Pour cost calculator: the drink, the period, and the variance between them

Pour cost is the cost of what is in the glass divided by what you charged for it. A 1.5 oz pour from a $24 bottle of 750 ml spirit costs $1.42, so a $12 cocktail built on it runs an 11.8% pour cost on the spirit. That is the per-drink question, and it is arithmetic. You can do it on a napkin, and the calculator below does it faster.

The period question is a different thing entirely. Take the inventory you started with, add what you bought, subtract what is still on the shelf, and divide by beverage sales. If your recipes say the bar should be running at 19% and the period says 26%, those seven points did not come from your pricing. They came from pours heavier than the recipe, drinks that were never rung, draft that foamed into a drain, and bottles that walked out of the building.

Per-drink pour cost is a pricing tool. Period pour cost compared against per-drink pour cost is an audit tool, and it is the one that recovers money. The calculator does both, and the second half of this page is about reading the gap.

Container sizes in ounces, and what they actually yield

Almost every pour cost mistake that is not a counting mistake is a units mistake. Invoices are in cases and liters, recipes are in ounces, kegs are in gallons and barrels, and the barrel in question is 31 gallons, which is not a number anyone has in their head. These are the conversions, and the yields assume a perfect container with nothing lost.

Container volumes in US fluid ounces, and theoretical yield before any loss.

Container	US fl oz	Pours at 1.5 oz	Pours at 2 oz	Pints at 16 oz
375 ml half bottle	12.7	8.4	6.3	—
750 ml bottle	25.4	16.9	12.7	—
1 L bottle	33.8	22.5	16.9	—
1.5 L magnum	50.7	33.8	25.4	—
1.75 L bottle	59.2	39.4	29.6	—
Sixth barrel keg, 5.17 gal	661	—	—	41.3
Quarter barrel keg, 7.75 gal	992	—	—	62.0
30 L import keg	1,014	—	—	63.4
50 L import keg	1,691	—	—	105.7
Half barrel keg, 15.5 gal	1,984	—	—	124.0

Two notes that cost real money. A wine pour is usually quoted at 5 oz, which gives five glasses from a 750 ml bottle and leaves 0.36 oz behind; pour six glasses from the same bottle and you have quietly moved the glass to 4.2 oz. And a 16 oz pint glass filled to the brim with beer holds less than 16 oz of beer, because some of it is head. If your recipe assumes 16 and the glass delivers 14, your draft yield is 12% worse than your spreadsheet thinks before anyone has spilled anything.

Theoretical against actual: the only pour cost number that finds money

Theoretical cost is what your recipes and your sales mix say you should have used. Every item sold has a recipe; every recipe has a cost; multiply the two across everything that rang and you have the cost of goods your point-of-sale system believes you incurred. Actual cost is what the shelves say: opening inventory plus purchases minus closing inventory.

Those two numbers are measuring the same liquor by two independent routes, which is exactly what makes the comparison useful. One is built from sales records, the other from a physical count. Neither can be quietly wrong without the other disagreeing. The difference between them is unaccounted product — liquor that left inventory without a corresponding sale.

A variance of a point or so is normal measurement noise. A variance of three points on $96,000 of beverage sales is $2,880 in a single period, and it recurs until somebody finds it. That is the number worth chasing, and almost nothing published about pour cost tells you to calculate it.

Nine places the variance comes from, and the different fix each one needs

Variance is a symptom, not a diagnosis. These are the causes, roughly in the order we find them, and each one has a different remedy. Lumping them together as “shrinkage” is how a bar stays 5% over target for years.

What creates a gap between theoretical and actual beverage cost.

Cause	How it shows up	Where the fix sits
Overpour	Spirits variance spread evenly across the whole well; no single item stands out	Operations: jiggers, measured pourers, spot checks
Drinks not rung	Variance on specific high-traffic items; cash shifts worse than card shifts	Controls: till discipline, shift reconciliation
Comps and staff drinks	Variance tracks promotional nights	Accounting: ring at zero to a comp account, never off-book
Draft foam and line loss	Isolated to draft; worse after a line clean or a gas change	Operations: pressure, temperature, line length
Breakage and spillage	Small, irregular, unglamorous	Accounting: a waste log, so it is measured instead of inferred
Receiving errors	Variance appears in the period of a large delivery	Controls: count the delivery against the invoice at the door
Wrong recipe in the POS	Consistent variance on one item that never moves	Data: the recipe, not the bar
Transfers to the kitchen	Beverage cost high, food cost low, prime cost unchanged	Accounting: transfer entries between cost centers
Theft	What is left once the eight above are ruled out	Controls, and it is the last conclusion, not the first

The order matters. Three of those nine are accounting problems, not bar problems — a comp booked off-book, a kitchen transfer never posted, and a recipe nobody updated will all produce a variance that no amount of watching the bartenders will fix. We look at those three first precisely because they are the cheapest to correct.

Why your pour cost and your cost of goods sold disagree

This is the part that belongs to the accountant, and it is the reason a bar owner can be told two different beverage cost percentages in the same week and both be arithmetically correct.

Your inventory system calculates beverage cost from counts. Your P&L calculates cost of goods sold from invoices posted in the period. They agree only when the period boundaries line up exactly, and in practice four things pull them apart:

• Invoice timing. A delivery that arrives on the last day of the period but is invoiced on the first day of the next one is in your closing count and not in your purchases. Beverage cost looks low this period and high next period, and nothing is wrong with the bar.
• The count date. Counting on Monday morning for a period that closed Sunday night includes Sunday’s deliveries and excludes Sunday’s service. A single day of drift on $96,000 of quarterly beverage sales is enough to move the percentage by more than most variances you are hunting.
• Transfers. Wine into a kitchen reduction, beer into a batter, spirits into a dessert. If no transfer entry is posted, beverage cost carries the cost of food production.
• Credits and returns. A keg returned or a corked case credited in a later period leaves the original cost sitting in the wrong one.

Reconciling those four is routine closing work, and it is the difference between a beverage cost percentage you can act on and one you argue about. Until the two sources agree, a variance investigation is chasing an accounting artifact. We do this reconciliation as part of the period close for the bars and venues we run, which is the honest reason this page exists: the calculator is free and the arithmetic is simple, and the reason operators still get the number wrong is almost never the arithmetic.

Pour cost, beverage cost and prime cost are three different numbers

The terms get used interchangeably and they should not be. Pour cost is conventionally the cost of a poured drink against its price, and is used both per drink and as a shorthand for the beverage category overall. Beverage cost percentage is the period figure across all beverage: the one the calculator above produces from inventory and sales. Prime cost is a different level of the P&L again — total cost of goods, food and beverage together, plus total labor, against total sales. It is the single number that decides whether the business works.

A bar can hold a perfect 18% pour cost and still fail, because prime cost includes the payroll that poured the drinks. Equally, a four-point beverage variance can be invisible at prime cost level if labor happened to be light that period, which is precisely why it should be watched on its own.

What a target should be, and why your own trend matters more

Conventional planning targets put spirits lowest, beer in the middle and wine highest, because the markup structure differs by category — a wine list priced at spirits multiples does not sell. Many operators plan around the high teens to low twenties for spirits, the low to mid twenties for draft and bottled beer, and the high twenties to low thirties for wine by the glass, blending to a bar-wide figure somewhere in the low twenties.

Treat those as planning conventions rather than a benchmark: they are the ranges the trade commonly works to, not a measured industry average, and they are not a standard we have audited. They also move with your category mix. A cocktail-led room, a craft draft house and a wine bar are three different businesses and should not share a target.

What is genuinely diagnostic is your own series. A bar-wide beverage cost that sits at 24% every period is a priced business, whatever the trade convention says. A bar-wide figure that moves between 21% and 27% period to period is not a pricing problem at all — it is a counting, controls or close problem, and the variance calculation above is how you tell which.

Common questions

How do you calculate pour cost?

For one drink, divide the cost of the liquor in the glass by the menu price. Cost of the pour is the container cost multiplied by the pour size divided by the container volume in the same units: a 1.5 oz pour from a $24 750 ml bottle is 24 multiplied by 1.5 divided by 25.36, which is $1.42. For a period, take opening inventory plus purchases minus closing inventory, and divide by beverage sales net of tax.

What is a good pour cost percentage?

There is no single correct figure, and it depends on category mix. Trade planning conventions commonly sit in the high teens to low twenties for spirits, the low to mid twenties for beer and the high twenties to low thirties for wine by the glass. Those are planning ranges rather than a measured industry average. A stable figure that covers your labor and overhead is better than a low figure that swings several points between periods.

What is the difference between pour cost and beverage cost percentage?

They are usually the same arithmetic at different scopes. Pour cost most often describes a single drink or a single product against its price. Beverage cost percentage describes the whole beverage category over a period, calculated from inventory movement and sales. Comparing the two is the useful exercise, because a gap between them is unaccounted product.

Why is my actual pour cost higher than my theoretical pour cost?

Because liquor left inventory without a matching sale. The common causes are overpour, drinks never rung, comps and staff drinks kept off the books, draft foam and line loss, breakage, receiving errors, an incorrect recipe in the point-of-sale system, untracked transfers to the kitchen, and theft. Three of those are accounting errors rather than bar behavior, so they are worth ruling out first.

How many 1.5 oz pours are in a 750 ml bottle?

A 750 ml bottle holds 25.36 US fluid ounces, which is 16.9 pours at 1.5 oz. In practice you should plan on 16, because the last fraction of a bottle is rarely poured cleanly. At 2 oz the same bottle gives 12.7, and at a 1.25 oz pour it gives 20.3.

Does this calculator store the figures I enter?

No. The calculation runs in your browser. Nothing is transmitted to us or saved, and closing the page discards everything. The CSV is generated locally on your device.