Chapter 22 — Instructor Material

Front of House Operations: Service Flow, Table Management, Reservation Systems, and Guest Experience

Part V opens here. Chapter 22 is the mechanics chapter: it teaches how a room is physically run and it resolves a capacity question the book has carried since Chapter 1. Chapter 23 takes hospitality, Chapter 24 takes revenue management. Keep those boundaries visible in class or students will try to solve everything at once.

Teaching Notes

Where this chapter sits

This is the third leg of a three-chapter capacity argument that most students will not have noticed is running.

• Chapter 7 drew the floor plan and said the room holds about 151.
• Chapter 14 measured the hearth and said the kitchen caps at about 144 — and explicitly handed pacing forward with the line "an hour is the wrong unit for a station that behaves like a queue."
• Chapter 22 answers it: the operating ceiling is about 132, by three independent routes, and both published ceilings assume a demand curve that has never existed.

If your students have read 7 and 14, open by putting 151 and 144 on the board and asking which one is the business. Most will say 144. The correct answer is neither, and the chapter is the derivation.

The chapter is also the last one before the book's hardest reframe. Chapter 23 will argue that the second visit is the entire profit model. Chapter 22's job is to establish that a guest whose entrées arrive thirty-six minutes late has had a bad night no matter how warmly anyone spoke to them — which is what makes Chapter 23 a consequence rather than a sentiment.

The single hardest point to teach

That a queue does not catch up during a rush.

Students — and most working managers — carry an intuitive model in which a busy period is followed by a catch-up period. It feels true because it is true when the average rate exceeds average demand over the whole night, which it usually does. What they miss is that the catch-up happens at the slack rate, and the slack rate during service is nearly zero.

Bellwether's on-plan Friday makes the point cleanly and I recommend walking it on the board line by line:

  FIRE HOUR      DEMAND   CAPACITY   SLACK      BACKLOG AT HOUR END
  5:30–6:30        18        29       +11              0
  6:30–7:30        30        29        −1              1
  7:30–8:30        36        29        −7              8      ← 14 minutes
  8:30–9:30        28        29        +1              7      ← still 14 minutes
  9:30–10:30        8        29       +21              0

The row students always mis-predict is the fourth. They expect the backlog to clear because the hour is "under capacity." It is under capacity by one cover. Fourteen minutes of delay survives to 9:30 and is paid by guests who arrived after the problem was created. That is the moment the concept lands, and it lands better on the board than on the page.

Second hardest, and closely related: that re-ordering a queue changes nothing. Figure 22.8 proves it — splitting the 40-top leaves the à la carte backlog at 26 covers either way — but students resist it because splitting feels like doing something. The unlock is the priority assumption: the party is contracted and the à la carte tables are not, so the party fires first in every window. Make students state that assumption out loud before they look at the numbers. Once they hold it, the identical result is obvious rather than surprising, and they have learned something durable about queues generally.

Common misconceptions

1. "We were only at 120 covers, we had loads of room." The chapter's central correction. Total covers describe aggregate fit; kitchens experience sequence. Drill it with Exercise 13, which reproduces the error on a different restaurant with different numbers so students cannot pattern-match to Bellwether.

2. "Turn time means turning tables faster." Students hear "turn time management" as "rush people." Figure 22.5 exists to break this: of a 113-minute cycle, 82 minutes are dinner and 19 are logistics. Put the segment table up and ask which rows the restaurant owns. The rule — you may take minutes from the restaurant; you may not take minutes from the guest — should be on the board for the whole session.

3. "The host is an entry-level job." Structurally true in most restaurants and analytically false. Ask the class to name the position whose decisions move the most revenue per hour worked. They will say server, then chef, then GM. The chapter's answer is the podium, and §22.3's arithmetic supports it.

4. "Rotation is about fairness." It is, but framing it that way makes it negotiable. Frame it as retention economics: \$145 a shift, roughly \$14,500 a year between Section 1 and Section 4, in an industry running ~75% turnover (Chapter 17). Rotation is cheaper than replacing the server who did the arithmetic.

5. "Splitting a big party into waves helps the kitchen." Universally believed, demonstrably false under a party-first priority rule. See above.

6. "The 40-top is just 40 more covers." It is 40 covers and 32 of 56 dining seats and 75 minutes of pacing inventory. Students who work the frozen Friday properly discover that a 142-cover book was never seatable — physical ceiling about 116 without the 9:00 reseat — which is a genuinely startling result and the best single argument in the chapter for doing arithmetic before saying yes.

7. Confusing seat turns, table turns, and cycle. Three numbers, one restaurant: 1.40 blended seat turns, 1.36 dining-room seat turns, 1.72 table turns. Insist students say which one they mean, every time. The business plan quotes the least useful of the three.

A demonstration that works

The seating simulation. Forty-five minutes, no materials beyond index cards, and it produces the chapter's findings experientially before students read the arithmetic.

Setup. Draw Figure 22.2's 17 tables on the board. Give one student the host stand and a stack of cards, each card a party (write party size and arrival time on it — build the deck to Figure 22.3's 120-cover Friday shape). Post the hearth's rate visibly: 29 covers an hour, and no more. Give a second student the pass: they tick off 29 covers an hour, no exceptions, and announce the ticket time every simulated half hour.

Round one — no rules. Let the host seat freely. They will fill the room by 7:30 and the pass will be twenty minutes behind by 8:00. Let it happen. Do not intervene.

Round two — same deck, one rule: eight covers per quarter hour. Same 120 covers get seated. The pass never exceeds about six minutes. Students see identical revenue and a completely different night.

Round three — insert the 40-top. Drop a card reading "PARTY OF 40, entrées fire 7:00" into the deck and let them try to save it. They can't, and discovering that is the point. Ask what they would have done at the moment of booking — which is where the memo in Exercise 29 comes from.

Debrief question: "In round one, at what moment was the night already lost?" The answer is about 6:45, roughly forty minutes before anyone at the pass could see it.

Timing

A single 3-hour session, or two 90-minute sessions split at §22.5.

If you have only 90 minutes: §22.3, the simulation (rounds one and two), and the Business Plan checkpoint. Everything else can be reading.

Assessment notes

Best short-answer items: Quiz 3, 5, 13, 18, 24. Quiz 5 (wave-splitting) and Quiz 13 (the three ceilings) discriminate well between students who computed and students who read.

Best problem set: Exercises 13, 17, 21, 27. Exercise 27 is the strongest single assessment in the chapter — it presents a table where every visible trend is good except two, and a student who congratulates the owner has failed regardless of how well they write.

Best written assignment: Exercise 29 (the pacing memo). It is short, it has a hard word limit, and it cannot be answered without the arithmetic. It also fails informatively: memos that wander into event pricing show a student who has not internalized chapter boundaries.

Capstone-adjacent: Exercise 30 is the actual Business Plan deliverable. Assign it if your course carries the progressive project; the "what would change my mind" paragraph is where the good answers separate.

A note on the Business Plan verdict

Students will want the chapter to either bless 1.40 or replace it, and it does neither cleanly. This is deliberate and it is worth defending in class. The chapter's finding is that 1.40 is a defensible forecast and a useless operating target, and that the right response to a misleading number is sometimes to keep it and add the number you should actually manage with.

Anticipate the objection: if the managed case is worth \$57,408, why not forecast it? The answer is conditionality — the upside exists only where demand exceeds the book, and no year-one plan should forecast a green floor team hitting a managed peak on the two nights that matter most. A student who argues the other way and handles the conditionality honestly should get full credit; this is a genuine judgment call and the book says so.

Sensitive material — please read before teaching §22.5

The efficient/rushed discussion touches real labor conditions. Turn-time pressure in restaurants is frequently experienced by servers as management pressure delivered downward with no corresponding authority — a server told to turn a table has no ability to make the kitchen faster or the check settle sooner, so the only lever they are left with is hurrying the guest.

Teach the chapter's rule as a protection for the server, not just for the guest. "You may take minutes from the restaurant; you may not take minutes from the guest" gives a server a defensible line to hold against a manager who wants tables turned, and it directs the pressure where it belongs: at the check settle, the reset, and the reservation book. If you have students who have worked the floor, they will have stories here. Make room for them; they are the best teaching material in the room.