How to Calculate Reorder Point for Manufacturing

To calculate a reorder point, multiply your average daily demand by the supplier lead time in days, then add safety stock: Reorder Point = (Average Daily Demand × Lead Time) + Safety Stock. The first term covers what you'll burn through while the replenishment is in transit; safety stock is the buffer for the days demand runs hot or the truck shows up late. When on-hand plus on-order falls to that number, you cut the purchase order.

That's the thirty-second answer. The real answer takes a week, because every number in the formula lies to you. Lead times drift. Demand spikes. Your ERP carries a static reorder point somebody set in 2019 and nobody has touched since. I ran planning at a $250M furniture manufacturer, and our single biggest source of both stockouts and dead inventory was reorder points that hadn't moved while the business did.

Here's the math that holds up on a factory floor, the variability you have to account for, and the way to keep these numbers honest across thousands of SKUs.

The reorder point formula

The core equation is two pieces stacked together:

Reorder Point = (Average Daily Demand × Lead Time in Days) + Safety Stock

The first half is your lead time demand — the quantity you consume between hitting the trigger and the replenishment landing on the dock. The second half, safety stock, absorbs the times reality runs hotter than your averages.

Keep the units consistent. If demand is in units per day, lead time must be in days. APICS guidance and the ASCM safety stock reference (2023) both hammer this point — a units-mismatch is the most common arithmetic error I see in a planning audit.

A simple worked example

You consume 80 units a day of a purchased component. Supplier lead time is 14 days. You hold 350 units of safety stock.

• Lead time demand = 80 × 14 = 1,120 units
• Reorder point = 1,120 + 350 = 1,470 units

When on-hand plus on-order drops to 1,470, you place the order. Clean. The trap is that both 80 and 14 are averages, and averages are where inventory goes to die.

Why the simple formula breaks in a real plant

Three things wreck the basic calculation:

• Demand isn't flat. That 80/day is a mean across days that ran 40 and days that ran 160. A promo, a big customer order, a seasonal build — any of them blows through a static buffer.
• Lead time isn't fixed. "14 days" is what the supplier quoted. The actual distribution might be 11 to 26 days once you account for their capacity, ocean freight, customs, and the week your buyer forgot to send the release.
• The two stack. A demand spike during a lead-time stretch is the double-whammy behind most stockouts. Safety stock has to absorb the combination, not each one alone.

Size safety stock by gut — "let's hold two weeks" — and you over-buffer the steady SKUs while under-buffering the volatile ones. You end up with stockouts and excess at the same time. I've seen both on the same shelf.

The cost is not abstract. IHL Group estimated inventory distortion — overstocks plus out-of-stocks — at $1.77 trillion in 2023, with stockouts alone near $1.2 trillion. Mis-sized buffers are a direct line into that number.

Calculating safety stock that holds up

The defensible method accounts for variability in both demand and lead time, then sizes the buffer to a service level you choose on purpose. The standard form, laid out in Peter King's widely cited MIT/APICS paper on safety stock equations (2011):

Safety Stock = Z × √(LT × σ_D² + D² × σ_LT²)

Where:

• Z = service-level factor (the inverse-normal value for your target)
• LT = average lead time
• σ_D = standard deviation of daily demand
• D = average daily demand
• σ_LT = standard deviation of lead time

It looks ugly. It's just saying: buffer against demand bouncing around and lead time bouncing around, scaled to how badly you want to avoid a stockout.

Where the Z factor comes from

Z is the point on the standard normal distribution that leaves your target service level below it. You can pull it from the NIST Engineering Statistics Handbook standard-normal table (2012) or any inverse-normal function:

Service level	Z factor
90%	1.28
95%	1.65
98%	2.05
99%	2.33

The Z factor is the dial. Push from 95% to 99% and safety stock jumps roughly 40% — that's the cost of those last four points, and it's exactly why you don't run every SKU at 99%. Worth knowing: a 95% cycle service level means you expect to not fully deplete cycle stock in about half your replenishment cycles, so safety stock sits idle most of the time. Service level and fill rate aren't the same thing either; the difference matters when you set targets, and I break it down in service level vs fill rate.

Pick service levels by SKU class, not company-wide

This is the move most teams miss. One blanket service level is wrong for everything. Tie it to the part's importance:

SKU class	Service level	Z factor	Logic
A — high value / high volume	98–99%	2.05–2.33	Stockout cost is brutal; protect it
B — mid	95%	1.65	Balanced buffer
C — low value / long tail	90–92%	1.28–1.41	Cheap to hold, cheap to stock out; don't over-invest

Run an ABC (or ABC-XYZ) analysis first, then assign service levels by class. A C-part stocked at 99% is working capital sitting in a bin doing nothing. The full method for splitting the catalog is in ABC-XYZ inventory analysis, and the deeper safety-stock variants live in how to calculate safety stock.

A worked manufacturing example

Purchased fastener feeding an assembly line:

• Average daily demand (D): 80 units
• Demand std dev (σ_D): 25 units
• Average lead time (LT): 14 days
• Lead time std dev (σ_LT): 4 days
• Target service: 95% (Z = 1.65)

Safety stock = 1.65 × √(14 × 25² + 80² × 4²) = 1.65 × √(8,750 + 102,400) = 1.65 × √111,150 = 1.65 × 333 = 550 units

Reorder point = (80 × 14) + 550 = 1,670 units

Notice σ_LT did most of the damage — the 102,400 term dwarfs the 8,750. Lead time variability, not demand variability, is the bigger driver here. That tracks with peer-reviewed work showing that reducing lead-time variability often cuts inventory more than shaving days off the average lead time (2024). For imported or single-sourced components, tightening supplier consistency frequently beats chasing a better forecast.

Don't forget the operational layers

The formula gives a number. Production reality adds constraints the math doesn't see:

• MOQs and lot sizes. If your supplier ships in cases of 500, your effective reorder behavior rounds up. Account for it or you'll over-order on every release.
• Order quantity (EOQ). Reorder point answers when; it needs a partner answering how much. The classic economic order quantity, EOQ = √(2DS / H), balances ordering cost against holding cost. A reorder point with no view of order size optimizes half the problem.
• Review period. Continuous review trips the instant you cross the point. Periodic review — you check weekly — needs extra buffer for the days between checks; add demand over the review interval to the lead-time term.

Continuous vs. periodic review, in one line each

• Continuous (s, Q): order a fixed quantity Q the moment inventory position hits reorder point s. Fixed quantity, variable timing.
• Periodic (R, S): every R days, order back up to a target level S. Fixed timing, variable quantity — and a larger buffer to cover the review gap.

Most ERP and MES systems run continuous review on transaction triggers but report on a periodic cadence, so know which policy your part is actually under before you trust the reorder field.

Keep reorder points alive

Here's the failure that costs the most: setting these once and walking away. Demand patterns shift quarterly. Lead times move with supplier capacity and freight. A reorder point is a living number, not a master-data field you set at go-live and forget.

The practical cadence:

• Recalculate quarterly at minimum, monthly for A-items.
• Recompute σ_D and σ_LT from a rolling 12-month window so you catch trend shifts instead of fossilizing last year's volatility.
• Flag any SKU where actual lead time has run 20%+ over plan for three consecutive orders — that buffer is now too small.
• Automate it. Across 5,000 SKUs you cannot do this by hand. The teams that win push the calculation into their planning system and review the exceptions, not the whole catalog.

That last point is where most mid-market manufacturers stall. The math isn't the bottleneck; the maintenance is. A spreadsheet recalc nobody owns degrades within two quarters, and you're back to stockouts and stranded stock side by side. Pushing the recompute into a planning engine — increasingly an AI-assisted one — is how it survives. We cover the approach in AI inventory optimization for mid-market manufacturers and the broader discipline in what is inventory optimization.

The maintenance gap also tracks a wider failure mode. McKinsey's 2024 global supply chain survey found nine in ten leaders hit disruptions that year, yet only 7% had end-to-end real-time visibility. Static reorder points are visibility debt — you can't react to a lead-time shift you never measured.

Get your numbers checked

If your reorder points haven't been recomputed in the last six months, you're almost certainly holding excess on your slow movers and stocking out on your fast ones at the same time. We'll run a free planning-maturity and stranded-inventory teardown on your actual SKU data — show you exactly where buffers are mis-sized and what that's costing in working capital. Book a call and bring your item master; you'll leave knowing how to calculate reorder point for your parts, with the variability that's actually in your supply chain.

Frequently asked questions

What is the basic reorder point formula?

Reorder Point = (Average Daily Demand × Lead Time in Days) + Safety Stock. The first term is the demand you'll consume while waiting for the replenishment to arrive, and safety stock is the buffer for demand spikes and late deliveries. Keep demand and lead time in the same time unit, or the math breaks.

How is safety stock different from the reorder point?

Safety stock is one component inside the reorder point — it's the buffer against variability, not the trigger itself. The reorder point is the total on-hand-plus-on-order level that signals a new order, which equals expected lead-time demand plus that safety-stock buffer. You can't calculate a defensible reorder point without first sizing safety stock for your chosen service level.

What service level should I use for the reorder point calculation?

Set it by SKU importance, not one company-wide number. High-value, high-volume A-items typically warrant 98–99% (Z of 2.05–2.33), B-items around 95% (Z of 1.65), and long-tail C-items 90–92% (Z of 1.28–1.41). Running every part at 99% wastes working capital, because the last few points of service cost roughly 40% more safety stock.

How often should reorder points be recalculated?

Quarterly at minimum, and monthly for your A-items. Recompute the demand and lead-time standard deviations from a rolling 12-month window so the buffer tracks current volatility instead of last year's. Manufacturers who set reorder points once at go-live and never revisit them drift into simultaneous stockouts and excess within a couple of quarters.

Does lead time variability or demand variability matter more?

For most purchased manufacturing components — especially imported or single-sourced parts — lead time variability dominates. In the standard safety-stock formula, the lead-time term (D² × σ_LT²) often dwarfs the demand term (LT × σ_D²), as the worked example above shows. That means tightening supplier consistency frequently cuts more inventory than improving the forecast.