Skip to main content

Pour Cost, Keg Yields, and Shrinkage: Bar Bookkeeping That Catches What Walks Out the Door

Published 11 min readMike ThriftMike Thrift
Pour Cost, Keg Yields, and Shrinkage: Bar Bookkeeping That Catches What Walks Out the Door

Your POS says Friday night brought in $4,200 at the bar. Your cooler says you burned through $1,400 of product to get there. That 33% pour cost just ate a third of your revenue before rent, payroll, or a single garnish — and if you cannot explain exactly where the gap between "should have used" and "actually used" went, some of it walked out the door as over-pours, spillage, comps nobody rang in, or outright theft.

Pour cost is the one number that separates bars that print money from bars that feel busy and still miss payroll. Here is how to calculate it correctly, what good looks like by category, how keg yields and shrinkage quietly distort your books, and the bookkeeping routine that turns a monthly mystery into a weekly control.

What Pour Cost Actually Measures

Pour cost is simple division with serious consequences:

Pour Cost % = (Cost of Goods Sold for beverages / Beverage Revenue) x 100

Cost of goods sold here means the wholesale cost of what you actually consumed — not what you purchased. The standard usage formula gets you there:

Beginning inventory + Purchases received - Ending inventory = Usage (COGS)

Divide that usage by what the POS recorded in beverage sales over the same period, and you have your pour cost. It is the complement of gross margin: a 24% pour cost means a 76% gross margin on drinks before labor and overhead.

The trap most owners fall into is measuring it blended and monthly. A blended 26% can look healthy while spirits run at 14% and wine-by-the-glass bleeds at 44%. And a monthly number arrives too late — by the time you learn March was bad, the over-pouring bartender has worked 20 more shifts. Track pour cost by category (spirits, draft beer, packaged beer, wine, non-alcoholic mixers and garnishes allocated sensibly), and run it weekly.

The Benchmarks Worth Targeting

Targets vary by concept, pricing power, and market, but industry ranges are remarkably consistent:

  • Spirits and cocktails: 15-20% is the goal for most bars; 18-22% is acceptable in high-rent markets or craft programs with expensive base spirits.
  • Draft beer: 20-25% for a well-run draft program.
  • Bottled and canned beer: 22-28%, since there is almost no over-pour but also less markup headroom.
  • Wine by the glass: 25-35% is normal; expensive by-the-glass programs with preservation systems can push higher.
  • Wine by the bottle: 30-40% or more, because bottle pricing is constrained by what guests will pay versus retail.
  • Blended bar program: 20-24% overall for spirits-forward concepts; 24-28% for beer- and wine-heavy rooms.

If your menu is cocktail-led and your blended number sits above 28%, you do not have a pricing problem alone — you have a control problem. If your wine bar runs 38% blended, that may be the business model, and the fix is menu engineering and glass-versus-bottle mix, not accusing the staff.

Price the Drink Before You Pour It

Pour cost starts at recipe costing, not at inventory day. Every drink on the menu needs a spec card with exact measures, and every spec needs a cost:

  1. Cost per ounce. A 750ml bottle holds 25.4 ounces. A $24 wholesale bottle of vodka costs about $0.95 per ounce. A 1.5-ounce pour costs $1.42 in spirits alone.
  2. Add the modifiers. Juice, syrup, bitters, garnish, and even the upcharge rim add up. A cocktail with $1.42 of vodka, $0.40 of liqueur, $0.25 of citrus and syrup, and a $0.15 garnish costs $2.22 all-in.
  3. Set price from target. Menu price = drink cost / target pour cost. That $2.22 cocktail at a 20% target prices at $11.10 — so you list it at $11 or $12 and know exactly what margin you chose.

Free-pouring destroys this math faster than anything else. A bartender who free-pours "1.5 ounces" but actually lands 2 ounces just raised that drink's spirits cost by a third — your 20% target becomes 27% on every one of those drinks, and no menu price fixes a heavy hand. Measured pourers, jiggers, or portion-control spouts pay for themselves in weeks on high-volume wells.

Revisit recipe costs quarterly. Wholesale prices move, bottle sizes change (a liter versus a 750ml changes cost per ounce by 25%), and seasonal menus with fresh or premium ingredients drift. A drink priced perfectly in January can be a loser by June if citrus prices spike and nobody re-costs it.

Keg Yields: Where Draft Profit Quietly Evaporates

Draft beer looks like easy margin until you count what never reaches a glass. A standard US half-barrel keg holds 15.5 gallons, or 1,984 ounces. In theory that is:

  • 165 twelve-ounce pours (before waste)
  • 124 sixteen-ounce pints (before waste)

Nobody gets the theoretical number. Foam, line purging, drip-tray loss, overfilled pitchers, and the first pint after a keg change all take their cut. A realistic yield is 5-10% below theoretical for a dialed-in system, and far worse for a neglected one.

The Math That Matters Per Keg

Take a half-barrel you bought for $150:

  • Theoretical: 124 pints at $7 = $868 revenue, 17.3% pour cost. Looks great.
  • Realistic (10% loss): 112 pints = $784 revenue, 19.1% pour cost. Still fine.
  • Sloppy (25% loss from foam, over-pours, and freebies): 93 pints = $651 revenue, 23% pour cost — and $217 of expected revenue never existed.

That missing $217 does not show up as an expense anywhere. It shows up as a pour cost that "feels high for no reason." Multiply it across four or six taps over a month and you have found the margin you thought you lost to rent.

Fixing Draft Yield

  • Temperature and pressure first. Warm kegs foam. Over-pressurized lines foam. Long, uninsulated runs foam. If every bartender complains about foam on the same tap, call the draft technician before you blame the staff — the fix is mechanical, not disciplinary.
  • Track per-keg. Log each keg tap date, kill date, and pints rung through the POS. A half-barrel that yields 90 pints twice in a row on the same line has a system problem; one that yields 90 on one bartender's shifts and 115 on everyone else's has a training problem.
  • Standardize the pour. Define the head (about an inch on a pint), the glass, and whether pitchers get a top-off. "A pint" should mean the same volume regardless of who is behind the bar.
  • Book keg deposits correctly. Keg deposits and shell deposits are refundable assets, not COGS. Park them in a deposit asset account and relieve them on return — expensing them inflates pour cost in the purchase month and understates it when the credit lands.

Shrinkage and Variance: Naming What Disappeared

Shrinkage is the umbrella term for everything consumed but not sold: spillage, breakage, over-pours, spoilage (stale opened wine, expired kegs), comps and voids that were never rung in, and theft. Variance is how you measure it.

A variance report compares two numbers over the same period:

  • Theoretical usage: what the POS says you should have consumed, given every drink rung in and its spec (48 gin-and-tonics at 1.5 ounces = 72 ounces of gin).
  • Actual usage: what inventory says you consumed (beginning + received - ending).

Variance = Actual usage - Theoretical usage. Positive variance means more disappeared than sales explain.

Well-run bars hold total variance under 3-5% of inventory value. Occasional spikes happen — a dropped bottle, a bad keg, a counted case wrong. Persistent variance above 10% on a category is systemic: over-pouring, unrecorded comps, or theft. The category breakdown tells you where to look. Spirits variance with beer clean points at the well, not the cooler. Draft variance with bottles clean points at the taps. Everything high at once points at counting errors or receiving fraud (short deliveries, substituted product) rather than the bar team.

The Five Usual Suspects

Over-pouring is the biggest quiet killer — usually generosity or habit, not malice. Heavy-handed regulars' pours, "buy-back" rounds that never get rung as comps, and free-poured call brands compound nightly.

Spillage and waste are real but bounded. A normal bar loses 1-2% to broken glass, dropped bottles, and genuine mistakes. When waste logs show zero entries for a month, the waste did not stop — the logging did.

Unrecorded comps and voids turn hospitality into shrinkage. Every comp, spill-replacement, and staff drink needs a POS button and a manager reason code. A comp rung properly raises pour cost honestly and shows up in the right expense bucket; a comp poured silently shows up as theft-shaped variance and poisons trust.

Spoilage hits wine-by-the-glass and craft draft hardest. Open bottles without a preservation plan, too many by-the-glass listings for your volume, and slow-moving kegs that go stale are purchasing decisions disguised as shrinkage.

Theft — bottles walking out, cash pints rung as water, friends' drinks never opened on the POS — is the smallest category in most bars and the first one owners suspect. Run the variance math before the accusation. Camera footage and POS void reports confirm; variance reports direct the search.

The Bookkeeping Routine That Makes It Stick

None of this works if the books cannot produce the two inputs — usage and beverage revenue — cleanly. Set up the chart of accounts so the answer falls out:

  • Separate revenue by category. Do not book all sales to one "Sales" line. Split POS revenue into spirits, draft beer, packaged beer, wine, food, merchandise, and cover or event charges. Without category revenue, category pour cost is impossible.
  • Separate COGS the same way. Spirits COGS, beer COGS, wine COGS, plus a small bar-consumables line for mixers, garnishes, and straws. Allocate shared items (tonic, citrus) by a consistent rule and revisit it — precision matters less than consistency.
  • Book comps and promos as their own expense. Ring them, then move their cost out of COGS into a comps/promotions expense. This keeps pour cost honest (you chose to give it away) and shows the true marketing cost of buy-backs and influencer tabs.
  • Count on a schedule and stick to it. Full bottle inventory weekly for spirits and wine, keg weights or flow-meter reads on the same day, packaged beer by case and single. Same day, same method, same units — preferably right before a delivery so receiving cutoffs are clean.
  • Reconcile receiving like cash. Match every invoice to the delivery ticket and the shelf: bottle counts, keg sizes, vintage substitutions, short cases. Invoice errors and short deliveries look exactly like shrinkage if they enter the books uncaught.
  • Close comps and voids daily. Manager approval for voids after the sale, reason codes required, and a daily void/comp report reviewed before close. The discipline is the control; the report is just the receipt.

This is where consistent bookkeeping pays for itself. When usage, category revenue, comps, and deposit credits each land in the right account every week, the variance report takes minutes and points at the tap, the well, or the wine station — not at "the bar is expensive." When everything lands in one sales line and one COGS line, every investigation starts with a spreadsheet rebuild, and most owners stop investigating. If your books already run on plain-text, version-controlled accounting, this category split is just a few more accounts and a weekly import from the POS and invoice stack — and the full history of every adjustment sits in the ledger for the next variance conversation.

A Weekly Cadence You Can Actually Keep

  • Daily (5 minutes): review comps, voids, and discounts from the prior shift. Anything without a reason code gets a same-day answer.
  • Weekly (1-2 hours): count inventory, post receiving, run category pour costs and the variance report. Flag any category more than 3 points above target or any single product with outsized dollar variance.
  • Monthly (half day): re-cost two or three menu items, review keg-yield logs per tap, audit the deposit asset balance against what the distributor holds, and adjust one thing — one price, one spec, one par, one training topic.
  • Quarterly: re-price the menu from current wholesale costs, review the by-the-glass list against spoilage logs, and service the draft system.

Start this week with one move: run last week's pour cost by category, even roughly. If spirits are over 24%, weigh the well bottles daily for a week and watch the variance by shift. If draft is over 28%, log pints-per-keg on your two fastest taps. You will usually find the leak in seven days — and once the team knows the numbers are watched, some of the leak fixes itself.

Simplify Your Financial Management

Once pour costs, keg yields, and variance each have their own accounts and a weekly rhythm, the bar stops leaking margin in the dark. Beancount.io gives you plain-text accounting that is transparent, version-controlled, and AI-ready — so your category COGS, comp tracking, and inventory adjustments stay clean and auditable week after week. Get started for free and keep every ounce accounted for.

Share this article

Source: https://beancount.io/blog/2026/09/09/pour-cost-keg-yields-shrinkage-bar-bookkeeping-guide

Published: September 9, 2026