Days of Inventory for Amazon Sellers Explained

Quick answer: Days of inventory equals sellable units divided by average daily sales velocity. It tells you how much time you have left, not just how many units, which is what actually prevents stockouts and overstock.

Stock level tells you how many units you have. Days of inventory tells you how much time you have left, and that difference is where most preventable stockouts and overstocks actually live.

Fifty units is a comfortable buffer for a product selling one a day, and a crisis for a product selling twenty a day. Sellers who watch unit count alone tend to discover the problem right around the time the listing hits zero.

The formula

Days of inventory equals sellable units divided by average daily unit sales. Three hundred sellable units selling 10 a day is about 30 days of cover. The same 300 units at 25 a day is only 12. The arithmetic is trivial. Choosing the right inputs is where the actual judgment starts.

Which inventory counts

Start with units that can realistically satisfy demand today. Available FBA inventory belongs in the calculation without question. Inbound stock is worth tracking separately rather than folding in as if it's already sellable, and reserved inventory needs context, since some of it is mid-transfer between fulfillment centers or tied to open orders. A useful dashboard shows available, inbound, and total pipeline stock as separate numbers, not one blended figure that hides which part is actually usable right now.

Which sales window to use

A seven-day window reacts fast, useful for launches, promotions, and sudden shifts, but noisy: one deal or one stockout-recovery spike can distort it badly. A thirty-day window is the practical default for most sellers, smoothing short-term noise while staying recent enough to catch a real trend. A ninety-day window is the steadiest of the three, but it can hide a recent acceleration or decline entirely. The most reliable read usually comes from comparing two windows at once: if seven-day velocity runs well above thirty-day, demand may be accelerating; well below, and the product may be cooling off.

Lead time changes what the number means

Forty days of inventory sounds comfortable until you add up that production takes 25 days, ocean freight takes 30, receiving takes 10, and the supplier needs a week just to confirm the order. True replenishment lead time here is over 70 days, which means 40 days of stock isn't comfortable. It's already late. A reorder point needs to cover supplier lead time, freight, customs, prep, Amazon receiving, and a safety buffer, not just factory production time, which is the piece sellers most often underestimate. Running too lean has its own penalty too: Amazon's Inventory Performance Index factors in low-inventory-level fees below a set days-of-supply threshold, covered in the stockout alert that actually matters.

Seasonality is a real adjustment, not an excuse

A garden product entering spring, a gift item approaching December, or a listing scheduled for a major deal event can all sell faster than the recent average suggests. Adjust the forecast when there's a real reason to, using previous seasonal performance and current conversion data. The key word is real: a hopeful sales target isn't demand data, and treating it like one is how overstock happens.

One-off inventory needs a different lens entirely

Used books, collectibles, and retail-arbitrage finds usually have one unit per SKU, where traditional reorder forecasting doesn't really apply since the seller often can't buy the same item again. The days-of-inventory idea still helps at the category or batch level, though: compare units listed, units sold per day, and average age by subject, condition, or sourcing batch. The question shifts from "when do I reorder this ISBN" to "which slice of the catalog is actually converting, and where is cash quietly sitting still."

A weekly routine that scales

Sort ASINs into four groups every week: urgent stockout risk, reorder soon, healthy cover, excess cover. Then work the exceptions, not the whole list. Urgent items get expedited shipping, a bid reduction, or a careful price increase. Reorder-soon items get a confirmed purchase order. Excess-cover items get inbound shipments paused and a look at price, Buy Box, and aging. Healthy products need no meeting at all, which is the entire point of sorting first.

Common ways the calculation gets quietly wrong

Counting canceled orders as sales. Letting out-of-stock days drag down apparent demand without adjustment. Blending parent and child ASINs that carry different stock levels. Assuming inbound stock lands exactly on its promised date. Using a ninety-day average for a product whose price changed yesterday. The formula is arithmetic; the quality of the decision depends entirely on clean inputs.

---

EcomSanity calculates days of stock from actual sales velocity against current inventory and flags listings approaching a low-stock threshold automatically, so the number is a standing gauge instead of something remembered right before it's already too late. For the underlying velocity formula and how to handle stockout days in the calculation, see Amazon sales velocity, and for what "usable inventory" actually means status by status, see Amazon FBA inventory statuses explained. For turning that number into an actual reorder decision, see the Amazon reorder point formula. A subscription product's days of inventory needs its own adjustment too, see Amazon Subscribe & Save for sellers for why a mismatched reorder cycle can drain stock faster than the standard velocity number predicts.