On The Line

Beverage Program Margin and Cost

The beverage program is where a full service restaurant makes or loses two to four points of contribution margin. Pour cost, spec discipline, non-alcoholic revenue, and the shrink you cannot see in the POS. Here is what a healthy bev P&L actually looks like.

Kiyong (Eric) Lee·Operations Executive·August 2026·9 min read

The short version

Blended beverage cost should run 20 to 25 percent of beverage sales. Anything higher is spec discipline or shrink.
Over pour is the biggest single leak. Not theft. A 0.25 ounce over pour on a spec drink gives up 17 percent of the cost on every serve.
Non-alcoholic beverages have quietly become 2 to 4 percent of total bev revenue at meaningful contribution margin. Do not skip it.
Theoretical pour cost minus actual pour cost is the shrink line. Under 1 percent is healthy. Over 3 percent is a real operational problem.
Beverage attach of 30 to 40 percent of food revenue is the target. Below 25 percent is a server training problem, not a menu problem.

Ask any restaurant operator where they make the most margin per dollar of sales and the honest answer is the same across concepts: the beverage program. A cocktail with $2.10 in cost sells at $14. A glass of wine with $3.20 in cost sells at $16. A bottle of sparkling water with $0.80 in cost sells at $8. There is nowhere else on the menu where those ratios exist.

Which is exactly why beverage is where the margin also leaks the fastest. A back of house that runs 30 percent food cost has fewer places to lose 3 points than a bar that should be running 20 percent pour cost. The pour cost line on the bev P&L is more sensitive to small behaviors, harder to observe from the manager station, and easier to fake with a well-designed menu that hides the leaks.

This piece is about what a healthy beverage program actually looks like on the P&L, where the leaks live, and how to fix them without turning the bar into a factory.

What blended bev cost should look like

The single most useful diagnostic on a bev program is the blended cost of goods as a percent of beverage sales, broken out by category. A healthy full service bev P&L looks roughly like this:

Fig. 1 · Target cost of goods by category on a healthy full service bev program.

The blended target is 20 to 25 percent of beverage sales. Every category has its own range, and understanding the composition matters more than hitting the top line number, because the composition tells you where to look when the number moves.

Wine by the glass runs the highest

Wine by the glass typically runs 28 to 34 percent pour cost because it is priced against a market anchor and because the bottle math forces a per-glass cost that is stubborn. There is a version of a wine list where you can push wine BTG below 26 percent, but it usually means either the pricing is aggressive versus market or the wines are ones your guests would not choose. Neither is a long-term win.

Cocktails are where the margin lives

A well-specced cocktail runs 15 to 20 percent pour cost. A cocktail with a signature ingredient (house infusion, unusual bitter, specialty syrup) can push to 22. A cocktail built around a premium spirit call runs higher, and that is fine as long as the price reflects it. If your cocktail pour cost is running above 25, either your bartenders are over pouring or your spec cards are wrong.

Beer is what it is

Beer pour cost is largely determined by your product mix and vendor terms. Draft costs run 20 to 24 percent. Bottle and can run 26 to 30 percent because the margin math on packaged beer is tighter. A well-curated beer list with 6 to 10 drafts that fit the concept beats a sprawling 20 tap list on both margin and inventory turns.

Non-alcoholic beverage has become a real line

Five years ago this was Coke and lemonade. Today, a full service restaurant with a real beverage program has non-alcoholic cocktails, house shrubs, alcohol free spritzes, quality tea, and quality coffee, and this category is 2 to 4 percent of total bev revenue at 12 to 18 percent pour cost. That is a higher contribution margin than the alcoholic average. Guests order these drinks without being asked now. Operators without a real NA menu are leaving contribution on the table.

Over pour is the biggest single leak

Here is the number that tends to shock operators when I show them the math. A standard cocktail spec calls for 1.5 ounces of base spirit. A bartender who consistently free pours to a 1.75 ounce measure is giving up 17 percent of the drink cost on every serve. Multiply that across 1,000 cocktails a week. The pour cost line on the P&L looks like a discipline failure. It is actually a training and measurement failure.

Over pour is not theft. It is a training gap that costs the same as theft. A bartender who does not know their measure is losing you a point of pour cost every week they are behind the bar.

Three interventions move this reliably:

Spec drink cards. Every cocktail on the menu has a printed spec card in the well: ingredients, exact quantities, glass, garnish, method. This is not for the guest. It is for the bartender who is new, the bartender who is tired, and the bartender who thinks they remember the recipe from three months ago.
Jiggers, not free pour. The old school pride of a bartender who free pours accurately is real, but the variance is real too. Jiggers add half a second to a drink and remove 12 to 15 percent of the pour variance. On a service bar in a full service restaurant, this is not close.
Monthly variance check. Compare theoretical pour cost to actual pour cost every month. If the gap is more than 1 percent of beverage sales, sit down with the bar manager and audit the top selling cocktails against spec. The gap will usually close inside 30 days if the bar manager knows the audit is happening.

Theoretical versus actual: the shrink line

Theoretical pour cost is what your inventory would have cost if every drink was made to spec. Actual pour cost is what your inventory actually consumed. The gap is bar shrink, and it is one of the single most useful diagnostics in the whole restaurant P&L because it isolates the operational leak from the pricing decisions.

A healthy bar keeps the gap under 1 percent of beverage sales. Between 1 and 3 percent, you have some drift and it is worth a monthly conversation. Above 3 percent, something is systematically wrong and it needs to be fixed inside 60 days. The usual suspects, in order of frequency: over pour on high volume cocktails, comps and buybacks not being recorded, spillage without a log, and in rare cases actual theft.

Almost every modern POS can compute theoretical pour cost if you enter the spec cards into the recipe module. Almost no restaurant does this because it is a two-day project and no one has time. It pays for itself in the first month.

Beverage attach: the real leading indicator

Beverage attach is beverage revenue divided by food revenue, expressed as a percent. In a healthy full service restaurant it lands between 30 and 40 percent. A wine focused concept can push above 50. A cocktail focused concept can push toward 45.

If your attach is below 25 percent, the issue is almost never the menu. The issue is server behavior. Servers who suggest a beverage to every table, mention a specific by-the-glass pairing with the appetizer, and offer a second round at the appropriate moment lift attach by 5 to 10 points without changing anything else. This is a training investment, not a menu redesign.

The one menu design mistake that hurts attach is pricing the low end too aggressively. If your least expensive glass of wine is 8 dollars and your sparkling water is 8 dollars, guests will choose water. If the low end glass is 12 dollars and sparkling water is 7 dollars, wine attach goes up. Guests are not just optimizing on price. They are looking at relative value.

The by the bottle question

Every restaurant with a wine program eventually asks whether the by-the-glass list or the bottle list should drive the program. The answer is both, and the balance shows up in the average check.

By the glass drives volume and captures the two top, the solo diner, and the guest who is not sure. By the bottle drives margin per cover and pushes the average check up 30 to 50 percent on tables that order one. A well-designed wine list uses the by-the-glass list to lead the guest into a bottle recommendation from the same producer or region, at the same style, at a per-ounce cost that makes the bottle look like a better deal.

The bottle list should have 5 to 8 selections in the $50 to $80 range that servers can recommend without needing sommelier-level training. That price band is where the bulk of the incremental margin comes from in most concepts, not the trophy bottles at the top of the list.

Happy hour: run the math before you run the promotion

Happy hour is one of the most abused levers in the bev program because operators discount the drinks without running the incremental math on food attach and cover count. Done right, happy hour fills 4 p.m. to 6 p.m. seats that would otherwise be empty, drives food attach on those seats, and produces contribution margin that would not have existed otherwise. Done wrong, it teaches your regular Friday guests to arrive at 5:45 and stop ordering full price cocktails.

The math that matters is not the pour cost on the happy hour drink itself. It is the incremental contribution per cover. A discounted cocktail with pour cost of 24 percent instead of 17 percent looks worse in isolation. But if that cocktail brought a guest into a seat at 5 p.m. who then ordered a $22 shared plate and a full price second round, the total contribution per cover from that seat is materially higher than the empty seat it replaced. Track happy hour on incremental cover margin, not on the standalone drink cost, or you will kill a working promotion for the wrong reason.

Comps and buybacks are a P&L line, not a courtesy

Every bar comps drinks. Regulars get a buyback on their third round. A guest sends back a cocktail that was not what they expected. A birthday party gets a round on the house. All of that is normal hospitality. What is not normal is when comps and buybacks are not recorded, in which case they show up on the P&L as inflated pour cost with no explanation.

Every comp gets rung in the POS as a comp, with a reason code. Not as a manual pour that never becomes a check. In a well-run bar, total comps and buybacks run between 1 and 2 percent of beverage sales, and the reason code distribution tells you something useful. Heavy service recovery comps mean the front of house is fixing kitchen or service issues. Heavy regular buybacks mean the bar manager is running loyalty for the venue. Both are fine. Both need to be visible.

The point

The beverage program is the highest margin per dollar area of a full service restaurant, which means it is also the fastest place to lose contribution margin when the discipline slips. Blended pour cost of 20 to 25 percent is the target. Wine by the glass runs highest, cocktails and non-alcoholic run lowest. Over pour is the biggest single leak and it is a training problem, not a theft problem. Beverage attach of 30 to 40 percent is a server behavior number, not a menu number.

Fix the spec cards. Put jiggers behind the bar. Run the theoretical versus actual monthly. Build a real non-alcoholic menu. Train servers on beverage suggestion instead of hoping the menu does it. Do that and two to four points of contribution margin move onto the P&L over a quarter, at the same food cost and the same labor line.

Rebuilding a beverage program that is quietly leaking?

I have coached bev programs across multi-unit full service groups and rebuilt spec discipline in bars where pour cost had drifted five points off target. If your bar P&L looks fine on the surface but the shrink line is telling a different story, I am happy to trade notes.

Get in touch

Questions people ask

What is pour cost in a restaurant?

Pour cost is the cost of goods for a beverage divided by the menu price, expressed as a percent. A cocktail with $2.10 in liquor, mixer, and garnish sold at $14 has a pour cost of 15 percent. A well-run full service bar targets a blended pour cost of 18 to 22 percent across all beverage categories combined. That is materially better than a healthy food cost of 28 to 32 percent, which is why bev margin matters.

What is a healthy beverage cost as a percent of sales?

For a full service restaurant with a real bar, blended beverage cost should land between 20 and 25 percent of beverage sales. Wine typically runs the highest at 28 to 34 percent depending on the by the glass mix. Spirits and cocktails run 16 to 22 percent. Beer runs 22 to 26. Non-alcoholic runs 12 to 18. Anything above the top of these ranges is a spec discipline problem or a shrink problem.

What is the biggest leak in a beverage program?

Over pour. Not theft. A bartender who consistently pours a 1.75 ounce measure into a 1.5 ounce spec is giving up 17 percent of the drink cost on every serve. Multiply across 1,000 cocktails a week and the pour cost line looks like a discipline failure when it is actually a training and measurement failure. The fix is spec drink cards, jiggers, and monthly variance checks against theoretical pour cost.

Should a restaurant sell non-alcoholic cocktails?

Yes, and not as a token. A well built non-alcoholic beverage program adds 2 to 4 percent to total beverage revenue with pour costs of 12 to 15 percent, meaning the incremental contribution margin is higher than the alcoholic average. The market has shifted materially since 2022 and guests are ordering these drinks without being asked. Restaurants without a real NA menu are leaving contribution on the table.

How do you catch bar shrink?

Compare theoretical pour cost to actual pour cost every month. Theoretical is what your sales mix would have cost if every drink was poured to spec. Actual is what your inventory says you actually used. The gap is shrink. A well-run bar keeps that gap under 1 percent of beverage sales. A gap above 3 percent points to over pour, comps not being recorded, spillage without a log, or in rare cases actual theft.

What is the right beverage attach rate?

Beverage attach is beverage revenue divided by food revenue, expressed as a percent. A healthy full service restaurant runs 30 to 40 percent beverage attach. A wine focused concept can push above 50. A cocktail focused concept can push toward 45. If your attach is below 25, the issue is almost never the menu. It is the server training around beverage suggestion, or the pricing at the low end being too aggressive so guests default to water.

Should wine by the bottle or wine by the glass drive the program?

By the glass drives volume. By the bottle drives margin per cover and average check. A balanced program has both, with a by the glass list designed to lead the guest into a bottle recommendation from the same producer or region. A program that leans hard on bottle sales alone loses the two top and the solo diner. A program that leans hard on by the glass alone leaves 15 to 20 percent of potential wine revenue and margin on the table.