A bestseller going out of stock on Amazon is not just a missed sales event. It can interrupt ranking momentum, waste ad spend, delay the next reorder, and force the finance team to explain why cash is tied up in slower SKUs. That is why ecommerce inventory KPIs need to do more than populate a dashboard. They need to tell your team what to buy, what to stop buying, and when to act.
For Amazon-centered brands, the best KPI set connects demand, available inventory, open purchase orders, supplier lead times, and unit economics. A single inventory number is rarely enough. The goal is to see risk early enough to change the outcome.
The ecommerce inventory KPIs that deserve daily attention
Not every metric belongs in the daily operating rhythm. Finance may review cash conversion over a longer period, while a replenishment manager needs immediate visibility into SKUs that will stock out before the next inbound shipment is available. Start with the KPIs that directly change replenishment and advertising decisions.
1. Days of supply
Days of supply estimates how long your available inventory will last at the current or forecasted sales rate.
Days of supply = Available units / Forecasted average daily unit sales
If a SKU has 900 sellable units and forecasted daily demand of 30 units, it has 30 days of supply. That sounds straightforward until you account for inventory that is inbound, reserved, stranded, or split across FBA, AWD, FBM, and Shopify. Looking only at one channel can create a false stockout alert or, worse, hide a real one.
Days of supply becomes useful when measured against total replenishment lead time, not an arbitrary target. If a supplier needs 35 days to produce and ship, and Amazon inbound processing adds another 10 days, 30 days of supply is already late. If a fast-moving seasonal item has 90 days of coverage, it may be consuming cash you need elsewhere.
2. Stockout risk and projected stockout date
A stockout-risk KPI flags SKUs projected to run out before their next replenishment can become sellable. The projected stockout date should reflect forecasted demand, on-hand inventory, confirmed inbound quantities, and realistic arrival dates.
This is the operating metric behind the question, “What needs action this week?” A good view separates true emergencies from items that only look risky because a purchase order is in transit. It also accounts for supplier-specific lead times, rather than applying one generic lead-time assumption across every SKU.
For high-margin or high-velocity products, use the signal to protect availability first. You may need to expedite a purchase order, redirect inbound inventory, or reduce advertising pressure until supply recovers. Slowing spend when inventory cannot support demand protects both conversion performance and customer experience.
3. Reorder point coverage
The reorder point is the inventory level that should trigger a replenishment decision. It generally combines expected demand during lead time with safety stock.
Reorder point = Forecasted demand during lead time + Safety stock
The KPI to watch is not only the reorder point itself, but whether each SKU is above or below it after considering open POs and expected receipt dates. A reorder point that ignores minimum order quantities, order cadence, or variable lead times creates recommendations your purchasing team cannot execute.
Safety stock should not be identical for every item. Stable, predictable sellers usually require less protection than volatile products with long supplier lead times. The trade-off is clear: more safety stock reduces stockout exposure but increases carrying costs and cash tied up in inventory.
4. Inventory turnover
Inventory turnover shows how many times inventory is sold and replenished over a period.
Inventory turnover = Cost of goods sold / Average inventory value
Higher turnover is often good, but it is not an automatic win. A SKU can turn quickly because it is understocked, repeatedly losing sales before demand is fully captured. Conversely, a strategic product with a long lead time may carry lower turnover by design.
Use turnover to spot capital that is not working hard enough, then pair it with in-stock rate and gross margin. A low-turn, high-margin SKU may deserve measured coverage. A low-turn, low-margin SKU with rising storage costs needs a tougher conversation about future purchase quantities.
5. Sell-through rate
Sell-through rate measures how much inventory sold during a defined period relative to the units available to sell.
Sell-through rate = Units sold / Beginning inventory plus receipts × 100
For Amazon sellers, sell-through is especially useful for identifying inventory that is moving too slowly relative to the capital committed. Review it by SKU, product family, and age bucket. A blended account-level rate can hide the fact that a few slow items are absorbing a disproportionate share of cash and storage costs.
Do not use sell-through alone to set a reorder quantity. Promotions, stockouts, listing changes, and seasonality can distort the period. Forecasted future demand remains the better foundation for the next PO.
6. Excess inventory value
Excess inventory value quantifies the dollars invested in inventory beyond your target coverage. It is one of the most commercially useful ecommerce inventory KPIs because it converts a vague concern about overstock into a cash number.
Calculate it by comparing current and inbound inventory against the forecasted demand you expect to cover over your planning horizon. The result should be valued at landed cost, not retail price. Landed cost gives finance and purchasing a clear picture of cash at risk.
Excess inventory is not always a mistake. A supplier price break, a known seasonal event, or an extended production shutdown can justify additional coverage. The key is that it should be deliberate. Inventory acquired without a documented demand case is not safety stock. It is a cash-flow decision made by accident.
7. Aged inventory percentage
Aged inventory tracks units that have remained unsold beyond defined thresholds, such as 90, 180, or 365 days. This KPI matters because a product can look adequately stocked in total while older units quietly build up.
Review aged inventory alongside the current forecast. If projected demand cannot clear the inventory within a sensible window, stop replenishing it before looking for ways to stimulate demand. The first win is preventing the problem from getting larger.
Set thresholds based on category behavior and lead time. A 90-day threshold may be too aggressive for a slower, planned-purchase category, while 180 days may be far too forgiving for a high-velocity Amazon product.
KPIs that connect inventory to profit
Operational availability protects revenue. Profitability metrics keep a growing catalog from becoming a growing cash problem.
Gross margin return on inventory investment
GMROII measures the gross margin earned for each dollar invested in average inventory.
GMROII = Gross margin dollars / Average inventory cost
This is particularly valuable when two SKUs have similar sales but very different margins, costs, or inventory requirements. Revenue can make a low-return product look important. GMROII reveals whether its inventory investment is actually producing enough gross margin to justify the capital.
Use it as a prioritization metric, not a reason to starve lower-margin products that anchor a broader assortment. Context matters. Some items support repeat purchases or create demand for profitable variations.
Forecast accuracy and forecast bias
You cannot improve replenishment decisions if your demand plan is consistently wrong. Forecast accuracy compares projected demand with actual sales. Forecast bias shows direction: are forecasts consistently too high or too low?
A forecast that is occasionally wrong is normal. A forecast that is systematically high creates overstock. One that is systematically low creates stockouts and emergency purchasing. Measure both at SKU level and at the aggregate level. Portfolio-level accuracy can look acceptable while high-revenue individual SKUs are repeatedly missed.
Also separate demand from constrained sales. If an item was out of stock for 10 days, its recorded sales do not represent full demand. Treating stockout-period sales as normal demand bakes the shortage into the next forecast.
Build a KPI cadence that produces decisions
The dashboard is not the process. Assign an owner and an action rule to each KPI. Stockout risk should lead to a PO, expedited shipment review, inventory transfer decision, or ad adjustment. Excess inventory should trigger a future-buy reduction, not another meeting with no purchasing change.
A practical cadence is daily review for stockout risk, days of supply, and exceptions on top-selling SKUs. Review replenishment recommendations and supplier constraints weekly. Review turnover, aged inventory, GMROII, and forecast bias monthly with finance and leadership. Agencies should standardize these definitions across accounts, while still allowing client-specific lead times, MOQs, and coverage targets.
The data model matters as much as the meeting schedule. Pulling sales from Amazon and Shopify while ignoring in-transit POs, AWD inventory, or supplier constraints produces polished but unreliable KPIs. Inventory Optimizer centralizes those inputs and applies SKU-level demand models so teams can move from reporting to supplier-ready reorder decisions.
The best inventory KPI is the one that changes a decision before cash is committed or a stockout occurs. Start with a small set, make the action rules explicit, and let every number earn its place in the weekly operating rhythm.
