How to use this safety stock calculator
This calculator sizes your stock buffer two ways: a service-level method built on Z-scores, and a quick peak-versus-average method. Enter average daily sales, lead time, and a 90-99% service level. You get buffer units, a reorder point, days of cover, and the capital the buffer ties up.
- Know your demand swings? Enter the daily sales swing (σ) for the Z-score method.
- No σ handy? Enter peak daily sales and worst-case lead time for the quick method.
- Add unit cost to see how much cash the buffer ties up.
The safety stock formula
The standard formula is safety stock = Z × √(lead time × σd² + (average daily sales)² × σLT²), where σd is your daily demand swing and σLT is your lead time swing. At a 95% service level, Z is 1.65. The two swings combine, so volatile suppliers raise the buffer even when sales are steady.
If your lead time barely moves, leave σLT blank. The formula collapses to Z × σd × √lead time, the version most inventory textbooks teach first. The calculator handles both cases and shows which one it used.
Service levels and Z-scores
A service level is the odds of covering all demand during one replenishment cycle. Raising it from 95% to 99% grows your buffer by 41%, because Z jumps from 1.65 to 2.33. Buffers scale with Z, so every extra nine gets more expensive than the last one.
Service level | Z-score | Buffer vs 95% |
|---|---|---|
90% | 1.28 | 22% smaller |
95% | 1.65 | baseline |
98% | 2.05 | 25% larger |
99% | 2.33 | 41% larger |
The ratios hold for any product, because the buffer scales linearly with Z. Pick the row that matches how painful a stockout is for that SKU, then run your own numbers in the calculator above.
The quick method: peak minus average
The peak method needs no statistics: safety stock = (peak daily sales × worst-case lead time) minus (average daily sales × average lead time). If you sell 32 units a day on average but plan for 45-unit days and a 42-day worst-case shipment, the gap is your buffer.
The catch: pairing your best sales day with your worst shipping delay plans for a perfect storm. That buffer would be 770 units in the example above, more than triple the Z-score answer. Use the quick method for rough sizing, then graduate to the service-level method once you can export 60-90 days of sales history.
Worked example: 32 units a day, 35-day lead time
Say you sell 32 units a day with a daily swing of 9 units. Lead time averages 35 days and wobbles by 4. At a 95% service level, the formula gives √(35 × 81 + 1,024 × 16) = 138.6, times a Z of 1.65 for roughly 229 buffer units.
- Reorder point: 32 × 35 gives 1,120 units of lead-time demand, plus the buffer = 1,349 units.
- Days of cover: 229 units at 32 a day is about 7 days of extra runway.
- Capital: at an $8 landed cost, the buffer ties up $1,832.
When to lower your service level
A high service level is a storage bill wearing a safety vest. Moving one SKU from 95% to 99% adds 41% more buffer inventory that mostly sits there, and FBA monthly plus aged-inventory fees charge you for every unit of it. Reserve 98-99% for products that fund your business.
- Hero SKUs that drive rank and reviews: 98-99%.
- Steady mid-tail products: 95% is the standard default.
- Slow movers and long-tail variations: 90% keeps cash free.
Common safety stock mistakes
Most bad buffers come from bad inputs, not bad math. The four errors below cause nearly every oversized or undersized buffer, and each takes minutes to fix once you know where to look. Check your inputs against this list before trusting any output.
- Computing σ from 14 days of history. Use 60-90 days so one promo spike does not set your buffer.
- Feeding a one-off viral day into the peak method. Plan for repeatable peaks, not records.
- Ignoring lead time swings. A supplier that ships in 30-45 days needs σLT; an average alone hides the spread.
- Setting 99.9% everywhere. Z hits 3.09 and your buffer nearly doubles versus 95%, for demand you may never see.
Frequently Asked Questions
What is safety stock?
Safety stock is extra inventory held to absorb demand spikes and late shipments. It sits on top of cycle stock and gets sized by your service level, typically with a Z-score formula.
What is a good service level for Amazon FBA?
95% is the standard default for most FBA products. Push hero SKUs to 98-99% and let slow movers run at 90% to free up cash and storage.
How do I find my demand standard deviation?
Export 60-90 days of daily unit sales and run STDEV in a spreadsheet. If you need a starting point today, 25-30% of average daily sales is a common range for established listings.
What Z-score matches a 95% service level?
A 95% service level uses a Z-score of 1.65. Common companions: 1.28 for 90%, 2.05 for 98%, and 2.33 for 99%.
Is safety stock the same as a reorder point?
No. Safety stock is the buffer; the reorder point is the trigger. Reorder point = average daily sales × lead time + safety stock, so the buffer is one input of two.
Should I include lead time variability?
Yes, whenever your lead time swings by more than a few days. A 4-day lead time swing raised the worked example buffer from 88 units to 229.
How often should I recalculate safety stock?
Quarterly, plus before Q4 and after any demand shift. Recalculate whenever daily sales move more than about 20% or a supplier changes its lead time.
Does this calculator work for Shopify or eBay sellers?
Yes. The formula is channel-agnostic. Demand swings, lead times, and service levels behave the same whether you fulfill through FBA, a 3PL, or your own garage.