Chapter 7: Designing the Restaurant: Floor Plan, Kitchen Layout, Equipment, and Capacity

"Every foot you give the dining room you take from the kitchen. Then you spend ten years living with whichever one you got wrong." — constructed; a kitchen designer, holding a tape measure

Overview

Chapter 6 ended with a signature and a problem. The signature committed two people to \$95,200 a year for ten years, personally guaranteed. The problem was found on day nine of due diligence, when a mechanical contractor looked up at a hood the previous café left behind, asked whether the wood-fired hearth was real, and then said the thing that costs money.

This chapter pays that bill. It also does something larger, and if you take only one idea out of Part II take this one: design sets your revenue ceiling and your labor cost before you hire a single person. Every number in the Chapter 4 forecast — 68 seats, 1.4 turns, 95 covers, a \$46 check — is a claim about a physical room. Either the room seats those people at that price or it doesn't; either the kitchen sends those plates in the ninety minutes when they are all ordered at once or it doesn't. A forecast the building cannot physically produce is not an optimistic forecast. It is a fictional one, and no amount of marketing or discipline later will fix a room that was drawn wrong.

So: two jobs, which are the same job seen from two ends.

The first is capacity — turning 2,800 square feet into 68 seats you can defend to a lender, a 16-seat patio that turns out to be carrying \$64,800 of the plan's revenue, and a kitchen that can put fifty plates through a pass in the busiest hour of a Saturday in July.

The second is the money. The hood, the make-up air unit, and the grease interceptor were carried in Chapter 6's construction contract as two allowances totaling \$26,000. The real scope comes in at \$75,000, spread across three trades, one of which nobody carried a dollar for. The \$310,000 construction line does not grow — it is one of five frozen components of a \$620,000 project — and the contingency is \$9,000, which is not enough to have this conversation with. So we do what an operator actually has to do: put the real number on the page, then decide, out loud and in order, what gets cut to pay for it.

That second exercise is the most useful thing in Part II. A budget is not a forecast. It is a set of promises you will have to break in a specific order, and the whole skill is knowing the order before the phone rings.

In this chapter, you will learn to:

• Compute seats from square feet, defend a density against a check average, and state the revenue ceiling your floor plan sets.
• Lay out a front of house — entry, host stand, bar, seating mix, service stations, restrooms, and an accessible route — and account for every square foot of it.
• Lay out a back of house around a single direction of travel, and prove the line produces the peak-hour plate count rather than the comfortable nightly average.
• Count steps, convert them to hours, convert the hours to dollars, and decide a layout question with a payback period instead of an opinion.
• Write an equipment schedule that functions simultaneously as a purchase order, a permit document, and a capacity statement.
• Size the invisible systems, discover the number is wrong, and re-cut a fixed budget honestly.

Learning Paths

🏗️ Opening — all of it, and §7.6 twice. This is the chapter where an abstract budget becomes a set of physical objects with lead times. Do the capacity arithmetic in §7.1 and the re-cut in §7.6 by hand. 📋 Managing — you inherit a room you did not draw. Weight §7.4 and §7.7: the fastest improvements available to a manager are almost always circulation and station placement, they cost a few hundred dollars, and nobody else in the building is looking at them. 🍸 Beverage — §7.2's bar geometry is yours, and it is worth more to you than any cocktail recipe: the working bar's dimensions set your speed, and speed sets your pour cost variance and your service standard. Note in §7.5 which piece of bar equipment got deferred and what that costs. 🚚 Small Format — a truck is a floor plan with the walls closer together, and every constraint in §7.3 and §7.6 applies to you with less room to solve it. §7.6's hood and grease sections apply to commissaries and ghost kitchens too, and §7.5's used-equipment discussion matters more to you.

7.1 Capacity math: square feet, seats, turns, and the revenue ceiling design sets

Start with vocabulary, because two terms run through the whole chapter and everything else hangs off them.

The front of house (FOH) is every part of the restaurant a guest may occupy or see from a seat: entry and vestibule, host stand and waiting area, dining room, bar, service stations, guest restrooms. The back of house (BOH) is everything else: the cooking line, prep, the dish pit, walk-in and dry storage, receiving, office, staff areas. That division is not merely descriptive. It is the fundamental design trade in the building, because every square foot is in exactly one of them, and the two want the same footage for opposite reasons.

Bellwether's split, frozen by the lease:

The common rule of thumb for full service puts FOH at 60–70% of the building and BOH at 30–40% (Tier 2 — trade convention, not a measured statistic, and it varies enormously with menu). Bellwether at 61/39 sits at the kitchen-heavy end of the band, and that is a decision, not an accident. A scratch kitchen with a wood-fired hearth, a short seasonal menu, and house butchery needs production space. The seats you give up to get it are the ones you were never going to fill on a Tuesday.

Square feet per seat

The seat-to-square-foot ratio is the total building area divided by the total seat count — or, in its more useful inverse, the square feet a designer allocates per seat within a specific zone. Operators quote it both ways and confuse each other constantly, so state which you mean.

Bellwether, both ways:

$$\frac{2{,}800 \text{ sq ft}}{68 \text{ seats}} = 41.2 \text{ total square feet per seat}$$

$$\frac{1{,}700 \text{ sq ft of FOH}}{68 \text{ seats}} = 25.0 \text{ square feet of front of house per seat}$$

Neither of those is the number a designer works to, because FOH is not all dining room. Here is the honest allocation.

Check the column: 130 + 140 + 200 + 890 + 60 + 190 + 90 = 1,700. ✓

The number that matters is 15.9 square feet per dining seat. Industry planning ranges, given as ranges because they are rules of thumb and not laws:

At 15.9, Bellwether is squarely in upscale casual. That is the correct band for a \$46 check, and here is why it is not a matter of taste.

🧮 Run the Numbers

The eighteen seats that aren't there — and why adding them loses money.

Bellwether's dining room is 890 square feet holding 56 seats. Compress to 12 square feet per seat — ordinary casual density, the spacing of a good neighborhood bistro — and the same room holds:

$$890 \div 12 = 74 \text{ seats}$$

Eighteen more seats, at a construction cost of essentially zero, and the upside arithmetic is intoxicating: 18 seats at 1.4 turns is 25 more covers a night, \$1,159 a service, \$5,795 a week, **\$301,340 a year** on a business whose entire operating profit on plan is \$261,020.

Three reasons it is wrong, in ascending order of cost.

One: the kitchen cannot make it. 900 square feet of back of house and one hearth produce a hard ceiling we compute in §7.3. More seats do not produce more plates; they produce longer ticket times, which is how a restaurant burns its reputation in month four.

Two: the seats would not fill. Chapter 2 established that Bellwether is taking share in a supplied market — its 68 seats are a ~40% increase on a 172-seat direct-occasion base. Adding capacity does not create demand; it creates a bigger room with the same people in it, which reads as empty. At 1.4 turns on 74 seats you need 104 covers to look as busy as 95 looks now.

Three, the expensive one: the density is the check average. Chapter 2 built the \$46 check from an occasion mix — 46% neighborhood dinner at \$41, 30% special occasion at \$57, 14% at \$36, 10% at \$50. A 12-square-foot-per-seat room, where a two-top hears the next table's whole conversation, does not sell a \$57 special-occasion check. Lose four dollars of average check across 36,140 covers and you have lost \$144,560 a year — half the upside, before the kitchen fails.

Seating density is a pricing decision wearing a furniture costume. Decide it against your check average, not against your square footage.

Turns, and the average nobody experiences

Seats are half of capacity. The other half is turns — how many times you rent each chair in a service. Chapter 22 owns turn time and the mechanics of moving a table; Chapter 24 owns yield management. What this chapter owns is the physical ceiling underneath both of them.

Bellwether's plan: 95 covers a night at 1.4 turns on 68 seats. Check it: $95 \div 68 = 1.397$. ✓

But 1.4 is an average across five nights, and no night is average. Here is the shape a neighborhood dinner house of this type actually runs — and this table is the reason the kitchen must be designed to something other than 95.

FIGURE 7.1 — The week the average hides                  [the Bellwether plan — constructed]

  68 seats, dinner Tuesday through Saturday. Bar length = covers.

  NIGHT        COVERS   TURNS   ▁▂▃▄▅▆▇█
  ────────────────────────────────────────────────────────────────────────
  Tuesday         62     0.91   ██████████████
  Wednesday       78     1.15   ██████████████████
  Thursday        92     1.35   █████████████████████
  Friday         120     1.76   ███████████████████████████
  Saturday       123     1.81   ████████████████████████████
  ────────────────────────────────────────────────────────────────────────
  WEEK           475     1.40 avg          475 ÷ 52 weeks = 24,700 dinner covers
  Brunch      2 × 110     1.62   (a shorter meal turns faster)  11,440 covers
  ────────────────────────────────────────────────────────────────────────
  ANNUAL COVERS                                                    36,140

Read that carefully, because it contains the chapter's central design instruction. The plan says 95. The building has to do 123, and on twenty weeks of the year it has to do 123 plus a patio. If you design the kitchen, the dish pit, the walk-in, and the service stations to 95, you have designed them to a number that occurs on approximately no nights of the year. Tuesday is under-loaded and Saturday is over-loaded, and it is Saturday that breaks things.

Check the arithmetic: 62 + 78 + 92 + 120 + 123 = 475, and $475 \div 5 = 95$. ✓ $475 \div (68 \times 5) = 475 \div 340 = 1.397$. ✓ And $475 \times 52 = 24{,}700$ dinner covers, plus $110 \times 2 \times 52 = 11{,}440$ brunch covers, for the 36,140 annual covers Chapter 2 established and Chapter 6 divided into \$95,200 of rent to get \$2.63 a guest. ✓

The patio is not decoration

Sixteen seats sit on the sidewalk in front of the building from roughly May to October. It is easy to read them as a pleasant extra. They are not.

Chapter 4 built a \$139,240 revenue bridge between Chapter 1's four-variable estimate (\$1,410,760) and the \$1,550,000 forecast. The patio carries:

$\$64{,}800 \div \$139{,}240 = 46.5\%$ of the entire revenue bridge, from sixteen chairs that are unusable for about thirty-two weeks a year. It is 4.2% of total forecast revenue. If those seats do not perform, the plan does not reach \$1,550,000 and Chapter 6's 6.1% occupancy becomes 6.4%.

Four design consequences follow, and every one gets skipped in first drafts:

1. A service path to the patio that does not cross the entry. A server carrying four entrées through the door people are arriving in is a collision and a cold plate.
2. A point-of-sale terminal and a water source within reach, or the patio becomes a station with a ninety-foot round trip and the section nobody wants.
3. Storage. Sixteen chairs, four tables, heaters, and umbrellas have to live somewhere for six months, and Bellwether has 200 square feet of storage and office in total.
4. The permit. Sidewalk-café approvals, fees, and whether the seats count toward occupant load are municipal questions with municipal timelines. Chapter 8 owns them; know now that they exist.

Those seats also carry essentially no additional occupancy cost — the rent is already paid, and the only marginal charge is a sidewalk-café permit fee. Per cover they are the highest-contribution sixteen seats in the business, which is exactly why they deserve engineering rather than an afterthought.

🧾 Read the Numbers

```text FIGURE 7.2 — "The capacity statement" [the Bellwether plan] THE ARTIFACT The one-page capacity summary that belongs in front of the floor plan in a business plan: seats, density, turns, covers, and the revenue the room can physically produce. THE CONTEXT 2,800 sq ft second-generation space, Rivermill District, after the lease is signed and before the drawings are finished. Dinner Tue–Sat plus weekend brunch. All figures constructed for teaching.

                 SEATS
                   Dining room                    56    15.9 sq ft/seat
                   Bar                            12    11.7 sq ft/stool
                   ─────────────────────────────────
                   INTERIOR SEATS                 68    25.0 sq ft of FOH per seat
                   Seasonal patio (sidewalk)      16    ~20 weeks a year

                 TABLE MIX (dining room)
                   8 two-tops                     16 seats   (combine to four-tops)
                   7 four-tops                    28 seats
                   2 six-tops                     12 seats
                   ─────────────────────────────────
                   17 tables                      56 seats

                 COVERS
                   Dinner, average night          95        1.40 turns
                   Dinner, Saturday              123        1.81 turns
                   Brunch, each weekend day      110        1.62 turns
                   Annual covers                 36,140

                 REVENUE THE ROOM PRODUCES
                   Dinner  24,700 × $46                   $1,136,200
                   Brunch  11,440 × $24                   $  274,560
                   ─────────────────────────────────────────────────
                   From seats and turns alone            $1,410,760
                   Revenue bridge (Ch. 4)                $  139,240
                     of which the patio                    $64,800
                   ─────────────────────────────────────────────────
                   YEAR-1 FORECAST                       $1,550,000

WHAT IT SHOWS A room whose density (15.9 sq ft/seat) is consistent with its price point, a table mix weighted to two- and four-tops that matches a neighborhood party-size distribution, and a physical capacity that reaches $1,410,760 without heroics. It also shows, in one line, exactly how much of the plan rests on things the four walls do not produce: $139,240, of which $64,800 is a sidewalk. WHAT IT DOESN'T It does not prove the kitchen can send these plates — that is §7.3, and it is a different question with a different answer. It does not prove demand exists at 1.4 turns; Chapter 2 says the market is supplied and the plan is taking share, and Chapters 22 and 24 have to earn the turns. It says nothing about the weather, which owns the $64,800. And it assumes every seat is sellable, which no room's seats ever entirely are — the two-top by the restroom door is a seat on this sheet and a complaint in service. THE DECISION Publish this page in the plan, directly facing the floor plan, and put the $139,240 line on it in the same type size as everything else. A lender who finds that gap themselves reads the whole plan differently than one who is handed it. THE LESSON A floor plan is a revenue forecast drawn to scale. Every seat you draw is a promise to fill it, and every seat you don't draw is a ceiling you have agreed to live under for ten years. ```

7.2 The front of house: entry, host stand, bar, seating mix, service stations, restrooms, ADA

Now the room itself. Bellwether's measured interior is about 34 feet wide and 82 feet deep — the 34-foot frontage Chapter 6 read off Mill Street, running back to the alley. The front 50 feet is front of house (34 × 50 = 1,700), the next 26.5 feet is the kitchen (34 × 26.5 ≈ 900), and the back 6 feet is storage, office, and receiving (34 × 6 ≈ 200). That is the whole building, and the proportions are worth holding in your head: the guests get the front sixty percent, the food gets the back forty, and the wall between them is where the money is made.

FIGURE 7.3 — Bellwether, front of house: 1,700 sq ft, 68 seats   [the Bellwether plan]

  Schematic; not to scale. ~34 ft wide, ~50 ft deep. Street at the bottom.
  KEY:  [] table    o seat    ### fixed millwork    /\ service station    > door

  #############################  T H E   P A S S  ##########################
  #   pick-up window, 10 ft, heated     > in    > out      dish drop       #
  ==========================================================================
  |                                        |                               |
  |  [] [] [] []    4 two-tops = 8 seats   |  RESTROOM 1     RESTROOM 2    |
  |  oo oo oo oo    on the banquette       |  (accessible)                 |
  |  ~~~~~~~ B A N Q U E T T E ~~~~~~~     |      >               >        |
  |                            /\ STA. 2   |-------------------------------|
  |                                        |                               |
  |   [][]     [][]     [][]               |  ### B A C K   B A R ###      |
  |  oooo     oooo     oooo                |  ##  taps  wells  glass  ##   |
  |     3 four-tops = 12 seats             |  ======== B A R ========      |
  |                                        |  o  o  o  o  o  o  o  o       |
  |   [][]     [][]     [][]               |     o  o  o  o                |
  |  oooo     oooo     oooo                |        12 stools              |
  |     3 four-tops = 12 seats             |                               |
  |            M A I N   A I S L E   42"   |                               |
  |  [] [] [] []    4 two-tops = 8 seats   |                               |
  |  oo oo oo oo    on the banquette       |     +---------------+         |
  |  ~~~~~~~ B A N Q U E T T E ~~~~~~~     |     |  HOST STAND   |         |
  |                            /\ STA. 1   |     +---------------+         |
  |   [][][]     [][][]     [][]           |                               |
  |  oooooo     oooooo     oooo            |                               |
  |     2 six-tops + 1 four-top = 16       |                               |
  ================================>  V E S T I B U L E  >===================
         [ 16-seat seasonal patio ]           M I L L   S T R E E T

Walk it the way a guest does, because that sequence is the design.

Arrival and the vestibule. The door opens onto a vestibule, not onto a table. This costs about 60 of the entry zone's 130 square feet and buys three things: a thermal break so the January draft does not land on your two best tables, an acoustic break, and somewhere for four people to stand while they are greeted. Chapter 6 named the cost of skipping the third — four people standing in the aisle, in the way of every server, for ten years.

The host stand sits to the right of the door with sight lines to the dining room, the bar, and the patio, because the host is running three inventories at once. It carries the reservation system, the waitlist, and a shelf. A host stand without a shelf becomes a host stand with menus on the floor.

The bar runs down the right-hand wall, twelve stools, back bar against the wall, working side facing the room. Two things about that geometry. It sits between the entry and the dining room, so every arriving guest passes it — worth real money in beverage attachment, and the reason a bar tucked behind a dining room underperforms. And the working side gets 200 square feet against the guest side's 140, which looks generous until you watch a bartender work: well, ice, glass storage, draft tower, the service-bar landing where servers pick up drinks, and a clear aisle a person can pass through with a case. Chapter 15 costs the pour; the bar's speed is set here, in inches.

Seating mix is the least-discussed high-leverage decision in a dining room. Bellwether's 56 seats are eight two-tops, seven four-tops, and two six-tops — seventeen tables. The two-tops sit on banquettes along both side walls, so they combine into four-tops in pairs, and combined down the center the room seats a 40-top private party without moving a fixed piece of furniture — which is what makes Chapter 29's private-events line physically possible. Chapter 24 owns table mix as a revenue-management problem; design owns optionality, and optionality comes from banquettes, square tables that push together, and not bolting anything down.

Two service stations, one at each end of the dining room, at 30 square feet each. Each carries a POS terminal and printer, water and glassware, silver, linen, bread, and coffee. Two stations rather than one is a labor decision, and §7.4 computes what it is worth.

🤝 Hospitality

The first ninety seconds are a floor plan, not a training problem.

Chapter 23 will argue that hospitality is a revenue model rather than a soft skill. Here is the part of that argument decided by a draftsman.

A guest's judgment is substantially formed before they sit down: whether the door was awkward, whether anyone saw them, whether there was somewhere to stand, whether they could hear the host, whether the walk to the table felt like being led or routed around obstacles. Every one of those is a dimension on a drawing.

• Nowhere to stand produces a huddle in the aisle, which produces a server who cannot pass, which produces a slow table, which produces a bad review about the service.
• A host stand facing the wrong way means arriving guests wait to be noticed. Ten seconds of not being seen reads as indifference and is not recoverable later.
• A restroom reached through the main aisle puts a person in the server's path twice a meal.
• A hard room at 82 decibels means your server leans in and half-shouts the specials, which reads as pushy no matter how warm they are — and it is why the acoustic treatment survived the value-engineering knife in §7.6 when the flooring did not.

None of these can be trained around. Walk any plan with one question: what will it feel like to arrive here, wait here, sit here, and leave here? Then price the fixes, because the fixes are part of the rent.

⚖️ Code and Compliance

Occupant load, egress, restroom fixtures, and accessibility — the four rules that set your seat count before you do.

You do not actually get to choose your seat count. You propose one, and four bodies of regulation tell you whether you may have it. All four vary by state, county, and city, all four change, and the only correct source is your local building department, fire marshal, health authority, and an architect who has permitted restaurants there. What follows is the shape of the questions.

Occupant load. The occupant load is the number of people a building official calculates a space is designed to hold, derived from floor area and a code-assigned factor for each use — dining, standing, bar, and kitchen areas all carry different factors. It is not your seat count; it is usually larger, and it drives everything below it. Have your architect compute it early, because a plan that seats 68 but computes to an occupant load requiring a second exit is a very expensive discovery at plan review.

Egress. Exit count, exit width, travel distance, door swing, and panic hardware are all functions of occupant load. Chapter 6's change-order log carried \$980 for exit signage and one panic device the fire marshal required — a small number, and a preview of a category that is not always small.

Plumbing fixture counts. Required restroom fixtures are generally a function of occupant load, and this is one of the most common reasons a seat count gets reduced late. Bellwether's two single-occupancy restrooms are inherited from the café, and Chapter 6's inheritance inventory listed their accessibility as unclear — which is why \$17,000 sat on the restroom line.

Accessibility. The Americans with Disabilities Act (ADA) is federal law applying to places of public accommodation, enforced independently of whatever your local building department approves. Chapter 8 covers compliance properly. What a designer must hold from day one of the drawings:

• An accessible route from the public way through the entrance to the dining areas, the bar, and the restrooms, with clear width maintained. The route is not a corridor you add later; it is the spine the tables get arranged around.
• A proportion of dining surfaces at an accessible height with knee and toe clearance, dispersed through the room rather than clustered at the worst table. This is why fixed banquette seating cannot be the only seating, and why the bar needs an accessible section of counter or an equivalent accessible dining surface.
• Restroom clearances, turning space, fixture heights, and grab bars to the published standards.
• The alterations trigger. Work on an existing facility generally brings altered elements up to current standards, and can carry an obligation to make the path of travel to the altered area accessible as well. That is the mechanism by which a modest renovation becomes an accessibility project — and it is why nothing in §7.6's value engineering touched a single accessibility dimension.

Accessibility is not a line item you value-engineer. It is a legal obligation and a hospitality obligation at the same time, and the operator who treats it as the first thing to cut has misunderstood both. Use a licensed architect, and verify locally.

7.3 The back of house: the line, the stations, the pass, prep, dish, storage, receiving

Nine hundred square feet, four cooks, and 123 covers on a Saturday. Here is how that works.

Kitchen workflow is the sequence and direction in which product, people, and plates move through a kitchen — from receiving through storage, prep, cooking, plating, and service to warewashing and waste. A good workflow has one dominant direction of travel, no crossings, and no backtracking. That sentence sounds like a platitude until you watch a kitchen where the dish return path crosses the hot line: it is a burn injury waiting to happen, a cross-contamination risk the health inspector will name (Chapter 25), and a permanent tax on every plate.

Bellwether runs the classic linear flow: deliveries in at the back, product moves forward through storage and prep to the line, and plates exit at the pass into the dining room. Waste and soiled ware move backward along a separate path on the opposite side.

FIGURE 7.4 — Bellwether, back of house: 900 sq ft + 200 storage   [the Bellwether plan]

  Schematic; not to scale. Product flows DOWN the page. Soiled ware flows along
  the right-hand path only, and never crosses the line.

  ============ ALLEY: receiving door, grease bin, waste, recycling ============
  |  DRY STORAGE   |  OFFICE  | STAFF /  |  MOP  |  RECEIVING & SCALE        |   200
  |  shelving      |  desk    | lockers  | SINK  |  (patio furniture, Nov–Apr)|  sq ft
  ============================================================================
  |                                    |                                      |
  |  WALK-IN COOLER   8 x 10           |          D I S H   P I T             |
  |  +--------------------+            |   +--------------------------------+ |
  |  |     80 sq ft       |  REACH-IN  |   | soiled landing > scrap > rack  | |
  |  |   (existing box,   |  FREEZER   |   | > door machine > clean landing | |
  |  |  new condensing)   |  2-door    |   | > shelving  ..... 3-COMP SINK  | |
  |  +---------->---------+            |   +--------------------------------+ |
  |                                    |                            [hand]    |
  |  P R E P   T A B L E S  (2)   ...  slicer, mixer, scale, board storage    |
  |                                    |                                      |
  |  ==================  T H E   L I N E  — 14 ft  ========================== |
  |   GARDE MANGER  |   SAUTÉ      |   HEARTH      |  FRY / SIDES / PLATE-UP  |
  |   low-boy,      |   6-burner   |   wood-fired  |  fryer, salamander,      |
  |   cold apps,    |   + oven,    |   + charbroil |  griddle plate, landing  |
  |   salads, dess. |   pasta      |               |                  [hand]  |
  |  ------------ under 16 ft of Type I hood, 2 sections ------------------   |
  |            [ 42-inch working aisle — two cooks can pass ]                 |
  |  ================  T H E   P A S S  — heated, 10 ft  ==================== |
  ============== to the dining room (Figure 7.3) : > in   > out ==============

The line, the stations, and the pass

The line is the row of cooking equipment where food is fired to order during service, together with the cooks working it. A station is one cook's defined territory on that line — their equipment, their refrigeration, their mise en place, and the specific list of menu items they are responsible for. Bellwether runs four:

The pass — also called expo, short for expediting — is the counter between the line and the dining room where finished plates land, get checked, get garnished, and go out together. It is a physical object and a job at the same time: the person working it reads the tickets, calls the courses, times the stations against each other, wipes the rims, and decides whether a plate leaves the kitchen. Chapter 14 owns expediting as a discipline. What design owns is that the pass must be heated, at least ten feet long, visible from every station, and reachable from the dining room without a server entering the kitchen.

That last clause is worth an extra sentence. If servers must walk into the kitchen to pick up food, you have added four steps and a door to every plate, you have put untrained bodies in the line's working aisle during service, and you have created the single most common source of friction between front and back of house in the industry. A pick-up window with the pass on the kitchen side and the servers on the dining side eliminates all three, and it costs nothing extra at the drawing stage.

👨‍🍳 On the Line

7:15 on a Friday, and what a 42-inch aisle is actually for.

The board says fourteen tickets. Four two-tops that sat at 6:45 are on entrées, five four-tops that sat in the last twenty minutes are on appetizers, three are just seated, and two are the six-tops the host warned about at four o'clock.

Here is what that looks like in the aisle. The hearth cook is turning eleven items on the deck with four more staged behind; every turn is a pivot and a reach. The sauté cook has six pans working, so both hands are committed and their body is a fixed obstacle. The fry/sides cook is moving constantly — low-boy, salamander, landing — and crosses behind both of them roughly every ninety seconds. Somebody will say "behind, hot" about forty times in the next hour.

At 36 inches, the fry cook cannot pass behind the sauté cook without both of them stopping, and every crossing becomes a two-to-four-second negotiation. At 48 inches nobody negotiates, but every reach from the range to the low-boy behind is a step instead of a pivot — and a cook makes that reach maybe a hundred and eighty times an hour.

Forty-two inches is the compromise, and it is a compromise, not a right answer. Trade convention puts a working line aisle in the 36–48 inch range; your equipment depths, your appliance door swings, and your local requirements decide it. Ask a kitchen designer to draw the door swings, because a fryer door and an oven door that open into each other will find each other on the third Saturday.

What actually goes wrong when the aisle is tight is not the collision. It is that the cooks silently reorganize the work to avoid crossings — the fry cook stops running food to the pass, the sauté cook stops helping plate, each station retreats into itself. Ticket times go up eight minutes and nobody can say why. That is a floor plan producing a labor problem, and no manager standing at the pass will ever diagnose it from the tickets.

Proving the line can produce the number

This is the arithmetic that makes the revenue forecast physically credible, and almost no business plan contains it.

The kitchen is not sized to the night. It is sized to the busiest hour of the night. Bellwether's seating curve concentrates 31 of 95 covers into the hour from 6:30 to 7:30, and entrées lag seating by roughly twenty-five to thirty-five minutes, which pushes plating into an even tighter band.

The planning rule this chapter uses: the peak hour delivers about 37% of the night's covers as entrées. That is a constructed planning number derived from the curve, not a measurement; Chapter 14 replaces it with your own ticket data within three weeks of opening.

Now the capacity side. Each station has a ceiling set by geometry and dwell time, and the hearth's ceiling is the tightest, which is exactly what you would expect: the concept's signature piece of equipment is also its throughput constraint.

FIGURE 7.5 — Where the ceiling actually is                [the Bellwether plan — constructed]

  THE HEARTH, from first principles:
    usable deck in the working heat zone .................  ~9 sq ft
    footprint of one plated item (pan or direct) .........   0.6 sq ft
    items on the deck at once, leaving room to work ......   12
    average dwell 14 min + 4 min to load, turn, and pull ..   18 min per cycle
    theoretical cycles per hour ..........................   3.33
    theoretical items per hour ...........  3.33 × 12 =      40
    derate 30% for fire management, uneven dwell, plating    × 0.70
    ──────────────────────────────────────────────────────────────
    SUSTAINED HEARTH CAPACITY ............................   28 items/hour

  PEAK-HOUR LOAD vs. STATION CAPACITY
  menu routed 50% hearth / 30% sauté / 20% cold

                       95-cover     123-cover    135-cover   STATION
                       night        Saturday     Sat + patio CAPACITY   PEAK USE
  ──────────────────────────────────────────────────────────────────────────────
  Hearth        50%      18            23            25          28      89%  ◄
  Sauté         30%      10            14            15          26      58%
  Cold / other  20%       7             9            10          45      22%
  ──────────────────────────────────────────────────────────────────────────────
  TOTAL ENTRÉES          35            46            50
  Tickets through the pass ~29         ~38          ~42          ~45     93%  ◄

  Two ceilings, and they arrive together: the hearth at 89% and the pass at 93%.
  Everything else has room. Constructed planning numbers — time your own line.

Read the right-hand column. On an ordinary 95-cover night the hearth runs at 64% and the kitchen is comfortable. On a Saturday in July with the patio full it runs at 89%, and the expediter handles roughly 42 tickets an hour against a practical ceiling near 45. That is the honest answer to "can this kitchen do the number": yes, comfortably, at 95; yes, at the edge, at 135; and the edge is the hearth and the pass, not the square footage.

Three instructions follow, and they must hold for the life of the concept:

1. No more than about half the entrée menu may route through the hearth. Chapter 10 writes the menu; this is the constraint it must be written against. A menu where 70% of entrées touch the fire turns 50 peak-hour plates into 35 hearth items against a 28-item ceiling, and the room stops.
2. The pass needs a second body on Friday and Saturday. One person expediting 42 tickets an hour while also garnishing is where mistakes enter the dining room. Chapter 19 costs the position.
3. Prep, not the line, absorbs growth. The only way this kitchen gets faster is more work done before four o'clock — a prep-space and prep-labor decision, and 900 square feet with two prep tables is adequate rather than generous.

Prep, dish, storage, and receiving

Prep is two tables between the walk-in and the line, placed so a prep cook's day runs walk-in → table → line without crossing the dish path. During service those tables become landing and staging space, which is why they are not against a wall.

The dish pit is the most under-designed room in most restaurants and the one that produces the most turnover. It needs, in order: a soiled landing large enough to hold a bus-tub surge without stacking on the floor, a scrap and pre-rinse station, the machine, a clean landing, and shelving within reach of where the clean ware is used. Give it a floor drain, light, ventilation, and a person who is not also expected to run food.

Storage is where Bellwether is genuinely tight, and the plan should say so. The inherited walk-in is 8 × 10 — 80 square feet — reused per Chapter 6's inheritance inventory. Common planning guidance puts a scratch kitchen at roughly 1 to 1.5 cubic feet of refrigerated storage per meal served per day, with an 80-square-foot box supporting somewhere around 100 to 150 covers a day on two or three deliveries a week (Tier 2 — trade convention; verify against your own menu). Bellwether runs 95 to 123 dinner covers plus brunch. The box is adequate and it is not spare.

The countermeasure is not a bigger box; it is more frequent, smaller deliveries, a purchasing decision (Chapter 13) with a labor cost attached. Design's contribution is to put the receiving door, the scale, and the walk-in on the same short path, which Figure 7.4 does.

Receiving shares the back 200 square feet with dry storage, the office, staff lockers, and — from November through April — sixteen patio chairs. Write that on the drawing. It is exactly the kind of thing that is obvious in October and a crisis in November.

🔍 Check Your Understanding

1. Bellwether's kitchen is designed to 50 peak-hour entrées, not to 95 covers a night. Explain why, and state where the 50 comes from.
2. The hearth's sustained capacity is 28 items an hour. What happens to the peak-hour picture if the menu is written so that 70% of entrées route through the hearth instead of 50%?
3. Why does the pass need to be reachable from the dining room without a server entering the kitchen? Name two costs of the alternative.

(1: Because the kitchen's constraint is instantaneous, not daily — the busiest hour delivers about 37% of the night's covers as entrées, and on the design night of 135 covers that is 50. A kitchen sized to the nightly average fails in the hour that matters. 2: 50 × 0.70 = 35 hearth items against a 28-item ceiling — the hearth is 25% over capacity, ticket times climb, and the whole room slows. The fix is menu design, not equipment. 3: Because a server entering the kitchen adds steps and a door to every plate, puts untrained bodies in a 42-inch working aisle during service, and creates the most common recurring FOH/BOH conflict in the industry.)

7.4 Flow: how many steps a plate and a person take, and why that is a labor number

Here is the idea this section exists to install: distance is money, it is payable weekly, and it compounds for the length of your lease.

Nobody thinks of a floor plan as a labor cost, because the cost never appears on a statement with a name. It shows up as an extra server on Friday, a table touched a minute late, a food runner you keep meaning to cut. It is invisible in exactly the way Chapter 1 said cost drift is invisible, and for the same reason: nobody measures it. So measure it.

The step-count audit

1. Count parties, not covers. At 95 covers and an average party of 2.4, about 40 parties.
2. Count the trips per party across a full meal — greet, drinks, order, bread, appetizers, clear, entrées, check back, clear, dessert, check, payment. Call it twelve round trips; most rooms run more.
3. Measure a representative round trip in paces, station to table and back, at 2.5 feet a pace.

Bellwether's floor, drawn as in Figure 7.3:

$$40 \text{ parties} \times 12 \text{ trips} = 480 \text{ round trips a service}$$

At an average round trip of 30 paces — 75 feet — that is $480 \times 75 = 36{,}000$ feet, or 6.8 miles walked across the floor team in a single dinner service. Multiply by 260 dinner services and two brunches a week and the number stops being cute.

🧮 Run the Numbers

What Chapter 6's \$2,400 change order bought.

Chapter 6's change-order log carried CO 003 at week four: relocate server well four feet to open the path from the pass to the dining room, +\$2,400, categorized as an owner-requested change — the only category that is optional. Chapter 6 said Chapter 7 would show you why it was the right call. Here it is.

Moving the station four feet shortens the average round trip by six paces, or fifteen feet.

$$480 \text{ trips} \times 6 \text{ paces} \times 2.5 \text{ ft} = 7{,}200 \text{ feet a service}$$

A server carrying plates covers about three feet a second. So:

$$7{,}200 \div 3 = 2{,}400 \text{ seconds} = \mathbf{40 \text{ minutes of walking per service}}$$

Across five dinner services a week, fifty-two weeks:

$$40 \times 5 \times 52 = 10{,}400 \text{ minutes} = \mathbf{173 \text{ hours a year}}$$

At a fully loaded server cost of \$22 an hour that is **\$3,813 a year against a one-time \$2,400. Payback in about seven and a half months, then \$3,813 a year for the remaining nine years of the lease: roughly \$34,000.**

And \$3,813 is the small half of the benefit. The forty minutes do not come back as standing around; they come back as tables touched sooner, drinks refilled before they are asked for, and a check dropped when the guest is ready — which is turn time (Chapter 22), average check (Chapter 24), and the second visit (Chapter 23) at once.

The general form, runnable on any layout question in any restaurant:

text annual hours saved = trips/service × paces saved × 2.5 ft ÷ 3 ft per second ÷ 3,600 × services per year annual dollars = annual hours × fully loaded hourly cost payback in months = one-time cost ÷ (annual dollars ÷ 12)

Run it before you argue about it. A layout debate settled with a payback period ends in four minutes; the same debate settled with opinions runs for the length of the project.

The principles the arithmetic produces

Two stations, not one. Bellwether's two service stations cost 60 square feet — three seats' worth of dining room — and they halve the average trip from the far end of the room. It is the cheapest labor reduction in the building.

No crossings in the back. The dish return path and the food-out path must not intersect the hot line. In Figure 7.4 soiled ware travels down the right-hand side and never crosses in front of the range. That is a safety decision before it is an efficiency decision, and Chapter 25 will show it is a food-safety decision too.

The cook's pivot. Inside a station, the three things a cook touches most — refrigeration, board, fire — should form a triangle they can work without stepping. A cook makes that reach on the order of a hundred and eighty times an hour at peak; converting one step into a pivot removes thousands of steps a week from one person.

The walk-in door faces the prep table. Sixty trips a shift at forty feet each way is 4,800 feet of walking for one prep cook. Turning the box is free at the drawing stage and impossible afterward.

Design the trip nobody counts. Somebody carries trash to the alley, changes the fryer oil, and gets the case of wine from dry storage during service. Those trips happen in the busiest hours and are never on anybody's list.

7.5 The equipment schedule: specifying, sizing, buying new vs. used, and what to never cheap out on

An equipment schedule is the numbered, itemized list of every piece of fixed and major movable equipment in the building, giving for each item a tag number, a description and model, quantity, dimensions and required clearances, utility requirements (gas input, electrical volts/phase/amps, water, drain), and who supplies and who installs it.

It looks like a purchasing document. It is three documents at once, which is why sloppy ones are so expensive:

• A purchase order — what you buy from, and what a lender or lessor finances against.
• A permit document. Health-department plan review and your mechanical, electrical, and plumbing engineers all design from it: the hood is sized from the appliances, the gas meter from the BTU column, the panel from the amps column. Change the schedule after the drawings are stamped and you have changed the mechanical design.
• A capacity statement. Figure 7.5's ceiling came off this list. The equipment schedule is the answer to "how many plates can you send."

FIGURE 7.6 — Bellwether equipment schedule (extract)      [the Bellwether plan — constructed]

  Prices are deliberately omitted from this extract: equipment pricing varies enormously
  by market, manufacturer, condition, and season, and any single figure would mislead.
  Utility figures are illustrative. Your engineer specifies; your AHJ (authority having
  jurisdiction) approves.

  TAG   ITEM                                   QTY  UTILITIES              NOTES
  ─────────────────────────────────────────────────────────────────────────────────────
  K-1   Wood-fired hearth oven, ~42" opening    1   gas assist 60 MBH;     SOLID FUEL —
                                                    120V/1ph controls      drives hood spec
  K-2   Range, 6 burner + standard oven         1   200 MBH                under hood
  K-3   Charbroiler, 24"                        1   80 MBH                 under hood
  K-4   Fryer, twin basket, w/ filtration       1   80 MBH                 3 ft from K-1
  K-5   Griddle plate, 24"                      1   60 MBH                 BRUNCH-critical
  K-6   Salamander broiler                      1   35 MBH                 wall mount
  K-7   Convection oven, full size              1   55 MBH                 prep + brunch
  K-8   Refrigerated chef base / low-boy        3   120V/1ph               one per station
  K-9   Prep top, refrigerated, 60"             1   120V/1ph               garde manger
  K-10  Walk-in cooler, 8 x 10 (EXISTING BOX)   1   208V/1ph               REUSE box;
                                                                           NEW condensing unit
  K-11  Reach-in freezer, 2 door                1   120V/1ph
  K-12  Dish machine, door type, high temp      1   208V/3ph + booster     floor drain
  K-13  Three-compartment sink                  1   H/C water, indirect
  K-14  Hand sinks                              4   H/C water              COUNT SET BY
                                                                           PLAN REVIEW
  K-15  Mop sink                                1   H/C water
  K-16  Ice machine, 400 lb + bin               1   120V/1ph, water, drain
  K-17  Slicer, mixer (20 qt), scales           1 ea 120V/1ph
  M-1   Type I hood, 16 ft, 2 sections          1   208V/1ph controls      ★ CONSTRUCTION
                                                                           LINE, not equipment
  M-2   Make-up air unit, tempered, rooftop     1   gas 400 MBH; 208V/3ph  ★ CONSTRUCTION LINE
  B-1   Back-bar refrigeration, 3 door          1   120V/1ph
  B-2   Draft system, 6 tap, glycol             1   120V/1ph
  B-3   Under-counter glasswasher               1   120V/1ph, water, drain ✗ DEFERRED — §7.6
  ─────────────────────────────────────────────────────────────────────────────────────
  Connected gas load, all items: on the order of 1.3 million BTU/hr against a meter
  and service sized for a café. See §7.6.

Three things on that schedule are worth a paragraph each.

The hood is not on the equipment budget. Tags M-1 and M-2 carry a star because in Bellwether's project they sit inside the construction contract with the mechanical trade, not inside the \$185,000 equipment line. That is a legitimate choice and not the only one — plenty of projects buy the hood as owner-furnished equipment. What is not legitimate is leaving it ambiguous. Items on the boundary between budgets get counted twice or not at all, and "not at all" is the common failure.

K-14, hand sinks, says the quantity is set by plan review. It is. Chapter 6's change-order log carried \$1,150 for a second hand sink the health plan reviewer required at the pass. Hand sinks, floor sinks, mop sinks, and indirect-waste requirements are among the most common plan-review additions — cheap when drawn, expensive when the walls are closed.

K-5, a 24-inch griddle plate, exists entirely for brunch. It has no dinner function. Bellwether serves two brunches a week at 110 covers each, and a brunch line without a flat-top cannot produce eggs at volume. This is exactly the item that gets value-engineered off a schedule by someone reading the dinner menu and then discovered on the second Sunday. Design for every daypart you have promised, including the ones you have not cooked yet.

New, used, and the constraint nobody mentions

The instinct to buy used is sound and saves real money. It has three limits operators discover late.

The reasons are not sentimental. Refrigeration is the one to never economize on: a compressor failure is not a repair bill, it is a walk-in full of product at 3:00 a.m., a temperature violation (Chapter 25), and a service you cannot run. The dish machine runs harder than anything else in the building and its failure closes you. Anything requiring a listing or label — a hood, a fire-suppression system, a gas appliance — may simply not be approvable by your authority having jurisdiction if it is used and the label is gone, and you find out at inspection.

And here is the constraint nobody mentions, straight out of Chapter 5. Bellwether's capital stack includes a \$60,000 equipment lease, and lessors generally finance new, titled, serial-numbered assets with a resale market. Used equipment often cannot be leased at all. The "buy used and save" instinct therefore runs directly into the financing structure: money you were going to borrow against equipment does not exist if the equipment came off a closed restaurant.

⚠️ Where the Money Leaks

The five specifications that look like savings and are loans from year two.

Every one gets cut in the third round of budget conversations, and every one is borrowed at a punitive rate.

1. Undersized electrical service and panel capacity. A full panel means every future addition — a second reach-in, a coffee machine, a patio heater — is a subpanel and an electrician. Chapter 6's log already carried \$6,200 for a subpanel found at week three.

2. Insufficient refrigeration at the stations. A cook walking to the walk-in mid-service is a cook not cooking. One low-boy per station is what makes a station a station.

3. Floor drains and floor slope. Adding a drain later means cutting concrete in a finished kitchen, which means closing.

4. Light-gauge stainless and cheap work surfaces. A table that flexes under a hotel pan gets replaced in year three at retail, plus the labor to swap it on a closed day you do not have.

5. The ice machine. Undersize it and you buy bagged ice on your busiest days forever — a recurring cost, a delivery, a storage problem, and a labor task on exactly the shifts with no labor to spare.

What they have in common: each converts a one-time, construction-priced capital cost into a permanent operating or labor cost. That is the wrong direction of trade in a four-to-six-point margin business, and it is the direction budget pressure always pushes. When you must cut, cut the fixed thing the guest can see — not the fixed thing that becomes a variable.

7.6 The invisible systems: hood and make-up air, gas and electrical load, grease trap, refrigeration

Everything expensive in a restaurant build-out is invisible when you are finished. Guests see flooring, light fixtures, banquettes, and a bar top. They will never see the thing that costs the most and that this section exists to price.

What the systems are

A Type I hood is a listed, grease-rated commercial exhaust hood installed over cooking equipment that produces grease-laden vapors — ranges, fryers, charbroilers, griddles, and solid-fuel appliances. It combines a capture canopy, grease-extraction baffles, grease-collection means, a fire-suppression system, and a grease-rated duct discharging above the roof. A Type II hood is a different animal for a different problem: heat and condensate only, over dishwashers, steamers, and ovens that do not produce grease. Putting a Type II where a Type I belongs is not a value-engineering choice; it is a failed inspection and a fire risk.

Make-up air is outdoor air mechanically supplied to replace the air the exhaust system removes. Common design practice supplies the large majority of it mechanically — often on the order of 80–90% — and lets the remainder transfer from the conditioned dining room, so the kitchen sits slightly negative relative to the dining room and cooking odors move toward the hood rather than the guests. The exact proportions are an engineered result, not a rule you can copy, and in a cold climate the supplied air must be tempered — heated — or you have installed a machine whose job is to blow winter into your kitchen.

A grease trap, or in its larger form a grease interceptor, separates fats, oils, and grease from wastewater before it enters the sanitary sewer. Small hydromechanical units live under a sink; large gravity interceptors live in the ground outside. Sizing is set by the local sewer authority or plumbing code, computed from fixture units, drainage flow, seating, or a local formula — and the rules vary so widely between jurisdictions that a sizing figure from a book is worse than no figure at all. Ask the sewer authority in writing, early, and have a licensed plumbing engineer calculate it.

FIGURE 7.7 — The air path, and why the whole building is one system

              ROOF
       [ EXHAUST FAN ]                    [ MAKE-UP AIR UNIT, tempered ]
              ^                                        |
              |  grease-rated welded duct              |  supplied outdoor air
              |  in a rated shaft                      |  (the large majority of exhaust)
   ===========|========================================|=====================
              |                                        v
        [ TYPE I HOOD, 16 ft, 2 sections ]      [ perforated ceiling plenum ]
              ^   ^   ^   ^                            |
         capture over the line                  low-velocity, so it does not
              |   |   |   |                      blow the capture off the hood
   ─────  GARDE  SAUTÉ  HEARTH  FRY  ─────────────────────────────────────
                          ^ solid fuel: its own provisions
                                                    ^
              KITCHEN slightly NEGATIVE  <---- transfer air ---- DINING ROOM
                                                                (rooftop units,
                                                                 slightly positive)

  If make-up air is short:  doors get hard to open · the hood stops capturing ·
  pilot lights blow out · the dining room gets a draft at the door · you heat
  outside air with your dining-room furnace and pay for it twice.

Three failure modes to hold on to, because they are what "the make-up air is undersized" actually feels like in a building. The hood stops capturing — smoke rolls out the front of the canopy at exactly the moment the line is busiest, a fire-code problem and a dining-room problem at once. Doors become hard to open, because the exhaust fan is pulling air through your front door. And you pay twice: the dining room's furnace conditions outdoor air the exhaust immediately throws away.

⚖️ Code and Compliance

Solid fuel is its own category, and this is where a book must stop being specific.

A wood-fired hearth is a solid-fuel cooking appliance, and in most American mechanical and fire codes that puts it in a category of its own, with requirements layered on top of ordinary commercial cooking provisions. In broad terms — and these are the questions to ask, not the answers:

• A dedicated exhaust system is frequently required for solid fuel, so the hearth may not simply share duct with the gas appliances beside it.
• Spark arrest and grease-extraction provisions differ, because what comes off a wood fire is different from what comes off a range.
• Duct construction, clearances to combustibles, and shaft enclosure are typically more demanding, and cleaning access and frequency are usually specified and higher than for gas.
• Fire suppression may require means in addition to the wet-chemical system protecting the line.
• Fuel storage — where the wood lives, how much, how far from the fire — is itself regulated in many jurisdictions.

This book will not give you a required exhaust rate, a duct dimension, a code section, or a clearance. They vary by jurisdiction, adopted code edition, appliance listing, and the specific configuration of your line, and a number copied out of a textbook into a real building is how people get hurt. What this book will tell you is who to hire: a licensed mechanical engineer, a kitchen designer who has permitted a solid-fuel appliance in your jurisdiction, a licensed plumbing engineer for the interceptor, and a pre-application conversation with your building department, fire marshal, health authority, and sewer authority — before you budget.

Chapter 6's mechanical contractor gave the partners the highest-return advice in Part II when he asked one question during due diligence: "Is that a hearth, or is that a decorative thing?" The cost of asking was a few hundred dollars. The cost of not asking is the rest of this section.

The real number

Chapter 6's construction contract carried two allowances — placeholder dollar figures inside a contract sum for scope not yet fully specified. \$18,000 for the hood and make-up air, inside a \$29,000 mechanical line. **\$8,000** for the grease interceptor, inside a \$34,000 plumbing line. Both assumed the café's existing equipment could be modified and reused. It cannot.

Here is what the re-drawn, re-bid scope actually costs. Every figure below is a constructed teaching number. Mechanical and plumbing pricing varies enormously by market, by whether the work is union, by roof access, by season, and by how many contractors want the job. The only real number is three local bids on a complete drawing set.

$$\$52{,}000 + \$6{,}500 + \$16{,}500 = \mathbf{\$75{,}000}$$

Against \$26,000 of allowances. **The gap is \$49,000, and it is worse than that, because the re-scope also drags soft costs with it: the mechanical and plumbing sheets have to be re-drawn and re-submitted (\$3,600 of additional architect and engineering fee), and the revised permit valuations plus a second plan-review cycle add \$900. **Total to find: \$53,500.

Notice something about how that \$75,000 is distributed. It lands across three trades, and the one nobody carried a dollar for — fire protection — is the one that is required because of the solid-fuel appliance. A hood is not a line item. It is a change to the building.

⚠️ Where the Money Leaks

Value engineering: how a budget actually gets cut, and the order that matters.

The \$310,000 construction line does not grow. It is one of five components of a frozen \$620,000 project, funded by a capital stack that closed in Chapter 5. There is no version of this conversation that ends with a bigger number. So \$53,500 has to come out of the work.

Value engineering is the disciplined substitution of a lower-cost means for the same function. It is not cutting. Cutting removes what the building has to do; value engineering removes what it costs to do it. The distinction sounds academic until you are in the meeting, where every proposal arrives wearing the same label.

First, the mistake almost everyone makes. The chef-owner opens with: "We have \$35,000 — the \$26,000 of allowances plus the \$9,000 contingency."

That is wrong, and it is the most expensive wrong sentence in a build-out. You have \$26,000. The \$9,000 is a construction contingency — money reserved for costs that are certain to occur and cannot yet be identified. The hood is no longer unidentified. It is known, priced, and drawn, and known scope is funded from scope. Spend the contingency here and you have solved nothing; you have deleted your only cushion at week zero on a project that still has an unverified gas service, eleven-year-old rooftop units, an unscoped waste line, and abandoned drains under the slab.

Chapter 6 named \$9,000 against a \$264,000 contract sum — 3.4% against a 10–15% rule of thumb — as a deliberate weakness. The discipline is to hold it there, not to spend it and pretend the problem is smaller.

Second, the order of cutting. There is one, and it is worth memorizing:

1. Never cut compliance. Accessibility, egress, fire protection, sanitation. Not negotiable, not once, not by a dollar.
2. Never cut capacity. Anything that lowers the plate count lowers the revenue the whole plan rests on. The hood is capacity. The electrical service is capacity. The refrigeration is capacity.
3. Never cut what becomes an operating cost. See the five specifications above. Converting capital into a permanent variable is the wrong direction of trade.
4. Cut what the guest cannot perceive — substrate, hidden construction, spec grade on things nobody touches.
5. Cut what the guest perceives but does not price — a floor finish, a ceiling treatment, a stone the guest could not name.
6. Defer what can genuinely be bought later out of operations, and write down when, with a number, or it will never happen.
7. Only then, reduce scope the guest experiences. And when you get here, stop and ask whether the project is the right size.

Bellwether's cuts, in that order — nine items, \$53,500:

#	Line	What changes	Saving
VE-1	Finishes	Delete the specified reclaimed-plank dining floor; grind, polish, and seal the existing slab; area rugs under the banquettes	\$11,900
VE-2	Finishes	Engineered quartz instead of the specified stone at the bar top and chef's counter; standard tile at the bar face	\$3,400
VE-3	Millwork	Delete custom millwork at the two service stations and the back-bar shelving; substitute modular stainless — moves to the equipment line	\$9,800
VE-4	Millwork	Banquettes: straight runs, no returns, one upholstery grade down, shop-standard bases	\$4,400
VE-5	Restrooms	Finish package down to a durable, code-compliant standard. No accessibility dimension, clearance, fixture height, or grab bar changes.	\$5,500
VE-6	Framing / ceilings	Leave the dining-room deck open and painted rather than installing a hung ceiling. The acoustic treatment stays in full.	\$7,400
VE-7	Electrical	Decorative lighting package reduced. Service upgrade, kitchen circuits, and controls remain at full spec.	\$4,700
VE-8	Demolition	Owner-supplied labor for non-structural demolition and disposal, two weekends	\$1,900
VE-9	Millwork	Office and dry-storage built-ins deleted; adjustable open shelving and a stock desk	\$4,500
		TOTAL FOUND	\$53,500

Check the column: 11,900 + 3,400 + 9,800 + 4,400 + 5,500 + 7,400 + 4,700 + 1,900 + 4,500 = \$53,500. ✓

Now look at what is not on that list, because that is the actual lesson.

The patio did not get cut. It was on the first draft — sixteen chairs and a surface upgrade look like the easiest \$8,000 in the building. Then somebody did the arithmetic in §7.1: those sixteen seats carry \$64,800, which is 46.5% of the revenue bridge the entire \$1,550,000 forecast depends on. A cut that removes 4.2% of revenue to save 2.6% of the construction budget is not value engineering. It is arson with a spreadsheet.

The acoustic treatment did not get cut, even though the ceiling above it did. A hard warehouse room at 82 decibels costs you check average, server effectiveness, and repeat visits — an operating cost forever, traded for a one-time saving. Rule 3.

The electrical service upgrade did not get cut. Rules 2 and 3.

A single accessibility dimension did not move. Rule 1.

And the \$9,000 contingency did not move. Which means Bellwether enters construction with 3.5% of contingency instead of 3.4% — a change that is not an improvement, it is arithmetic, because the contract sum got smaller. That weakness is still the plan's, and the plan still has to say so.

🧾 Read the Numbers

```text FIGURE 7.8 — "The re-cut construction budget" [the Bellwether plan] THE ARTIFACT The construction line of the project budget, re-cut after the mechanical and plumbing scope was re-drawn and re-bid as specified scope rather than as allowances. Compare directly against Chapter 6's Figure 6.5. THE CONTEXT Pre-construction. 2,800 sq ft second-generation conversion. The $310,000 construction line is frozen; the $620,000 project is frozen; the capital stack closed in Chapter 5. All figures constructed for teaching.

               HARD COSTS — the contractor's stipulated sum
                                                   AT BID    CHANGE    RE-CUT
                 Demolition and disposal          $ 13,000   - 1,900  $ 11,100
                 General conditions, super., fee    36,000        --    36,000
                 Plumbing — interceptor NOW
                     SPECIFIED (was $8,000 allow.)   34,000   + 8,500    42,500
                 Electrical                          41,000   - 4,700    36,300
                 Mechanical — hood + MUA NOW
                     SPECIFIED (was $18,000 allow.)  29,000   +34,000    63,000
                 Fire protection — solid-fuel prov.  12,000   + 6,500    18,500
                 Framing, drywall, ceilings, doors   24,000   - 7,400    16,600
                 Restrooms, incl. accessibility      17,000   - 5,500    11,500
                 Millwork                            31,000   -18,700    12,300
                 Finishes                            27,000   -15,300    11,700
                 ────────────────────────────────────────────────────────────
                 CONTRACT SUM                      $264,000   - 4,500  $259,500

               SOFT COSTS — not in the contractor's number
                 Architect and MEP engineering     $ 24,000   + 3,600  $ 27,600
                 Permits, plan review, expediting      9,000   +   900     9,900
                 Utility connection and meter fees     4,000        --     4,000
                 ────────────────────────────────────────────────────────────
                 SOFT COSTS                        $ 37,000   + 4,500  $ 41,500

                 Contingency  (HELD — not enlarged,   9,000        --     9,000
                               and not spent)
                 ────────────────────────────────────────────────────────────
                 CONSTRUCTION LINE                 $310,000        $0  $310,000
                                                   = $110.71 per rentable sq ft

               THE SOLID-FUEL PACKAGE, ISOLATED
                 Mechanical: hood, duct, roof, fan, MUA, controls      $52,000
                 Fire protection: suppression + solid-fuel provisions    6,500
                 Plumbing: in-ground grease interceptor                 16,500
                 ──────────────────────────────────────────────────────────────
                 REAL COST                                            $75,000
                 Allowances carried at bid ($18,000 + $8,000)          -26,000
                 Additional soft cost (re-draw, re-submit, permits)     +4,500
                 ──────────────────────────────────────────────────────────────
                 FOUND BY VALUE ENGINEERING                           $53,500

WHAT IT SHOWS A budget that did not grow and a scope that did. The line lands on $310,000 exactly and the cost per square foot is unchanged at $110.71 — which is the point: the price of a square foot did not move, but what a square foot buys did. Three trades absorbed a $49,000 increase and six lines absorbed a $53,500 decrease, and nothing that produces revenue, protects a guest, or satisfies a code was touched. WHAT IT DOESN'T It does not fix the contingency. $9,000 against $259,500 is 3.5% where 10–15% is the rule of thumb — the plan is still $16,900 to $29,900 light on cushion before anything goes wrong, and the known unknowns are named: an unverified gas service and meter, a waste line whose interior nobody has scoped, two eleven-year-old rooftop units, and a slab with abandoned drains under it. It does not price the schedule: a hood and a make-up air unit are long-lead items, and a re-drawn hood does not ship on the original date. It does not show the equipment line, which absorbed the modular stainless from VE-3. And it does not show what the room now looks like — a polished concrete floor and an open painted deck are a legitimate design, and they are not the design that was drawn. THE DECISION Sign the revised contract at $259,500 with zero open allowances, publish both budgets side by side in the plan, and write one sentence under them: "the construction line did not move; the scope did, and here is exactly what came out." Then put the schedule risk in the risk register, because the long-lead items just got longer. THE LESSON A budget is not a forecast. It is a set of promises you will have to break in a specific order — compliance last, capacity second to last, finishes first — and the operators who survive are the ones who wrote the order down before the phone rang. ```

What the equipment line had to absorb

VE-3 deleted \$9,800 of custom millwork from the construction contract and replaced it with modular stainless service stations and back-bar shelving. Those are equipment, and equipment is a different frozen line: \$185,000, which does not grow either.

The modular units cost \$7,400 installed — which is why the substitution saved money at all. So \$7,400 has to come out of the equipment schedule, and this is exactly the moment Chapter 6 warned about: construction runs over, so the operator quietly takes \$20,000 out of the equipment budget. The difference between the disciplined version and the disaster version is one word: declared.

And then you say what the deferral costs, out loud, with a number. Without a bar glasswasher, glass goes to the main dish machine — which means the bartender or a barback leaves the bar with a rack and comes back, roughly eleven minutes of additional labor per bar shift across seven bar shifts a week:

$$11 \times 7 \times 52 = 4{,}004 \text{ minutes} = 67 \text{ hours a year} \times \$19 = \mathbf{\$1{,}273 \text{ a year}}$$

Against \$7,400 of deferred capital that is a payback of about five and a half years — so on the labor arithmetic alone, the deferral is defensible, and it is the item on the list that will be least regretted. Two costs do not appear in the number, though: glass breakage in transit, and the bartender being off the bar during service, which is a hospitality cost Chapter 23 would price higher than \$1,273.

The point is not that the answer is obviously right. The point is that we can call it defensible because we computed it. The undisciplined version defers the glasswasher because it was the cheapest thing on the schedule — a reason that happens to correlate with the right answer here and would have felt just as satisfying if it had been wrong.

Project total, unchanged:

Gas, electrical, and refrigeration — the three that are still open

Gas. Figure 7.6 totals on the order of 1.3 million BTU per hour of connected load against a one-inch service and a meter sized for a café's griddle. Interior re-piping sits in the plumbing line and the meter upsize in the \$4,000 utility connection fee. The exposure is not price, it is calendar: a service upgrade goes into the utility's queue, not yours, and a queue is not something a change order fixes. This is precisely the unknown the \$9,000 contingency exists for — which is why it was not spent on the hood.

Electrical. Chapter 6 found 200 amps with a full panel. The service upgrade and distribution survived value engineering untouched inside the revised \$36,300; everything on the equipment schedule with a plug depends on it, and Rule 2 says capacity does not get cut.

Refrigeration. The walk-in box is reused; the condensing unit is not. A nine-year-old compressor on a box you depend on for every service is a false economy, and the replacement is on the schedule as K-10. The box remains adequate-not-generous at 80 square feet — a constraint that becomes a purchasing frequency question in Chapter 13 rather than a construction question here.

7.7 Designing for the shifts you'll actually run — and for the ones you hope to

A floor plan is a bet on a schedule. Bellwether's schedule is five dinners, two brunches, one dark day, and a patio for about twenty weeks. Each of those is a different building.

Dinner, Tuesday through Thursday. Sixty to ninety covers in a 68-seat room. The design question is not capacity; it is how not to look empty. The countermeasures are physical: seat toward the front and the bar first, keep the back section unset until it is needed, and give the host a seating map that fills the room in a deliberate order. Chapter 22 owns the seating discipline; design owns the fact that the room needs a front worth filling — which is why the bar sits between the door and the dining room.

Friday and Saturday. 120 to 123 covers, the hearth at 82%, the pass at high load, a waiting area that has to hold real people, and every service station in use at once. This is the design night. Everything in §7.3 was sized here.

Brunch, Saturday and Sunday. 110 covers each at 1.62 turns and a \$24 check — and here is what operators miss: the constraint moves. Nobody orders a hearth-roasted chicken at 10:30 in the morning. The brunch bottleneck is the flat-top and the plate-up station, which is why tag K-5, a griddle with no dinner function, is on the schedule. Brunch also runs a much heavier coffee and non-alcoholic load at the service stations — a floor-plan consequence, because a station sized for water and silver is not a station sized for coffee service to 110 people.

Monday, dark. A dark day is not idle capacity; it is the deep clean, the hood cleaning solid fuel requires more often, the big deliveries, and — if you want it — the private-event day (Chapter 29). Design for it: tables that move, nothing bolted, and somewhere to put what you pull out.

Patio season. Twenty weeks, sixteen seats, twelve incremental covers a night, \$64,800 of revenue, and a peak-hour load that rises from 46 entrées to 50. It holds — at 89% on the hearth. Patio season is the season with no slack, and it arrives with the weather that makes the kitchen hottest.

Designing for the shifts you hope to run

Three cheap things, done now, that keep options open:

1. Extra electrical capacity and stubbed circuits where future equipment might go — a second reach-in, an espresso machine, patio heaters.
2. A design that matches your permitted use. Chapter 6 negotiated takeout, delivery, catering, private events, and retail sale. Where does a takeout order get staged so it does not sit in the server aisle (Chapter 28)? A three-foot shelf near the entry, drawn now, costs nothing and solves a problem you will otherwise solve badly for years.
3. Knock-out space and access — a wall you could open, a path equipment could come through, a mechanical route with room for one more unit.

And one expensive thing not to do: do not build for a volume you have not proven. The temptation is "let's carry the bigger hood so we can add equipment later." Sometimes that is right and an engineer should tell you. But Chapter 1's undercapitalization argument applies directly — money spent on capacity you do not yet need is money not in the working-capital reserve, and the reserve is what survives February.

🔍 Check Your Understanding

1. The re-cut budget lands on \$310,000 and \$110.71 per square foot — exactly the same figures Chapter 6 published. What actually changed?
2. Why was the \$9,000 contingency not used to help pay for the hood, given that \$9,000 would have reduced the value engineering from \$53,500 to \$44,500?
3. The patio appeared on the first value-engineering list and came off it. State the arithmetic that removed it.

(1: The scope. Three trade lines absorbed a \$49,000 increase, six lines absorbed \$53,500 of reductions, and \$4,500 moved from hard cost into soft cost for re-drawing and re-permitting. The price of a square foot did not move; what a square foot buys did. 2: Because a contingency is for costs that are certain to occur and cannot yet be identified, and the hood is now identified, drawn, and priced. Spending it would have left the project with no cushion at week zero against an unverified gas service, an unscoped waste line, eleven-year-old rooftop units, and abandoned slab drains. 3: The patio carries \$64,800 — 46.5% of the \$139,240 revenue bridge and 4.2% of forecast revenue. Cutting roughly \$8,000, about 2.6% of the construction line, to endanger 4.2% of revenue is a trade that loses money in year one and every year after.)

🍽️ The Business Plan

Checkpoint 7 of 40 — the floor plan, the equipment schedule, and the hood, priced.

Chapter 6 gave the plan an address and a problem. This chapter gives it the two documents that make every revenue number in the plan physically credible, and it closes the problem.

What the plan gains

The Design & Capacity section, containing:

Attach the front-of-house plan (Figure 7.3), the back-of-house plan (Figure 7.4), the capacity statement (Figure 7.2), the station-capacity analysis (Figure 7.5), the equipment schedule (Figure 7.6), and both construction budgets — Chapter 6's Figure 6.5 and this chapter's Figure 7.8 — side by side. A reader should be able to see, on facing pages, what was promised and what was delivered.

What this section settles

The 68 seats are justified, not asserted: 15.9 square feet per dining seat is the correct density for a \$46 check and the occasion mix Chapter 2 built, and going denser would buy eighteen seats at the cost of the check average that makes the whole plan work.

The kitchen can make the number. 95 covers comfortably, 123 on a Saturday, 135 with the patio at the edge — and the plan now names the edge, which is the hearth at 89% and the expediter at roughly 42 tickets an hour. That is what converts \$1,550,000 from a spreadsheet claim into an operational one.

The hood is priced and paid for. \$75,000 of real scope, funded by \$26,000 of allowances and \$53,500 of declared reductions, landing the construction line on exactly \$310,000. Nothing that protects a guest, satisfies a code, or produces a plate was cut.

What this section does not settle

The contingency is still 3.5%. \$9,000 against a \$259,500 contract sum where the rule of thumb is 10–15% — \$16,900 to \$29,900 light on cushion, with four named unknowns in front of it: an unverified gas service and meter, an unscoped waste line, two eleven-year-old rooftop units the lease may make the tenant's problem, and abandoned drains under the slab. Chapter 6 called this a deliberate weakness. It still is, it did not get better, and the plan must keep saying so.

The schedule got longer. A Type I hood, a tempered make-up air unit, and a rooftop fan are long-lead items, and a hood redesigned in month three does not ship on the original date. The re-scope protects the budget and exposes the calendar. Chapter 9 owns the countdown; the risk register owns this line.

The room is not the room that was drawn. A polished concrete floor, an open painted deck, engineered quartz, and simplified banquettes are a coherent design — but they are not the design in Chapter 3's brand section, and somebody has to confirm the room still reads as a \$46 room.

The patio is weather. \$64,800 of revenue depends on twenty usable weeks, and nothing here makes it rain less.

And the turns are still unproven. The room seats 68 and the kitchen sends the plates. Whether 1.4 turns of demand actually walks through the door is Chapters 22 and 24, and it is the largest open assumption in the plan.

Open questions carried forward

1. Can the building permit, the certificate of occupancy, the health permit, and the sidewalk-café approval all be obtained inside the schedule the lease assumes — and what does the re-scoped mechanical package do to plan review? (Chapter 8)
2. What do the long-lead items do to the opening date, and what does a six-week slip cost against a \$35,000 pre-opening budget? (Chapter 9)
3. Can a menu be written where no more than half the entrées route through the hearth, and does that menu still express the concept? (Chapters 10, 14)
4. Does the value-engineered room still support a \$46 check and a \$57 special-occasion occasion? (Chapters 3, 24)
5. Is \$9,000 of contingency survivable, and if the gas service comes back as a utility project, where does the money come from without touching the \$45,000 reserve? (Chapters 9, 33)
6. Does demand deliver 1.4 turns and 110 brunch covers on a room this size in this trade area? (Chapters 22, 24)
7. When does the bar glasswasher actually get bought, out of what, and who is accountable for it? (Chapters 31, 33)

Conclusion

Two thousand eight hundred square feet became 68 seats, 17 tables, four stations, one pass, and a number: fifty plates in the busiest hour of a Saturday in July. That is what design does. It converts a lease into a ceiling, and every revenue forecast in the rest of this book lives underneath it.

The seats are defensible because the arithmetic is on the page. Fifteen point nine square feet per dining seat is the density a \$46 check requires; compressing to twelve would have bought eighteen seats and cost four dollars of average check — \$144,560 a year — against an upside the kitchen could not have produced anyway. The kitchen is defensible because we sized it to the hour rather than the night, found the ceiling exactly where you would expect it, at the concept's signature piece of equipment, and wrote down the menu constraint that keeps it from being reached.

And the hood got paid for. Seventy-five thousand dollars of real scope across three trades against \$26,000 of allowances set on a false assumption, plus \$4,500 of re-drawing and re-permitting. Fifty-three thousand five hundred came out of finishes, millwork, restroom finishes, a ceiling, a lighting package, two weekends of the owners' own labor, and one piece of bar equipment deferred to year two with the labor cost of the deferral computed. The line landed on \$310,000 exactly. The cost per square foot did not move. What a square foot buys did.

Notice the order of those cuts, because it is the most portable thing in this chapter. Compliance never. Capacity never. Anything that turns capital into a permanent operating cost never. Then substrate, then finish, then deferral, and only then scope the guest experiences — and if you reach that last step, stop and ask whether the project is the right size. Every operator eventually faces a version of this meeting. The ones who survive it are not the ones with a bigger contingency. They are the ones who wrote the order down before the phone rang.

Two things remain open and the plan says so: \$9,000 of contingency against four named unknowns, and a schedule that just got longer.

Chapter 8 takes the building and asks whether you are allowed to open it. Entity, business license, health permit, certificate of occupancy, fire, signage, insurance — and the liquor license, frequently the longest, most expensive, and most binary item on the entire timeline. You have a room that works. Now find out what it costs to be permitted to use it.

Key Terms

Front of house (FOH) — every part of a restaurant a guest may occupy or see from a seat: entry, host stand, waiting area, dining room, bar, service stations, guest restrooms. Typically 60–70% of a full-service building. (Ch. 7)

Back of house (BOH) — everything a guest does not occupy: the line, prep, dish, storage, receiving, office, staff areas. Typically 30–40%; the share you give it is the share you take from seats. (Ch. 7)

Station — one cook's defined territory on the line: their equipment, refrigeration, mise en place, and the menu items they are responsible for during service. (Ch. 7)

The line — the row of cooking equipment where food is fired to order during service, together with the cooks working it. Bellwether runs four stations: garde manger, sauté, hearth, fry/sides/plate-up. (Ch. 7)

The pass (expo) — the counter between the line and the dining room where finished plates land, are checked, garnished, and sent; also the position that reads tickets, calls courses, and times the stations against each other. Chapter 14 covers expediting as a discipline. (Ch. 7)

Type I hood — a listed, grease-rated commercial exhaust hood over equipment producing grease-laden vapors, combining capture canopy, grease extraction and collection, fire suppression, and grease-rated duct discharging above the roof. Distinct from a Type II hood, which handles heat and condensate only. (Ch. 7)

Make-up air — outdoor air mechanically supplied to replace the air a kitchen exhaust system removes; usually the large majority of exhaust volume, tempered in cold climates, balanced so the kitchen sits slightly negative relative to the dining room. Undersized make-up air stops a hood from capturing. (Ch. 7)

Grease trap (grease interceptor) — a device separating fats, oils, and grease from wastewater before it enters the sanitary sewer; small hydromechanical units under a sink, large gravity interceptors in the ground. Sizing is set by the local authority and varies enormously by jurisdiction. (Ch. 7)

Kitchen workflow — the sequence and direction in which product, people, and plates move through a kitchen, from receiving through storage, prep, cooking, and plating to warewashing and waste. A good one has a single dominant direction of travel, no crossings, and no backtracking. (Ch. 7)

Seat-to-square-foot ratio — total building area divided by total seats (Bellwether: 41.2 sq ft per seat), or the square feet allocated per seat within a zone (its dining room: 15.9). State which. (Ch. 7)

Equipment schedule — the numbered, itemized list of every fixed and major movable item, with tag, description, quantity, dimensions and clearances, utility requirements, and who supplies and installs it. Simultaneously a purchase order, a permit document, and a capacity statement. (Ch. 7)

Occupant load — the number of people a building official calculates a space is designed to hold, from floor area and code-assigned factors by use. Not your seat count, usually larger, and it drives exit requirements and plumbing fixture counts. Varies by jurisdiction. (Ch. 7)

Value engineering — the disciplined substitution of a lower-cost means for the same function, to return a project to budget without reducing what the building has to do. Distinct from cutting, which reduces function. Never compliance, never capacity, never anything that converts capital into a permanent operating cost. (Ch. 7)

Spaced Review

1. Chapter 6 published a construction line of \$310,000 at \$110.71 per rentable square foot. Chapter 7 publishes the identical two figures after absorbing \$75,000 of solid-fuel scope against \$26,000 of allowances. Explain, in three sentences, how both statements are true — and what a reader would have to see to verify it.
2. Chapter 6 argued that "rent is fixed; the denominator is a hope." Bellwether's rent per cover is \$2.63 on 36,140 covers. Using this chapter's capacity work, name two design decisions that change the denominator and one that cannot.
3. From Chapter 2: the \$46 average check was built from an occasion mix in which 30% of covers are special-occasion guests spending \$57. Explain how §7.1's density arithmetic connects to that number, and what would have happened to it at 12 square feet per seat.
4. From Chapter 1: prime cost is the sum of cost of goods sold and total labor as a percentage of sales. Using §7.4's step-count method, explain why a floor plan is a prime-cost document, and compute the annual dollar value of a layout change that removes four paces from an average round trip on a 40-party night, five nights a week, at \$22 an hour. (Twelve trips per party; 2.5 feet per pace; three feet per second.)
5. The recurring question: the partners defer the \$7,400 bar glasswasher to year two and absorb roughly \$1,273 a year of additional bar labor. Does this decision move prime cost, and in which direction? When does it hit the bank account as opposed to the P&L — and what would you have to measure a year later to know whether it was the right call?