Reorder Point and Safety Stock, Worked Through With a Real Example
The formula is one line. Getting the inputs right is where most people go wrong.
Figures and typical ranges described here reflect US manufacturing, retail and distribution practice.
If you searched "reorder point safety stock explained," here is the worked example. Say you sell a component at an average of 20 units a day, your supplier's average lead time is 14 days, and in your worst month you sold 35 units a day while your supplier once took 20 days to deliver.
Step one: demand during lead time
Demand during lead time at the average case is 20 units/day × 14 days = 280 units. This is the inventory you'd need on hand, at minimum, to cover the gap between ordering and receiving — assuming everything runs exactly at average, which it won't.
Step two: the worst case you've actually seen
Worst-case demand during lead time is peak demand × peak lead time: 35 units/day × 20 days = 700 units. That's the number that would have kept you in stock during your worst historical stretch. The gap between this and the average case — 700 minus 280, or 420 units — is your safety stock under the max-demand method.
Step three: your reorder point
Reorder point is average demand during lead time plus safety stock: 280 + 420 = 700 units. In this example, the reorder point and the worst-case demand figure happen to land on the same number, which is a property of the max-demand method — it's deliberately sized for the worst case, not for a statistical probability of stockout.
Why the max-demand method is a reasonable starting point
The max-demand method has one big advantage over more statistically rigorous approaches: it's transparent. Anyone can check your two worst historical numbers and verify the math. A service-level method using standard deviation of demand produces a more precisely calibrated safety stock for a target fill rate, but it requires more data, more assumptions, and is harder to explain to someone who isn't comfortable with statistics. For a lot of small and mid-size operations, transparent-and-slightly-conservative beats precise-and-opaque.
Where the method understates risk
The max-demand method only protects against the worst case you've already seen. If your business is growing, a new customer could push demand past your historical peak, and your reorder point — calibrated on old data — would leave you short. Recalculating this figure at least quarterly, and after any major change in customer mix, keeps it from quietly going stale.
Where the method overstates the buffer you need
If your worst historical case was a one-time anomaly — a single large one-off order that skewed your peak-demand figure — using it as your ongoing planning assumption ties up more cash in safety stock than the actual risk justifies. It's worth looking at whether your peak-demand number came from a repeatable pattern or a one-off event before locking it into your reorder point.
Segmenting by value and criticality changes the answer
Not every item deserves the same safety stock treatment. A high-value, low-criticality item (something you could substitute or delay without real cost) can run leaner safety stock than a low-value, high-criticality item (a cheap part that stops an entire production line if it's out of stock). ABC/XYZ segmentation — categorizing items by value and demand variability — is the standard way to apply different reorder logic to different parts of your inventory rather than one blanket formula.
What changes when you have multiple suppliers for the same item
If you dual-source a component, your effective lead time and its variability both usually improve, because a delay from one supplier doesn't automatically mean a delay in receiving inventory at all. The reorder point math still applies, but the "peak lead time" figure should reflect your ability to lean on the second supplier during a disruption from the first — which is one of the concrete ways dual sourcing pays for its extra unit cost.
Putting it into practice
The reorder point calculator on this site runs this exact calculation — demand per day, lead days, peak demand, peak lead time and unit cost — and also shows the cash value tied up at your reorder point and how many days of cover that represents. Running your real numbers through it, rather than a hypothetical example, is the useful next step.
What to do when you don't have enough historical data yet
For a new product with no sales history, the max-demand method has nothing to calculate from. In that situation, a reasonable starting approach is to use a comparable existing product's demand pattern as a proxy, deliberately erring toward more conservative (higher) safety stock until you've accumulated enough of your own sales data to replace the proxy with real figures — typically after two to three reorder cycles.
Recalculating on a schedule, not just when something goes wrong
It's tempting to only revisit a reorder point after a stockout forces the issue. A better habit is recalculating on a fixed schedule — quarterly for stable products, monthly for anything volatile or growing quickly — so the number stays current with actual demand and lead-time patterns rather than responding only after the fact to a problem that's already happened.
What changes for items with a minimum order quantity
When a supplier enforces a minimum order quantity larger than what your reorder-point math alone would suggest ordering, the practical reorder point needs to account for carrying that extra quantity until it's consumed. In that case, comparing the carrying cost of the excess against any price break the MOQ earns you is worth doing explicitly, rather than treating the MOQ as a fixed constraint you simply accept without question.
General information for supply chain and procurement decisions, not consulting advice — your industry, scale and specific contracts may change what applies.