Calculate safety stock, reorder point, stockout risk and buffer cost. Use a statistical model, the max/average method, or paste historical demand and lead-time data.
Best when you know demand variability. Add lead-time variability if supplier delivery time is not stable.
Useful when you only have average and peak values. It is easy to apply but does not correspond to a specific service level.
Fastest route from raw data to a usable result. Paste one demand observation per period and, if available, historical supplier lead times.
Results
—| Service | Safety Stock | ROP |
|---|
| Method | Safety Stock | Interpretation |
|---|
How to Calculate Safety Stock
Safety stock is inventory held above expected lead-time demand to absorb uncertainty. Which formula fits depends on the data actually available. This calculator supports three approaches: a service-level method for variable demand, a combined model for variable demand and lead time together, and a max/average method for cases where only simple historical extremes are on hand.
1. Statistical Safety Stock: Variable Demand, Stable Lead Time
Safety Stock = Z × σd × √L
Z is the Z-score for the target cycle service level, σd is the demand standard deviation per period, and L is the average lead time expressed in that same period.
2. Variable Demand and Variable Lead Time
Safety Stock = Z × √(L × σd² + d² × σL²)
This version layers supplier lead-time variability on top of demand variability. It assumes the two sources of uncertainty are independent, and that lead-time demand is close enough to a normal distribution for the model to hold.
3. Max / Average Method
Safety Stock = (Maximum Demand × Maximum Lead Time) − (Average Demand × Average Lead Time)
This one is easy to apply when detailed statistics aren’t available. The trade-off is that a single extreme observation can inflate the result, and the output doesn’t correspond to any defined service level.
Which Safety Stock Formula Should You Use?
| Your data | Best starting method | Why |
|---|---|---|
| Demand varies; lead time is stable | Z × σd × √L | Ties the buffer directly to a target service level |
| Demand and supplier lead time both vary | Combined statistical model | Accounts for both major sources of uncertainty at once |
| Only averages and historical peaks are known | Max / Average | Works without standard deviations |
| Frequent zero-demand periods or highly lumpy orders | Use standard formulas with caution | Intermittent demand often calls for a Poisson, empirical, or simulation-based model instead |
Safety Stock vs. Reorder Point
Safety stock is the buffer itself. The reorder point is the inventory level that triggers a new replenishment order:
Reorder Point = Average Demand × Average Lead Time + Safety Stock
WareStat calculates both figures together, so the result can be applied directly within an inventory policy.
How the Historical Data Mode Works
Paste recent demand figures and the calculator derives the sample average, standard deviation, and maximum on its own. Add supplier lead times as well, and it measures their average and variability too. Short histories and highly intermittent or lumpy demand get flagged, since those conditions can make a standard normal safety-stock formula unreliable.
Frequently Asked Questions
What service level should I use?
No single percentage works best across the board. A higher service level lowers the modeled risk of a stockout, but it also raises the inventory needed to reach it. Critical, high-margin, or hard-to-replace items often justify a higher target than low-value or easily replenished products.
What if I don’t know demand standard deviation?
Switch to Historical Data mode and paste the demand observations directly. If only average and maximum values are on hand, the Max / Average method is the more appropriate choice — better that than estimating a standard deviation from insufficient data.
Can I use weeks instead of days?
Yes. Demand, demand standard deviation, and lead time all need to share the same time basis — weekly demand, for instance, should be paired with lead time expressed in weeks.
Does a 95% service level mean 5% of units will be lost?
No. Here it refers to a cycle service level: roughly a 95% probability of avoiding a stockout during a replenishment cycle, given the model’s assumptions. That’s a different measure from a 95% fill rate.
How often should safety stock be recalculated?
Recalculate whenever demand volatility, supplier lead time, seasonality, or service targets shift meaningfully. Fast-moving and seasonal SKUs typically need more frequent review than stable, slow-moving ones.
Sources and Assumptions
These statistical formulas follow the standard inventory-planning treatment of demand and lead-time uncertainty. Results are planning estimates, not guarantees.