This page is not about accusing anyone. In a small shop, money and goods leak in several different ways, and not all of them are theft. Below: when the numbers really do signal something, what to do in the first week, and which remedies work with no software at all. Most of them are about routine, not about an app.
If you have no time to read, the whole page boils down to these five lines.
When stock does not add up, the first thing that comes to mind is usually the rarest cause. The five below are far more common, and checking them takes about an hour.
If the gap survives all five checks, then it is time to look at the numbers. How the count itself is done, and how to read and correct a difference, is covered in the stocktaking guide; this page is only about repetition.
Theft never shows up labelled “theft”. It appears in one of five familiar shapes — and each of them can have an innocent cause too. So what follows is not proof, but where to look.
A gap between the counted cash and the expected figure at shift close happens to everyone. The signal is when it always leans the same way — short — and always on the same person’s shift.
Random error spreads across the whole shelf. Theft sticks to the same goods: small, expensive, pocket-sized, uncoded or sold by weight.
The staff report shows who refunded how much, and the amount comes off that seller’s own sales. If one person’s returns stand out sharply, asking why is fair.
Shifts are optional: sales are still recorded without one, but the cash reckoning is not — so no difference appears either. Shifts suddenly going unopened is a signal in itself.
One stocktake’s difference says nothing. Line up three, though, and the direction shows: the gap is shrinking, holding steady, or growing each time.
Plenty of articles say “check the cancelled receipts”. BuySale has no such list, and no general change log of “who edited a price, who deleted an item” — you will not find that signal in our software.
One signal is a coincidence. Two signals repeating on the same person for three weeks is a reason to talk. There is a practical reason not to rush: a conversation backed by three weeks of records does not turn into an argument, while one backed by a single day’s hunch nearly always does.
This plan is not designed to catch anyone. Its purpose is that after one week you hold records instead of suspicions. All of it together takes half an hour a day.
Saying “checks start now” on day one stops everything, and you learn nothing. Export your stock to Excel and keep the file — that is the week’s starting point.
Cash in the drawer is written down at the start and counted at the end. This is not surveillance but also protection for the seller: when there is no difference, it is on record that there is nothing to answer for.
Choose the fifteen most valuable and smallest lines: cigarettes, energy drinks, batteries, phone accessories, a perfume bottle, a piece of jewellery — whatever fits your shop. You count only that list, at the same time every day. Five minutes, not a full stocktake.
For one week, receive and enter incoming goods yourself. Very often the “shortage” is born right here and has nothing to do with theft. It is also one of the strongest remedies: the person who receives goods is not the person who sells them.
Returns, price changes, haggling at the till, credit sales and seeing cost prices are five separate permissions. Withdraw the ones that are not needed: a revoked permission stops working on the very next tap.
Drop in mid-day, count the cash and compare it with the expected figure in the app. Not once — once or twice a week, at a different hour each time. The end-of-day count is predictable; a mid-day one is not.
A week of records gives you one of three answers: no gap; a gap whose cause is intake or write-offs; or a gap that falls on one person’s days. Even in the third case the conversation is a question, not an accusation.
None of these needs software, a camera or money. In most shops theft becomes possible precisely because these four things are missing.
If one person receives the goods, enters them and sells them, only that person knows the count. Take receiving on yourself or give it to someone else. This single change closes the biggest source of shortages.
At handover the cash is counted in front of both people and the figure written down. After that, “it was already like that when I came” disappears: every shift has a clear beginning and end.
Everyone knows about the one big annual count in advance. A daily count of fifteen items at varying times cannot be predicted or planned around — that is where its power lies.
Where price and discount live in the seller’s mouth, the books never reconcile: the amount taken from the customer and the amount recorded can differ, and there is no checking it afterwards. In BuySale the app takes the price from the product card itself. Haggling at the till — typing in the total agreed with the customer — is unlocked by a separate “haggling at the till” permission that is not part of the ordinary seller set: until you grant it, a seller cannot lower a price at the till. Even when it is granted, every discount stays recorded on the sale and profit is counted from the reduced price. A discount on the product itself depends on the “change price” permission; who gets either is your decision.
These get written about rarely because they are awkward. Which is exactly why they stay unfixed for years.
A nephew, a daughter-in-law or a neighbour working in the shop is the commonest blind spot: no shift, no permissions, no counting, because it feels awkward. A rule is only a rule when it applies to everyone — and the person made an exception ends up in an awkward position too.
An item the seller takes off the shelf meaning to pay later is the biggest source of argument. The rule is simple: what is taken is recorded there and then — as a cash sale or as a credit entry in that person’s name. What is not recorded turns into a shortage at month end that nobody can account for.
Broken, expired or opened-by-a-customer goods are a shop’s normal losses. Unrecorded, though, they disappear into the shortage and can no longer be told apart from theft. A plain notebook is enough: date, item, quantity, reason.
A camera in plain sight deters people and settles disputes. But it does not count the drawer, does not count stock, and nobody watches the footage — someone has to open it at some point. A camera does not replace the numbers; it comes after them.
Every remedy above can be done without software — just by hand. What software does is two things: it keeps the reckoning for you and narrows what a seller can do. Below is what exists in BuySale and bears directly on the plan above.
Better to know this list up front — otherwise you expect something the software does not do, and lose faith in it.
And one important boundary: deducting a shortage from wages, holding an employee liable or dismissing them is a matter of labour law. We do not answer that, and nothing on this page is legal advice. What the software gives you is a record: when, how much, and on whose shift the difference arose. Take the next step with a lawyer or an accountant.
This part usually gets skipped, and that is where the whole effort falls apart. The goal is not a confession — it is to stop the gap.
Not from one signal but from repetition. For a few weeks record three things: the difference at shift close, a daily count of fifteen chosen items, and the number of returns in the staff report. If the gap always leans short, always falls on the same person’s days and always on the same goods, that is where to look. Before that, check the intake records, the units, mixed-up near-identical items and unrecorded write-offs: in practice most shortages come from those four.
No. BuySale has no cancelled-receipt list and no general change log of “who edited a price, who deleted an item”. Every sale, credit entry and expense is stored with whoever entered it, and the staff section keeps the last 30 sign-ins, but there is as yet no screen showing every change. A mistaken return cannot be cancelled or deleted either.
That is a labour-law question and we do not answer it — nothing on this page is legal advice. Note that most answers the internet gives for this query are written for Russian law and do not apply directly to a shop in Uzbekistan. What the software gives you is a record: when, how much and on whose shift the difference arose, with the seller’s own note. Take the next step with a lawyer or an accountant.
A camera in plain sight deters people and helps settle a dispute, but it does not count the drawer or the stock. The real problem is different: someone has to watch the footage, and in practice almost nobody does — because you do not know which day and hour to open. That is exactly what the numbers tell you: which day and whose shift the gap fell on. So a camera does not replace the numbers; it comes after them.
Yes, and this is the commonest blind spot of all. If shifts, permissions and counts do not apply to a relative, then when a shortage appears there is nobody to check it against and the suspicion spreads across the whole shop. A rule is only a rule when it is the same for everyone. In practice it protects the relative too: when a gap does appear, it is on record that their shift was clean.
In practice it is five minutes, and it does not replace a full stocktake — you still need one. A short daily count has two advantages: a gap shows up the next day rather than a month later, so everyone still remembers the reason; and it cannot be predicted. Ten items instead of fifteen is fine — what matters is that the list is short and done every day. After a week most people fold it into tidying the shelf.
Not by itself, and we do not promise that. Software does two things: it keeps the reckoning (the shift difference, who sold and refunded what, the stocktake gap) and it narrows what is possible (cost price hidden, prices and discounts behind a permission, returns behind a permission). The rest is routine: who receives deliveries, whether shifts are handed over, whether counts happen. Without the routine, software is just a tidy report. BuySale can be tried free for 7 days, after which it costs from 6 630 UZS a day by shop type (from 198 900 UZS per 30 days).
To start today, one action is enough: count the cash this evening and write the figure down. The rest can be added next week.
A POS app on your phone: camera barcode scanning, cash, card and credit sales, 58/80 mm receipts and shift cash counts. Not a fiscal till. 7 days free.
Learn more →How to start with BuySale: sign in by phone number with a code from Telegram, set up your shop, add your first products and try 7 days free.
Learn more →BuySale pricing by shop type: from 6 630 UZS a day, 198 900 UZS per 30 days. 7 days free and no setup fee.
Learn more →Each seller’s net sales, receipts, returns and shift shortages, with a monthly target and bonus worked out for you. Sales history shows who sold what.
Learn more →BuySale assistant
Answers from what is published on this site
AI can make mistakes — check anything important via Telegram.
Writing an answer