What Is Safety Stock?
Safety stock is extra inventory held as a buffer against two sources of uncertainty: demand variability (customers buy more or less than forecast) and supply variability (suppliers ship late or deliver short). Without safety stock, any deviation from the plan causes a stockout.
Safety stock is not a fixed number — it should be calculated per SKU based on that item's actual demand variability, forecast accuracy, and lead time uncertainty. Integer Demand computes optimal safety stock automatically for every item in your catalog.
The Standard Safety Stock Formula
The most widely used formula for safety stock accounts for both demand and lead time variability:
Safety Stock = Z × √(Lead Time × σdemand² + Avg Demand² × σlead time²)
- Z — Service level Z-score. 90% = 1.28, 95% = 1.65, 99% = 2.33, 99.9% = 3.09
- σdemand — Standard deviation of demand per period (week or month)
- σlead time — Standard deviation of lead time in periods
- Lead Time — Average lead time in periods
- Avg Demand — Average demand per period
For brands with stable lead times (σlead time ≈ 0), the formula simplifies to: Safety Stock = Z × σdemand × √Lead Time
Safety Stock by Service Level Target
90% Service Level (Z=1.28)
1 in 10 replenishment cycles may stockout. Appropriate for C-class slow-movers with low margin impact.
95% Service Level (Z=1.65)
Industry standard for B-class items. Balances cost of carrying extra inventory against cost of stockouts.
99% Service Level (Z=2.33)
Recommended for A-class best-sellers. Higher carrying cost is offset by not losing revenue on top SKUs.
99.9% Service Level (Z=3.09)
For mission-critical items where a stockout would cause severe customer or business impact. Expensive to maintain.
Safety Stock Calculation Methods
- Fixed days of supply — Simplest method. Safety stock = N days of average demand. Easy to explain but ignores actual variability. Use as a starting point only.
- Statistical (Z-score) method — Uses historical demand standard deviation. Most accurate for items with sufficient sales history (12+ months).
- Forecast error-based — Safety stock = Z × MAE (Mean Absolute Error of the forecast). Directly ties safety stock to the quality of the demand forecast. Better than raw demand deviation for items with strong seasonality.
- ABC-differentiated — Apply different service level targets (and therefore different Z-scores) to A, B, and C class items. Reduces total inventory investment while protecting revenue-critical SKUs.
Safety Stock for Seasonal and Intermittent Demand
Standard safety stock formulas assume relatively stable demand. Two common exceptions require special treatment:
- Seasonal items — Calculate safety stock separately for peak and off-peak periods. A product that sells 5x more in Q4 needs 5x the safety stock during that window.
- Intermittent demand — Items that sell infrequently (many zero-demand periods) have inflated standard deviations. Use Croston's method or SBA for the demand forecast, then apply the Z-score formula to the non-zero periods only.
- New products (NPI) — No historical data means no demand deviation to measure. Use analogous product history, assume worst-case variability, and set a conservative fixed days-of-supply target until sufficient data accumulates.
Integer Demand detects intermittent demand patterns automatically and adjusts safety stock calculation accordingly.
Reorder Point and Its Relationship to Safety Stock
Safety stock feeds directly into the reorder point (ROP) — the inventory level that triggers a new purchase order:
Reorder Point = (Average Daily Demand × Lead Time in Days) + Safety Stock
Integer Demand's Inventory Projection table shows the projected stock level week by week against the reorder point, so you can see exactly when each SKU will breach the threshold and when a PO needs to be sent.
Start Forecasting Smarter
Join eCommerce brands using Integer Demand to reduce stockouts, cut excess inventory, and protect margins with AI-powered demand planning.
Try Integer Demand Free