Chapter 35 — Instructor Material

Teaching Notes

What this chapter is actually for

Chapter 35 opens Part VIII and it is the first chapter in the book whose answer is no. That is the pedagogical event. Every prior chapter taught a technique the student was meant to adopt; this one teaches a decision the student is meant to decline, and to decline with arithmetic rather than caution. Students who have absorbed thirty-four chapters of "here is how to do it well" arrive here primed to ask "how do I open the second one," and the chapter's job is to make them ask a different question and to make the different question feel like the ambitious one.

The single sentence to get onto the board in the first ten minutes: is the business profitable without you in it?

The three most common misconceptions

1. "Two restaurants means twice the money." This is the belief the chapter exists to dismantle, and it survives explicit refutation because it is arithmetically almost right — a second unit does add revenue and does add unit-level profit. Students accept the argument in class and reproduce the error on the exam. The reliable fix is to make them build the comparison themselves rather than read it: give them unit two's P&L alone and ask "is this good?" (they say yes, correctly), then hand them the four drag items one at a time. The moment the group cash figure goes below the single-unit figure is worth letting land in silence.

2. "The bench is a nice-to-have you build after you grow." Students consistently sequence this backwards, because in their experience organizations hire when the work arrives. The chapter's claim is that the bench must exist, in seat, for twelve months before signing. Expect pushback: "you can't afford a GM you don't need yet." That objection is correct and is exactly the point — §35.5 says plainly that building a bench makes a single unit less profitable for about eighteen months, and that this is the real price of optionality. Do not soften it. The honest teaching is that the cost is real and the alternative is worse.

3. "Cannibalization is a small percentage, so it's a small problem." Students model it as a uniform haircut across the week. The chapter's finding is that it concentrates on the nights with no waitlist, which are precisely the nights nearest break-even. Figure 35.7 is the cure and it is best taught by drawing the before-week on the board without the break-even line, then adding the line, then adding the transfers. The reaction when Tuesday drops from eleven-below to seventeen-below is the reaction you want.

A fourth, smaller one worth catching: students routinely conflate cushion and headroom. The 95/77/132 triple is the cleanest available drill. Ask the question cold — "how much room does this business have?" — and let half the room say 37.

The hardest point to teach

Owner-adjusted unit profit (§35.2). It is arithmetically trivial and conceptually slippery, and it fails to land for a specific reason: students who have never run a business do not viscerally understand that an owner's compensation is a residual, not a salary. The idea that a restaurant's reported 16.8% margin might be a repeatable 5.5% depends entirely on grasping that the owners are absorbing \$175,680 of management cost by working for whatever is left.

Two things help. First, make it personal and small before making it corporate: ask the room what their own labor would cost to replace at market, and what would happen to their household's "margin" if they had to buy it. Second, do not present the four-row fork as a table to read — build it live, one row at a time, and stop after each row to ask "so is this a business you would copy?" The answer changes between row two and row four, and watching it change is the lesson.

Expect the sharp student to ask why the chapter doesn't just say what the partners draw. The honest answer — that most real operators genuinely do not know, and that this is itself the finding — is worth giving.

A demonstration that works

The legal-pad exercise. Twenty minutes, no slides, and it produces the chapter's whole argument from the students rather than at them.

Put this on the board and nothing else:

  Second restaurant.  Revenue ~$1,240,000.  Operating margin ~10%.
  Debt service ~$85,000.
  So it adds about $42,000 a year.   Should we sign?

Take a show of hands. Most rooms split roughly evenly, with the yes votes slightly ahead.

Then hand out four index cards, one at a time, thirty seconds apart, and re-poll after each:

• Card 1: "Your first restaurant now needs a general manager. \$80,520 loaded."
• Card 2: "You now have a company. Bookkeeping, insurance, technology, a vehicle, recruiting. \$87,000."
• Card 3: "Thirty-six percent of your existing covers live closer to the new site. Net loss after backfill: 28 covers a week."
• Card 4: "Your Tuesday was already eleven covers below cash break-even. It's now seventeen."

By card three the room has flipped. By card four somebody usually asks, unprompted, what the money would do somewhere else — which is the cue for §35.7 and §35.8, and it is much better arriving as a student's question than as the instructor's next slide.

The transferable meta-lesson, and it is worth naming explicitly at the end: the answer changed because the analysis got wider, not because the numbers got worse. Every card was true before the first vote.

Timing

Roughly five to seven hours of student work; three contact hours if you are teaching it in blocks.

If you have only one session, teach §35.1, the legal-pad demonstration, and §35.8. That trio carries the argument. §35.6 is the most compressible section and §35.7's licensing subsection can be handed entirely to Chapter 36.

Assessment notes

Exercises 35.17, 35.19, 35.21, 35.23, 35.25, and 35.27 are the computational spine and are the right basis for a quantitative assessment; 35.23 is the single best exam question in the set because it requires the student to construct the counterfactual rather than be handed it. Exercises 35.32 and 35.36 are the best discussion-graded items. Exercise 35.41's stress test rewards students who separate fixed from variable costs and punishes those who apply a uniform haircut — a useful discriminator.

Watch for one specific error in grading: students who compute the group figure correctly but compare it to zero instead of to the single-unit outcome. That is the chapter's central error and it should cost real marks even when the arithmetic is flawless.