How to Calculate Pour Cost for Liquor, Beer and Wine: Bottle, Keg and Glass Math

A bar has two pour costs, just as a kitchen has two food costs. The theoretical one comes from each drink's recipe and price. The actual one comes from counting the bottles and kegs and dividing what was used by what was sold. The gap between them is overpouring, spills, comped and staff drinks that were never rung in, breakage, keg and line losses — the parts of the bar that never appear on an invoice. Compare them category by category; the coffee shop guide applies the same drink-by-drink costing to espresso and milk.

Bottle math: ounces per bottle

Check the size on every invoice. Federal standards of fill allow many sizes: spirits may be bottled in 25 sizes, including 700 mL and 1.8 L as well as 750 mL, 1 L and 1.75 L (27 CFR 5.203), and wine in 25 sizes from 50 mL to 3 L, plus even-liter containers of 4 L and larger — none of them 1.75 L (27 CFR 4.72). A 700 mL bottle at the same price as a 750 mL one costs 7.1% more per ounce. NIAAA's "standard drink" of 1.5 oz of 40% spirits is a health measure, not a legal pour size; it notes those amounts "may not reflect customary serving sizes" (NIAAA).

Cost a spirit pour and a cocktail

At $14, that margarita runs 24.6%. Cost what you buy, not what the drink contains: shaking and stirring add water from the ice, which has no cost unless you buy ice. Batched syrups and juices are sub-recipes with their own yield; squeezing a case of limes and measuring the juice gives you the cost per ounce better than any chart.

Draft beer: keg math

Federal rules define a barrel of beer as 31 U.S. gallons (27 CFR 25.11) and allow halves, quarters and sixths among its fractional parts (27 CFR 25.156). A half barrel is therefore 15.5 gallons, or 1,984 fluid ounces; a sixth barrel is about 661 fluid ounces. The Brewers Association's Draught Beer Quality Manual prints the half barrel as 165 twelve-ounce servings — capacity divided by glass size, with no allowance for loss.

Use the beer actually poured, not the glass size. When researchers bought and measured 480 drinks at 80 Northern California venues in 2007, without the servers knowing, draught "pints" averaged 14.05 fl oz of liquid (Kerr et al., 2008). The Brewers Association manual gives no loss percentage, but it does name the causes: foam at the tap, the beer discarded from the lines at every cleaning — which it recommends at least every two weeks — and a line volume that grows with length and diameter, from about a sixth of an ounce per foot of 3/16-inch vinyl tubing to three-quarters of an ounce per foot of 3/8-inch line.

Wine by the glass

A 750 mL bottle holds 25.36 fl oz: 5.07 five-ounce glasses or 4.23 six-ounce glasses. At a hypothetical $15 per bottle, a 5 oz glass costs $2.96, or 24.6% of a $12 glass. The number that moves wine pour cost is how much of each opened bottle is sold. If only four glasses are sold before the rest is poured out, each glass sold carries $15 ÷ 4 = $3.75, or 31.3%. The same 2007 study measured wine servings averaging 6.18 fl oz — 23.6% more than a 5 oz pour — which is why the glass size on the menu needs a matching measure behind the bar.

Actual pour cost from inventory

The blended 20.5% hides three different problems. A dollar gap per category tells you where to look: overpours and unrung drinks in spirits, foam and line loss in draft, open-bottle waste and generous pours in wine. The method is the same one the ideal versus actual food cost guide uses for the kitchen: resolve the data before you name a cause.

Measure pours before choosing jiggers or free pour

How accurately people pour by eye is an empirical question, and the evidence says to measure rather than assume. In a study of 86 Philadelphia bartenders asked to pour 1.5 oz (44.3 mL), pours into short, wide glasses averaged 55.5 mL and into tall, slender glasses 46.1 mL (Wansink & van Ittersum, 2005); a later set of four studies (one in a laboratory, three at public events) in which non-bartenders poured what they considered a typical drink found "no credible effect of glass shape" on the size of that serving (Knibb et al., 2018). What holds across the research is that free pours vary. Test yours: have each bartender free-pour your spec into a measure ten times, record the average and the spread, and cost the gap.

• —Jiggers give the most consistent measure and cost some speed.
• —Measured pourers stop the pour at a set volume; check them against a measure, because they wear.
• —Batching the spirit portion of high-volume cocktails removes pour-to-pour variation for those drinks.
• —Mixed approach: measure the expensive bottles and let tested, consistent bartenders free-pour the well. Decide from your test results, not from a rule of thumb.

Marji's recipe costing is built around kitchen units — pounds, weight ounces and counts — and does not convert milliliters or fluid ounces, so drinks belong in the free pour cost calculator, which works from bottle size in milliliters. Marji does not count bar inventory or import POS sales.