Where hotel bar stock actually disappears
Without a ledger, shrinkage and consumption look identical. A count that disagrees with what should be on the shelf is the only thing that tells the two apart.
A bottle that got poured into a paying guest's drink and a bottle that walked out the back door look identical on a monthly count — both are simply gone. That is the actual problem with bar stock: without a record of every movement in between, shrinkage and consumption are indistinguishable, which means shrinkage is invisible by default, not because it isn't happening.
The places it actually happens
A free round poured for a friend and never rung up. A heavier pour than the recipe calls for, repeated often enough to matter. A bottle broken and never logged as waste. A delivery counted in at the wrong quantity and never corrected. None of these are dramatic on their own — the damage is that each one is invisible individually, and invisible things compound over a season.
Why a monthly count alone cannot catch it
A count tells you what is left. It does not tell you what should be left, which means a count on its own has nothing to compare against — a shelf that is short is just a shelf that is short, with no way to say whether that is normal consumption or something else.
What should be left is a fact only a ledger can produce: purchases in, sales out at the recipe's stated quantity, waste logged as waste, adjustments logged as adjustments. The count then either matches the ledger's expected balance or it does not — and a mismatch is the first honest signal that something needs investigating.
Movements, not a single number
The useful record is not one on-hand figure that gets overwritten every time someone updates it — it is every change kept as its own row: a purchase, a sale-driven usage, a waste entry, an adjustment, a physical count. Nothing edited, nothing deleted; a correction is a new entry, so a discrepancy carries a date and, where relevant, a person, rather than being silently absorbed into whatever number happened to get typed in last.
What logging waste actually looks like day to day
A bottle dropped and broken, a keg that turned before it was finished, a batch of garnish that spoiled before service — each one is a thirty-second entry: item, quantity, reason. It feels like extra admin on a busy night, and it is exactly what makes the count at the end of the month mean something. A ledger with no waste entries at all is not a bar with no waste; it is a bar where waste is invisible instead of accounted for, which is a worse position than knowing the real number.
What a real discrepancy looks like
A bar that should have 40 bottles of a house spirit on the ledger — purchases minus recipe-driven usage minus logged waste — but counts 33 has a seven-bottle gap that is either a counting error, an unlogged waste event, or genuine shrinkage. All three are worth knowing which one it is, and none of them are visible without the expected number to compare against.
A month, reconciled
Opening count 40 bottles, 60 purchased through the month, 71 sold through recipe-linked orders, 4 logged as waste. Expected closing count: 40 + 60 − 71 − 4 = 25. An actual count of 25 means the ledger and reality agree — nothing to investigate. A count of 19 means a six-bottle gap that a waste-only explanation cannot cover, which is exactly the kind of number that turns a vague feeling that 'stock runs short some months' into a specific, dated question worth asking.
Why this catches heavy pours too, not just theft
Shrinkage is not only about stock walking out the door. A bartender consistently pouring a slightly heavier measure than the recipe calls for produces the exact same signature as theft — recipe-expected usage does not match what actually left the shelf — without anyone doing anything dishonest. A ledger does not distinguish motive; it distinguishes expected from actual, which is precisely the distinction that matters for fixing the problem, whatever its cause turns out to be.
What this does not fix by itself
A ledger does not stop shrinkage — it makes it visible, checkable, and attributable to a period rather than a vague annual sense that things run short. Fixing it once it is visible is a staffing and trust question the record cannot answer for you; what it can do is turn a suspicion nobody could ever prove into a specific number for a specific month.
How often to count
A full physical count every week is more effort than most small bars can sustain, and a count once a year is too infrequent to catch anything before it has compounded into a large gap. Monthly is the realistic middle ground for most properties — frequent enough that a discrepancy is still traceable to a recent, rememberable period, infrequent enough to actually happen consistently rather than being skipped when the bar is busy.
The habit that makes the ledger honest
None of this works if waste and adjustments only get logged when someone remembers, days later, working from memory. The habit that actually holds is logging a waste event at the moment it happens — the broken bottle gets entered before the shift moves on, not reconstructed at count time from a hazy recollection of what went wrong that week.
Who should be doing the count
A count done by the same person who does the pouring, alone, has an obvious weakness — it is not an independent check on that person's own work. Rotating who counts, or having a second person present for the count even briefly, is a small change that makes the reconciliation mean more, without accusing anyone of anything specific; it is simply a better-designed check than one person marking their own homework. The same rotation is worth applying to who logs waste, for the same reason.
Turn the answer into a working view
See it on your own property
See the bookings, rooms and team workflow on your own property.