Beverage cost control in hotel bars: measures, par stock and the monthly count

Beverage cost control is the loop that compares what your recorded sales say should have been poured with what your stock count says was actually used. It rests on three things: a standard measure for every drink, a par stock for every bar, and a count taken the same way each time.

The month closes, the bar count is done, and the number does not match the sales report. Nobody carried out a case of whisky. No one can say which bottle, which night or which drink explains the difference, and the report has no column where that answer would live. This article builds the missing column. After reading it you will be able to define a standard measure, set a par stock, run a repeatable count, and calculate the gap between expected and actual use in any bar, including one where nothing is sold at the counter. It extends the control loop from our guide to food and beverage cost control in hotels to the specific mechanics of drinks.

Where does the money actually go missing in a hotel bar?#

In small amounts, at the point of pour, and almost never in a form the sales report can show. A bar loses money wherever a quantity is used but not recorded.

Free pouring is the first leak. A bartender counting in their head pours a different amount every time, and the difference per drink is invisible. Over a month, across hundreds of serves, it becomes material. The second leak is unrecorded comps: a drink given to a guest, a guide, a supplier or a colleague with no button pressed and no line on a sheet. The stock leaves the bar; the record does not follow it.

Transfers between outlets are the third. A bottle carried from the lobby bar to the pool bar or to banquets without a transfer note leaves one outlet showing a shortage and the other showing nothing at all. Breakage and spillage written off without a witness or a count form the fourth leak, and it is the simplest place to hide a missing bottle, because the write-off is expected. Staff drinks and tasting sit in the same family: legitimate as a written policy, expensive as an unrecorded habit. The control is the record, not the ban.

Why the sales report cannot see any of this#

Revenue reports measure what was sold. Bar losses live in what was used. The two never meet unless someone builds the bridge, and that bridge is the count, the recipe and the calculation described below. Until it exists, the sales report will keep saying business is fine while the store keeps issuing more bottles than the till can explain.

What is a standard measure, and why does it decide your cost?#

A standard measure is the fixed quantity of alcohol a drink is built on, written down and used by every bartender in every outlet. Without it, the cost of a drink can only be guessed, not calculated.

The choice is between a measuring device and free pouring. A measure costs a few seconds per drink and gives a known quantity; free pouring is faster and gives an unknown one. Whichever you choose, write it down as policy, because a policy you can verify is worth more than a faster pour you cannot cost. If you allow free pouring, verify it with spot checks and accept a wider variance.

Every mixed drink needs a recipe card: each ingredient, its quantity in the same unit the store uses, the glass, the garnish and the method. A drink without a card cannot be costed or counted. The same discipline that governs standard recipes and portion cost in the kitchen applies at the bar. Cost per serve is the sum of quantity times unit cost for every ingredient, not just the spirit: mixers, juice, garnish and the ice you pay for all belong in the number.

Bottle yield tells the team what a bottle is worth in serves. For example, a 70 cl bottle poured at 5 cl gives 14 serves before any allowance for spillage; the same bottle at 6 cl gives 11. Glassware must be sized to the measure, because a drink poured correctly should look right in the glass, or the bartender will top it up to fill the space. Spot checks keep the measure honest: weigh a poured drink, or compare empty bottles with till records for one shift, often enough that nobody knows which shift will be checked. Finally, set the menu price from the cost you calculated, not from what the bar across the street charges.

What a recipe card for a mixed drink must contain#

Illustrative only, not a benchmark. At a menu price of, for example, 20.00, the drink costs 5.80 to build, or 29 percent of the price; your own numbers come from your own invoices and your own recipe cards.

How do you set par stock for a bar?#

Par stock is the quantity of each item a bar must hold to serve expected demand between deliveries, set per outlet and reviewed when demand or the menu changes.

There are two stocks, not one: the central store and the bar. Par applies to the bar; the store carries the buffer behind it. Build the par from demand, not from shelf space: expected serves per item over the delivery cycle, divided by serves per bottle, plus a margin for a busy weekend or a late delivery.

Par is a ceiling as well as a floor. A bar holding more than its par hides losses in the back bar and ties cash up in bottles that are not selling. The requisition is the paper that moves stock: the bar asks the store for a quantity, the store issues it, both sign, and the bar's stock rises in the same record where the store's falls. Transfers between bars need the same paper as a requisition, because without it, one outlet's shortage becomes another outlet's unexplained surplus. This connects directly to the receiving side described in hotel purchasing and receiving, from order to storeroom.

Set a reorder trigger per item, so the bar asks for stock when the closing count reaches a level, not when the shelf looks empty. Review par after every menu change, every season and every large group booking.

Requisition, transfer and write-off: three different papers#

A requisition moves stock from the store to the bar. A transfer moves it between two bars or to banquets. A write-off removes it from the books because it was broken, spoiled or given away, and it needs a witness. Three papers, three signatures, three different effects on the count. If your team uses one sheet for all three, you have no count.

How should you count bottles and kegs?#

Count at a fixed time, in a fixed order, by two people, and record every open bottle at its actual content rather than as full or empty.

Timing decides whether the count is data or a guess. Count before service opens, when the bar is clean and nothing is being poured. A count taken during or after service measures the shift, not the stock. Frequency follows value: a full count monthly for the report, a short count of high-value items weekly or daily, and open bottles recorded at every shift handover.

Open bottles need a method, not an eye. Weigh them, decant into a measuring container, or use a marked sleeve. "About half" is not a count and cannot be compared with last month. Kegs are counted by weight or by a flow meter, because a keg recorded as "one" says nothing about the beer left inside it. Two signatures belong on the sheet: the person who holds the stock does not count it alone. Count sheets follow location, not product: main bar, pool bar, restaurant, banquets, store, and any bottle sitting on a trolley or in a store room corner. Record in the same unit as purchasing, so the count can be set against invoices without conversion.

What makes a count repeatable#

Expected versus actual use: how do you find the gap?#

Expected use is what the recorded sales say should have been poured. Actual use is opening stock plus receipts minus closing stock. The difference between the two is the number worth investigating.

The usage formula, in words, per item, per outlet, per period: opening stock plus what was received, minus closing stock. Expected use per item is the sum, across every recipe sold, of drinks sold multiplied by the quantity in the recipe, plus a recorded allowance for comps and known spillage. Compare in money as well as in quantity: a three-centilitre gap on a house spirit and the same gap on a premium malt are not the same problem, and the report should say so.

Read the gap as a question, not a verdict. It can be a measure, an unrecorded comp, a transfer, breakage, a counting error, or a recipe that changed without the card. Rank by value and investigate the top five items, not the whole list; a bar that chases every item at once chases none of them. Run the calculation monthly for the report and weekly for the highest-value items, and keep the history, because a bar's own trend is more useful than any external benchmark.

What the monthly page per outlet should show#

Illustrative only, not a benchmark. The closing note on the page carries the total gap in money, with a line for recorded comps and breakage to explain part of it, and a note on each of the top five variances. In this example, the 550 cl gap on house gin is the first item to investigate, and the 300 cl on the malt is worth more money per centilitre even though it is smaller in quantity.

How do you control an all-inclusive bar where nothing is sold?#

Replace the sale with a recorded serve. Every drink poured is counted, so the same expected-versus-actual calculation still works without a price at the counter.

The till receipt is gone, so the recipe card and the serve count become the only record of use. If nobody records the serve, no report can tell you what the bar used. The simplest reliable record is one button per drink on the bar's system, or a tally sheet by drink and by hour. The point is the count, not the technology.

Shape the included list deliberately: local brands, draft beer and house wine included; premium brands limited by outlet, by hour or by a supplement, so the expensive item is not the default. Cost per guest night is the useful unit here, not cost per drink sold, because the guest has already paid; compare it with occupancy and with the food side of the same guest night, as described in our guide to all-inclusive hotel economics and cost per guest night. Breakage, spillage and staff drinks still happen and still need a log; the absence of a price does not remove the cost. Watch the demand curve too: the same guest drinks more at nine in the evening than at three in the afternoon, so par levels and staffing follow the curve rather than the average.

Why the count survives when the price disappears#

The count never depended on revenue in the first place. It depends on quantities: what the store issued, what the shelf holds, what the recipe says. A recorded serve, a measured pour and a monthly bottle count give you the same gap calculation a paid bar uses.

What to do next#

Start with one bar and ten items. Write the recipe cards, set the par, and count those ten items at a fixed time every month for three months. The gap you find in month two tells you which of the four leak points is yours, and the same loop then extends to the next outlet without any new method.