Chapter 4 — Teaching Notes

What this chapter is doing

Chapter 4 is where the course turns from description to construction. Chapters 1 through 3 told students what a restaurant is, what a concept is, and what a brand is. Chapter 4 makes them build something that can be wrong.

That is the pedagogical point and it should shape how you teach it. The deliverable is not a forecast. It is a register. A student who leaves this chapter with a beautiful spreadsheet and no assumptions register has learned the wrong half. A student who leaves with a crude forecast and a rigorous register has learned the right one, and will produce the beautiful spreadsheet later without help.

The chapter's spine is a single deliberate tension. Chapter 1 built a bottom-up estimate for Bellwether from four variables and got \$1,410,760**. The plan's headline is **\$1,550,000. The gap is \$139,240, and the chapter refuses to resolve it — it names it, itemizes it, registers it as the lowest-confidence row in the plan, and then shows in the sensitivity analysis that this single unproven assumption carries about 27% of the plan's operating profit.

Do not let students "fix" the gap. Several will try, and the attempt is the most teachable moment available to you. See the misconceptions below.

Timing

  • A 75-minute session: §4.3 built live on a projected spreadsheet (25 min) → the gap and the bridge, Figure 4.3 (15 min) → the register, §4.4 (20 min) → the tornado, Figure 4.6 (15 min). Assign §4.1, §4.2, §4.5, and §4.7 as reading. This ordering works because the arithmetic earns the document, rather than the document introducing the arithmetic.
  • Two sessions: split after §4.4. Session two opens by having students bring one register they wrote, then does §4.5 and §4.6 in full and closes on the executive summary. The executive summary is far better taught after the sensitivity work, for exactly the reason the chapter gives.
  • Self-paced: 5–7 hours including exercises. Exercise 16 and Exercise 26 are the minimum computational set.

Common misconceptions

1. "The plan says \$1,550,000, so the forecast should say \$1,550,000." This is the big one, and students reach for it immediately. They will adjust turns from 1.4 to 1.57, watch the arithmetic land on the headline, and feel that they have solved something. They have committed the exact error Case Study 2's pattern-one composite demonstrates, and they have done it in under a minute.

Teach it by letting them do it. Have the room solve for the turns figure that closes the gap (1.57). Then ask: who decided 1.57? What evidence produced it? The answer is "the number we wanted," and the moment that lands is the moment the register becomes obviously necessary rather than bureaucratic.

2. "A conservative forecast is a good forecast." Students conflate honest with low. A forecast deliberately depressed to look prudent is just as unattributed as one inflated to look attractive, and it has a specific operational cost: you build a cost structure to a number you do not believe, and then you are understaffed on a Saturday. The correct standard is not conservative or aggressive. It is traceable.

3. "Occupancy is 6.1%, so if we can just get it to 5.5% we'd be better off." Students read every percentage on a P&L as a lever. Occupancy is $95,200 in dollars and does not move; the percentage moves only when revenue moves. This confusion is worth an explicit board exercise — put $95,200 over three different revenue figures and let them watch a "cost improvement" appear from nothing. It also sets up the operating-leverage discussion in §4.5 and pays off again in Chapter 32.

4. "A point is a point." Students will compare a point of food cost to a point of pour cost to a point of labor as though they were commensurable. They are $11,160, $4,340, and $15,500 respectively, because the bases differ. Exercise 19 exists for this and should be assigned to everyone, not just to the students who like arithmetic.

5. "Naming the risks weakens the ask." Nearly universal, and it survives the reading. The counter-argument in §4.7 is logical, and logic loses to instinct here. What works better is the role-reversal in the discussion guide below: make them be the underwriter.

6. "Sensitivity analysis gives you a range of likely outcomes." It does not; it gives you a ranking of consequences. The distinction matters because a range invites students to average the scenarios into a "realistic case," which is exactly the wrong output. Emphasize that consequence and probability are different columns and the register carries both.

The hardest point to teach

That the largest risk in a plan can be a number that isn't in the plan.

Students can compute all seven bars of the tornado in Figure 4.6 and still not see why the top bar — "bridge revenue never materializes, −\$69,954" — is a risk at all. It does not appear as a line on the pro forma. It lives in the difference between two numbers on two different pages, which is precisely why it was invisible until the register put it in a row.

What works: print the year-one pro forma alone and ask the class to find the plan's biggest risk in it. They will pick food cost or labor, because those are the visible lines. Then hand them the bottom-up worksheet from Figure 4.2 and ask again. The realization that the largest exposure was not findable from the financial statement — that you needed the build to see it — is the whole chapter compressed into two handouts.

Follow it with the general principle: a financial statement shows you the conclusions; only the assumptions show you the risks. Every later money chapter depends on students having internalized this, and Chapter 31 will assume it.

A demonstration that works

Break the plan live, in a spreadsheet, in front of them.

Build the model on a projector from the §4.6 split: fixed \$517,700, variable 49.76%, so contribution 50.24¢. Show it reproducing \$261,020 exactly at \$1,550,000 of revenue. That reproduction is worth lingering on — students who watch a model land on the frozen number to the dollar trust the rest of the session.

Then take requests. Let the room call out changes — "drop turns to 1.25," "raise labor three points," "lose the patio" — and put each result on the board next to the plan. Six or seven of these takes ten minutes and produces the tornado organically, and students remember a ranking they generated far better than one they read.

The turn that makes the session: after five or six single-variable moves, ask them to pick the three most likely to happen together. They will name the combination in §4.6 (soft turns, soft check, high labor) without being led there, because the causal link is intuitive once you ask. Then run it: \$154,854, 11.4%. Now ask whether that reassures them.

The strongest students will get to the §4.6 answer on their own — that an 11.4% "downside" sitting above the industry-typical range means the downside case is not really a downside case. When someone says it, stop and let the room sit with it. That is a graduate-level reading of a business plan and it happened in week four.

Assessment notes

  • Exercises 16 and 19 are the computational core. A student who can complete a pro forma that foots and compute exposure per point can follow the rest of the book's money chapters.
  • Exercise 26 (ranking four adverse moves) is the single best diagnostic in the set. Watch for students who notice that ranks 2 and 3 are within \$930 of each other, and for those who add the unprompted observation that consequence is not probability.
  • Exercise 13 is the most instructive to grade, because the bridge deliberately does not close — three plausible items reach \$86,880 of a \$107,600 gap. Students who invent a fourth item to force it to zero have missed the point; students who name the \$20,720 residual and offer three honest options have got it.
  • Exercise 30 (rewriting the risk statement) is the best writing prompt in the chapter and works well as a short graded assignment. Look specifically for the trigger — the condition at which action is taken. Most students supply a consequence and an action and omit the trigger, which is the part that turns a paragraph into a management system.
  • Exercise 34 (the lease escalation) rewards students who go past the arithmetic to the three structural reasons the three-year figure understates the problem. The compounding one is findable; the "escalation and revenue miss are not independent risks" one separates the top of the class.
  • Exercise 32 is the best group exercise: assign different categories to different teams and let them discover independently that "other operating, 14.0%" was concealing eight separate assumptions.

Connections forward

Flag these explicitly so students build the map: the register → every subsequent Business Plan checkpoint, and finally Chapter 40, which assembles the whole document. The turns and check assumptions → Chapters 22 and 24. The labor line → Chapters 17, 19, and 20. The bridge's three channels → Chapters 24, 28, and 29. The fixed/variable split and the break-even preview → Chapter 32. The ramp and the working-capital reserve → Chapters 9 and 33. The occupancy escalation → Chapter 6. The ask → Chapter 5, immediately.

Tell them plainly that the register they start now is a live document they will edit in fifteen more chapters. Students who treat it as a one-time assignment get much less out of Part VII.