Beverage Cost Percentage Formula: Pour Cost Guide for Bars and Restaurants

The two formulas (drink level and period level)

Drink-level pour cost = (liquid cost + garnish + mixers) ÷ menu price × 100. A cocktail with $1.90 of spirits, $0.45 of fresh juice and syrup, and $0.15 of garnish costs $2.50; at a $12 menu price the pour cost is 20.8%. Every component counts — bars that cost only the base spirit understate true cost by 3 to 6 points.

Period-level beverage cost = (opening inventory + purchases − closing inventory) ÷ beverage sales × 100, with inventory valued at cost and counted in the same units it is purchased. Count full bottles, weigh open spirits bottles or use a calibrated strip, and value kegs by remaining weight. The period formula catches what the drink-level math cannot: spillage, over-pouring, comps, voids, theft, and receiving errors.

Run both every week. Drink-level costing tells you whether the price is right; period-level costing tells you whether the operation honors it. When actual exceeds theoretical by more than 2–3 points on spirits, the cause is almost always process — free-pouring without jiggers, unrecorded comps, or keg waste — not pricing.

Keep beverage sales separated from food sales in the POS. A blended cost percentage hides a disciplined kitchen behind a leaking bar, or vice versa. The food-cost control method at https://restaurantmargin.com/blog/food-cost-variance-restaurant applies the same variance discipline to the bar.

Bottle yields: the math that sets every price

Yield = bottle volume ÷ pour size, minus realistic waste. A 750ml spirits bottle at 44ml (1.5oz) yields 17.0 theoretical servings; at 30ml (1oz, common for high-proof or premium pours) it yields 25. A 750ml wine bottle at 150ml pours yields 5 glasses; at 120ml (tasting or lunch pours) it yields just over 6. A 58.7-liter (1/2 barrel) keg at 473ml (16oz) pours yields about 124 pints before foam loss — budget 5–10% waste on draft depending on line length, temperature control, and staff technique.

Cost per serving = bottle cost ÷ realistic yield, where realistic yield is theoretical minus your measured waste factor. A $24 bottle of well vodka at 17 theoretical pours costs $1.41 per pour at zero waste; with 12% measured waste (15 effective pours) it costs $1.60. That $0.19 difference is 1.6 points of margin on a $12 cocktail — earned or lost entirely in pour discipline.

Coffee follows the same logic with different units: an espresso-based program should cost shots per kilo of beans (a 1kg bag at 18g doubles yields about 55 shots before dial-in waste), plus milk cost per drink by recipe volume. Milk, not beans, usually dominates the cost of a latte — measure it.

Price from cost per serving with the target formula: menu price = cost per serving ÷ target cost share. A $2.50 cocktail cost at a 20% target prices at $12.50; round to the menu architecture ($12 or $13) and recheck the resulting percentage rather than pretending the rounded price hits the target exactly.

• Spirits 750ml ÷ 44ml pour = ~17 servings; ÷ 30ml = ~25 servings.
• Wine 750ml ÷ 150ml = 5 glasses; ÷ 120ml = ~6 glasses.
• Half-barrel keg ÷ 473ml pints = ~124 pours before foam loss.
• Always convert bottle cost to cost per realistic serving before pricing.

Target percentages by category (planning references)

Targets below are widely used planning references, not universal laws — verify against your concept, local taxes, and service model. Spirits and cocktails typically target 18–24% pour cost; the range exists because well drinks subsidize premium call brands poured at higher cost shares. Wine by the glass usually targets 25–35% because bottle cost is transparent to guests who know retail prices; wine by the bottle often runs 30–45% since the markup multiple, not the percentage, carries the profit. Draft beer commonly targets 20–30% with keg deposits and line waste accounted for; bottled and canned beer runs higher, 25–35%, because there is no draft margin to harvest. Coffee and non-alcoholic beverages target 15–25% on paper but demand strict milk and syrup measurement to hold it.

Set targets per category, not one blended bar target. A single 25% target lets cocktails subsidize a sloppy wine program indefinitely. Review category percentages monthly and the overall bar percentage weekly.

Taxes and service model shift every target: jurisdictions with high alcohol excise or strict responsible-service pour limits compress achievable margins, while bottle-service or tasting-menu concepts play a different game entirely. Calibrate to your P&L, not to an internet benchmark.

Worked example: cocktail, wine glass, and pint

Cocktail: 2oz bourbon from a $28 bottle ($1.65 per 2oz at 12.7 realistic pours per 750ml), $0.40 demerara and bitters, $0.10 expressed citrus, $0.12 ice program allocation — total $2.27. At $13 the pour cost is 17.5%. Garnish and ice allocations look fussy until you multiply them by 400 servings a week ($88 weekly, $4,500 yearly).

Wine by the glass: $14 bottle wholesale ÷ 5 pours = $2.80 per 150ml glass, plus $0.10 preservation gas allocation = $2.90. At $11 the cost is 26.4%. Open-bottle spoilage is the silent killer: a bottle that yields 3 sold glasses and 2 discarded ounces costs $4.67 per sold glass (42.5%). Preservation systems and by-the-glass list sizing pay for themselves in weeks.

Draft pint: $145 half-barrel ÷ 112 realistic pints (124 theoretical minus 10% foam and line loss) = $1.29 per pint, plus $0.15 glass-breakage and CO2 allocation = $1.44. At $6.50 the cost is 22.2%. Long glycol lines, warm spots, and untrained first-pour technique each add a point — walk the lines monthly with a thermometer and a scale.

Weekly bar control checklist

Pair this with the kitchen’s variance review so food and beverage get equal discipline:

• [ ] Theoretical vs actual pour cost computed per category; gaps over 2 points investigated, not averaged away.
• [ ] Open-bottle and keg inventory weighed; receiving checked against invoices and credits claimed for shorts.
• [ ] Jiggers, scales, and portion tools in use on every station — free-pouring audited with a mystery measure.
• [ ] Comps, voids, and staff drinks logged with a reason; unlogged drains treated as variance, not mystery.
• [ ] By-the-glass list matched to preservation capacity; slow movers reduced before spoilage, not after.
• [ ] One menu price or pour corrected per week; the bar improves by single corrections, not overhauls.

Limitations and assumptions

Yields and targets here are planning references computed from standard volumes; your bottles, pour sizes, waste rates, taxes, and supplier prices differ, so recompute with measured figures. Alcohol regulation — licensing, pour limits, happy-hour rules, labeling, and responsible-service obligations — varies by jurisdiction and overrides any pricing suggestion in this guide. Theft and loss control involves employment and privacy considerations; apply your policies and local law, not this article, when investigating discrepancies. Illustrative prices are teaching figures in US dollars, not market data. Build the weekly review into the same routine as food costing — the operating scorecard method at https://restaurantmargin.com/blog/restaurant-weekly-p-and-l-flash-report covers the combined cadence, and plan options at https://restaurantmargin.com/pricing systematize it across periods.