What shrinkage actually covers.
Shrinkage is the variance between book inventory and physical inventory at any point in time, expressed as a percentage of sales over the period being measured. A retailer with EUR 4 million in annual sales and a EUR 80,000 negative variance at year-end has 2% shrinkage. The number is industry-typical — apparel runs 1.5-2.5%, electronics 1-2%, pharmacy 0.5-1.5%, supermarket 1-2%.
Shrinkage has four sources, in rough order of frequency for SMB retail: internal theft (staff, including back-of-house), process and paperwork errors, external theft (shoplifters, organized retail crime), and damage or expiry not booked properly. Most owners assume it is mostly external theft. The data almost always shows otherwise.
The first move: cycle counts.
You cannot manage shrinkage you cannot measure. A working cycle count program (covered in detail in our cycle count post) is the foundation. Without it, you discover shrinkage once a year as a single big number with no useful detail. With it, you see shrinkage as a pattern: which SKUs, which branches, which days of the week.
Pattern is what makes shrinkage findable. If the same 30 SKUs go short every month at the same branch, it is not random shoplifting — it is targeted, internal or external, at predictable items. If shortages spread across 200 SKUs at one branch but not others, it is a branch-level process or staffing issue. Cycle count data is the input to every investigation that follows.
Camera coverage and the back door.
Most retailers have cameras at the front of house. Few have meaningful coverage of the back door. Receiving, return area, breakroom, manager office — these are where most internal shrink happens. A EUR 200 IP camera per zone, 30-day storage, and a written policy that staff are aware of it dramatically reduces back-door shrink.
When a cycle count finds a variance, the camera trail is the next stop. Pull footage of the back door for the period since the last clean count of that SKU. You are looking for asymmetric activity: items going out without paperwork, late-night entries by staff who should not be there, frequent solo trips to the stock room from particular cashiers. You are not building a court case; you are finding patterns to act on.
Process audits — paperwork without goods.
A surprising amount of "shrinkage" is process error. Goods received but never booked. Transfers shipped but never received. Returns processed twice. Damaged goods written off as missing instead of as damage. Staff buys booked at full price, not employee discount. Each of these creates a phantom variance with no actual missing goods.
The audit: take a sample of large variances and trace each one through the document trail. Did the GRN match the supplier invoice? Did the transfer dispatch match the destination receipt? Did the customer return get processed once or twice? About 30-40% of "shrinkage" variances disappear under document review. That is not theft — that is process the system never closed.
- GRN-to-supplier-invoice match: any unbooked receipts?
- Transfer dispatch-to-receipt match: any unbooked receipts at destination?
- Returns: any duplicates or returns posted as new sales?
- Damage write-offs: are they posted to the correct account?
External theft — what actually works.
Shoplifting is real but smaller than most owners think for SMB retail. The high-loss categories are obvious: razor blades, batteries, branded perfume, mobile accessories, baby formula. Move these behind the counter, into locked cases, or near a staffed checkout. The friction of having to ask reduces casual theft enormously.
For higher-end items (electronics, branded apparel), security tags and exit alarms still work. For grocery and FMCG, mirror placement and aisle layout matter — long sight lines from the cashier to every aisle reduces opportunity. None of this needs to be expensive. A few thousand euros of well-placed cameras and locked cases beats a vague "security upgrade" twice the price with no specifics.
A worked investigation.
A clothing chain in Amsterdam had 2.8% shrinkage at one branch versus 1.2% chain-wide. Cycle counts showed the variance was concentrated in 60 SKUs out of 4,200 — branded denim and a specific shoe line. Footage review found two staff making frequent breakroom trips with bags. Document review found three transfers from the warehouse marked "in transit" that had never been received and were 6+ months old.
Outcome: the two staff were terminated after a documented investigation. The three phantom transfers were the warehouse forgetting to clear them; the goods had actually arrived but the receipt was missed for some, and others had been short-shipped without a variance booked. After both fixes, branch shrinkage dropped to 1.0% within 6 months. Half the original "shrinkage" was theft, half was process. This is typical.
How nonari helps surface shrinkage early.
Nonari produces a shrinkage report per SKU per branch per period, sourced from cycle count variances and inventory adjustments with reason codes. The report ranks branches by variance rate, SKUs by absolute shrinkage value, and time periods by trend. The patterns that show up — the same SKU shorting at the same branch on weekend evenings — are the leads for investigation.
Nonari also enforces document-level discipline that prevents the most common process-shrinkage causes: GRNs must match supplier invoices, transfer receipts must reconcile to dispatches, returns reference original invoices, and overrides on negative inventory are all logged with reasons. This kills the paperwork half of shrinkage at source so the remaining variance is actually worth investigating.