Every stockout and every pallet of slow stock comes back to two numbers: when you reorder and how much you keep in reserve. The reorder point tells you when to buy. Safety stock is the cushion that absorbs the days when demand runs hot or the supplier runs late.
This guide gives you the formulas, from the simplest to the statistical, with worked numbers you can copy into a spreadsheet.
In short: reorder point = demand during lead time + safety stock. Measure demand only on days you were in stock, measure lead time from order to sellable, and size safety stock from how much demand and lead time actually vary, not from a flat percentage.
The Reorder Point in One Line
Reorder point = (Average daily demand x Lead time in days) + Safety stock
When available stock plus stock already on order falls to this number, you place the next order. The first part covers the sales you expect while you wait for the delivery. Safety stock covers the sales you did not expect.
Step 1: Measure Average Daily Demand Properly
Use real orders, not guesses, and fix the two errors that quietly ruin most forecasts:
- Exclude days you were out of stock. If you sold 288 units in 30 days but were out of stock for 6 of them, demand is 288 / 24 = 12 a day, not 9.6. Counting stockout days as zero demand makes every future order too small, which causes the next stockout.
- Combine all channels. Demand is the total across Amazon, eBay, Shopify and wholesale for the stock pool that serves them.
Choose a window that reflects how the product sells now. Thirty to ninety days suits steady products; for growing or seasonal ones, weight recent weeks more heavily or use last year's same period adjusted for growth.
Step 2: Measure Lead Time from Order to Sellable
Lead time is not what the supplier quotes. It is the days from placing the order to the units being available to sell:
- supplier production or picking time,
- transit and customs,
- receiving, counting and putaway in your warehouse,
- and for Amazon FBA, the extra shipment to Amazon and its receiving time.
Record it on every purchase order you receive. After a few orders you have an average and a spread, and the spread matters as much as the average.
Step 3: Choose a Safety Stock Method
Method 1: Days of Cover
Safety stock = Average daily demand x Safety days
Simple and easy to explain. At 12 units a day and 7 safety days, keep 84 units of safety stock. Its weakness is that it treats a steady product and a spiky one the same way.
Method 2: Max Minus Average
Safety stock = (Max daily demand x Max lead time) - (Average daily demand x Average lead time)
It uses your worst observed day and your slowest delivery. It protects you well, but because it assumes the worst demand and the worst lead time happen together, it usually holds far more stock than you need.
Method 3: Statistical, Using a Service Level
This method asks how often you are willing to run out during a replenishment cycle and sizes the buffer to match. If only demand varies:
Safety stock = Z x Standard deviation of daily demand x Square root of lead time in days
If both demand and lead time vary, which is the common case for imported goods:
Safety stock = Z x Square root of ( (Lead time x Demand deviation squared) + (Average demand squared x Lead time deviation squared) )
Z comes from the standard normal distribution and sets your service level, the chance of not running out in a given cycle:
| Service Level | Z |
|---|---|
| 90% | 1.28 |
| 95% | 1.65 |
| 97.5% | 1.96 |
| 99% | 2.33 |
| 99.9% | 3.09 |
Going from 95% to 99% raises safety stock by about 40% for the same product. Reserve the highest service levels for your best sellers and products where a stockout costs you rank, not every SKU.
A Worked Example
A product sells an average of 12 units a day with a standard deviation of 4 units. Lead time averages 21 days with a standard deviation of 4 days. You want a 95% service level, so Z is 1.65.
If only demand varied:
Safety stock = 1.65 x 4 x Square root of 21 = 1.65 x 4 x 4.58 = 30.2, so 31 units
With lead time variation included:
(21 x 4 x 4) + (12 x 12 x 4 x 4) = 336 + 2,304 = 2,640 Square root of 2,640 = 51.4 Safety stock = 1.65 x 51.4 = 84.8, so 85 units Reorder point = (12 x 21) + 85 = 252 + 85 = 337 units
Notice where the extra stock comes from. Demand swings alone call for 31 units; allowing for late deliveries takes it to 85. For many importers the cheapest way to cut safety stock is a more reliable supplier, not a better forecast.
For comparison, the max minus average method with a worst day of 20 units and a slowest delivery of 28 days gives (20 x 28) minus (12 x 21), which is 560 minus 252, or 308 units, more than three times the statistical answer.
Try it on your own data: start free, no card needed.
How Much to Order Each Time
The reorder point tells you when. The order quantity is a separate decision.
Order Up to a Target
The most practical rule for sellers is to order enough to reach a target level of cover:
Order quantity = (Target days of cover x Average daily demand) + Safety stock - Available - On order
Set target days of cover to your lead time plus the time until you would normally order again.
Economic Order Quantity
If ordering has a real fixed cost (freight minimums, inspection, admin time), the economic order quantity balances that against the cost of holding stock:
EOQ = Square root of ( (2 x Annual demand x Cost per order) / Holding cost per unit per year )
At 4,380 units a year, $60 per order and $2.50 a year to hold each unit (for example 25% of a $10 cost), EOQ = square root of (2 x 4,380 x 60 / 2.50) = about 458 units, or roughly 38 days of demand. Then round to the supplier's case pack and minimum order quantity, and check it against your cash plan.
If You Only Check Stock Periodically
If you review stock on a schedule (every Monday, say) rather than continuously, a product can cross its reorder point the day after you looked. Cover the review period too:
Order up to level = Average daily demand x (Lead time + Review period) + Z x Demand deviation x Square root of (Lead time + Review period)
The longer the gap between reviews, the more safety stock you need. Daily review, by a person or a system, keeps buffers smaller.
Adjust for Trend, Seasonality and Promotions
- Trend. If sales are growing 5% a month, demand during a 45 day lead time is higher than last month's average. Forecast the lead time period, not the past.
- Seasonality. Before a peak, raise demand for the weeks it covers and order early enough that the extra stock is sellable before the peak starts.
- Promotions and ads. A planned deal or a new campaign is known demand. Add it to the forecast instead of hoping safety stock absorbs it.
- Marketplace limits. For FBA, space limits may cap what you can send. See the FBA capacity guide.
New Products with No History
Formulas need history, and a new product has none. Until it has a few weeks of sales:
- Borrow demand from a similar product you already sell, scaled for price and expected interest.
- Order smaller and more often at first, even at a higher cost per unit, so a wrong guess does not leave you with months of stock.
- Use days of cover for safety stock (Method 1) with generous safety days, then switch to the statistical method once you have eight to twelve weeks of data.
- Review weekly instead of monthly until demand settles.
Several Suppliers or Several Locations
- Several suppliers for one product. Calculate the reorder point with the lead time of the supplier you will actually order from next. If you split orders, use the slower supplier's lead time for safety stock.
- Several locations. Calculate demand and reorder points per location when each location serves its own orders. A product can be fine overall and still run out in the warehouse that ships most of its orders.
- Amazon FBA. Treat Amazon's network as its own location with its own lead time, which includes your shipment to Amazon and its receiving time.
Common Mistakes
- Counting stockout days as zero sales. Forecasts shrink and stockouts repeat.
- Using the quoted lead time. Measure it from your own received orders.
- One percentage for every product. A flat "20% extra" over-stocks steady items and under-protects spiky ones.
- Ignoring stock already on order. You reorder twice for the same gap.
- Never revisiting the numbers. Recalculate at least monthly, and after any supplier change.
- Starting from a wrong count. The best formula fails on a wrong starting number. Keep counts accurate with cycle counting.
Reorder Point Checklist
- Average daily demand per product, across all channels, excluding stockout days.
- Lead time per supplier, measured from order to sellable, with its spread.
- A service level per product tier (for example 98% for top sellers, 90% for the long tail).
- Safety stock calculated with the method that matches how the product behaves.
- Reorder point checked against available plus on order, not on hand alone.
- Order quantity rounded to case packs and minimums, and checked against cash.
- Numbers recalculated monthly and before every peak.
How Invechar Does This for You
Every hour, Invechar's Brain checks each product against its demand. Demand comes from your real orders with the trend applied, lead time is learned from when each supplier's purchase orders actually arrive, safety stock is sized by how volatile each product's demand is, and stock already on order is counted so nothing is ordered twice. When a product reaches its reorder point you get a restock task, one draft purchase order per supplier, with every quantity editable before you approve. Your cash plan can hold large orders back, and each received order teaches the Brain the supplier's real lead time.
Read more about restocking and the Restock Planner and demand forecasts, or include freight and duty in your numbers with the landed cost guide.
Sources
- NIST/SEMATECH e-Handbook of Statistical Methods, Cumulative distribution function of the standard normal distribution, for the Z values. Accessed October 8, 2026.
- The reorder point, safety stock, order up to and economic order quantity formulas are standard inventory theory; the worked examples use illustrative numbers.