Retail supply chain KPIs measure whether stock is where shoppers want it, at a cost the business can carry. The core set covers availability (on-shelf availability, in-stock rate, lost sales), inventory (turns, days of cover), clearance and freshness (sell-through, markdown percentage, waste) and plan quality (forecast accuracy, forecast bias, plan acceptance). Each one needs a fixed formula, a stated grain, a named data source and an owner.
How to Define Retail Supply Chain KPIs
Before a number goes on a dashboard, write down four things about it: the formula, the grain (SKU by store by day, or category by region by week), the data source and the person who owns the result. Two teams that report “availability” from different grains will disagree every week, and both will be right.
Keep the list short and tie each KPI to a decision someone makes: an order, a transfer, a price change or a forecast edit. A measure with no decision behind it tends to end up in a report that nobody acts on.
These are operational measures. If you need to put a money value on a planning project, our post on supply chain planning software ROI covers baselines and attribution.
Availability: On-Shelf Availability, In-Stock Rate and Lost Sales
Availability is what the shopper sees, and retailers measure it in three ways:
- On-shelf availability (OSA): the share of store and SKU checks where the item is on the shelf and sellable. It comes from store audits or shelf images, or it is estimated from point-of-sale gaps, such as an item that normally sells every day showing a run of zero-sale days.
- In-stock rate: the share of store, SKU and day combinations with positive system stock, usually taken at store opening. It is cheap to calculate from inventory records, though it misses stock that is in the backroom or wrongly counted.
- Lost sales: demand you could not serve because the item was out. You estimate it from the forecast, or the usual rate of sale, for the hours or days the item was missing. Agree the method once and keep it, or the trend means nothing.
Report in-stock rate and OSA side by side. A wide gap between the two points to shelf filling or inventory accuracy in the store, and ordering more will not close it.
Inventory: Turns and Days of Cover
Inventory measures show what availability costs:
- Inventory turns: cost of goods sold for a period divided by average inventory at cost for the same period. Annualize it so months and categories compare.
- Days of cover (days of supply): stock on hand divided by forecast average daily sales for the coming days. Use forward demand. Dividing by last month’s sales overstates cover going into a peak and understates it coming out.
- Excess and aged stock: units above a maximum cover target, or older than a set age, by store and SKU.
Turns and cover are averages, and a healthy chain figure can hide stores holding months of an item while others run out. Look at the spread by store and SKU, and read inventory next to availability. A drop in inventory that comes with a drop in availability has moved the cost into lost sales.
Sell-Through, Markdown Percentage and Waste
These three show how stock leaves the business, and which one matters depends on the category:
- Sell-through: units sold divided by units available (opening stock plus receipts) over a period. Fashion and seasonal teams track it weekly against the curve they planned for the season.
- Markdown percentage: markdown value (original price minus selling price, times units sold) divided by sales at original price. Some retailers divide by net sales instead, so state the denominator.
- Waste: the value of units written off for expiry or damage, as a share of sales in the same category. For fresh food, read it next to availability.
Each of these can be improved by giving up another. Overstocking fresh lifts availability and waste together, and heavy early discounts raise sell-through by pushing up markdown percentage. Targets for one should always name the limit on its pair.
Forecast Accuracy, Bias and Plan Acceptance
The last group measures the quality of the plan itself:
- Forecast accuracy: commonly 1 minus WMAPE, where WMAPE is the sum of absolute errors divided by the sum of actual sales. Measure it at the grain where decisions are made, such as store and SKU by day or week. Chain-level monthly accuracy looks far better and says little about store orders.
- Forecast bias: the sum of forecast minus actual, divided by the sum of actuals. A positive figure means over-forecasting, which builds stock and waste. A negative figure means under-forecasting, which shows up as lost sales. Accuracy can hold steady while bias drifts, so track both.
- Plan acceptance: the share of recommended orders, transfers or prices that planners and store managers execute without editing. Solvoyo calls this the user acceptance rate.
If planners rewrite a large share of recommended orders, availability and inventory reflect their edits as much as the system’s logic, so acceptance is what lets you credit a KPI change to the plan. Solvoyo explains how it measures this on its user acceptance rate page.
How Should Retail KPI Targets Be Set and Reviewed?
Set targets by segment. A fast-moving staple and a slow seasonal item should not share an availability target, and fresh categories need a waste ceiling next to their availability goal. Read these pairs together:
- Availability with days of cover or turns
- Availability with waste, for fresh categories
- Sell-through with markdown percentage
- Forecast accuracy with forecast bias
In Solvoyo’s published case study, the online grocer İstegelsin raised its fresh fill rate from 72% to 86% within a year of go-live with no increase in its waste ratio.
A workable cadence: daily exception lists for stock-outs and availability at store level, weekly reviews of cover, sell-through and bias, and monthly or seasonal reviews of turns, markdown percentage and waste.
How Solvoyo Tracks Retail KPIs
Solvoyo is a supply chain decision automation platform for retailers and manufacturers. Forecasting, replenishment, allocation and markdown decisions run on one data model, so every plan uses the same KPI definitions. The KPI cockpit reports forecast accuracy, plan adherence and service levels, and acceptance and reason codes are tracked for each decision, so a change in a KPI can be traced to the decisions behind it.
Each order is checked against pack sizes, shelf space and delivery days before planners see it, so the edit log stays short: over 95% of plans execute untouched. You can see the reporting side on the Retail Analytics & Collaboration page, and how these KPIs connect to forecasting and replenishment in our guide to retail supply chain planning. To see the KPIs on your own data, request a demo.
Frequently Asked Questions
The core retail supply chain KPIs are on-shelf availability or in-stock rate, lost sales, inventory turns or days of cover, sell-through and markdown percentage for seasonal goods, waste for fresh categories, forecast accuracy and bias, and plan acceptance. Read them in pairs, such as availability with inventory or waste, because each can be improved at the expense of another.
On-shelf availability is the share of store and SKU checks where the item is on the shelf and sellable. Retailers measure it with store audits or shelf images, or estimate it from point-of-sale data by flagging items that normally sell but show a run of zero-sale days. In-stock rate, based on system stock, is cheaper to calculate but misses stock that is not on the shelf.
Forecast accuracy measures the size of the errors, commonly as 1 minus WMAPE (absolute errors divided by actual sales). Forecast bias measures their direction: whether the forecast runs consistently above actual sales, which builds stock and waste, or below, which causes lost sales. A forecast can be fairly accurate and still biased, so track both.





