every warehouse. If you hold stock in two places, each place sells at its own rate and needs its own level, or one location runs dry while the other overflows.
• Treating a stock-out as demand that never existed. Days when the item was out of stock show zero sales, which pulls the average down and makes the next reorder point too low. Exclude those days, or the calculation repeats the shortage.

The common thread is that the formula is only as good as the history behind it. Ten minutes spent checking the inputs is worth more than any refinement of the formula itself.

How Nextora handles minimum stock

In Nextora's ERP area every product holds a cost price, a sell price and a minimum stock level, and stock is counted per product and warehouse. The product card shows three totals, on hand, reserved and available, with a warning when available stock is at or below the minimum. Enter the reorder point you calculated as the minimum and the warning appears at exactly the moment to order.

The ERP home has a Needs attention list whose rows include stock below its minimum and customer orders that want more than is in stock, each opening its document. On the purchasing side, the supplier record carries a typical lead time, and each line of its price list has its own minimum order quantity and lead time, which are the inputs the formula needs. Posted receipts recalculate the average cost, so the extra units you hold are valued correctly. The ERP page shows stock, purchasing and sales together.

A quarterly routine

1. Export daily sales for your top items for the last 90 days.
2. Calculate the average and standard deviation for each.
3. Update lead times from the last few deliveries, not from the price list.
4. Recalculate the reorder points and update the minimums.
5. List items that ran out in the quarter and check whether the buffer, the lead time or the review habit was the cause.

Reorder points only work if the stock figures beneath them are honest; on hand, reserved and available explains the three numbers. When the level is reached, the purchase order process takes over, from order to receipt to paid invoice.