Restaurant stock control in South Africa: catching shrinkage before it eats your margin

Shrinkage is the polite word. In a South African kitchen it covers everything from a barman being generous with doubles, to portion creep on the grill, to stock walking out the back door during a load-shedding blackout. The problem is that all three look identical on a profit and loss statement: your food cost went up and nobody can say exactly why.

A weekly stocktake is what separates those causes from each other. Not a perfect stocktake — a consistent one.

Why weekly, and why the same day

A monthly count tells you that something went wrong four weeks ago. By then the person responsible has forgotten, the supplier price has changed again, and you are reconstructing history rather than managing a kitchen. A weekly count gives you a number you can still act on.

Same day, every week, after the till period closes and after every delivery is entered. Consistency matters more than accuracy here: a line you always miss in the same place looks exactly like theft, and a count taken at a different point in the trading week is not comparable to the last one.

Before anyone counts anything

1. Close the till period. Counting while sales are still landing guarantees a wrong figure.
2. Enter every delivery that has physically arrived — and nothing that has not. An unentered delivery reads as if you lost the stock.
3. Record transfers between the bar and the kitchen, or between sites. Untracked transfers are the single most common cause of a variance that turns out to be nothing.
4. Print the count sheet in the order you will walk, not alphabetically. Alphabetical sheets make people criss-cross the building and skip lines.

The counting route

Walk the same route every week: dry store first, then the walk-in, then the freezer, then the kitchen line, then the bar and cellar, left to right without doubling back. Two people — one counting aloud, one recording — is faster and materially more accurate than one person doing both, because the counter never sees the expected number and so never anchors to it.

The most expensive habit in stocktaking is guessing in halves. 'About half a bottle' is where consistent errors hide, because everybody rounds the same direction.

Reading the variance in rands, not units

Your report should show, per line, what you should have — opening stock plus deliveries minus what your recipes say you sold — against what you actually counted, with the difference converted to rands. Then sort by value.

This is the step most operators skip, and it is the one that changes decisions. Forty units short on cooldrink cordial matters far less than four bottles of premium gin. Sorted by units, the cordial looks like the crisis. Sorted by rands, you can see where the money actually went.

Telling theft from over-portioning

The pattern tells you more than the size. Look for direction over several weeks, not the number in any single one:

• Short one week, fine the next — almost always a counting error or an unrecorded transfer. Do nothing.
• Consistently short by a small percentage across many lines — over-portioning. Scoops, ladles and scales on the section fix it, and it is not a discipline problem: nobody can eyeball 80g accurately through a Friday service.
• One or two high-value lines short, consistently, while everything else is fine — worth a closer look. Spirits, prime cuts and cooking oil are the usual candidates.
• Everything short after a specific shift or a specific delivery day — a process problem at a known point, which is the easiest kind to fix.

Theft is real and it is also the least common of these. Most variance in most kitchens is portioning, wastage that nobody logged, and recipes that no longer match what the kitchen actually does. Start there before you start suspecting people.

Load-shedding is a stock control problem too

Extended outages mean fridge and freezer losses that never appear on a waste sheet because they happen overnight. If you are not logging spoilage after an outage, that loss shows up in your variance as unexplained shrinkage — and you end up looking for a thief when the real culprit was six hours without power. Log it as waste, with the reason, so the number tells the truth.

What software actually adds

All of the above works on paper, and for a small stable site a spreadsheet is genuinely enough. Software earns its place at the point where prices move faster than you can re-cost, where a second person needs to edit the same numbers, or where you open a second site.

South Africa has established local options — GAAP and Pilot have been serving the market for decades and integrate with tills in ways an overseas newcomer does not. If you need a till and stock control as one system from a supplier with local support and a long track record, they are the sensible starting point, and we would rather say so plainly.

Where FoxEra fits, and where it does not

FoxEra is free stock control and recipe costing: photograph a supplier invoice and David AI reads every line into your stock, dishes re-cost themselves when a supplier price moves, and waste and variance report in your own currency. There is no subscription and no contract.

The honest limits, stated plainly: we are a new company with no South African customers yet, so there are no local case studies to point at and we will not invent any. We do not do tills or bookings. And South Africa is not yet one of our configured billing countries — the free tier works and needs no card, but there is no rand payment path today. If that matters to you, one of the established local systems is the better answer right now.

The part that decides whether any of it works

Correct pack sizes, real yields after trimming, and recipes that match what the kitchen actually plates. Get that groundwork wrong and every system on the market — ours included — will give you precise, confident, wrong numbers. Budget a day or two for it before you blame the software.