Bar Profit: Pour Costs, Par Levels and a Beverage Program That Pays

A bar's profit margin comes down to three numbers: what each drink costs to make, how much stock sits on your shelves, and how far the stock count drifts from what the till says you sold. Pour cost, par levels and variance are the three. Bars tend to watch the first and ignore the third, which is where the money leaks.

This guide gives you the formulas, a worked example and a weekly routine.

What is pour cost, and how do you work it out?

Pour cost is the cost of a drink as a share of its selling price. Divide the ingredient cost by the price. Provi's guide gives a simple case: a beer costing 3 that sells for 12 has a pour cost of 25%.

For the whole bar, divide inventory usage by beverage sales. Usage is opening stock, plus purchases, minus closing stock. Keep spirits, wine, beer and soft drinks in separate lines, because their targets differ.

Ranges that operator guides publish for typical bars are, in rough terms:

These come from one distributor-backed guide and carry no named survey. A second guide, from Backbar Academy (2025), puts a healthy bar average at 18% to 24% and says fine-dining or wine-heavy programmes may run higher. Use them as orientation only.

How do you price a drink to hit a target?

Divide the ingredient cost by your target pour cost. A cocktail whose ingredients cost 2.50, at a 20% target, prices at 12.50. Add every ingredient, including the garnish and the share of the ice, and the consumables.

The same Provi guide adds a caution: it assumes about 20% variance from breakage, spillage and theft, taken from unnamed industry analysts. On that assumption, the 25% beer above really runs near 31%. The number is not a standard, but the lesson is. Your printed pour cost is a best case. The real one includes everything that goes wrong between the bottle and the guest.

The recipe tells you what a drink should cost. The stock count tells you what it did.

What does variance tell you?

Variance is the gap between what your sales say you poured and what your stock count says you used. BevSpot's inventory guide describes it as the difference between product cost and usage, and says some variance is normal. The Backbar guide links high variance to over-pouring, spills, unrecorded comps and theft.

Think of a small worked example. Imagine your sales system says you poured 100 measures of a gin from one bottle's worth of drinks, and the count says you used the equivalent of 110. That is a 10% gap on that bottle. One bottle is trivial. Across a whole spirits shelf, week after week, it is not.

Work from the biggest losses first. Backbar suggests focusing on the products with the highest loss percentage. In practice, that usually means a short list: the expensive spirits, the free-poured cocktails and the draft lines.

Why does the sales mix matter more than any single target?

Because your bar-wide pour cost is a blend. BevSpot's pricing guide uses targets of 14% for liquor, 20% for beer and 22% for wine, and shows how the split of sales moves the overall figure. In its illustration, half of sales are spirits, a quarter wine and a quarter beer.

The practical point: if guests drift towards the category with the highest cost, your overall pour cost rises even though no single drink changed. Watch the mix, not just the recipes. A bar that sells more wine in winter and more beer in summer will see its blended figure move with the season, and that is not a leak.

How do you set par levels?

A par level is the stock you need on hand to meet normal demand until the next delivery. BevSpot gives a weekly formula: par level equals weekly usage plus safety stock, divided by the number of deliveries per week. It suggests safety stock of about 20% to 30% of weekly usage.

There is a catch. Dividing the safety stock by the number of deliveries shrinks the buffer for each delivery cycle. A late truck or a busy night puts the whole buffer at risk, whatever the delivery count. A more cautious version works out usage for one delivery cycle and adds the full safety stock on top.

A worked example, with invented numbers. A bar uses 40 bottles of its house vodka a week and has two deliveries.

1. Usage per delivery cycle is 40 divided by two: 20 bottles.
2. Safety stock at 25% of weekly usage is 10 bottles.
3. Add them: the cautious par level is 30 bottles.

The published formula gives 25, because it divides the safety stock by two as well. Use the cautious figure for fast sellers, where a stock-out costs sales. For expensive, slow-moving spirits, where cash on the shelf is the bigger cost, trim it as your counts show how much buffer you really use.

Raise it before a busy event and lower it in a slow season. The point of par is to cut two losses at once: the cash sitting on a shelf, and the lost sale when a product runs out.

Which habits protect the margin?

A few plain habits do most of the work, and the evidence behind them is practical rather than academic.

• Measure every pour. Backbar recommends jiggers for cocktails and wine, and regular pour tests. An extra quarter or half ounce per drink is an unrecorded loss.
• Track variance weekly. A monthly view hides the week things went wrong.
• Check invoice prices. Compare each invoice with the previous one or with an agreed price list.
• Look after draft. Train staff to manage foam, consider keg scales to track yield, and clean lines on a regular schedule.
• Standardise wine pours. Provi notes that wine is hard to cost unless you fix the pour size first.

Backbar's guide says results from this work can show within a quarter if tracking is consistent. That is the publisher's own claim, so test it in your bar.

How do you build a programme that pays?

Design the menu around the drinks you can make well and cost cleanly. A shorter list is easier to pour consistently, easier to count and easier to train.

Three tests help:

1. Can every drink be made with a standard recipe and measure? If not, you cannot cost it.
2. Do several drinks share ingredients? Overlap lowers par levels and waste.
3. Does the list include a few drinks that earn well and sell well? The same sort of review used in menu engineering works for a drinks list, sorted by cash margin and popularity.

If your bar sits inside a restaurant, the same logic extends to the wine list, covered in wine lists that sell. For a bar and cafe format, see the notes on the cafe-bar hybrid.

What mistakes cost bars the most?

Habits cause most of the damage here, rather than decisions.

• Costing the drink but not the extras. Garnish, ice, syrups made in house and bar consumables all count. Provi puts consumables such as napkins and limes at a few percent of sales.
• Setting par once. A level set in spring is wrong by December.
• Counting only when something feels off. By then the trail has gone cold.
• Letting one person order, receive and count. Split the roles where you can, even if it is only a second pair of eyes on the count.
• Treating the numbers as a verdict on staff. Variance is information about a system, so explain it that way and the team will help you close it.

A weekly routine that fits in an hour

Keep it short so it actually happens.

1. Count the top 15 products by cost.
2. Compare usage with sales for each and note the biggest gaps.
3. Check any product near or below par.
4. Compare the week's invoices with the last.
5. Pick one gap to investigate and one change to try next week.

Key takeaways

• Pour cost is ingredient cost divided by price, and published targets vary by category and source.
• Your recipe pour cost is a best case. Variance shows the real one.
• Set par levels from usage per delivery cycle plus a full safety margin, and review them with the season.
• Measure pours, count weekly and start with the most expensive products.
• Build a list you can cost, pour consistently and count easily.