Safety stock formula: how to avoid stockouts and overstock
The safety stock formula tells you how many extra units to hold so a surprise in demand or supplier timing does not leave shelves empty. The basic version is safety stock = z x standard deviation of demand x square root of lead time. This guide shows each part, a worked example and a simpler rule for when your data is thin.
What safety stock does
Safety stock is a buffer. You hold it on top of the units you expect to sell while you wait for an order to arrive. Too little and you run out. Too much and cash sits on the shelf.
The buffer is linked to the reorder point. The reorder point is average demand during lead time plus safety stock. When stock falls to that point, you place the next order.
Sources: Wikipedia, "Safety stock" (reorder point and safety stock definitions).
The basic safety stock formula
Use this when lead time is steady and demand moves up and down:
Safety stock = z x σ x √L
- z is the z-score for the service level you want.
- σ is the standard deviation of demand per period, for example per week.
- L is the lead time, counted in the same periods.
The square root is there because random swings partly cancel out over time. Four weeks of lead time does not mean four times the risk. It means twice the risk.
If lead time also varies, the textbook form adds a second term: z x the square root of (average lead time x demand variance + average demand squared x lead time variance).
Sources: Wikipedia, "Safety stock" (general formula and the case with fixed lead time).
Pick your service level and z-score
Service level is the chance you will not run out during one order cycle. A higher level needs a bigger z-score and more stock. These are standard normal values:
- 90 percent service: z = 1.282
- 95 percent service: z = 1.645
- 97.5 percent service: z = 1.960
- 99 percent service: z = 2.326
Going from 95 to 99 percent lifts your buffer by about 41 percent. Pick the level by what a stockout costs you, not by habit. A fast-moving item with angry customers may deserve 99 percent. A slow, low-value item may not.
Sources: NIST/SEMATECH e-Handbook of Statistical Methods, critical values of the normal distribution.
A worked example
Say a product sells 100 units a week on average. The standard deviation of weekly demand is 20 units. Your supplier takes 4 weeks. You want 95 percent service.
- Square root of 4 weeks = 2
- Safety stock = 1.645 x 20 x 2 = 65.8, so round up to 66 units
- Average demand in lead time = 100 x 4 = 400 units
- Reorder point = 400 + 66 = 466 units
When stock hits 466 units, order again. Each time you do, the odds of running out before the delivery lands are about 5 in 100.
Sources: Wikipedia, "Safety stock" (reorder point formula); NIST/SEMATECH e-Handbook (z = 1.645). The numbers are the editors' own illustration.
A simpler rule when data is thin
Not every team has clean demand history. A common shortcut is the max-minus-average method. Take your highest daily use times your longest lead time, then subtract average daily use times average lead time.
Example: peak use is 20 units a day and the longest lead time is 14 days. That is 280. Average use is 15 a day and the usual lead time is 10 days. That is 150. Safety stock is 280 minus 150, or 130 units.
This rule is easy to explain, but it assumes the worst case on both sides at once. It tends to hold more stock than the formula. Use it for new items, then switch to the formula once you have a few months of sales data.
This section reflects the editors' own practical judgment and common inventory practice, not a single study.
Common mistakes
- Mixed time units. If demand is weekly, lead time must be in weeks too.
- Using the average instead of the spread. Safety stock depends on how much demand swings, not on how big it is.
- Ignoring supplier delays. If deliveries are often late, add the lead time variance term.
- Setting it once. Recheck the inputs every quarter, or when a supplier or season changes.
This section reflects the editors' own practical judgment.