Menu engineering for bars: pouring for margin

Bars respond to menu engineering faster than kitchens do. A cocktail list is short, the recipes are exact, the cost of a change is a printed card, and guests choose from a list rather than from a craving. Get the ranking right and the margin moves within a fortnight.

Cost every pour, including the invisible ones

A drink's cost is not just spirit. It is spirit at your actual invoice price, modifiers, mixers, juice, syrup, garnish, ice dilution, and the glass and straw. On a $15 cocktail, garnish and fresh juice frequently add a dollar that never appears in anyone's costing sheet.

Then account for pour variance. Free pour typically costs a bar three to six points against jigger or metered service. Until variance is measured with a weekly count by category, every theoretical drink cost on your list is fiction.

Rank the list by contribution and popularity

Sort each drink into four groups using twelve weeks of POS data: high margin and high volume, high margin and low volume, low margin and high volume, low margin and low volume.

The high-margin, high-volume drinks go at the top of each section and into every bartender recommendation. High-margin, low-volume drinks usually have a naming or description problem, not a taste problem, and often move with a rewrite alone. Low-margin, high-volume drinks get re-specified — a different base spirit at the same perceived quality, or a small price move. The rest come off the list; they consume prep, backbar space and menu attention for nothing.

Design decisions that change what gets ordered

Keep sections to five to seven drinks; longer lists push guests to default to what they already know, which is usually a well pour. Anchor each section with one premium item so the drink beside it reads as reasonable. Describe drinks by experience rather than by ingredient list. Avoid a column of aligned prices — it turns a list into a price comparison.

Batch the top three sellers where the recipe allows. Batching cuts ticket time at the peak, which sells more drinks per hour and improves labor productivity at the same time.

Wine and beer are separate problems

Wine by the glass lives or dies on the yield you actually get from a bottle and on how consistently the pour is measured. Five ounces from a 750ml bottle gives five glasses; an unmeasured hand gives four and a half, which is a ten percent cost increase nobody sees. Beer needs line yield and keg waste tracked; foam and line cleaning losses of eight to twelve percent are common where nobody is counting.

Hold the gain

Re-cost the list whenever a major supplier price moves, run a weekly beverage count with variance by category, and review drink mix monthly. Bars that do this typically add four to eight points of beverage margin without raising a single headline price.

If you want the list read against your own POS mix, our bar and restaurant consulting engagements start with exactly this analysis.