Reorder Point and Safety Stock for Amazon Sellers

Running out of stock costs you sales and organic position. Holding too much costs you cash and storage fees. Both mistakes usually come from guessing how long stock will last and how long it takes to replace. Four numbers remove most of the guesswork: your daily sales rate, your days of cover, your total lead time and a safety buffer. This article shows how they combine into a reorder point you can calculate for every SKU, how storage costs limit the buffer, and how a short weekly routine keeps you out of trouble.

Daily sales rate and days of cover

Your daily rate is units sold over a recent window divided by the days in that window. A window of four to eight weeks usually balances noise against freshness, but choose one that suits the product. Count units, not revenue, so price changes do not distort it.

Two adjustments matter. First, exclude days you were out of stock. A zero-sales day with no inventory says nothing about demand and will understate your rate. Second, decide what to do with spikes. If a promotion tripled sales for a week, a plain average will overstate normal demand, so either remove that week or track it separately. If sales are clearly rising or falling, weight the recent half more heavily. For seasonal products, last year's pattern around the same dates is a better starting point than last month.

Days of cover

Days of cover is sellable units divided by the daily rate. With 1,200 sellable units and a rate of 20 units a day, you have 60 days of cover. Count only what can be sold today. Units still in transit to Amazon are not sellable until they are received, so track them as a separate line with an expected date.

The useful comparison is days of cover against your lead time. If you have 60 days of cover and replacement takes 40 days, you have a 20-day cushion. If replacement takes 70 days, you are already late.

Lead time runs from order to sellable

Many sellers measure lead time as the supplier's production time. The number that matters is the time from placing the order until the units can be sold on Amazon. That includes:

• Production or preparation at the supplier.
• Pickup and freight to wherever you ship from.
• Transit, and customs if the goods cross a border.
• Any prep or relabeling.
• Check-in and receiving at the Amazon fulfillment center, which can take a few days or much longer depending on season and destination.
• The gap between your review dates. If you only look at a SKU every week, add up to a week.

Record the actual dates from each purchase order and keep two figures per SKU: the typical lead time and the worst recent one. Both will be used below.

Reorder point and safety stock

The reorder point is the inventory position at which you must place the next order. Inventory position means sellable units plus units already on order and on the way, so you do not reorder twice for the same gap.

In its basic form, reorder point equals daily sales × lead time, plus safety stock. Safety stock is the buffer that protects you when demand runs above average or the shipment runs late. A simple and conservative way to size it is to take the worst case for both, then subtract the average case. Suppose a SKU averages 20 units a day, peaks at 26, typically takes 40 days to become sellable and has taken as long as 50. Safety stock is (26 × 50) minus (20 × 40), which is 1,300 minus 800, or 500 units. The reorder point is then 800 plus 500, or 1,300 units.

At 20 units a day, 1,300 units is 65 days of cover, which is the 40-day lead time plus 25 days of buffer. Expressing the trigger in days is handy because it works for any SKU: reorder when days of cover, counting inbound, falls to lead time plus safety days.

The worst-case method is cautious. For an item that is cheap, fast-moving and easy to expedite, you may accept a smaller buffer. For a best seller with a long supply chain, you may want more. The buffer is a decision about how much risk of a stockout you are willing to carry, and it should follow what a stockout costs you.

Storage costs cap the buffer

Safety stock is not free. Amazon charges for the space your inventory occupies, usually by volume and more in busy seasons, and it can add charges for units that sit too long. It also limits how much inventory you can send in, based on your sales and your product types. Your buffer has to fit inside those limits, so check your current limits and fee schedule in Seller Central rather than assuming.

A quick test shows the trade-off. Say holding a unit costs $0.05 a month in storage, and you carry 500 spare units for three months. That is 500 × $0.05 × 3, or $75. A single week without stock at 20 units a day is 140 lost sales, and at $9 profit before ads each that is $1,260, before counting any rank you lose along the way. For a profitable, fast SKU the buffer is cheap insurance. For a slow SKU with thin margins, the same arithmetic points the other way.

A weekly exception routine

You do not need to review every SKU in depth each week. You need to find the ones that need a decision. Build one sheet with a row per SKU and let it flag exceptions.

Set your own upper limit for days of cover, based on product shelf life and storage cost. Anything above it is an overstock exception just as anything below the trigger is a shortage exception. Consider three hypothetical SKUs against the 65-day trigger. SKU A has 38 days of cover and needs an order now. SKU B has 70 days and is fine for the moment. SKU C has 210 days and should not be reordered, and may need ads dialed back or a markdown.

A rule for retiring weak SKUs

Every catalog has items that take up space and cash without earning much. Decide in advance how you will spot them, because deciding in the moment tends to favor keeping everything.

A workable test uses monthly contribution: units sold × profit per unit before ads, minus ad spend, minus storage, minus the cost of the cash tied up in stock. Say a SKU sells 6 units a month at $9 profit before ads, which is $54. It spends $40 on ads and $6 on storage, leaving $8. The 90 units on the shelf cost $12 each to land, so $1,080 is tied up, and at an assumed 1.5% a month that costs $16. The SKU loses $8 a month and has 15 months of cover.

If a SKU fails that test for two reviews in a row, stop reordering and sell it down. Then choose between clearing it with a price cut, removing it from the fulfillment center, or liquidating it, and compare what each costs. Check first whether the SKU supports a variation family or a bundle before you pull it.

The short version

• Calculate daily sales from a clean window, excluding stockout days and promotion spikes.
• Measure lead time from order to sellable, using your own dates.
• Set the reorder point as daily sales times lead time, plus safety stock, using inbound units in the count.
• Size the buffer against what a stockout costs and what storage costs.
• Flag shortages and overstock every week and decide on exceptions only.
• Retire SKUs that fail a monthly contribution test two reviews in a row.