How much safety stock is enough, and when should we reorder?
Safety stock covers the demand and delivery surprises you cannot forecast. Choose the service level you want, enter how much demand and lead time vary, and the tool returns the buffer, the reorder point and the cash the buffer ties up.
Your assumptions
Worked-example values — replace with yoursAssumes demand and lead time are independent and roughly bell-shaped.
This is an estimate. Check it on your own data.
A single service level for every item is the usual mistake. Setting it by class on your real demand history, and seeing the cash effect by family, is what a workspace does.
Check it on your own data → Your inputs go with you in the link and stay in this browser; nothing is sent to ADEXMA until you submit a form.
Starter checkout and carrying these inputs into the Scan and your workspace are proposed (Wave 1, not live yet): until they ship, the buttons lead to the plans page and the free Scan, and you enter your figures there.
The formula
- z = the normal-distribution factor for your service level (95% → 1.645)
- safety stock = z × √( lead time × (demand variation)² + (average demand)² × (lead-time variation)² )
- reorder point = average demand × lead time + safety stock
- cash in safety stock = safety stock × unit cost
Worked example · the values pre-filled above
Service level: 95 % of cycles without a stockout · Average demand: 200 units · Variation in weekly demand (standard deviation): 40 units · Supplier lead time: 4 weeks · Variation in lead time (standard deviation): 0.5 weeks · Cost of one unit: $25 per unit
- z for a 95% service level = 1.645.
- Variance = 4 weeks × 40² + 200² × 0.5² = 6,400 + 10,000 = 16,400, so the square root is 128.1 units.
- Safety stock = 1.645 × 128.1 = 211 units. Reorder point = 200 × 4 + 211 = 1,011 units.
- Cash tied up = 211 units × $25 = $5,268.
Result: 211 units · reorder at 1,011
What this assumes
- Demand and lead time are in the same period (weeks here) and vary independently.
- Variation is a standard deviation measured on recent history; with too little history it is unreliable.
- A cycle-service level is not the same as a fill rate. This tool targets the chance of no stockout per cycle.
- Intermittent or lumpy demand breaks the normal approximation — the tool warns when variation exceeds the average.
- It is an estimate from your own inputs. Results from different tools overlap and must never be added together into one savings figure.
Sources
- Silver, E., Pyke, D. & Thomas, D., Inventory and Production Management in Supply Chains (CRC Press, 4th ed.) — Safety stock under demand and lead-time uncertainty; cycle service level.
- Chopra, S. & Meindl, P., Supply Chain Management: Strategy, Planning, and Operation (Pearson) — Safety stock and reorder point formulas.
- Croston, J., “Forecasting and Stock Control for Intermittent Demands”, Operational Research Quarterly 23(3), 1972 — Why lumpy demand needs a different method.
Questions people ask
How do I calculate safety stock?
Multiply the service factor z by the square root of (lead time × demand variance + average demand squared × lead-time variance). For a 95% service level z is 1.645.
What is the reorder point?
Average demand during the lead time plus the safety stock. When stock on hand plus on order falls to this level, place the order.
Does this work for slow, lumpy items?
Not well. When demand varies by more than its average, the normal-distribution assumption understates the buffer. The tool shows a warning; such items need an intermittent-demand method.
Related tools
On your own data, Chloé picks up this finding in a workspace and your advisor confirms it. See the plans →