Service level is a probability
A 95% service level means you expect to meet demand in 95 out of 100 cycles.
Stats
Safety Stock Calculator: Demand, Lead Time and Service Level
Calculate safety stock with a fixed-lead-time formula, a combined demand-and-lead-time variability formula, or a max-versus-average method. The result includes the inventory position at which to reorder, plus a service-level comparison for planning inventory trade-offs.
Inputs
Try a preset to see how each source of uncertainty changes the formula and result.
Calculate averages and deviations from history
Paste one observation per line, or separate values with commas or spaces. Use demand and lead times measured in the selected daily frequency.
Each value is demand for one complete selected period.
Each value is one purchase order's elapsed lead time, in the selected period.
Results
Safety stock—
Reorder point—
Expected lead-time demand:—
Buffer coverage:—
Cycle service level:—
Formula selected:—
Show calculation breakdown
Service-level comparison
Compare the extra buffer required as the cycle service target rises.
| Service level | Safety stock | Reorder point | Incremental units | Incremental value |
|---|
Advertisement
Which safety stock formula should you use?
Fixed lead time: SS = Z × σd × √L
Stable supplier Use when demand varies but replenishment lead time is effectively fixed. You need average demand, demand standard deviation at the selected frequency, average lead time in that same frequency, and a cycle service target. It assumes period demand is independent and demand over lead time is approximately normal.
Demand and lead time both vary: SS = Z × √(L × σd² + d² × σL²)
Variable supplier Use when purchase-order lead times vary. Here d and σd are average demand and demand deviation per period, while L and σL are average lead time and its deviation in the same period. The formula assumes demand and lead-time variability are independent and lead-time demand is approximately normal. Correlated delays and demand surges require a more advanced model.
No standard deviations: SS = max demand × max lead time − average demand × average lead time
Rough operational method Use only when you have representative maxima but insufficient data for standard deviations. It does not map the buffer to a cycle service probability, and a single outlier can overstate stock. Review which observations count as credible maximums.
For every method: expected lead-time demand is d × L, and reorder point is d × L + SS. Reorder when inventory position—on hand plus on order minus backorders—reaches the reorder point.
Worked examples you can reproduce
1. Variable demand, fixed lead time — “Stable supplier” preset
Average daily demand d = 120, daily demand deviation σd = 35, lead time L = 14 days, lead-time deviation σL = 0, and 95% cycle service Z = 1.644854.
Lead-time variance is L × σd² = 14 × 35² = 17,150; its square root is 130.9580. Safety stock is 1.644854 × 130.9580 = 215.4068. Expected lead-time demand is 120 × 14 = 1,680. Reorder point is 1,680 + 215.4068 = 1,895.4068. Operationally, hold 216 units of safety stock and place an order when inventory position reaches 1,896 units.
2. Demand and lead time both vary — “Variable supplier” preset
Average daily demand d = 80, demand deviation σd = 18, average lead time L = 10 days, lead-time deviation σL = 2 days, and 95% cycle service Z = 1.644854.
Combined variance is (10 × 18²) + (80² × 2²) = 3,240 + 25,600 = 28,840; its square root is 169.8234. Safety stock is 1.644854 × 169.8234 = 279.3347. Expected lead-time demand is 80 × 10 = 800. Reorder point is 800 + 279.3347 = 1,079.3347. Operationally, hold 280 units of safety stock and reorder at 1,080 units.
Safety stock FAQs
Which safety stock formula should I use?
Use the fixed-lead-time formula when supplier lead time is stable, the combined formula when both demand and lead time vary, and max versus average when you have credible maximums but no standard deviations. Automatic mode chooses fixed or combined from the lead-time standard deviation.
How do I calculate demand or lead-time standard deviation?
Use observations measured at one consistent frequency, calculate their average, subtract the average from each observation, square those differences, divide their sum by n − 1, and take the square root. The history analyzer performs this sample-standard-deviation calculation. Range ÷ 4 is only a rough estimate.
Should I use daily or weekly data?
Use the frequency at which your demand is reliably recorded, then express average demand, demand standard deviation, average lead time, and lead-time standard deviation in that same period. Do not combine daily demand with lead time in weeks.
How should I choose a cycle service level?
Choose it from stockout impact, customer promise, item criticality, replenishment flexibility, and carrying cost. Compare several levels because the extra inventory rises quickly near 99%.
Is cycle service level the same as fill rate?
No. Cycle service level is the probability of no stockout during a replenishment cycle. Fill rate is the proportion of demand supplied immediately. A 95% cycle service level does not necessarily mean a 95% fill rate.
What is the difference between safety stock and reorder point?
Safety stock is the uncertainty buffer. Reorder point is the inventory-position trigger and equals expected demand during average lead time plus safety stock in this calculator.
How should I handle seasonal or intermittent demand?
Segment seasonal periods or use a forecasting and simulation method that represents seasonality, promotions, intermittency, skew, and correlation. Normal-distribution formulas can underperform for these patterns.
How often should I recalculate safety stock?
Recalculate when demand, supplier performance, service policy, or review frequency changes. Monthly is common for active items; stable items may be reviewed quarterly, while volatile or critical items may need more frequent review.
Can safety stock be zero?
Yes. It can be zero when measured variability is zero, lead time is zero, or management accepts the stockout risk. Verify that zeros reflect enough representative history rather than missing data.
Methodology, data handling and limitations
Methodology updated: 16 July 2026. Formula notation follows established inventory-management treatments including Edward A. Silver, David F. Pyke and Douglas J. Thomas, Inventory and Production Management in Supply Chains, and Sunil Chopra, Supply Chain Management: Strategy, Planning, and Operation. No claim of independent professional review is made.
Rounding: calculations use full-precision browser numbers. The breakdown retains unrounded values; operational safety stock and reorder points round upward to whole units so the displayed policy does not round protection down.
Data handling: calculation inputs and pasted history are processed locally in your browser. Sharing puts scenario values in the URL, so inspect the link before sending it.
Known limitations: statistical formulas assume independent demand periods, independence between demand and lead time in the combined model, and approximately normal lead-time demand. Intermittent, seasonal, promotion-driven, correlated, or highly skewed demand may need empirical quantiles, forecasting, bootstrapping, or simulation. This planning estimate does not replace item-level policy review.
5 Fun Facts about Safety Stock
Variability grows with lead time
Demand variance accumulates over lead time, which is why the formula uses sqrt(L).
Lead time
Safety stock is not cycle stock
Cycle stock is for regular demand; safety stock is for uncertainty.
Inventory math
Higher service levels are costly
The Z-score grows quickly near 99%, so safety stock can rise fast.
Costs
Forecast improvements reduce buffers
Lower demand variability directly reduces safety stock requirements.
Forecasting
Disclaimer
These outputs are planning estimates, not purchase instructions. Statistical methods approximate lead-time demand with a normal distribution; the combined method can include measured supplier lead-time variability but does not model demand–lead-time correlation. Validate policies against actual stockouts, order constraints, seasonality, and supplier performance.