How to Calculate Pour Cost for an Indian Bar

Quick answer

Pour cost is your liquor cost divided by liquor sales, and a healthy bar runs 18 to 24 percent once you measure it by the peg instead of the bottle.

Pour cost is what your liquor costs you, expressed as a percentage of what you sell it for. The formula is simple: cost of liquor used, divided by liquor sales, times 100. A well-run bar lands between 18 and 24 percent. The hard part isn't the formula. It's that most Indian bars calculate it from bottles bought rather than pegs poured, and the gap between those two numbers is where the money goes.

Key takeaways

What is pour cost?

Pour cost, also called liquor cost or beverage cost, is the percentage of your drink revenue that gets eaten by the liquor itself. If a peg costs you ₹60 and you sell it for ₹250, that pour ran at 24 percent.

Two numbers go into it, and only one of them is obvious.

Liquor sales is the easy one. It's what your bar rang up on alcohol over the period, before tax.

Cost of liquor used is where bars go wrong. It isn't your purchase bill. It's:

Opening stock + purchases − closing stock

Buy nothing in a slow month and your purchases are zero, but you still poured drinks from the shelf. Buy heavily in December and your purchases overstate what you actually used. Only the opening-and-closing method tells you what left the bottle.

The formula, with real numbers

Take a mid-sized Bengaluru pub over one month.

Cost of liquor used is ₹2,40,000 + ₹1,80,000 − ₹2,10,000 = ₹2,10,000.

Pour cost is ₹2,10,000 ÷ ₹9,00,000 × 100 = 23.3 percent.

That's inside the healthy band, and most owners would stop there and feel fine. They shouldn't. A 23.3 percent pour cost at the bar level can hide a whisky category bleeding at 34 percent, propped up by beer running at 15. Category-level pour cost is where decisions live. Bar-level pour cost is where they hide.

If you already track this on the kitchen side, the logic is identical to food cost percentage. Liquor is just the half with better margins and worse discipline.

Why the bottle is the wrong unit

In India the serving unit is the peg, not the ounce. A chota (small) peg is 30ml and a bara (large) peg is 60ml as the working standard, though it genuinely varies: regional practice puts a standard peg at 40ml with an 80ml large in parts of Delhi and Rajasthan, while Goa commonly pours 60ml. Whatever your house measure is, write it down. You cannot compute pour cost against a measure nobody agreed on.

Here's why it matters. Take a 750ml bottle of whisky costing you ₹1,500, poured as 30ml pegs sold at ₹250.

Poured at 30ml Free-poured at 40ml
Servings per bottle 25 18.75
Your cost per peg ₹60 ₹80
Pour cost on a ₹250 peg 24% 32%
Revenue per bottle ₹6,250 ₹4,687

That's ₹1,563 of revenue gone per bottle, and not one rupee of it was stolen. It was given away, generously, by a bartender pouring by eye during a Saturday rush. Run twenty bottles a week through that and you've lost more than ₹30,000 a week to a wrist.

This is the single biggest reason pour cost drifts, and it's why bars that switch from free-pouring to measured pours see the number move within one cycle.

The tax bit almost everyone gets wrong

You'll read, often on pages selling bar software, that alcohol in India attracts 18 percent GST. It doesn't. Alcoholic liquor for human consumption is carved out of GST at the constitutional level: the Constitution (101st Amendment) Act, 2016 defines GST as a tax on supply of goods and services except the supply of alcoholic liquor for human consumption. Taxing it stays with the states, under Entries 51 and 54 of the State List.

In practice that means your liquor carries state excise duty and state VAT, at rates your state sets and changes in its own budget, while your food carries GST. Two tax systems, one bill, one evening.

Three things follow, and they all hit pour cost:

  1. Your liquor cost is state-specific. The same bottle costs materially different amounts in Karnataka, Maharashtra and Goa, so a pour cost benchmark from a US bar blog is close to meaningless for you.
  2. You can't claim input tax credit on liquor. Excise and VAT paid on stock are a hard cost baked into your COGS, unlike GST on many other inputs.
  3. Your billing has to split the two cleanly. Food GST and liquor VAT on one table's bill is a compliance requirement, not a formatting preference. Our guide to GST on restaurant food covers the food side of that split.

If a vendor's pricing page tells you liquor is taxed at 18 percent GST, they have never billed a bar.

The number that actually matters: variance

Here's the opinion, and it's the one thing worth taking from this piece.

Pour cost is a lagging indicator. It tells you, at month end, that something went wrong. It doesn't tell you what, or where, or who. By the time you compute it, the liquor is drunk and the month is closed.

The number you can actually act on is variance: what your sales say you should have poured, against what the shelf says you actually poured.

The math is not complicated:

That 15 percent is your real problem, and it has a small number of causes: over-pouring, drinks comped without being rung up, breakage nobody logged, and theft. Indian F&B operators commonly report losing 10 to 20 percent of inventory to poor control and pilferage, and bar-side losses sit at the worse end of that range because liquid is easy to give away and impossible to see.

Notice that variance points at a bottle and a week. Pour cost points at a spreadsheet.

Why most Indian bars can't compute variance

Not laziness. Structure.

To calculate theoretical usage you need two facts joined together: what was sold, and how many millilitres each sale represents. Most bar setups have the first and not the second. The billing system records "1 × Old Monk Large, ₹220" and stops. It doesn't know a large is 60ml, doesn't know the bottle is 750ml, and therefore can't tell you that the sale should have consumed one-twelfth of a bottle.

So the owner counts bottles on Sunday, compares against a number nobody can produce, and gives up. That's the actual failure, and it's a data-model problem rather than a discipline problem.

A system that carries the serving size on the item can close that loop by itself. In dineomAI, each beverage carries its peg size and bottle size, so every sale draws down stock in millilitres through a ledger, and a variance report puts theoretical against counted per item, in rupees, for whatever date range you pick. We build this, so weigh that accordingly. But whether you use our software, a competitor's, or a disciplined spreadsheet, the requirement doesn't change: the peg size has to live on the item, or variance is guesswork.

Running a stock count that means something

Counting badly is worse than not counting, because it produces a number you'll trust.

Once you have two clean counts in a row, variance becomes trustworthy, and only then is pour cost worth quoting to anyone.

Bringing pour cost down without watering the drinks

Four levers, in the order they usually pay off.

  1. Measure every pour. Jiggers or measured pourers on the fast-moving bottles first. This is the cheapest and biggest win available to most bars.
  2. Fix the menu mix, not just the price. Pour cost is a weighted average. Selling more of your 15 percent items moves it faster than repricing your 30 percent ones. That's a menu engineering job.
  3. Ring up every comp. A free drink is a marketing cost, and it's fine. A free drink that never touches the till is shrinkage that will show up as a mystery.
  4. Raise the ticket, not the pour. A guest who orders a second round because reordering was easy is worth more than a guest who got a heavy first one. That's the argument for an ordering system built for bars, and it's the same logic behind average order value work generally.

Note what isn't on that list: pouring less than you advertise. Short-pouring a 60ml peg to 50ml is the one lever that works on a spreadsheet and fails in the room. Regulars notice, and you'll lose more in covers than you save in whisky.

FAQ

What is a good pour cost percentage?

Most well-run bars sit between 18 and 24 percent overall, with beer typically cheaper to pour than spirits and cocktails the most expensive. Anything above 30 percent points at over-pouring, unrecorded comps, or theft rather than at your pricing. Judge each category separately, because a healthy blended number can hide one bad one.

How do you calculate pour cost?

Divide the cost of liquor used by liquor sales for the same period, then multiply by 100. Cost of liquor used is opening stock plus purchases minus closing stock, not simply what you bought that month. Use the same period for both figures, and count stock on the same day of the cycle each time.

Is alcohol taxed under GST in India?

No. Alcoholic liquor for human consumption is excluded from GST by the Constitution (101st Amendment) Act, 2016, so it carries state excise duty and state VAT instead, at rates each state sets. Food on the same bill is taxed under GST. Any source telling you liquor attracts 18 percent GST is wrong.

What is the difference between pour cost and variance?

Pour cost is a percentage that tells you how expensive your drinks were last period. Variance compares what your sales say you should have poured against what you actually poured, so it points at a specific bottle and a specific week. Pour cost diagnoses the bar; variance diagnoses the problem.

How much liquor does a typical bar lose to shrinkage?

Indian F&B operators commonly report 10 to 20 percent inventory loss from weak controls and pilferage, and bars often sit at the higher end because over-pouring is invisible. The largest single cause is usually free-pouring rather than theft, which is good news, because a measured pour fixes it in a week.

What to do next

Pick your three fastest-moving bottles and do one thing this week: work out how many pegs each should yield at your house measure, then count what actually went. If the gap is more than about 5 percent, you've found real money, and you found it without buying anything. Then decide whether your billing system knows your peg sizes, because if it doesn't, you'll be doing that arithmetic by hand every week. Book a short dineomai demo if you'd rather it ran itself.

See it on a real table

A 15-minute demo: watch dineomAI take an order, speak five Indian languages, and upsell the right pairing. No hard pitch.

Book a 15-minute demo →