What the terms used in grocery chain operations meetings actually mean, how they are calculated and when they mislead. Every entry opens with a one-sentence definition.
What is on hand, how fast it turns and how much of it is capital sitting still.
Inventory turnover
Stock turn · Stock rotation
Inventory turnover is the ratio showing how many times your average stock is sold and replaced within a period.
Inventory turnover = Cost of goods sold / Average stock at cost
Four turns a year means average stock was depleted and replaced four times over the year; the same thing can be read as stock sitting on the shelf for about 90 days. The equivalent in days is 365 divided by the turnover figure.
Fix the measurement scope first. The same product gives different turnover figures at store level and at chain level, and whether warehouse stock is included changes the result again. So when turnover is reported, which locations and which stock types were counted must be written down.
On its own it misleads. The easiest way to raise turnover is to cut stock, and that produces lost sales. It has to be read alongside margin; the measure that combines the two into one number is GMROI.
GMROI is the ratio showing how much gross profit each unit of money tied up in stock produces.
GMROI = Gross profit / Average stock at cost
A GMROI of 2.4 means every lira tied up in stock produced 2.4 lira of gross profit over the year. The value of the measure is that it folds turnover and margin into one number: a fast-turning, low-margin product becomes comparable with a slow-turning, high-margin one on the same scale.
It is meaningful at category and product level, not at chain total. A GMROI calculated across the whole business lets the categories that work hide the ones that do not.
Watch which stock value sits in the denominator. If average stock at cost is taken from just two points — period start and period end — it will not reflect reality in a seasonal category such as ice cream; an average of monthly closing values gives a truer figure.
Dead stock is stock that has not sold at all over a defined period, or whose sales rate is too slow to clear the quantity on hand.
The threshold is set per aisle: in fast-moving food, 90 days without movement counts as dead stock, while in non-food aisles such as kitchenware the same period may be entirely normal. What matters is that the threshold is written down and that every report uses the same one.
Dead stock is not only a loss of value; it is also a cost of space and capital. Every unit on the shelf holds the place of another product that could have turned. That is why a dead stock report is usually read together with the amount of capital it ties up.
Safety stock
Buffer stock
Safety stock is the buffer quantity held against the possibility that demand or lead time deviates from what was expected.
The amount depends on two uncertainties: how volatile demand is, and how late the supplier runs. If both fluctuate, the buffer grows. Setting it as a fixed number of days is common, but the correct approach is to measure past deviation and calculate against a target service level.
Safety stock is a cost line. Raising the service level from 95 to 99 per cent more than doubles the buffer under most distributions. The decision is made by comparing the cost of the lost sale with the cost of the tied-up capital.
Fill rate
Service level · Order fill
Fill rate is the ratio showing how much of incoming demand can be met from stock without delay.
There are two different definitions and they get confused. The order-based measure asks whether an order was met in full; the line-based measure asks how much of the quantity requested could be supplied. For the same period the two come out noticeably different.
When unfulfilled demand is not recorded, fill rate looks higher than it is. A customer generates no demand record for a product that is not on the shelf; the system only sees the sale that happened, and the loss stays invisible.
Shrinkage
Waste · Wastage · Known loss
Shrinkage is the quantity of goods written off stock without being sold because of spoilage, breakage, expiry or loss during handling.
In retail, shrinkage is usually discussed together with count variance, but they are different things. Shrinkage is a known loss with a recorded cause; count variance is a deviation whose cause is unknown. Adding the two into one figure makes theft and spoilage impossible to tell apart.
At the butchery and deli counter, shrinkage is part of the process: how many kilos remain after a kilo of meat is trimmed and portioned is defined in advance. The gap between defined shrinkage and actual shrinkage is the earliest indicator of a process problem.
Expiry date
Use-by date · Best-before date
The expiry date is the last date a product can safely be consumed, and in stock tracking it is a property of the batch, not of the product.
Expiry tracking only works when stock is held at batch (lot) level. A system that keeps a single quantity at product-code level cannot tell you which 40 of the 200 units on hand expire this week.
The correct dispatch order is built on expiry: the batch closest to expiring leaves first. A system that ships in order of receipt leaves a later-arriving but shorter-lived batch at the back, and produces waste.
Cycle count
Perpetual inventory count · Partial count
A cycle count is the practice of counting defined groups in rotation and continuously, instead of counting the whole inventory in one exercise.
A full count done once a year only reveals the error twelve months later; at that point it is impossible to find where the deviation came from. In cycle counting, fast-moving products are counted more often and slow movers less often, and the deviation is caught within weeks.
For counting not to stop operations, it has to work on mobile and offline. In parts of the warehouse where the connection drops, holding records on the device and transferring them when it returns is what keeps the count running.
What the product actually cost you, and which figure belongs in which decision.
Weighted average cost
Average cost · WAC
Weighted average cost is the unit cost found by dividing the total value of all batches on hand, bought at different prices, by the total number of units.
Weighted average = Total batch value / Total units
It is the figure that describes the past. Stock valuation, cost of goods sold and closing inventory value are all built on it; this is the cost accounting uses.
It is not the whole of the invoice unit price. Freight, customs, insurance, price-difference invoices that arrive later and currency movements are all absorbed into the average. Because these items arrive after the goods have entered the warehouse, how the costing routine handles retroactive correction determines whether the report stays consistent.
Last purchase cost is the figure showing the unit price at which the product was most recently bought.
It is the figure that prices the future. When you print a new ticket, the fact that the old batch in your hand was cheap is irrelevant; what matters is what it costs to resupply the product today. Pricing and purchasing decisions are made with this number.
The gap between it and average cost is not an error, it is time. If the gap can be explained, the system is healthy. In periods of sharp currency movement, designing a discount against average cost turns yesterday's cheapness into today's loss.
COGS is the total cost to the business of the products sold in a period.
COGS = Opening stock + Purchases in period − Closing stock
It is the other side of gross profit: sales revenue minus COGS gives gross profit. It is also the figure that sits in the numerator of inventory turnover.
Which costing method produced it changes the result directly. In a period of rising prices, COGS calculated on weighted average comes out lower than COGS calculated on last purchase cost, and profit looks higher than it is.
Gross margin
Gross profit margin
Gross margin is the share of revenue left after product cost has been deducted from sales revenue.
Gross margin = (Sales − COGS) / Sales
The same product can show two different margin figures on the same day, and both can be right: one was calculated on weighted average cost, the other on last purchase cost. The gap can reach several points. That is why every margin screen should state which cost definition it used.
In any comparison made on margin, whether discounts, returned goods and sales rebates were included is decisive. Two reports built from different line items cannot have their margins compared.
Recipe cost
Bill of materials cost · BOM cost
Recipe cost is the unit cost that comes out of adding up the raw materials and processing steps that make a product, at defined quantities.
In aisles that produce their own goods, such as the in-store bakery and deli, stock is depleted through the recipe at the moment of sale: how much of each raw material comes off when a tray of pastries is sold is written in the recipe. If the recipe is out of date, both cost and stock are wrong.
Adding shrinkage, labour and allocated overhead on top of the raw material value gives the full cost. Profitability calculated on raw material value alone appears markedly above reality in any aisle that produces its own goods.
Landed cost is the allocation of purchase expenses outside the product price — freight, customs, insurance — into the unit cost of the product.
The choice of allocation key changes the result: is the expense distributed by unit, by value or by weight? If a heavy, cheap product and a light, expensive one travel in the same shipment, the choice of key moves their costs in opposite directions.
These expenses are usually invoiced after the goods have entered the warehouse. The correction has to be written as a separate movement type, with both the period it belongs to and the date it was processed recorded; otherwise last month's margin changes overnight and nobody can say why.
Planning and allocation
Which product goes where, in what quantity and when.
Allocation
Distribution
Allocation is the process of deciding how stock on hand, or stock about to arrive, is divided between stores.
Equal distribution is the most common and the most expensive method: send every store the same quantity and stock piles up in some while sales are lost in others. Correct allocation reads the store's past sales velocity, its current stock, its shelf capacity and regional differences in demand together.
The allocation decision and the transfer decision feed each other. A wrong initial allocation is corrected later by store-to-store transfer, and that produces a logistics cost every time.
Initial allocation is the first shipment sent to stores before a product has built up any sales history.
It is the hardest decision because there is no data to lean on. It is built from the past performance of comparable products, the store's share of the category and its shelf capacity.
The quality of the initial allocation shapes the result of a new product or a promotional campaign. Stock that starts in the wrong branch is dragged around for the whole campaign with transfers and discounts; both corrections eat margin.
Demand forecasting
Demand planning
Demand forecasting is the work of predicting future sales quantity from past sales and the variables that affect them.
When only past sales are used as input, the forecast is systematically wrong: sales recorded on days when stock ran out sit below true demand. A correct model flags days with a stock-out and adjusts demand for those days.
The calendar effect matters especially in Turkey. Religious holidays shift about eleven days each year, so comparing against the same calendar week of the previous year gives the wrong answer; the comparison has to be built on distance from the holiday, not on the calendar week.
A store-to-store transfer moves a product from a branch holding surplus stock to a branch that is short of it.
The benefit of a transfer is measured by comparing the probability that the goods will sell at the destination against the logistics cost. Transfers made for a product nearing its expiry date, or one whose campaign has ended, usually do not cover their own cost.
A single cost definition must be used in valuation. If transferred goods are recorded at one cost on the way out and another on the way in, an unexplained leak appears in stock value.
Capacity management
Capacity management is the planning of an available resource — space, shelf, staff, time — against demand, and the monitoring of how full it runs.
In a grocery chain, shelf metres, chiller and warehouse space, the number of tills and staff shifts make up the capacity. The question is the same for each: how full is the resource, and what is the empty part costing?
The capacity decision is made together with the demand forecast. A capacity plan made without seeing demand produces either idle resource or unmet demand, and the cost of both can be measured.
Where the number is calculated, where it is stored and why it has to come out the same on every screen.
KPI
Key performance indicator
A KPI is a measure that shows, in a single number, how a process is performing against its target.
For a KPI to be useful, three things have to be written down: how it is calculated, which data feeds it, and who owns it. If any of the three is missing, the number gets reported but produces no decision.
The same KPI coming out differently on different screens is usually not a formula error but a definition difference: one screen deducted returns, the other did not. The fix is to hold the definition in one dictionary and have every screen read it from there.
ERP is enterprise software in which purchasing, inventory, production, sales and finance run on a single system and shared data.
The ERP records the work; business intelligence analyses what was recorded. The two are often confused: an ERP has reporting screens, but they run on transactional data and are not designed for cross-cutting analysis.
What decides an ERP selection is not the number of modules but whether processes can be configured to the business's own flow. Conforming to the template a packaged product imposes is the real cost item in most projects.
Would you like to see what a term means in your own data?
The definitions are the same in every business; the numbers are not. The first call is about your data structure.
