Retail shrink: how to measure it in a supply store

Every hardware store, feed store, and lumberyard runs on two numbers. The first is the number in the system, the one that says there are 240 bags of layer pellets in the warehouse and 38 boxes of 1-5/8" deck screws on the shelf. The second is the number that is actually out there, which you only learn when somebody walks the aisle and counts. If you have been in this business more than a season or two, you already know those two numbers disagree, and they disagree in the same direction almost every time. The count never matches, and it is almost always short.

Most owners feel that gap as a low-grade annoyance rather than a measured loss. A box of blades here, a pallet of feed there, a receiving mistake nobody caught until the vendor's invoice was long paid. What almost nobody does, and what this article is about, is write down by how much. Retail shrink is simply the difference between what your books say you own and what you actually own, measured over a period of time. It is one of the few numbers in your store that tells you whether your paperwork, your receiving dock, and your register are all telling the same story, and it is worth an hour of arithmetic once you know how the arithmetic works.

What shrink is, and the number to aim for

Shrink is book inventory value minus physical inventory value for a given period. If your system says a department should hold $50,000 in product and your count finds $48,500, your shrink for that period is $1,500, whether the missing goods walked out the door, spoiled in the rain, or never arrived in the first place despite what the packing slip claimed. The word covers every cause, which is exactly why the measurement matters: you cannot fix a cause you have not separated from the pile.

To compare shrink across stores and across years, the trade divides that dollar figure by net sales for the same period, which gives you a shrink percentage. So what should yours be? The most useful benchmark for your kind of store comes from the NHPA 2025 Cost of Doing Business Study, which drew on 1,082 independent home improvement stores for fiscal year 2024. It put shrink at an average of 1.7% of sales, with a median of 1%, as reported in Hardware Retailing's highlights.

Pay attention to that median, because it is the more honest number of the two. An average gets dragged upward by the stores having a genuinely bad year, the ones with a receiving dock that counts the packing slip instead of the pallet, or a theft problem they have not found yet. The median tells you that half of the independent stores in the study came in at or under 1% of sales. That is the realistic target for a well-run independent: not zero, because zero shrink in a store selling cut chain, bulk seed, and bagged concrete is a fairy tale, but something near or under 1%.

You will see bigger figures quoted from the National Retail Federation, and they are worth knowing for context rather than as your yardstick. The NRF's 2023 National Retail Security Survey put average shrink at 1.6% of sales for fiscal 2022, up from 1.4% the year before, roughly $112.1 billion in losses across US retail. That survey covered all of retail, including apparel and grocery, where theft patterns look nothing like a farm and feed store. The NRF has since stopped publishing an annual shrink rate, and its 2026 report on theft and violence took a different shape entirely, reporting that retailers saw a 12.4% decrease in shoplifting incidents in 2025 versus 2024 while other fraud and scams rose. National numbers move around and change methods; the NHPA figures are the ones drawn from stores that look like yours.

The shrink formula, worked

Here is the whole calculation, done at a scale that will feel familiar. Say you count one department, maybe power tool accessories, over one period. Your books say the department should be holding $42,000 in product at cost. Your count finds $41,300. Over that same period, you sold $52,000 worth of product out of that department at cost, which is about $87,000 at your retail prices.

Two things to notice. First, shrink is stated either at cost or at retail, and you have to say which. The same missing goods are worth about $1,170 at your shelf prices but only cost you $700 to replace, so the dollar figure changes with the valuation even though the missing product is identical. Industry studies quote shrink at retail unless they say otherwise, which is worth remembering any time you compare your number to a published one.

Second, and this is where owners trip, the numerator and the denominator must use the same valuation. Divide your $700 of shrink at cost by your $87,000 of retail sales and you get 0.8%, a flattering fiction that will convince you nothing is wrong. Done consistently, cost against cost or retail against retail, the percentage comes out the same either way, so pick one and stay with it. Cost is cleaner for an independent because your counts and your purchase orders already live at cost. Retail matches the industry surveys if you want to benchmark.

One more practical point: a single full-store count once a year gives you exactly one shrink number a year, and by the time you have it, the cause is twelve months cold. Counting a category at a time on a regular schedule gives you a number per category per month, which is a number you can actually act on. More on that in a moment.

Where shrink hides in a supply store

Ask a general retailer where shrink comes from and they will say shoplifting. Ask someone who runs a hardware store or a lumberyard and you get a longer, more interesting list. Theft is on it, but so is the scale at the feed bin and the cost file nobody updated since the last price increase. Here are the seven places it hides, roughly in the order you should look.

External theft: the pocketable stuff

The classic shoplifting target in a supply store is small, high in value, and easy to pocket: drill bits, saw blades, batteries, hand tools, specialty fasteners, and pet medications in a farm store. These are also the items that move fast and get restocked often, which makes a slow leak hard to spot without regular counts of just those categories.

Internal theft and sweethearting

It happens, and pretending otherwise is naive. The common form at the register is sweethearting, a cashier passing merchandise to a friend at a discount or at no charge, or voiding a line item after the cash drawer closes. It shows up in your counts as an ordinary shortage, which is one more reason per-category numbers matter: a theft pattern concentrated in one department at one register looks different from a storewide paperwork problem.

Paperwork shrink

This is the quiet giant. Receiving counts the packing slip instead of the pallet, so the system adds forty cases when thirty-six arrived. A vendor short ships and nobody files the claim. A return goes back on the shelf under the wrong SKU. A transfer between your two locations leaves one store on paper and never arrives at the other on paper. None of this involves anyone leaving with anything. The product simply exists in your system and nowhere else.

Unit-of-measure shrink

Supply stores sell in units that would baffle a clothing shop, and every conversion is a chance to lose track. Bulk feed goes out by weight, chain, rope, wire, and lumber get cut to length, mulch moves by the yard. When the sale unit and the stocking unit differ, a wrong conversion factor or a scale that has drifted out of calibration bleeds inventory that no thief ever touched. A scale that reads light by a few percent is generous to every customer who crosses it, all day, every day. Setting those conversions up once, on the product, is the subject of our piece on units of measure.

Unrecorded store use

The mistinted gallon that cannot go back on the shelf. The demo unit on the counter. The part pulled for a repair, the cleaning supplies from your own aisle, the box of screws handed to a good customer as a thank-you. Every one of these is a legitimate cost of doing business, and every one is shrink if nobody rings it through to a store-use account.

Damage and spoilage

Feed past its date, plants that did not survive the week, the broken bags of concrete at the bottom of the pallet. This is shrink with a smell. LBM Advantage, which works with lumber and building materials dealers, lists damage in handling and storage alongside mis-pulls, short ships, receiving errors, and theft as the usual causes of yard shrink, and anyone who has swept up behind a forklift knows why.

Price-file shrink

The strangest entry on the list, because nothing is missing at all. If a vendor's cost went up and your file never updated, your book value is wrong before anyone counts a single item. The physical inventory might match the shelf perfectly and your shrink number still comes out wrong, which is why the measurement below starts with the file, not the floor.

How to measure it without closing the store

The annual full count has its place, mostly for taxes and for peace of mind, but as a management tool it is a thermometer you read once a year. Cycle counts are the working instrument: you count one category at a time on a schedule weighted by risk and value, so the fast-moving, pocketable, high-dollar categories get counted often and the slow ones get counted occasionally. The lumberyard consultants mentioned above recommend weekly cycle counts for high-risk categories, and that cadence is realistic when a count covers an aisle rather than the building. If you want the full method, our guide to cycle counting for retail stores walks through scheduling and thresholds, and it pairs well with a pass at finding and clearing dead stock, since stale inventory distorts the same numbers.

A few rules make the difference between a count that teaches you something and a count that just adjusts the file. Pick your valuation first, cost or retail, and write it on the top of the worksheet so nobody mixes the two. Freeze receiving and transfers for the area while you count, because a truck being unloaded mid-count will corrupt every variance it touches. Set a threshold, a percentage or a dollar figure, and recount anything over it before you accept the number; counters are human and pallets are ambiguous. Record the reason for every large variance before you adjust it, in plain words, while the person who found it still remembers what they saw.

That last rule deserves emphasis, because the moment you adjust on hand, the evidence disappears. The corrected system now agrees with the shelf, the variance is gone from every report, and next month you will have no way to know whether the shortage in fasteners was a receiving error or a pattern. Keep the variance report itself, dated, with the reasons written on it. That stack of reports is your real shrink record, and it is what lets you track shrink percentage per category over time instead of rediscovering the same leak every January. In a store carrying thousands of SKUs, the category view is the only view that tells you where to look.

Reading the number

Once you have a few months of per-category numbers, they start talking to you. A department that runs at or under 1% is telling you its paperwork is sound: receiving is counting the pallet rather than the slip, returns are landing on the right SKU, and the scale is honest. A storewide average near the NHPA median of 1% means your processes work and your losses look like everyone else's.

A department that sits at 3% month after month is telling you something else, and here is the part that surprises owners: it is almost never a thief. A thief moves around, takes what is valuable this month, and leaves gaps that wander. A process problem sits in one place and produces the same shortage on the same items like clockwork. The feed room that is always short by the same percentage has a scale problem or a conversion problem. The paint department that is always short has mistints going into the dumpster without a store-use entry, and a feed store running bulk product by weight alongside deliveries has even more conversions to audit. Read the pattern before you read the people. Fix the process first, and if the number is still 3% after the process is fixed, then you start asking harder questions.

And keep the whole thing in proportion. Shrink at the NHPA average of 1.7% of sales is real money, but it is not chicken feed in either sense: it is a manageable cost with known causes, most of them fixable with a clipboard, a calibrated scale, and an honest receiving count.

Getting a shrink number out of Rundoo

If your store runs on Rundoo, everything above maps onto tools you already have, with a couple of traps worth knowing about. Counts live in the POS under Counts, and they come in two scopes: a Full count of the whole store, and a Cycle count you can scope by vendor, category, product line, or tag. Each count opens with "to count" and "counted" tabs so you can see what is left while you work. On the Staff app, counting runs by camera scan on iOS and Android, several counters can work the same count at once, and counts run blind, meaning the person scanning never sees the expected quantity. That last detail matters more than it sounds: a counter who can see the expected number tends to find the expected number.

The count table shows Expected, Counted, Variance, Cost, and Variance ($) for every product. When you complete the count, Rundoo adjusts on hand by the variance and posts a journal entry for the dollar difference to an inventory-variance expense account. That posted amount is your shrink number for the counted scope, at cost, ready to drop into the formula from earlier.

Now the trap: the Variance report can be exported to CSV or Excel, but only before you click Complete. After completion the export is no longer offered, although the Completed tab still shows the table. So the order of operations is export first, complete second, which conveniently matches the rule about keeping your variance reports with the reasons written down.

Two plain limitations. Rundoo has no dedicated shrink transaction and no shrink reason codes, and a manual edit to on hand does not create a shrink entry, so measured shrink comes from counts, not edits. The reasons stay in your notes rather than in the system. Stores typically record mistints and demo use through a store-use account rather than letting those items fall into the variance. For the paperwork side of the list, three related reports do real work: Negative inventory, which catches sales that drove on hand below zero, Write-offs, and Transfer Discrepancies, which is where the transfer that left one store and never arrived on paper finally surfaces. Rundoo runs in 500+ independent stores across the US and Canada, and these count and variance tools are the same ones those stores use to keep their books honest.