Retail KPIs for a Store Chain: Ten Measures, Their Formulas and Where Each One Misleads

Retail KPIs are the few measures a retailer reads every week to see whether it is selling more, keeping its margin and putting its stock to work. A chain with twenty stores can produce hundreds of numbers; it needs about ten, each with one formula that means the same thing every week and in every store.

This guide lists those ten in three groups, with the formula in one line, the mistake that most often makes each one misleading, and a link to the detailed guide where there is one. Two rules apply to all of them. Define the basis once (VAT in or out, returns netted or not, stock at cost or at retail) and keep it. And read each KPI store by store against comparable stores, not only as a chain total.

1. Sales KPIs

Like-for-like sales growth is the sales growth of the stores that traded in both periods, so it separates what the existing estate did from what openings and closures added. Total growth and like-for-like growth are both true; they answer different questions, and listed retailers publish the bridge between them.

The common mistake is a base that is not like for like. NEXT's definition excludes store closures, openings and refits¹; others count a refitted store or a relocation differently, so a figure is only comparable with one built on the same rule. The like-for-like guide compares six published definitions.

Transactions are the number of completed sales receipts in the period. The mistake is counting returns, voids and staff sales as customer visits, or mixing online orders into a store's count. Some retailers call the same figure visits: Tokmanni defines customer visit development as the development in the number of customer transactions².

Average transaction value is net sales ÷ transactions. The mistake is averaging the stores' own averages, which weights a quiet store the same as a busy one; the chain figure is total sales over total transactions. Its value is in the split: sales growth is transaction growth times ticket growth. In the second quarter of 2026, Tokmanni's like-for-like customer visits fell 0.7% while the like-for-like average basket rose 1.0%, and like-for-like revenue grew 0.3%². The average transaction value guide works through the aggregation.

2. Margin KPIs

Gross margin % is (net sales − cost of goods sold) ÷ net sales, with sales excluding VAT. Three mistakes recur: comparing a margin with VAT in the sales figure against one without; confusing margin with markup (a 50% markup on cost is a 33% margin on sales); and comparing stores whose margin includes different shares of markdown and waste. A store's margin is best read against the median of stores of the same format, as the retail operations dashboard guide explains.

GMROI, gross margin return on inventory, is gross margin in money ÷ average stock at cost: how many pounds of margin each pound of stock earns in a year. It combines margin and turnover, and it shares turnover's weakness. Thin stock raises it, so a category with frequent empty shelves can show an excellent GMROI while it loses sales. Read it with availability, and keep the stock at cost: margin over stock at retail is a different and smaller number.

3. Stock KPIs

Stock turnover is cost of sales ÷ average stock at cost, and days of stock is 365 ÷ turnover. The mistake is dividing sales by stock at cost, which overstates the turn by the inverse of one minus the margin. Days of stock is the easier figure to act on store by store, because it reads directly as how long the shelf would last.

Sell-through rate is units sold ÷ units received, the measure for seasonal and one-time buys rather than lines that are reordered. The mistake is reporting one figure that mixes full-price and markdown sales, and reading it without the week of the season it was taken in.

Availability is the share of product and store lines with stock to sell, and the out-of-stock rate is its complement. In Italian fast-moving consumer goods, GS1 Italy's monthly barometer put the out-of-stock rate at 3.4% in September 2025, with 4.2% of sales lost³. The mistake is measuring it from the stock record alone. In a study of ten drugstores, counted in 2017, 31% of items showed a difference between the record and the shelf, and the record was more often too high than too low in every store⁴, so a line the system shows in stock can be empty on the shelf. The out-of-stock guide shows how to find those lines from sales that stop, and the safety stock guide how to size the buffer that prevents them.

Dead stock share is the value at cost of the lines with no sale in a set number of days, divided by the value of all stock. The mistakes are one threshold for every category (a quarter without a sale means something different for batteries and for garden furniture) and a chain-level view: a line dead in one store can be selling in another, and the fix is a transfer before a markdown. The dead stock guide sets out the rules by store.

Shrinkage is the stock that is recorded but not there, usually (recorded stock − counted stock) ÷ sales over the period, at cost, split into known loss such as damage and write-offs and unknown loss such as theft and errors. The mistake is treating it as visible only at the count: between counts the record drifts, and the difference already distorts availability, turnover and orders. National figures do not replace a chain's own count either. Police in England and Wales recorded 519,516 shoplifting offences in the year to June 2026, 2% fewer than a year earlier, but the ONS notes that reporting rates for shoplifting are known to be very low⁵. Our guides to the stock take and to stock control cover the count and the routine around it.

4. How to read them together

Each KPI on its own can be improved by damaging another, so read them in pairs that check each other.

• Turnover and GMROI with availability: a rise that comes from empty shelves is a loss, not an efficiency.
• Sell-through with the full-price share of sales: a buy cleared by markdowns is not a buy that sold.
• Gross margin with shrinkage and markdowns: a margin held by not counting the losses will fall at the next count.
• Like-for-like sales with transactions and ticket: a rise from higher prices alone and a rise from more customers call for different responses.

A workable rhythm for a chain: sales, transactions and ticket by store every day or week; gross margin and availability weekly; stock days, dead stock share and sell-through weekly by category and store; GMROI monthly; shrinkage at every count. In each case the chain total is the summary and the store list is where the decisions are, so the first question of any KPI review is which stores moved it.