Demand and lead time
Use a consistent time unit. A model using daily demand must express lead time in days. Review the variation around each average, rather than treating the average as a promise.
Safety stock, reorder points and purchase quantities, explained.
A practical inventory-planning guide for distributors and purchasing teams, with a worked example that makes the assumptions visible.
SCMSigma product guide · Updated November 3, 2026
Three different inventory decisions
Safety stock is a buffer against uncertainty. The reorder point tells you when to replenish. The purchase quantity tells you how much to order. They work together, but they answer different questions.
| Measure | Decision it supports | What to review |
|---|---|---|
| Safety stock | How much buffer should we hold? | Demand variability, lead-time variability and the chosen service level |
| Reorder point | When should we replenish? | Expected lead-time demand plus safety stock |
| Purchase quantity | How much should we order? | Target stock, inventory position, pack size and minimum order quantity |
Start with reliable inputs
Review sales history at the SKU level, supplier delivery times, current stock, stock already on order and committed demand. Check whether a promotion, a new product or a supply disruption makes the recent history less representative.
Inventory position
Review stock on hand, incoming supply and demand already committed. A physical stock count alone does not describe the position available to support future demand.
A worked safety-stock example
This illustrative calculation uses the independent demand and lead-time variation model reflected in SCMSigma’s planning examples. It assumes daily demand and lead time are independent, demand is sufficiently stable for a normal approximation, and the input estimates are representative.
| Input | Illustrative value |
|---|---|
| Average daily demand (d) | 20 units |
| Daily demand standard deviation (σd) | 5 units |
| Average lead time (L) | 14 days |
| Lead-time standard deviation (σL) | 2 days |
| Target cycle service level | 95%; z ≈ 1.645 |
Safety stock = z × √(L × σd² + d² × σL²)
With these inputs, safety stock is 1.645 × √(14 × 5² + 20² × 2²), or approximately 72.6 units. Rounding the buffer up gives 73 units of safety stock.
The cycle service level is the probability of avoiding a stockout during a replenishment cycle under the model assumptions. It is different from the fill rate, which measures the share of demand supplied. Neither figure guarantees real-world service.
Calculate the reorder point
Expected demand during lead time is 20 × 14 = 280 units. Adding the rounded buffer gives a reorder point of 353 units.
Reorder point = expected lead-time demand + safety stock
Use the relevant inventory position when reviewing a replenishment trigger. Otherwise, you may order again even though an incoming delivery already covers the need. SCMSigma exposes planning inputs so the purchasing team can inspect the recommendation.
Translate the plan into an order quantity
Suppose an item’s inventory position is 120 units, its planning target is 500 units, its order pack is 24 units and its minimum order is 48 units. The unrounded need is 500 − 120 = 380 units. Rounding that need to the next complete pack gives an illustrative order quantity of 384 units.
| Step | Calculation |
|---|---|
| Unrounded need | 500 − 120 = 380 units |
| Minimum-order check | 380 is above the 48-unit minimum |
| Pack rounding | 16 packs × 24 = 384 units |
This example illustrates a target-stock policy. Your policy may also use review intervals, fixed quantities, supplier constraints and existing orders. Review the policy before treating a suggested quantity as an instruction to buy.
When the basic model needs more care
- Intermittent or new-item demand: limited or sparse history can make a simple normal approximation unreliable.
- Seasonality and promotions: a historical average may understate the demand expected during a future lead time.
- Related disruptions: demand spikes and supplier delays may happen together, violating the independence assumption.
- Data quality: stock errors, incorrect units and missed purchase orders can produce a mathematically consistent but impractical recommendation.
Use the interactive planning lab to explore assumptions, then review the result against your operational constraints and actual data.
Explore your next planning decision.
Try the sample workspace to inspect demand, lead times, stock positions and purchase recommendations before deciding how SCMSigma fits your business.