Reorder point and safety stock calculator for Amazon FBA and 3PL stock

How do you use this reorder point calculator?

Work one SKU at a time, in three groups. Demand: average daily units on Amazon and on any other channel that draws on the same stock, each with its day-to-day spread. Lead time: five legs, from the supplier PO to units available at FBA, plus how many days deliveries usually miss by. Stock: what sits at FBA, on the way to FBA, at AWD, your 3PL or your warehouse, and on open supplier POs. Every field starts with the worked example below, an illustrative SKU rather than a client's numbers, so you can see each line move as you type your own.

The result gives two answers from the same formula. The supplier reorder point uses the full lead time, every channel and every place, and tells you whether the next purchase order is due. The FBA transfer point uses only the last leg (transit to Amazon plus check-in), Amazon demand and FBA stock, and tells you when to send the next shipment from your 3PL or warehouse.

This is the trap the calculator is built to catch. FBA shows more than 32 days of cover, which feels comfortable, yet the supplier order is already due, because the lead time is 73 days and the whole position is 77 units under the reorder point. The reverse happens too: thousands of units at the 3PL, a supplier order nowhere near due, and FBA running dry because nobody sent a transfer 11 days ahead.

How is the reorder point calculated?

Reorder point = average daily demand × total lead time + safety stock

The reorder point glossary entry covers the definition and common mistakes; here is how each term is filled in. Daily demand is the average for days the SKU was in stock, because days at zero drag the average down and make the reorder point too low. Total lead time runs from placing the supplier PO to units being available to sell at FBA, not to the day the container leaves the port. The result is compared with the inventory position, meaning stock in every place plus open POs, and it is rounded up to a whole unit: 4,176.2 means reorder at 4,177.

Run the same formula on the last leg and you get the FBA transfer point. Amazon does something similar inside AWD: its AWD replenishment help says auto-replenishment triggers when available units in the FBA network plus units in transit to it fall below the target level by at least half a carton, and that target uses forecast demand, its volatility, the lead time from AWD to FBA and any minimum or maximum you set (checked September 25, 2026). Our transfer point is the plain version of that rule for stock you move yourself. It treats the last leg as fixed, so enter the long end of your check-in times, not the average.

The transfer point says when to send, not how much. Size each transfer to the days of cover you want at FBA, within your FBA capacity limits. Amazon's capacity limits page says open shipments count toward capacity usage, so a transfer takes space from the day you create it (checked September 25, 2026). In the US, the low-inventory-level fee applies when both the 30-day and 90-day historical days of supply are under 28 days, measured per FNSKU since January 15, 2026, with exemptions such as SKUs that sold fewer than 20 units in the past 7 days (checked September 25, 2026). Thin FBA stock has a price even before it runs out.

How much safety stock do you need?

Safety stock = z × √(L × σD² + D² × σL²)

Here L is the total lead time in days, σD the standard deviation of daily demand, D average daily demand and σL the standard deviation of the lead time. When the lead time never varies, σL is zero and the formula shrinks to z × σD × √L. This is the textbook version given on Wikipedia's safety stock page, which also states its assumptions: demand on each day is an independent, normally distributed draw, and the averages and spreads stay constant (checked September 25, 2026). The safety stock glossary entry covers other methods and where to hold the buffer.

The z value sets the service level: the chance that one replenishment cycle, from placing an order to receiving it, ends without a stockout. It is not the share of orders filled, which is a different measure. We checked each z below against the NIST/SEMATECH table of the standard normal distribution (checked September 25, 2026): the table gives an area of 0.39973 between 0 and 1.28, and adding 0.5 for the half below zero gives about 90%. The last column is the worked example SKU at each level.

Two things stand out. First, the top of the table is expensive: 99.9% needs almost twice the buffer of 95%, so keep the highest levels for SKUs where a stockout really hurts. Second, the lead-time term often matters more than demand noise. With a fixed 73-day lead time the example needs 183 units of safety stock; let deliveries vary by 6 days and it needs 527, almost three times as much. For ocean-freight SKUs, a supplier and forwarder who deliver on time save more buffer than smoother sales would.

The normal assumption fits steady sellers best. For slow, lumpy SKUs with many zero-sale days, the formula can misstate the buffer. For those, add up what actually sold during each of your past lead times and compare the reorder point with the busiest of those stretches. If the busiest stretch is higher, the formula is too optimistic for that SKU.

Why are FBA lead times longer than your supplier's?

Because your supplier quotes one leg and FBA needs all five. The number that matters is days from PO to units available to sell, and each leg has its own place to measure it:

Amazon publishes guidance for the last leg. Its help page on tracking shipments says a delivered shipment typically takes 2 to 5 days to check in and up to 1 more day to reach Receiving, and that units become available for sale as soon as they are received at the first Amazon facility (checked September 25, 2026). Prime eligibility typically follows 6 to 15 days after receipt, depending on transfers to the final fulfillment centers. The worked example's 6 days sits at the top of the check-in range, which is where a planning number belongs.

The same page says weather disruptions can add 3 to 7 days and post-holiday peaks, such as after Black Friday or Prime Day, 5 to 10 days. Measure your own from the Shipment events tab: Shipped to Delivered is the transit leg, and Delivered to Receiving is the check-in leg. Use a longer last leg for shipments that land in late November, December or after Prime Day, and re-measure after each peak, because averages from a quiet month run short.

Amazon's own tool splits lead time too. Its Restock Inventory help offers three supply chain settings: the supplier ships to Amazon, where only supplier lead time counts; the supplier ships to your facility and you forward the full order, where supplier and your own lead time add up; or you hold bulk stock and replenish Amazon as needed, where only your lead time counts (checked September 25, 2026). Our calculator follows the same logic: the supplier reorder point uses all five legs and the FBA transfer point only the last two. If your supplier ships straight to Amazon, put that trip in the freight leg, enter zero for the 3PL and transit legs, and ignore the transfer point.

What changes when you sell on more than one channel?

Demand adds up, and stock splits across places. On the demand side, add each channel's daily units that draw on this stock: Walmart, Shopify, TikTok Shop or wholesale. Orders from other channels that FBA ships through multi-channel fulfillment draw on FBA stock, so enter them with Amazon demand. The calculator applies the full five-leg lead time to all of this demand. Orders your 3PL ships straight to customers only wait for the supplier, freight and receiving legs, so for a SKU sold mostly off Amazon the result errs on the high side.

The spreads do not add up the same way. If the channels move independently, the combined spread is the square root of the sum of their squares, so one pooled buffer needs fewer units than separate channel buffers added together. In the worked example, separate buffers for Amazon and the other channels add up to 551 units against 527 pooled. With a fixed lead time the gap is wider: 239 separate against 183 pooled, about 23% less.

Pooling helps with demand noise but not with late deliveries, because one late container delays stock for every channel at once. If your channels spike together, for example when a viral video lifts Amazon searches and Shopify orders in the same week, treat them as moving together: plan on the result's "Separate buffers per channel, added up" line rather than the pooled safety stock.

On the stock side, count every unit once, in the right check:

Amazon's FBA inventory report reference shows the FBA Manage Inventory report with separate columns for fulfillable, inbound working, inbound shipped and inbound receiving quantities, and the Restock Inventory report with reserved units split into FC transfer, FC processing and customer orders (checked September 25, 2026). Together they make the split above easy to pull. Working quantities are the usual double count: they sit in a shipment plan and on your 3PL's shelf at the same time. The replenishment guide covers how the full plan fits across FBA, AWD and 3PLs.

When is another tool the better choice?

If you sell only on Amazon, ship each SKU on one route and keep no stock outside Amazon, Amazon's Restock Inventory is the better choice: it is built into Seller Central and uses Amazon's own demand forecast, a probabilistic model with a mean and an optimistic level that demand should stay at or below 90% of the time, updated weekly (checked September 25, 2026).

If you have no daily sales history, only your worst day and your longest lead time, Omni's safety stock calculator is the quicker choice: it offers a max-minus-average method alongside the statistical one (checked September 25, 2026). And if you want off-the-shelf forecasting and purchasing across many warehouses and channels, and your process fits a standard tool, planning software such as Inventory Planner by Sage, whose replenishment feature promises you will "know what to order, when and where", is the better choice (checked September 25, 2026).

This calculator is for the case in between: one SKU, a lead time you can split into legs, stock in several places, and a formula you can see and check line by line. Unlike Restock's documented settings, it asks for other channels, 3PL stock and open POs. Unlike a one-line calculator, it shows which leg of the lead time drives the number.

Want reorder points from your own data?

A calculator works one SKU at a time from numbers you look up. Doing it for every SKU, every day, from real receipts and shipment events is the replenishment problem, and when it slips it shows up as stockouts. The stockout cost calculator estimates what one costs you from your own numbers, so you can weigh it against the extra safety stock.

Ecomsellertool builds and deploys an inventory and replenishment agent on your own accounts, starting from Growth OS, whose inventory module gives "reorder points from real lead times, draft POs and shipment plans before you run out". Every purchase order comes back to your team for approval. You keep your accounts, your data and the custom code we build; the Growth OS base is licensed to you.

We have built this kind of multi-location stock tracking before. Ecomsherpa, the warehouse management system we built for a client that had run its stock in Excel, manages both Amazon FBA inventory and local warehouse inventory, along with purchase orders, replenishment quantities and sales velocity. It also imports and exports Excel sheets, so the team could keep its spreadsheets while orders and stock moved into one dashboard.

To see where your own replenishment leaks, start with the free 24-hour diagnostic. You connect Amazon with Login with Amazon, no password is shared, and we only read data; we never change listings, prices, stock or ads. You get a report within 24 hours of connecting, on business days. Prefer to talk first? Schedule a call from the card below, or try the other free calculators.