Chapter 26 — Discussion Guide

1. The chapter claims the POS is "the system of record" and that this matters more than any feature comparison. Is that overstated? What to listen for: students should be able to trace at least four specific metrics back to specific captured fields — menu mix from item-level sales, theoretical usage from items times recipes, sales per labor hour from timestamps, ticket time from the KDS, RevPASH from cover counts. The stronger move is noticing that the quality of the record is a training problem, not a software problem, which reframes the POS decision as a people decision. Watch for students who want to argue features; ask them which metric the feature produces.

2. Seventy-three cents of every processing dollar goes to banks you will never speak to. Does that make the industry's twenty-year fight over interchange a waste of effort? What to listen for: a genuine tension, not a slogan. Good answers separate collective action (worth supporting through an association, operates on a decade horizon) from individual attention (operates this quarter and is entirely in the operator's control). The best answers name the opportunity cost explicitly: an hour spent on the interchange fight is an hour not spent requesting your own interchange category detail. Push back on anyone who concludes the political effort is pointless — Case Study 26.2 also shows merchants won real, durable rights, including the ability to discount for cash.

3. Should Bellwether surcharge? Take the decision seriously in both directions. What to listen for: the arithmetic first (a compliant credit surcharge recovers a large fraction of \$43,573), then — and this is the better half — the measurement asymmetry. Recovery is countable; guests who quietly do not return are not. Push toward the variable that actually decides it: not the money, but whether the concept's whole thesis is the second visit. Then make them name the compliance steps their answer requires. Anyone who says "just surcharge, everyone does it" should be asked what state they are in and what the network rules say about debit.

4. A guest note reads: "older couple, husband is loud and slightly deaf, wife orders for both, big spenders, do not seat near the bar because he complains." Where exactly is the line? What to listen for: run this one live, without warning. Students split fast, and the split is instructive — some see operational usefulness, others see a dossier. Steer toward the over-the-shoulder test and then toward the harder question the test does not settle: who audits guest notes, and how often? The strongest answers notice that the note contains genuinely useful information (seating preference) buried in material that would be indefensible if read aloud, and rewrite rather than delete. This is a good place to connect Chapter 23's recognition systems to Chapter 21's culture material — what gets written in a guest record is a cultural artifact.

5. The chapter says switching a POS costs three years of the subscription, so you will accept the price increase. Is "lock-in" a market failure, a fair reward for building something sticky, or just the buyer's fault for not negotiating? What to listen for: students usually arrive at "the buyer's fault" and should be pushed past it. The useful distinctions are between lock-in that is inherent (your two years of sales history genuinely lives there and moving it is genuinely hard) and lock-in that is manufactured (proprietary hardware, 90-day auto-renewal notice windows, export in PDF only). The first is a cost of doing business; the second is a design choice. Ask what a well-functioning market would look like and whether any contractual term could produce it. Good answers land on the data-export clause as the one lever that converts manufactured lock-in back into inherent lock-in.

6. Bellwether's technology costs \$73,273 — more than utilities, more than insurance, more than marketing, and 28% of the plan's operating profit. Fifteen years ago most of this line did not exist. What did restaurants get for it? What to listen for: this is the synthesis question and it deserves the last fifteen minutes. Weak answers list features. Strong answers argue that the stack bought measurement — that a 1995 restaurant could not compute weekly prime cost, ticket time, or menu mix at any reasonable cost, and that the entire management system this book teaches is only economically possible because the marginal cost of recording a transaction fell to nearly zero. The best answers push back on that and note that a great many restaurants now pay 4.73% of sales for measurement they never look at — which is the most expensive possible outcome and, in my experience, the most common one. End there. It is the right note on which to close Part V.