Reorder enough to cover sales until the order after this one can land. Multiply the weekly selling rate by production and transit weeks, plus the weeks until you next reorder, plus your buffer. Subtract the stock you hold, inbound shipments and orders placed but not shipped. Round what is left up to whole cartons.
An order does not have to last until it arrives. It has to last until the next one does.
The formula, one line at a time
The arithmetic
Weeks to cover = production + transit + review period + buffer
Target = weekly selling rate × weeks to cover
Short by = target − (stock held + inbound + ordered, not shipped)
Order = short by, rounded up to whole cartons
The review period is how often you sit down and reorder. If you check every two weeks, it is two. The buffer is the cover you want left when the next order arrives, in weeks of sales. Every term is per SKU and per sales channel: stock sitting in another channel does not cover this one's sales.
A reorder, worked
One SKU selling 70 units a week. The factory needs 5 weeks, freight takes 5, you review every 2 weeks and you want 4 weeks of buffer. You hold 300 units at FBA, and a shipment of 400 lands in three weeks. Cartons hold 24.
| Line | Working | Result |
|---|---|---|
| Weeks to cover | 5 + 5 + 2 + 4 | 16 weeks |
| Target | 70 × 16 | 1,120 |
| Supply already yours | 300 held + 400 inbound + 0 ordered | 700 |
| Short by | 1,120 − 700 | 420 |
| Order | 420 ÷ 24 = 17.5, so 18 cartons × 24 | 432 |
Why the weeks until your next order belong in it
A reorder point formula stops at the lead time: sales during production and transit, plus safety stock. Used as an order size, it leaves out the two weeks between this review and the next. Run the example without them: 70 × 14 is 980, less 700 is 280, which rounds to 12 cartons, or 288 units.
The next order cannot be placed until your next review, two weeks from now, so it lands two weeks after this one would. Those two weeks of sales, 140 units, have to come from somewhere. With the smaller order they come out of the buffer: half of the 4 weeks you meant to keep is gone before anything unusual has happened. A late container or a good fortnight then finds a 2-week cushion, not a 4-week one.
The rule generalises. Every week you add between reviews adds one week of sales to every order. Reviewing every four weeks instead of every two means each order carries 140 more units here, and that is a cash decision worth making deliberately.
Subtract what is coming, once
Supply already yours is everything you will not have to order again: stock on hand, shipments on their way, and purchase orders the factory is still making. Leave out an order that has not shipped yet and you buy those units twice. Count a shipment both as inbound and again inside the stock figure after it has been received and you buy too few.
The quieter double count is a second calculation added to a first. Say you drafted 300 units of this SKU last week and today the arithmetic says 432. The order needs 132 more, not 432 more: the 432 is the total the SKU needs, and the draft is part of it. Vendlen's Order Planning works this way. A proposal is the total target for the shipment plan it goes into, so a line already in the plan is topped up to the target and never added to it.
One limit on subtracting inbound: a shipment counts toward the quantity whenever it lands, but it only protects the weeks after it lands. If you will run out before it arrives, the order size is fine and the date is the problem. How days of supply hides a stockout before the inbound lands works through that case.
Rounding and minimums come last
Round up to whole cartons at the end, so the partial carton is the only extra: 432 against 420 here, 12 units. Rounding earlier in the chain, or rounding each channel separately before adding, stacks partial cartons.
Leave the factory's minimum out of this number. A SKU that needs 432 needs 432 even if the minimum run is 500. Whether to raise it to 500, combine it with other SKUs or bring a later order forward is a separate decision about cash, and how to meet a supplier MOQ without overbuying works through it. Vendlen keeps the two apart the same way: a minimum never holds a proposal back, and the per-SKU minimum and the total-order minimum are both checked when the plan becomes a purchase order.
Where this still needs a person
The selling rate is the input that moves the answer most. Every unit a week of error becomes 16 units in this order. In the sales-velocity example, nine days out of stock make a 30-day rate read 30% low before any arithmetic starts, which is why how to calculate Amazon sales velocity is its own question.
The formula projects the current rate forward. It does not know about a promotion you have planned, a seasonal peak, or a launch. If you know demand is about to change, change the rate first and say why.
The buffer is a commercial decision rather than a calculation: how much cash you will tie up against how much a stockout costs you. Vendlen does not guess it. Its Why this quantity panel shows the target, what is on hand, any units already in production and the proposed order, so the buyer can check each line before approving. When the rate or the stock is missing, the proposal reads Missing data instead of a quantity. A product with no measured rate has not stopped selling. See Inventory and Supplyfor how the proposal fits the rest of the chain.
What to do with this
- Pick one SKU on one channel. Write down its weekly selling rate, measured on days it was in stock.
- Add up the weeks to cover: production, transit, the gap until your next review, and the buffer you actually want.
- Multiply. Then list every unit already yours: on hand, in transit, and ordered but not shipped. Check that a received shipment is not also still listed as inbound.
- Subtract the list from the target. If a draft order for this SKU already exists, the result is what the draft should total, not what to add to it.
- Round up to whole cartons. Then, separately, check the result against the factory's minimums and against the week you will run out.
Questions sellers ask
What is the difference between a reorder point and a reorder quantity?
The reorder point is when to order: the stock level, or the date, at which a new order still lands before you run out. The reorder quantity is how many to order. Both start from the selling rate and the lead time, but the quantity also has to cover the weeks until your next order and subtract what is already coming.
How much safety stock should I keep for FBA?
Set it in weeks of sales rather than units, so it scales with the product. A longer or less reliable lead time, or a seasonal product, argues for more weeks. Each extra week costs one week of sales in cash, so the buffer is a trade-off you choose, not a number to copy.
Should inbound shipments count when I calculate a reorder?
Yes, every unit already on its way or on order counts once, or you buy it twice. Then check separately whether that inbound lands before you run out. A shipment that arrives after the stockout is the right quantity at the wrong time.
How often should I reorder from my supplier?
As often as you can review properly. The time between reviews is part of what each order has to cover, so every extra week between reviews adds a week of sales to every order. Reviewing less often means larger orders and more cash tied up.