Pour Cost: How to Control Beverage Cost Behind the Bar

What pour cost actually measures

Pour cost is beverage cost divided by beverage sales, expressed as a percentage. Calculate the cost side the same way you calculate food cost: opening stock plus purchases minus closing stock, for the same period the sales figure covers. Get those two windows out of alignment — a delivery counted after the stocktake, a week of sales measured against a month of purchases — and the result is fiction.

The number answers one question: of every dollar taken across the bar, how much went back out to suppliers. At 22%, seventy-eight cents of each dollar is gross profit before a single wage or rent payment. That is why the bar carries so much of the margin in most restaurants, and why a few points of drift matter more than the same drift on food.

It is also the fastest number in the business to change. Food cost moves over weeks as menus and suppliers change. Pour cost can move three points in a fortnight because one person started free pouring, and it can move back just as quickly. That responsiveness is what makes it worth measuring often.

Why one number is not enough

The most common mistake is tracking a single blended beverage cost. Spirits, wine, beer and soft drinks have different natural cost structures and different failure modes, and averaging them buries the failures inside the averages.

Consider a bar at $40,000 a month: spirits $14,000 at 18%, wine $12,000 at 32%, beer $11,000 at 24%, soft drinks $3,000 at 12%. Blended, that is 23.4%. Push spirits to 26% — the equivalent of losing roughly a bottle in four — and the blended figure reaches 26.2%. Eight points of damage in the highest-value category registers as under three points on the line most operators watch, and easily gets dismissed as noise or a counting error.

Split it and the same event is unmissable. Four numbers instead of one costs you nothing extra to produce, because you are already counting each category separately at stocktake. The only additional work is recording sales by category, which every till system already does.

Setting a target for each category

Useful starting ranges: spirits and cocktails 15–22%, wine 28–38%, beer and cider 22–30%, soft drinks 8–15%. Treat these as orientation, not standards. A wine-led restaurant deliberately running 40% on bottles to sell serious wine is trading percentage for a much larger cash margin per sale, and that is a defensible decision.

The better target is your own theoretical cost. Take your recipe specifications, multiply by the quantity of each item sold, and you have what the period should have cost if every drink went out correctly. That number is your target, because it is derived from your actual pricing and your actual sales mix rather than from someone else's bar.

The gap between theoretical and actual is where the diagnosis lives. If both sit high, the problem is pricing and the answer is your list. If actual runs well above theoretical, the problem is operational — over-pouring, unrecorded drinks, waste or loss. These two situations produce an identical P&L line and need opposite responses, which is why so much effort gets spent on the wrong fix.

The four ways pour cost leaks

Over-pouring

A 700 ml bottle at $28 costs $2.00 per 50 ml measure and should yield 14 pours. At $11 a drink that is 18.2% and $154 of revenue. A free-poured average of 57 ml — a 15% over-pour, entirely normal for a confident bartender and invisible to the eye — drops the bottle to 12.3 pours and $135, at 20.9%. Nineteen dollars per bottle, twelve bottles a week, and one back-bar line is costing you close to $12,000 a year.

Unrecorded drinks

Comps, staff drinks, manager tastings, remakes and the round that got poured while the till was busy. Individually trivial, collectively the second-largest category of loss in most bars. The important point is that unrecorded giveaways are mathematically identical to theft in your numbers, so failing to record them makes it impossible to tell an over-generous culture from an actual problem.

Waste and oxidation

This is overwhelmingly a wine issue. A bottle opened for one glass and discarded two days later costs the full bottle against the revenue of one glass. On a fifteen-wine by-the-glass list this happens continually and is almost never written down. The fix is a shorter list where every wine sells fast enough to finish the bottle, not better stoppers.

Costing errors

Frequently the bar is running correctly and the target was wrong from the start. A 50-litre keg costed at 88 pints that actually yields 82 after line cleaning, foam and the tail end of the keg will show a permanent beer variance that no amount of staff training will close. Measure real yield across three or four kegs before concluding anything about behaviour.

Counting the bar properly

Pour cost is only as good as the stocktake underneath it, and bar counts fail in predictable ways:

• Count open bottles consistently. Tenths by eye, or weigh them — either works, but the method must not change between counts or the movement is meaningless.
• Count at the same point in the trading week. Comparing a Sunday-night count to a Wednesday-morning one builds the weekend into the variance.
• Record transfers in both directions. Wine to the kitchen, juice from the kitchen, stock moved between bars. Untracked transfers are the single most common reason a pour cost figure looks impossible.
• Reconcile deliveries against invoices before counting. A short delivery paid in full shows up as consumption you never had.
• Two people, one counting and one recording. Faster and materially more accurate than one person doing both.

None of this is difficult, but all of it has to be the same every time. A consistent method with a small systematic error still produces a usable trend; an inconsistent method produces noise that looks like signal, and you will chase it.

A weekly routine that holds

Full monthly counts, plus a weekly spirits count. Spirits hold the most value in the least volume and drift fastest, and counting that one category takes about twenty minutes. Fifty-two chances a year to catch a problem beats twelve, and the cost of the extra time is trivial against a fortnight of unnoticed over-pouring.

Post the category numbers where the bar team can see them. Pour cost falls measurably from visibility alone, before any procedural change, for the same reason waste falls when the bin log goes on the wall. People manage what they can see being measured.

Recost your top ten drinks quarterly against current invoice prices. Supplier costs move continuously and recipe cards do not, so a list priced correctly last year will contain two or three items quietly earning less than you think. This is the same discipline as recosting plates, and it belongs on the same schedule as your food cost review.

What not to do

Do not raise prices to fix a control problem. It works on the spreadsheet and hides the leak, so the leak keeps growing behind a better-looking number while you lose the volume the price rise costs you. Fix the cause, then price the list on its merits.

Do not optimise pour cost without watching cash margin. A $9 cocktail at 22% pour cost returns $7.02 of gross profit. A $16 cocktail at 28% returns $11.52. The second looks worse on every pour cost report and is substantially better for the business. Percentages control for scale, which is useful for spotting change and misleading as an objective.

Do not treat a single bad period as a trend. Bar counts are noisier than kitchen counts because open-bottle estimation is inherently imprecise. Two consecutive periods moving the same direction is a signal; one is usually a counting artefact, and reorganising the bar around it wastes credibility you will need when the signal is real.

Finally, keep the bar connected to the rest of the P&L. Beverage cost sits alongside food cost inside cost of goods, and both feed prime cost. A bar running three points better than last year is worth having, but only if you can see it survive the trip down to the bottom line.

Frequently asked questions

How do you calculate pour cost for a single drink?

Divide the bottle cost by the bottle size to get a cost per millilitre, multiply by the pour size, then divide by the selling price. A $28 bottle of 700 ml poured at 50 ml costs $2.00 per drink; at an $11 price that is an 18.2% pour cost.

How often should pour cost be calculated?

Monthly for all categories, weekly for spirits. Spirits carry the most value in the smallest volume and drift fastest, and counting that category alone takes about twenty minutes. Anything less frequent means paying for a problem for weeks before you see it.

Do jiggers and measured pourers actually work?

They close a 10–15% over-pour reliably, which is the most common size of problem. They do not address unrecorded drinks or waste, and they slow service slightly on cocktails, so use them where the measurement showed a pouring problem rather than everywhere by default.

Should staff drinks be counted in pour cost?

The stock has left inventory either way, so it is already in the number. Record it separately so you can see how much of the variance is a known policy cost. Otherwise an expanding staff-drink habit is indistinguishable from loss in the figures.

Why is my wine pour cost so much higher than spirits?

Partly structural — wine is normally priced on a lower multiple than spirits — and partly oxidation on the by-the-glass list. If it sits above 38%, the usual cause is glass pours that sell too slowly to finish the bottle before it turns.

What causes a sudden jump in pour cost with no obvious reason?

In order of likelihood: a counting error, an untracked transfer, a delivery landing on the wrong side of the stocktake, or a price increase you have not yet reflected in the recipe cards. Verify the inputs before investigating the team.

Is a lower pour cost always better?

No. Pour cost is a ratio, and you bank cash rather than ratios. A higher-percentage drink with a larger gross profit per sale beats a cheap drink with a flattering percentage. Use pour cost to spot what changed, and cash margin to decide what to sell.

Run the numbers

Use the free Beverage Cost Percentage Calculator to apply everything above to your own figures.