Chapter 15 — Key Takeaways
The core claims
-
The bar is what pulls blended cost below the food-cost line. Bellwether's 72/28 mix at 30% food and 22% pour gives a blended COGS of 27.8% and a prime cost of 60.1% — on target. Without the beverage subsidy the same restaurant runs 62.3%. The bar is not a nice extra; it is how the plan holds its line.
-
A beverage dollar returns 78 cents; a food dollar returns 70. Eight cents on \$434,000 is \$34,720 a year.
-
\$12.88 per guest is a target, not an observation. A four-point miss on beverage attachment costs \$46,624 of contribution — \$897 a week. It is the largest unverified number in the beverage plan.
-
Cost the whole drink or don't bother. Spirit, modifiers, citrus at tested yield, house syrup, bitters, garnish, ice, service items, spillage allowance. The four lines people omit are exactly the four that never appear on an invoice you would notice.
-
Split pour cost by category, always. Spirits 17%, wine 28%, beer 24%, non-alcoholic 12%. A mix shift produces a worse blended number from a better-run month, and reading only the blend sends you to interrogate a bar that did nothing wrong.
-
Your beer pour cost is mostly a purchasing decision. \$2.53 a glass from a sixth barrel and \$1.16 from a half barrel, same bartender, same night. No amount of careful pouring fixes a tap list priced wrong.
-
The leaks are the biggest quantifiable opportunity in the chapter. A half ounce of drift is \$7,631 — more than a full week of the bar's entire contribution. Five ordinary habits together are \$16,169, enough to push the whole restaurant past its prime-cost target with nobody stealing anything.
-
Recording a comp does not save a dollar. It buys visibility — which is the difference between a management conversation and a three-week investigation.
-
A discount removes a share of contribution margin, not of price. That is why one-third off requires 73.5% more volume just to stand still, and why below-cost pricing is a marketing spend rather than a price.
-
Responsible service sits above every number in this chapter. If your incentives push volume and your policy limits it, the incentives win. Design them to point the same way.
The key formulas
Pour cost % = beverage COGS ÷ beverage sales
Usage = beginning inventory + purchases − ending inventory
Liquor cost % = spirits COGS ÷ spirits sales (spirits only)
Blended COGS = (food share × food cost %) + (bev share × pour cost %)
Bottles on hand = full bottles + Σ tenths
Draft yield = sellable ounces ÷ keg ounces
Cost per sellable = keg cost ÷ (keg oz ÷ pour size ÷ (1 + loss adjustment))
glass more simply: keg cost ÷ (theoretical glasses × (1 − loss %))
Menu price = drink cost ÷ target pour cost %
Break-even = original CM ÷ discounted CM
multiplier
The numbers worth memorizing
| 750 mL | 25.4 fl oz |
| 1 liter | 33.8 fl oz |
| 1.75 liter | 59.2 fl oz |
| Half barrel | 15.5 gal = 1,984 oz |
| Quarter barrel | 7.75 gal = 992 oz |
| Sixth barrel | 5.16 gal = 660.5 oz |
| Typical draft loss | 8–15% (Bellwether plans 12%) |
| Line cleaning cadence | every two weeks |
| Shake dilution | ~20–25% of volume; stir 15–20% |
| Tenths-method accuracy | ±5% per bottle; random, not systematic |
| Variance to investigate | >1.0 point, or a category >10% over ideal, two weeks running |
The Bellwether beverage plan, on one line each
- Beverage 28% of \$1,550,000 = **\$434,000 · pour cost 22.0% = \$95,480 · contribution \$338,520 = \$6,510/week**
- Rivermill Sour: \$3.20 cost · \$15.00 price · 21.3% pour cost · \$11.80 contribution
- Categories: spirits \$173,600 @ 17.0% · wine \$164,920 @ 28.0% · beer \$69,440 @ 24.2% · N/A \$26,040 @ 11.5%
- Draft: six lines, 14 oz pour, ~98 kegs a year, 12% loss, biweekly cleaning at ~\$1,560
- Constraint carried: no glasswasher until year two — \$1,273/yr of barback labor, 5.8-year payback
- Happy hour proposal: Tue–Fri 4:30–6:00, bar and patio, +\$747.92/week ≈ \$38,900/year, subject to regulatory verification and an eight-week cannibalization measurement
What you should be able to do Monday morning
Walk behind your bar with a count sheet, value every open bottle in tenths, compute last week's pour cost by category from the usage formula, compare it to what your cost cards say it should have been, and know — in dollars and in ounces — exactly how big the gap is and which of five ordinary explanations you have to rule out before you say a word to anybody.
Then post a written pour standard, put a jigger in a dedicated slot, put a transfer clipboard by the walk-in, and put a reason code on every drink that leaves without a sale.