Retail loss prevention software brings together the events where stock or money leaves the business without a matching sale: count variance, damage and write-off, voids and refunds at the counter, manual discounts, and stock that went missing between locations. It records each with enough context to be examined and reports the patterns across them. With Pentoggle, a retailer can describe where their losses actually arise and generate the starting application around it.
This page is about seeing the pattern. The individual controls live on their own pages: Retail POS Software for voids and refunds at the counter, Retail Inventory Audit Software for count variance and its reasons, Retail Goods Receiving Software for short deliveries, Retail Stock Transfer Software for stock lost between locations.
This page describes an operational and reporting workflow. It takes no position on what monitoring, investigation or action is permitted where you operate, and it names no rules. Employment and privacy requirements differ by place and change; establish what applies to your business with a suitable adviser before using any of this in relation to an individual.
Most retailers already run an accounting system such as QuickBooks, Tally or Xero, which will eventually show the loss as a stock adjustment or a reduced margin. That stays where it is. What is often still managed outside it is the detail: which shift, which department, which transaction type, and whether the same thing happened last week.
Key takeaways
- Retail loss has several sources and they are frequently confused. Process error, damage, supplier shortage, administrative mistakes and theft all show up as the same missing stock at the end of a count.
- Process failures are usually the larger and more fixable share in stores that have not measured them, and looking there first is both more productive and fairer than assuming theft.
- The counter controls, voids, refunds without an original sale, and manual discounts, are the transactions worth recording with a name and a reason, because they are where money can leave without a sale.
- Patterns are evidence to look at, not conclusions about people. A figure can reflect the shift, the department, the customer mix or who was given the difficult work.
- A useful number is shrinkage as a share of sales, read alongside the reason breakdown behind it.
The spreadsheet is often not the problem
An owner who knows the store, watches the counter and notices when something is off is a real control, and in a small store with the owner present it catches a great deal.
The trouble starts at identifiable points.
When the loss is one number a year
The count produces a shortfall. It is written off as shrinkage. Whether it came from the receiving door, the counter, the stockroom or the shelf is not known, so nothing specific changes.
When each control is in a different place
Voids are in the POS log. Damages are on a clipboard. Count variance is in a spreadsheet. Transfer discrepancies are in a chat thread. Each is individually small and no one sees them together.
When exceptions leave no trace
A refund without an original sale, a discount typed in by hand, a void after payment. Each is a legitimate action in some circumstances, and if none of them is recorded with a name and a reason, the legitimate and the irregular look identical.
When suspicion arrives before evidence
Something is clearly wrong and the response is a view about a person. Without records, that view cannot be tested, and both a wrong accusation and an undetected problem are possible outcomes.
What retail loss prevention software holds
Stock variance
Count results with variance by item, location and reason, drawn from the audit process rather than entered separately.
Damage and write-off
What was damaged or written off, where, when, by whom, with a reason, a value and a photo where useful.
Counter exceptions
Voids before and after payment, refunds with and without an original sale, price overrides and manual discounts, each with the cashier, the shift, the reason and any approver.
Cash variance
Expected against counted cash by shift and by cashier, from the daily close.
Receiving discrepancies
Short and damaged deliveries recorded at the door, with whether they were claimed and credited.
Transfer discrepancies
Differences between what was dispatched and what was received between locations, by route.
Access and approvals
Who is permitted to void, refund, discount or adjust, up to what value, and the record of each approval given.
Pattern reporting
The above brought together by store, department, shift, person, transaction type and time, with trends rather than single events.
Loss has several sources and they look alike
At the end of a count, the store is short. That shortfall can come from at least six places, and treating it as one thing produces the wrong response.
Administrative error is stock that was never really there or was recorded twice: a receiving entry at the wrong quantity, a sale rung against the wrong variant, a transfer received but not booked, a return refunded and never put back into stock. This is often a large share in stores that have not measured it, and it is entirely fixable with process changes.
Supplier shortage is stock that was paid for and never arrived, which shows up as missing at the count if the delivery was not counted at the door.
Damage and spoilage is stock that existed and became unsellable, and it is only distinguishable from other losses if someone records it when it happens.
Process failure in transfers is stock that left one location and did not arrive at another, with the cause anywhere along the route.
Customer theft is stock removed from the shop floor.
Internal theft is stock or money removed by someone working in the store.
The order in which a store investigates matters. Starting with the process explanations is usually more productive, because they account for a substantial part of the shortfall in stores that have not previously separated them, and because the fixes are cheap and permanent. It is also fairer: assuming theft first and being wrong damages a relationship with someone who did nothing, and in a business that depends on staff who handle stock and cash all day, that cost is real.
The counter exceptions are worth recording
Three transaction types at the counter allow money to leave without a corresponding sale, and each is entirely legitimate in ordinary use.
A void cancels a sale. Voids before payment are routine. A void after the customer has paid in cash is the one worth attention, because the cash is then unaccounted for by the system.
A refund without an original sale returns money for an item the system has no record of selling. Receipts get lost and stores make judgement calls, so this happens honestly and often.
A manual discount or price override lowers the price outside the promotion rules. Again, ordinary: a damaged item, a price match, a long-standing customer.
The control is not to prevent these but to record them: who did it, on which shift, for what reason, and who approved it where a threshold applies. Once recorded, the ordinary cases become invisible in the reporting and the unusual patterns stand out, without anyone needing to watch the counter.
What makes this work is that the reason takes one tap and the approval takes a manager's code. A control that slows the counter will be worked around, and a workaround leaves no record at all, which is worse than the situation it replaced.
Patterns are where to look, not what happened
The reporting on this page produces patterns: a cashier whose void rate is well above their colleagues', a department whose count variance is consistently high, a shift where cash is regularly short, a transfer route that loses stock.
Each of those is a reason to look. None of them is a finding.
A high void rate can mean the cashier works the busiest lane, handles the returns counter, is new and correcting their own scanning errors, or is doing something wrong. A department with high variance can have a receiving problem, a barcode problem, a genuinely difficult product to count, or a theft problem. A shift with short cash can have a counting habit, a float practice, a busy period with no relief, or something else.
The useful sequence is to treat the pattern as a question, check the process explanations first, look at whether the pattern follows the person or the place, and only then consider what else it might be. When a pattern does follow a person across locations, shifts and departments, that is a stronger signal than any single figure, and it is still a reason for a careful conversation rather than a conclusion.
Two practical cautions. Data of this kind should be used consistently rather than selectively, because applying it to one person and not another turns a control into a grievance. And what you may do with it, in terms of monitoring, investigation and any action affecting someone's employment, differs by place and changes over time. Establish that with a suitable adviser before you rely on it.
There is also a design point worth stating. A system that only ever surfaces negatives will be experienced as surveillance, and staff will work around it. Recording the ordinary alongside the exceptional, and being open with the team about what is recorded and why, tends to produce better data and better relationships than a system nobody was told about.
Where loss prevention looks different by business type
- Supermarket Software, where losses are spread across many small items, fresh waste is a category of its own and the count is continuous.
- Convenience Store Software, where shifts run round the clock, cash handling is frequent and one person is often alone in the store.
- Liquor Store Software, where high-value, easily concealed items justify tighter counting of specific shelves.
- Electronics Retail Software, where units are individually valuable and serial numbers make a specific unit traceable.
- Apparel Retail Software, where fitting rooms, high return volumes and size variants create both real losses and administrative ones.
- Multi-Store Retail Software, where comparing the same figures across stores is the fastest way to see which location is different.
Why retailers choose Pentoggle for loss prevention
One view across every leak
Count variance, damages, counter exceptions, cash variance, receiving shortages and transfer differences in one place.
Reasons recorded at the point of the event
Each exception carries a person, a shift and a reason, so ordinary cases are visible as ordinary.
Process causes separated from the rest
The reason breakdown makes administrative error, damage and supplier shortage distinguishable from what remains.
Patterns over time, not single events
Trends by store, department, shift and transaction type, so a look is prompted by a pattern rather than by an impression.
Sits around your accounting and your POS
QuickBooks, Tally, Xero and comparable systems continue recording adjustments and write-offs. Your POS continues processing transactions. Pentoggle adds the context that explains them.
A useful number for loss prevention
Shrinkage as a share of sales, read alongside the reason breakdown behind it.
The share matters more than the amount, because it can be compared across stores and over time as the business grows.
The breakdown is what makes it actionable. Two stores with the same shrinkage figure, one mostly administrative error and one mostly unexplained, have different problems and need different work. A store whose figure is falling because administrative errors are being fixed is making real progress even if the total is still high.
Read it by store and by department. Unexplained variance concentrated in one department is a place to look, and it is worth exhausting the process explanations for that department before drawing any other conclusion.
Ready to build retail loss prevention software?
You know what your annual count wrote off.
You may not be able to say how much of it was theft, how much was a receiving error nobody caught, and how much was a return that never went back on the shelf.
Describe where you think stock and cash go missing to Pentoggle in plain English and generate a working first version in hours, then refine it around your process.