A note up front: what follows is a composite, built from the patterns we see across stock counts at retail and distribution businesses, not one specific company's numbers. But every figure here is realistic, and if you have never run a full physical count against your records, the shape of it will likely look familiar.
The Setup
Picture a mid-sized retailer with a few hundred SKUs, a small team, and a stock spreadsheet that has been "mostly right" for years. Nobody has done a full count in over a year — it always felt like something that could wait, and the top-line numbers never seemed bad enough to force the question. Then, for one reason or another — an investor asking for clean numbers, a new hire, a slow quarter — the count finally happens.
How to Run a Physical Stock Count
For context on what follows, a basic physical count works like this: freeze all stock movement for the count window, count every SKU by hand against its recorded quantity, log every discrepancy rather than just correcting it silently, and only then update the books to match reality. Skipping the "log every discrepancy" step is the single most common shortcut businesses take — and it is the one that erases the chance to learn anything from the count beyond the final corrected number.
What a First Full Count Typically Turns Up
A shrinkage rate nobody had a number for. Across businesses like this, unexplained shrinkage of 3–8% of stock value is common on a first count — breakage, theft, expiry, and simple miscounts that were never logged individually but add up to a real figure once totaled.
Phantom stock. Items the sheet says exist that physically do not — sold, damaged, or moved without the record catching it. These are the SKUs a sales team confidently promises to customers, right up until someone goes to the shelf.
Ghost shortages. The reverse also happens: stock physically present that the sheet says is out, sitting unsold and unnoticed because nobody thought to look for something the system claimed did not exist.
A handful of SKUs responsible for most of the gap. The discrepancy is rarely spread evenly. Usually a small number of fast-moving or high-value items account for most of the value lost — the ones touched most often, by the most people, are the ones most likely to drift.
"The spreadsheet wasn't lying. It just stopped being told the truth months ago."
What It Actually Costs
Put a number on it and the abstract "our stock numbers are a bit off" starts to look different. On a business carrying, say, $150,000 in inventory value, even a conservative 4% unexplained shrinkage is $6,000 quietly gone — not stolen in one event, not a single mistake, but the accumulated cost of a system that could not keep up with a business moving every day.
Reading the Results: What to Do With the Gap You Find
Once a count is complete, the discrepancy list itself is the useful part — not just the corrected total. Group the differences by SKU and look for concentration: if three or four items account for most of the shrinkage value, that points to something specific — a handling issue, a theft pattern, or a supplier discrepancy — worth investigating directly rather than treated as generic loss. If the gap is spread evenly and thinly across hundreds of SKUs instead, that points to a process problem — the recording habit itself, not any one product.
Why the Count Doesn't Fix the Problem
This is the part that surprises owners most: doing the count does not solve anything on its own. It produces an accurate number for one moment in time. The day after the count, the drift starts again — the same untracked shrinkage, the same manual updates falling behind — because nothing about the process that caused the gap has changed. A count is a diagnosis, not a cure.
What Actually Closes the Gap
The fix is not counting more often. It is removing the need to reconcile in the first place — inventory that updates itself with every sale and purchase, so the recorded number and the shelf number never get the chance to drift apart. This is what Axis does: stock moves automatically with every transaction, so a full physical count becomes a confirmation, not a correction.
Frequently Asked Questions
What is a normal inventory shrinkage rate? For businesses relying on manual tracking, an unexplained shrinkage rate of 3–8% of stock value on a first full count is common. Rates vary by industry and product type, but shrinkage above that range, or a rate that keeps climbing count over count, usually points to a process gap rather than bad luck.
What's the difference between shrinkage and phantom stock? Shrinkage is stock that is genuinely gone — sold, broken, stolen, or expired — but never recorded as such. Phantom stock is a subset of the same problem: it is what the records still claim exists even though it physically does not, which is what causes staff to promise stock to customers that isn't there.
How long should a physical stock count take? It depends heavily on SKU count and team size, but a few hundred SKUs typically takes a small team a full day when done properly, including logging discrepancies rather than just correcting totals. If counts are taking noticeably longer each time, that is itself a sign the underlying gap is growing.
The Bottom Line
If you have never run a full count against your records, the numbers above are a reasonable preview of what you will find. The real question is not whether the gap exists — it almost certainly does. It is whether you find out on your own terms, or during a moment when you can least afford the surprise.
Related Reading
Ready to optimize your systems?
Our engineers are ready to discuss your architecture and how we can help you build institutional-grade infrastructure.