Reorder Point Calculator

The reorder point is the inventory level that triggers a new purchase or production order — low enough to keep stock lean and capital free, high enough that replenishment arrives before the shelf empties — and it can state, as a probability, how confident that arrival plan is. It has two ingredients: the demand you expect during the supplier’s lead time, and a safety stock that absorbs the surprises. Enter your average daily demand, lead time, demand variability and target service level: this calculator returns the reorder point with both ingredients itemised, so you can see exactly how much of it is expectation and how much is insurance — and which lever, pulled, would shrink it cheapest.

Two Quantities Wearing One Number

The formula is ROP = (average daily demand × lead time) + safety stock, and its pedagogy is the split. The first term, lead-time demand, is pure expectation: if the world behaved on average, stock would touch zero the moment the replenishment lands. The second term prices the world's refusal to behave: safety stock = z × σ_d × √L, where σ_d is demand's daily standard deviation, L the lead time, and z the service-level factor (1.65 for 95%, from the same normal table as every fence on this site). Each factor teaches: variability enters at full strength, lead time at square-root strength (daily surprises partially cancel over longer windows — the same √n law that governs sampling error), and the service level as an explicit policy choice — the probability of riding out a replenishment cycle without a stockout, which management must choose because mathematics cannot.

Worked Example: Fifty a Day, Six Days Out

The defaults describe a healthy mid-size SKU: demand 50 units/day, lead time 6 days, daily σ of 8, service target 95%. Lead-time demand = 50 × 6 = 300 units. Safety stock = 1.65 × 8 × √6 ≈ 13.2 × 2.449 ≈ 32.3 units — notice how small the insurance is beside the expectation, because six days of averaging forgives most daily wobble. Reorder point = 333 units (rounded up: reorder points round toward safety). Run the levers, as the calculator invites: raise the service level to 99% (z = 2.33) and safety stock becomes 45.6, ROP 346 — thirteen units buy four points of stockout protection. Double the lead time to 12 days and lead-time demand doubles to 600 while safety rises only to 45.7 (× √2) — ROP 646, the square root quietly subsidising the slower supplier. And halve σ to 4 (a forecasting or pack-size win) and the insurance halves with it — variability is the one factor that pays its reduction at 100 cents on the dollar.

What the Formula Assumes, and What Practice Adds

The textbook formula presumes demand is roughly normal and independent day to day, lead time is constant, and review is continuous (the stock level is watched constantly, so the order fires the moment the point is crossed). Practice edits each assumption. Lead-time variability — often the larger sinner — has its own extended formula (σ over lead time combines demand and lead-time variances: safety = z × √(L·σ_d² + d²·σ_L²)); when suppliers are erratic, that version's second term dominates and the honest calculator is this one plus a hard conversation about the supplier. Periodic review (ordering every Monday) adds the review interval to the exposure window — order-up-to levels run higher than reorder points by a cycle's worth of cover. Lumpy, intermittent demand (spare parts) breaks the normality the z-table presumes, and belongs to Croston-style methods. None of this indicts the formula: it is the correct spine, and the extensions are vertebrae added where the business's shape demands them.

Reorder Point Is Not Order Quantity

The commonest confusion in inventory's first week: the reorder point answers when, not how much. The quantity decision lives elsewhere — classically in the EOQ's balance of ordering cost against holding cost, practically in case packs, minimum orders and truck geometry. A perfect reorder point wedded to a foolish quantity still bleeds money (in ordering labour or in warehousing), and the two are tuned independently in a continuous-review (Q, R) system: Q from economics, R from this page's insurance arithmetic. The pairing matters in one place — the service level: the z chosen here is a promise about cycles, and a cycle is defined by Q's rhythm (more, smaller orders mean more cycles exposed to stockout risk per year at the same per-cycle level). Set the two numbers in sight of each other, revisit them when demand's mean or σ drifts, and treat any reorder point older than its demand data as a rumour.

A Second SKU: When Insurance Dominates

Contrast teaches proportions: take a slow mover — demand 6 units/day, a 14-day lead time, daily σ of 4, the same 95% service target. Lead-time demand is just 84 units, but safety stock is 1.65 × 4 × √14 ≈ 24.7 units, a reorder point of 109 in which insurance is about 23% of the whole — more than double the fast mover's share in the main example. The reasons are structural: long lead times stretch the √L, and low-volume demand is proportionally noisier (σ of 4 against a mean of 6 is a wild series; 8 against 50 is a tame one). Slow movers are where reorder-point discipline pays hardest, precisely because intuition — trained on fast movers' small insurance shares — systematically underinsures them.

The Quantity Beside the Point: One EOQ Paragraph

To make the when/how-much split concrete, price the worked example's order quantity with the classical EOQ: annual demand 50 × 365 = 18,250 units, ordering cost $40 per order, holding cost $2 per unit-year. EOQ = √(2 × 18,250 × 40 / 2) = √730,000 ≈ 854 units per order — about 21 orders a year, each triggered when the inventory position touches this page's 333. The two numbers govern different risks (333 guards the lead-time window; 854 balances the paperwork against the warehouse) and are set by different costs, which is the whole doctrine in one example: never let the quantity conversation set the trigger, or the trigger conversation inflate the quantity.

Related Tools

More from the Other Calculators hub: the Takt Time Calculator — the production pace the stocked line must sustain — and the Finance hub's break-even tools, where the cost of holding this stock is priced.

Frequently Asked Questions

What service level should I choose?

By consequence, not by habit: 95% (z = 1.65) is the common default for ordinary SKUs; 98–99%+ for items whose stockout stops a line, loses a contract or empties a promise; 90% or below for cheap, substitutable stock where a brief gap costs little. The asymmetry to price is stockout cost against holding cost — and note the shape of the trade from the worked example: protection beyond 95% gets expensive in units quickly, because the normal tail is long.

How do I find my demand standard deviation?

From history: take daily (or per-period) demand over a representative window — a quarter is a common minimum — and compute its standard deviation, the descriptive-statistics staple. Use the same period length as the demand figure you entered (daily σ with daily demand). Beware two contaminations: stockout days record sales, not demand (truncated history understates both mean and σ), and promotions manufacture variability that the safety stock will then insure forever — clean the series, or segment the SKU.

Does the reorder point include stock already on order?

The trigger compares the inventory position — on hand plus on order, minus backorders — against the reorder point, not the physical shelf count alone. Otherwise a slow supplier invites duplicate orders every day the shelf sits below the point while replenishment is already in transit. Any system (or spreadsheet) implementing this page's number must track position, not just presence; the arithmetic is trivial, the omission is the classic double-order bug.

What if lead time varies more than demand?

Then this page's safety stock underprices the risk, and the extended formula belongs in its place: safety = z × √(L·σ_d² + d²·σ_L²), where σ_L is the lead time's own standard deviation. The d² multiplier on lead-time variance is the reason supplier reliability dominates for high-volume items — a day of supplier wobble at 50 units/day swings exposure by 50 units, dwarfing most daily demand noise.

How often should the reorder point be recalculated?

Whenever its inputs' truth expires: demand trends and seasonality move the mean, assortment changes move σ, supplier changes move L. In practice: fast movers reviewed monthly to quarterly, the long tail annually, and all of them on any structural event (new channel, new supplier, pack change). A reorder point is a forecast wearing a trigger; forecasts have shelf lives, and expired ones stock out or pile up with equal indifference.

Is safety stock just wasted inventory?

It is insurance, and like insurance it looks wasteful in every period it goes unclaimed. The worked example's proportion is the argument: 32 units of insurance guarding a 300-unit exposure — about a tenth of cycle stock — buys a 95% ride-through rate. The waste question is properly asked of the level (is 95% the right promise for this SKU?) and of the variability that makes insurance necessary at all; the existence of a buffer against a variable world is not the inefficiency.