Safety Stock Calculator
Set safety stock from a target service level, using how much your daily sales and lead time actually vary, not just a fixed cushion.
Your numbers
Your result · updates as you type
Safety stock for a 95% service level
43 units
At a 95% chance of staying in stock through the reorder cycle (a 5% chance of a stockout), keep 43 extra units beyond what average demand and lead time would need. That's about 3.6 days of average sales.
- Z-score used
- 1.64
- Combined demand/lead-time variability (σ)
- 25.81
- Stockout risk at this level
- 5%
Higher service levels cost more in carrying cost for a shrinking reduction in stockouts, so they make sense for your best sellers, not everything. Feed this number into the reorder point calculator, and check what it costs to hold with the inventory carrying cost calculator.
Breakdown
- Service level → safety stock
- 90%
- 34 units
- 95%
- 43 units
- 97.5%
- 51 units
- 99%
- 61 units
- 99.9%
- 80 units
The service-level method
Instead of a worst-case guess, this method uses how much your daily sales and lead time actually bounce around their averages (their standard deviation), and a target service level: the odds you stay in stock through the reorder cycle.
Example: the default numbers, 95% service level
- Average daily usage, std. dev.
- 12 units/day, 3 units/day
- Average lead time, std. dev.
- 10 days, 2 days
- Combined variability (σ)
- √(10×3² + 12²×2²) = √666 ≈ 25.81
Z for 95% is 1.6449, so safety stock = 1.6449 × 25.81 ≈ 42.45, rounded up to 43 units (about 3.6 days of average sales).
Picking a service level
| Service level | Z-score | Stockout risk |
|---|---|---|
| 90% | 1.28 | 1 in 10 cycles |
| 95% | 1.64 | 1 in 20 cycles |
| 97.5% | 1.96 | 1 in 40 cycles |
| 99% | 2.33 | 1 in 100 cycles |
Higher service levels don't cost proportionally more: going from 95% to 99% only raises safety stock by about 41%, but it more than triples how rarely you're caught out. Reserve 99%+ for your best sellers and anything with a long lead time.
Where to get the standard deviations
- Daily usage. Export the last 60–90 days of orders per SKU and run Excel's
STDEVfunction over the daily totals. - Lead time. Look at your last 5–10 purchase orders for that supplier and take the standard deviation of days from order to received.
- No history yet? Start with a rough guess (a std. dev. around 20–30% of the average is common for demand) and tighten it once you have real data.
Questions people ask
What service level should I use?
95% is a common default. Use 99% for your best sellers or anything where a stockout costs you a lot, and consider 90% for slow, low-margin items where carrying cost matters more.
What's a Z-score?
A statistics term for how many standard deviations above the average you need to cover a given probability. Higher service levels need a higher Z-score, which needs more safety stock.
Why does lead time variability matter as much as demand variability?
A supplier that's sometimes 20 days late can blow through your safety stock just as easily as a sales spike, especially for slower-moving items where lead time swings dominate the formula.
How do I turn this into a reorder point?
Add it to expected usage during your average lead time (average daily usage × average lead time), or use the reorder point calculator, which does that step for you.
Related tools
Want this builtinto your store?
I'm Emmanuel, a Shopify and WooCommerce developer. If this tool pointed at something you'd rather have fixed for you (a faster store, cleaner product data, a migration, a custom feature), send me a note.
Emmanuel Asika
Shopify & WooCommerce developer, Dublin
I build and fix online stores: new builds, migrations, speed, SEO and the custom bits apps can't do.
Last reviewed 30 September 2026.