Amazon FBA Reorder Point: How to Set It From Actual Lead Times

An Amazon FBA reorder point is the stock level that tells you to place the next order: daily sales velocity times lead time in days, plus safety stock. The formula is the easy part. The lead time inside it is where a plan quietly goes wrong, because it is easy to treat it as one number from a supplier quote. Build it from each step of the trip instead, using the actual durations of your recent shipments.

Below: the formula, a worked example where a 60-day plan meets an 80-day shipment, a step-by-step way to rebuild lead time, what we found in our own shipments, the traps that skew each input, and how much to order without running into Amazon's fees on either side.

How to calculate an Amazon FBA reorder point

An Amazon FBA reorder point equals your expected sales during the lead time plus a safety buffer. In formula form, reorder point = demand during lead time + safety stock, and it takes three inputs:

1. Daily sales velocity: the units you sell per day, averaged over a lookback period.
2. Lead time: the days from placing the order until the units are available to sell in FBA.
3. Safety stock: extra units that cover delays and sales spikes.

Propamp's Restock Suggestion uses the same shape and shows its math:

1. Safety stock = velocity × 14, or 14 days of sales.
2. Reorder point = safety stock + velocity × lead time.
3. Days until stockout = current stock ÷ velocity.

When stock falls to the reorder point, you order. Each input can be wrong in its own way; the sections below start with lead time.

Why one lead time number makes you stock out

A single lead time number hides a chain of steps, and it usually reflects the plan rather than what happened on your last shipment. Stock made overseas and sold through FBA typically passes through these steps:

1. Production at the factory
2. Ocean or air freight
3. Customs clearance
4. Domestic freight to a warehouse or to Amazon
5. Prep and labeling
6. A transfer from AWD (Amazon Warehousing and Distribution), if you store units there first
7. Amazon receiving, until the units show as available

Most steps start only when the one before them ends, so a late step pushes every step after it. A supplier's quote may cover production alone, depending on the terms you agreed, and a forwarder's transit time covers only its freight leg. If your lead time is the sum of the quotes, every slip downstream comes straight out of your safety stock.

Worked example: 60 days planned, 80 days actual

Here is how a 20-day gap between plan and reality turns into a stockout. The numbers are illustrative round figures, not data from any real account.

Say a SKU sells 20 units a day. These are the planned and actual days for each step of its last shipment:

1. Production: 30 planned, 34 actual
2. Ocean freight: 20 planned, 26 actual
3. Customs: 3 planned, 6 actual
4. Domestic freight: 4 planned, 5 actual
5. Prep: 1 planned, 2 actual
6. Amazon receiving: 2 planned, 7 actual

The plan, built from the quotes, adds up to 60 days. The shipment took 80.

With safety stock at 14 days of sales, or 280 units:

1. Reorder point on the plan: 280 + 20 × 60 = 1,480 units.
2. Reorder point on the actual: 280 + 20 × 80 = 1,880 units.

If you reorder at 1,480 units, you hold 74 days of stock for an 80-day trip. The shelf goes empty on day 74 and stays empty for 6 days, about 120 units of missed sales. The 14-day buffer absorbed 14 of the 20 late days, and the other 6 became a stockout.

No single step looked alarming: the biggest slip was 6 days on the water. Together the slips came to 20 days, and only the step-by-step view shows that ocean freight (6 days late) and Amazon receiving (5 days late) are the ones to plan around.

How to build lead time from actual shipments

Build lead time as a sum of steps, each based on what your recent shipments actually took. The routine works in a spreadsheet or in a tool:

1. List the steps for each supplier and route. Sea and air shipments from the same factory need separate lead times.
2. For each step, record the planned and the actual days on your last few shipments.
3. Use the typical actual for steps that stay stable, such as prep.
4. Use the slower end of the recent range for steps that swing, such as customs or receiving.
5. Add the steps. That sum is the lead time for the reorder point.
6. Update it every time a shipment lands.

Don't take the worst case for every step at once. That plans for a shipment where everything goes wrong together, which inflates the reorder point and the cash tied up in stock.

The math is simple; the work is collecting the actual dates. If your POs (purchase orders) already live in one place, such as Propamp's purchase order tracking, dates like placement, production start, ready and ETA are already recorded instead of scattered across email threads.

What we found in our own shipments

On our own shipments, the arrival date showed how far each trip drifted, but not which step caused it. When we read the lead time plans of our open shipments on our own brand's account (as of October 26, 2026, through Propamp's MCP server rather than the screens), we saw this:

1. Each plan was a list of steps with a planned duration: ocean freight, two or three ground freight legs, a customs hold, warehouse and prep, and FBA check-in. Planned time to check-in ran from 31 to 60 days.
2. Our plans had no actual dates entered for their steps. The actual dates we had sat on the shipment: an actual arrival date, a customs clearance date on some shipments and a first receipt date.
3. Of the four linked shipments that had arrived, two came in later than planned, by 4 and 8 days, and two came in earlier, by 4 and 11 days.
4. On the two recent shipments with a customs date we could use, customs cleared ahead of the planned customs-hold day, by 9 and 13 days.

Four arrivals are a small sample, and neither late one slipped by more than the 14 days of safety stock that Restock Suggestion holds. What the arrival dates could not tell us is which step to plan around. That is why the routine above depends on your own records: log milestone dates such as customs clearance and first receipt as they happen, because those were the only step-level actuals we had.

Velocity moved the answer as well. When we divided stock by sales velocity for seven stocked products on the same account (US marketplace, same date), the 7-day velocity gave the product with the least cover about twice the days of cover that the 30-day velocity gave. A restock call changes with the velocity window you divide by and with whether the shipment due this week actually arrives, so check both before you order.

Where the usual reorder point advice breaks

The usual advice breaks when an input is skewed: velocity measured during a stockout, a fixed buffer, a seasonal peak, or a stock count that leaves out units already on the way.

Velocity measured during a stockout is too low

A stockout drags down the velocity you use for the next reorder, so the next reorder point comes out too low. Say the same SKU sells 20 units a day when it is in stock, but it was out of stock for 10 of the last 30 days. It sold 400 units, so the 30-day average reads 13.3 a day. Plug that into the 80-day lead time and the reorder point drops to about 1,253 units instead of 1,880, a third lower. The next stockout is already scheduled.

Before you trust a lookback average, check how many days the SKU was out of stock in that window. Then pick a window without the stockout, or enter the in-stock rate by hand.

A fixed safety buffer covers only its own length

A safety stock set in days of sales covers a delay of exactly that many days, as long as sales hold steady. A fixed buffer, such as the 14 days in Restock Suggestion, does not grow when a route becomes less predictable. So put the risk into the lead time instead: if customs or receiving keeps slipping, the slower actual belongs in the lead time number.

Seasonal peaks need the velocity you expect on arrival

Lookback velocity describes the past, so before Prime Day or Q4 you need the sales rate you expect while the new units are selling. With an 80-day lead time, an order placed in mid-August becomes sellable in early November. Raise velocity by hand for the weeks the order has to cover, or start from the same period last year.

Current stock should include units already on the way

Compare the reorder point with everything that will be sellable before a new order could land: available FBA units, inbound shipments, units in AWD and stock at a 3PL (third-party logistics warehouse). Leave out the units on the water and the reorder point triggers a second order for stock you have already paid for.

How much to order: Amazon's fee floor and ceiling

The reorder point tells you when to order; the order quantity decides how long the stock lasts, and Amazon's fees bound it on both sides. Restock Suggestion computes order quantity = max(0, velocity × (coverage days + lead time) − current stock), where coverage days is how long the new order should last after it lands.

In the example, with 60 coverage days: 20 × (60 + 80) − 1,880 = 920 units. The 920 new units alone cover 46 days of sales. They land on top of the 280 units of safety stock still on hand, so the next cycle starts with 1,200 units, or 60 days of sales.

Two Amazon fees mark the edges:

1. The floor: the low-inventory-level fee. As Amazon staff explained in the Seller Forums, Amazon measures historical days of supply, your inventory relative to actual sales, over the last 30 days and the last 90 days, and charges the fee only when both are below 28 days. A plan that keeps FBA stock thin between shipments can pay it even without a stockout. Exemptions apply and Amazon updates these rules, so check the current terms in Seller Central.
2. The ceiling: the aged inventory surcharge. Amazon charges it monthly on items stored in a fulfillment center for more than 181 days, as of October 2026. If an order takes longer than 181 days to sell through after it lands, its last units start paying the surcharge.

Between those edges, coverage days is a cash decision: a longer cover means fewer orders and more money tied up in stock.

Setting it up in Propamp

In Propamp, the reorder math lives in Restock Suggestion and the lead time history lives in your shipments. A setup that follows the method above:

1. Open Restock Suggestion. For each product it shows the reorder point, order quantity and days until stockout, and it ranks products as Critical, Urgent, Soon or Planned.
2. Set velocity: pick the lookback period, use the year-over-year option, or enter your own number, for example before a promotion. Override templates can replace the velocity-based reorder math for specific products.
3. Check the Out of Stock view for how many days each product was out of stock before you trust a lookback that covers a stockout.
4. Pull the lead time plan. Propamp keeps the lead time of a shipment step by step: production, air or ocean freight, customs, domestic freight, prep and AWD. Today you read it through Propamp's Amazon MCP server, so an AI agent such as Claude can lay the planned days next to the actual dates you recorded. Here is how to connect Claude to your purchase orders and shipments. If Amazon check-in on recent shipments ran long, add that time on top.
5. Check stock by location across FBA, FBM (fulfilled by merchant) and AWD, including AWD on-hand and available units, before you decide what counts as current stock.

MCP (Model Context Protocol) access is included in the trial; on paid plans, outside AI clients need the MCP access add-on. Claude needs a Pro plan or higher.

Restock Suggestion does the math on the inputs you give it. It has no seasonal forecast and does not send stockout alerts, so the velocity and lead time you feed it decide the result.

Try it on your own catalog

Start a two-week free trial (no credit card) from the Amazon inventory management page, connect Seller Central, and run your top SKUs through Restock Suggestion with lead times built from actual shipments.

How we made this: drafted with AI assistance and checked against Propamp's public documentation, the sources linked above and our own brand's account (observations as of October 26, 2026).