Ch15 Discussion

Discussion Guide

Prompt 1 — "The bar is what holds the plan on its line." Ask: Bellwether's prime cost is 60.1% with the bar and 62.3% without it, and nothing about the kitchen changes between those two numbers. What does that imply about a chef-owner who says "I'm not really a bar person, we'll keep it simple"?

Listen for: students recognizing that "simple" is fine and "uncosted" is not — that the objection is not to a small program but to an unexamined one. Push anyone who treats this as an argument for a bigger bar; §15.5's short-list discipline argues the opposite. The best answers connect it to Chapter 1's leverage point: in a four-to-six-point margin business, 2.2 points is not a detail.

Prompt 2 — The comp that costs nothing to record. Ask: If ringing a comp with a reason code does not change net sales, COGS, or pour cost by a single dollar, why is it the first thing this chapter tells you to do?

Listen for: the distinction between the money and the knowledge. The strong answer is that an unrecorded comp is indistinguishable from over-pour, spillage, and theft, so it converts a solvable management question into an unsolvable investigation. Extend it: what else in the book works this way? (Waste logs in Ch.13, the transfer sheet in §15.2, the shift log in §15.7, the DSR in Ch.34.) The generalization — controls buy visibility, not savings; the savings come from what you do with the visibility — is worth writing on the board.

Prompt 3 — When the control does not pay. Ask: Line cleaning returns \$1,129 against \$1,560. Should you skip it? Now give me a rule for when you fund a control that fails its own payback calculation.

Listen for: students identifying the unmeasured term (the guest who quietly stops ordering draft) rather than just asserting "quality matters." The sophisticated answer distinguishes controls that reduce cost from controls that protect revenue and notes that the second category is chronically underfunded because its benefit never appears on a P&L line. Push back on anyone who says "just do everything" — the \$936 in-house alternative is in the chapter precisely so the decision stays real.

Prompt 4 — A quarter ounce, and what you say on Monday. Ask: Week 19 shows spirits \$76 over ideal — about a quarter ounce a drink. Your bar has three bartenders and one of them is new. What do you do first, and what do you not do at all?

Listen for: the sequence from Exercise 15.37 — check the shift mix, check the station, check the training — before checking the person. Listen especially for anyone who proposes a staff meeting on one week of data; the chapter says do not, and the reason (one week is noise plus a hypothesis) is worth drawing out. A good extension: what would three consecutive weeks change? What would a fourth?

Prompt 5 — The discount you cannot make up on volume. Ask: A \$1 margarita costing \$1.60 has a negative contribution margin, so the break-even multiplier is undefined. Is the promotion therefore irrational?

Listen for: the reframe — it is not a price, it is a marketing spend denominated in tequila, and it must be accounted for and judged as one (cost per cover acquired, food attachment, incrementality). Then push to the design question: every rule in a well-built dollar-drink program (dine-in only, one drink, all month, food messaging) exists to force the attachment. Ask what an independent could steal from that architecture without ever running a dollar drink.

Prompt 6 — The happy hour that worked too well. Ask: In the composite, the happy hour produced \$52,000 of incremental contribution and \$112,320 of check erosion. The operator did nothing wrong tactically. What did they get wrong?

Listen for: scope, not depth. The window ran to 7:00 — thirty minutes into peak — and ran in the dining room, not just the bar. The shoulder-hour logic only works on seat-hours that were otherwise worthless. The best answers name the measurement that would have caught it (the adjacent 7:00–9:00 window, tracked against a comparable prior period) and observe that it is the one thing nobody looked at. Close the discussion by asking what they would do now, eighteen months in, when guests have already learned the schedule — there is no clean answer, which is the point.

Prompt 7 — The line that outranks the arithmetic. Ask: This chapter says responsible service "is not a cost-control topic and never becomes one," and that if your incentives push volume while your policy limits it, the incentives win. Design a beverage program in which the incentives and the policy point the same direction.

Listen for: concrete mechanics rather than sentiment — fixed prices rather than rounds, long windows rather than compressed ones, food available throughout, per-guest limits, pace tracking, a written refusal protocol, and management publicly backing a cutoff. Push anyone whose answer is only training: training that fights an incentive structure loses. Ask directly: what does a manager do in the four seconds after a bartender cuts someone off and the guest complains? The answer to that question is the culture, and it is worth ending the session on.