Appendix J — The Twelve-Month Opening Timeline

"Nobody's first restaurant opened on the date they told their mother." — constructed; the thing operators say about the second one

J.1 How to use this timeline, and why it will be wrong

This appendix is a planning instrument. It is not a prediction, and if you treat it as one it will cost you money.

Here is the honest structure of the problem. About half the elapsed time between signing a lease and serving a paying guest is spent waiting on people who do not work for you and have no financial interest in your opening date: a building department plan reviewer, a health department plan examiner, a fire inspector with a queue, an alcohol authority working a statutory clock, a landlord's contractor who has not finished the demising wall, a factory with a sixteen-week backlog on the one piece of equipment your concept depends on.

You cannot compress those. You cannot escalate them. In most jurisdictions you cannot pay to jump the line, and where you can, what you buy is measured in days.

So the correct response to schedule risk in a restaurant build is not a tighter schedule. It is a later announced opening date and more working capital. Those are the only two levers that actually work. Everything else — the aggressive Gantt chart, the "we'll train in week one" plan, the contractor who promises twelve weeks when everyone else said eighteen — is optimism wearing a hard hat.

The most expensive scheduling mistake in the business

One error costs more than all the others combined, and it comes in two parts, usually committed by the same person in the same quarter.

Part one: signing a lease whose rent commences before you have permits in hand. Rent commencement is a negotiated term. It can be tied to lease execution, to delivery of the space, to a fixed calendar date, to issuance of your building permit, or to the earlier of certificate of occupancy and a fixed outside date. Those are radically different deals wearing identical language, and a first-time tenant frequently signs the worst version without noticing, because the rent per square foot is the same in all of them.

Part two: announcing an opening date to the public. Once a date is in a press release or on a neighborhood feed, you have converted a schedule risk into a reputational one — and you will then make bad decisions to protect the date. Opening without your liquor license. Opening with two stations you have not trained. Opening before the health inspector has actually issued the permit rather than merely walked the space.

🧮 Run the Numbers

What rent commencement is worth.

The Bellwether plan carries 2,800 square feet at \$28 per square foot base plus \$6 in NNN — triple net, meaning the tenant pays a share of taxes, insurance, and common-area costs on top of base rent. That is \$95,200 a year, or **\$7,933 a month, or about \$1,830 a week. The lease grants three months of free rent**.

Version A — rent commences at lease execution. You sign at T-12. Three months free carries you to T-9. You then pay rent for eight months on a space that is a job site, a dumpster, and eventually a very clean empty restaurant: $8 \times \$7{,}933 = $ **\$63,467**.

Version B — rent commences at the earlier of certificate of occupancy or a fixed outside date. You take occupancy at T-1 with the certificate in hand, and the free-rent period covers your first two months of operation, when you need it most. Pre-opening rent paid: \$0.

The difference is roughly **\$63,500** on an identical lease at an identical rate. That is 1.8 times the entire \$35,000 pre-opening budget in §J.9, and more than the \$45,000 working-capital reserve. It is the largest number in this appendix, and it is decided by one clause, in one meeting, eleven months before you cook anything.

You will not always win Version B; landlords resist open-ended commencement for the obvious reason. The realistic middle is commencement at the earlier of CO or a fixed outside date generous enough to absorb one full permit-review cycle. But know which version you are signing, price the difference, and have your attorney put the number in front of you before you initial the page.

The months below are not promises. They are a structure for knowing, in any given week, which single thing is actually holding up your restaurant.

J.2 The critical path

A critical path is the sequence of dependent tasks whose total duration sets the completion date. Anything on it, if it slips a week, slips your opening by a week. Anything off it can slip and cost you nothing but sleep. Most first-time operators over-invest, dramatically, in the things that are not on it.

FIGURE J.1 — The critical path and what runs beside it              [the Bellwether plan]

  MONTH            T-12 T-11 T-10  T-9  T-8  T-7  T-6  T-5  T-4  T-3  T-2  T-1  OPEN
  -----------------------------------------------------------------------------------
  THE CRITICAL PATH  (a week lost here is a week lost on the date)
  lease signed       ##
  design + MEP            ####:####:####
                                    ^-- undersized hood + grease trap found here
  permit submitted                   ##
  permit review                      ....:....:....
  permit issued                                  ##
  construction                                   ####:####:####:####:####
  equipment deliv/install                                       ####:####
  health + fire final                                                     ##
  certificate of occupancy                                                ##
  liquor license                     ....:....:....:....:....:....:....:###
  hire + train                                                            ####
  soft open                                                                 ##
  OPEN                                                                        >>
  -----------------------------------------------------------------------------------
  RUNS BESIDE IT  (has a deadline, does not set the date)
  entity / EIN / insurance ##:##
  menu R&D + cost cards          ####:####:####:####:####
  purveyor accounts                                      ####:####
  POS + reservations build                                     ####:####
  brand, signage, website                        ####:####:####
  marketing + PR                                                     ####:####
  -----------------------------------------------------------------------------------
  ##  active work        ....  waiting on someone else        >>  revenue begins

The path in prose

Lease signed → design. Nothing begins until you control the space; no architect will draw against a floor plan you might not get. The lease also fixes what you are allowed to build — use clauses, hours, venting rights, roof access for makeup air, the landlord's approval rights over your drawings.

Design → permit submission. You cannot submit what is not drawn. Construction documents for a restaurant include architectural, mechanical, plumbing, and electrical sheets plus an equipment schedule, and health department plan review usually needs its own set showing finishes, hand sinks, warewashing, and food flow.

Permit submission → permit issuance. The longest waiting block on the chart and the one you control least. Plan an initial review cycle and a correction cycle: reviewers issue comments, you respond, they review again. First-time applicants clear on the first pass rarely.

Permit issuance → construction. No permit, no work — and a contractor who offers to start "just the demo" without one is offering to put your certificate of occupancy at risk to make his own schedule look better.

Construction → equipment delivery and installation. The hearth, hood, walk-in, and warewasher have to land in a building that is ready for them, in the right order, each connected by a licensed trade and then inspected.

Installation → health and fire inspections → certificate of occupancy. The certificate of occupancy (CO) is the building department's statement that the space may be occupied for its intended use. It issues only after every other final passes, which makes it the hardest gate in the sequence and the one most operators discover late.

In parallel, the liquor license. It does not depend on construction, which is why it can start early — and why failing to start it early is the most self-inflicted delay in this appendix.

CO → hiring, training, soft open, open. You can hire before the CO. You should not start hourly staff before it, because paid staff standing in an unfinished building is the fastest way to convert a schedule slip into a payroll problem.

What feels urgent and is not

These matter. None of them sets the date, and every one will try to eat a month of your attention: the logo and the menu font; the website and the "coming soon" post; chairs, plateware, glassware, the napkin fold; the final wine list (the license is critical path, the list is not); the playlist, the uniforms, the neon. The menu itself has to be right by T-2 — it does not have to be right at T-9.

A separate category: things not on the critical path that carry a hard, unforgiving deadline anyway. Workers' compensation coverage in force before your first employee works an hour. Payroll onboarded before the first pay period. Food-handler certification for everyone who touches food. Insurance bound before the contractor mobilizes. These do not move the date; they stop you cold if you miss them.

J.3 The twelve months

Read the role names, not the people. At Bellwether the chef and the general manager are the two owning partners, so "you," "the chef," and "the general manager" may collapse into two humans doing seven jobs. That is normal in an independent, and it is also why first-time openings slip: the same two people are the bottleneck on the design, the menu, the hiring, and the permits at once.

Funding is closed and in place before T-12. Everything below assumes the money exists.

T-12 · Control the space and assemble the professionals

This month converts a concept into a legal and physical position. You are not designing yet; you are acquiring the three things design requires — a signed lease, a formed entity, and professionals under contract. The lease negotiation is the consequential work, and it is nearly finished before the month starts. Beyond commencement (§J.1), the terms that matter to this timeline are delivery date and delivery condition, venting and roof rights, tenant-improvement allowance mechanics, and the landlord's approval rights over your drawings — including how long the landlord has to respond and whether that clock runs concurrently with the permit clock or after it.

Money that moves. The security deposit lands the day you sign — commonly one to three months' rent, roughly \$8,000 to \$24,000 here — and it appears nowhere in the \$620,000 project cost as a named line. That is not a flaw peculiar to this plan; it is the normal condition of restaurant budgets. Add the attorney's lease work, the architect's retainer, entity fees, and the first insurance binder: about **\$22,000** in a month that produces no visible progress at all.

T-11 · Schematic design and the first real budget

Now the space becomes a drawing. Schematic design tests whether 56 dining seats, 12 bar seats, a wood-fired hearth, and a functioning kitchen actually fit inside 2,800 square feet — 1,700 front of house, 900 back of house, 200 storage and office — with legal egress, accessible routes, and a food flow that does not cross dirty and clean. The hard conversation is between the chef and the architect about kitchen square footage: every foot the kitchen takes is a seat the dining room does not get, and every foot it does not take is a cook working in a space that produces slower tickets for ten years.

Money that moves. Schematic design fee, MEP retainer, and the survey — about **\$16,000**. Nothing ordered, nothing built, roughly \$38,000 into the project.

T-10 · Design development, and the month the hood problem surfaces

This is the month the plan discovers that the existing Type I hood and grease interceptor left by the previous café tenant are undersized for a wood-fired hearth. Solid fuel is a different regulatory animal than a gas range: it typically demands higher exhaust volume, often its own dedicated duct, spark arrestance and grease-removal provisions, and matched makeup air so the building does not go negative and pull the flue backward into your dining room. The interceptor, sized for a café dish pit, is not sized for a full kitchen. Discovering this in design is expensive; discovering it in construction is expensive and slow, because by then it is a redesign, a permit revision, and a stopped job. The remediation cost is worked in Chapters 6 and 7 and sits inside the \$310,000 construction line. What concerns this appendix is the schedule: redrawing the mechanical package, re-coordinating the makeup-air unit against the roof structure, and confirming gas capacity pushes permit submission out by roughly four weeks.

Money that moves. Design development, the MEP redesign, and the landlord submission package: about \$19,000. Still nothing built.

⚠️ Where the Money Leaks

"Second generation" is not a discount. It is a bet.

A second-generation space — one that was already a restaurant — is genuinely cheaper to build than raw shell, because the grease interceptor, hood, gas service, floor drains, and three-compartment sink may already exist. That is why operators chase them.

The bet is that what exists is adequate for your concept and current under today's code. Both halves fail routinely. Equipment installed to the code in force fifteen years ago does not have to be replaced until you touch it — and a permit for your build is you touching it. A hood that legally served a café can be illegal the moment a hearth goes underneath it.

What the disciplined operator does: pay a mechanical engineer and a plumber to survey existing conditions before the lease is executed, or make the lease contingent on a satisfactory survey. A \$3,000 to \$6,000 pre-lease investigation is the cheapest insurance in this appendix. Bellwether caught it in design, which cost money. Catching it in construction costs money and a season.

T-9 · Permit submission

Construction documents are complete, coordinated, stamped, and submitted. This is the most important date on the chart that nobody celebrates, because from the moment of submission your project sits in someone else's queue and the only thing you control is how fast you answer comments. Submit everything you can simultaneously. Building, mechanical, plumbing, and electrical usually travel together; health department plan review is typically a separate submission to a separate agency; fire review may be separate again. The sign permit is almost always its own track, frequently the slowest, and frequently forgotten until the sign is fabricated.

Money that moves. Permit and plan-review fees — in most jurisdictions calculated off declared construction valuation and often substantial — plus the liquor application, the sign permit, and the contract deposit to the general contractor: about \$32,000.

T-8 · Waiting, and ordering the things that cannot wait

Nothing on the critical path advances this month except a reviewer's queue position. This is where impatient operators do damage: they start "temporary" work without a permit, or they decide that since nothing is happening they may as well pick chairs for three weeks. Do the parallel work instead, and do the one genuinely hard thing — order the long-lead equipment. It is uncomfortable, because you are committing five figures before you hold a permit that could theoretically force a design change. The rule that resolves it: order the items whose specification cannot change (hearth, walk-in box, hood and makeup-air unit, warewasher), and hold everything that could still move.

Money that moves. Equipment deposits are the shock. Thirty to fifty percent at order is common, and on a \$185,000 package the long-lead portion alone can be **\$80,000** out the door before a wall is framed. This is where most first-time operators genuinely understand what a working-capital reserve is for.

T-7 · Permit issuance and mobilization

The permit issues — in this plan at the end of T-7, after one correction cycle. If it does not, construction does not start and everything below shifts right by exactly as long as the delay. With the permit in hand the contractor mobilizes: site protection, dumpster, temporary power, permits posted, and a schedule of values that becomes the basis for every progress payment for the next five months. This is also the month to open the search for salaried kitchen leadership. A sous chef worth hiring is currently employed somewhere, gives notice, and has a life; sixty to ninety days from first conversation to first shift is normal, which is why hiring this role late is item ten in §J.8.

Money that moves. Permit issuance fees, contractor mobilization, and the insurance down payment: about \$26,000.

T-6 · Demolition and the discovery month

Walls open, and roughly a third of what you assumed about the space turns out to be wrong. Concealed conditions — a slab that is not where the drawing says, a sanitary line running the wrong direction, a structural member in the path of the new duct, adhesive that has to be abated — surface in the first three weeks of demolition and produce the first change orders. Budget emotionally for this. The construction contingency exists precisely for this month, and an owner who fights every change order in T-6 will have spent the goodwill they need in T-2.

Money that moves. The first construction progress payment (draw) against the schedule of values, the first change orders, and utility upgrade deposits: about \$54,000.

T-5 · Rough-in

The invisible half of the build: mechanical, plumbing, and electrical rough-in, the hood curb set and the duct run, the makeup-air unit placed, framing, and blocking for everything that will hang on a wall. This month carries an inspection milestone first-timers forget — rough inspections, where the building department examines rough-in before the walls close. A failed rough inspection in T-5 is cheap. The same defect found at final in T-2 means opening finished walls.

Money that moves. Construction draw 2, the sign deposit, and purveyor account setups: about **\$68,000**. Cumulative cash out is now roughly \$307,000, five months from a dollar of revenue.

T-4 · Close the walls; do not announce anything

Walls close, floors go down, finishes begin, the walk-in shell is assembled in place. The restaurant starts to look like a restaurant, which is exactly why you will be tempted to announce a date. Do not. The reason is arithmetic, not superstition: at T-4 you have passed no final inspection, hold no certificate of occupancy, may not have a liquor license, and have not commissioned the hearth. Four of your ten biggest slip risks (§J.8) are still entirely ahead of you. What you do instead is the employment infrastructure — unglamorous, deadline-bound, and completely fatal if late.

Money that moves. Construction draw 3, legal review of the handbook, payroll setup, and equipment balances coming due: about \$66,000.

T-3 · Equipment lands

The month the kitchen becomes real. The hearth is delivered and set — for a masonry or refractory hearth this is multi-day work involving flue connection, structural support, and curing. The hood and makeup-air unit are connected, the walk-in runs, line equipment is placed, plumbed, gassed, and wired. Your salaried hires start: the sous chef, the assistant manager. They are on payroll now, working in a construction site, which is expensive and correct — they are building the systems that will carry the opening.

Money that moves. The largest month of the project: construction draw 4, the equipment balance on delivery, and the FF&E package — about \$112,000.

👨‍🍳 On the Line

A hearth is not an oven you plug in.

Solid-fuel equipment has a commissioning period no schedule ever includes. A masonry or refractory hearth needs a controlled curing and burn-in sequence over several days before it holds temperature. Then your cooks have to learn this hearth: where the hot zone actually sits, how long the coal bed takes to build for the first order at 5:30, how much wood holds through a 96-cover Friday, what happens to the fire when the makeup air kicks on.

Nobody learns that in an afternoon. On a hearth-driven menu the practical rule is two weeks of live fire before you serve a paying guest — which means the hearth must be operational and legal at T-2, not T-1. Build that backward into your delivery date and tell your contractor why, because "the chef wants to practice" does not survive a schedule negotiation and "the hearth is not commissioned and we cannot execute the menu" does.

T-2 · Punch list, deep clean, first inspections

Substantial completion. The punch list gets written and worked, and then the building is cleaned in a way it will never be cleaned again, because you cannot pass a health inspection through construction dust. This is when inspections start, and the order matters: fire final, hood suppression test and certification, then mechanical, plumbing, electrical, and building finals, then health — and only then the certificate of occupancy. Request a pre-inspection walkthrough from the health department if your jurisdiction offers one. Many do. It is free, it is not graded, and it converts a failed first inspection into a punch list.

Money that moves. Construction draw 5 with retainage held back, the smallwares package, and final inspection fees: about \$46,000.

⚠️ Where the Money Leaks

The tenant-improvement allowance is a reimbursement, not a fund.

Bellwether's \$75,000 tenant-improvement allowance reads like capital. It is not. In the overwhelming majority of leases it is paid by the landlord after the work is complete, after you deliver paid invoices, after unconditional lien waivers from every contractor and subcontractor, and often after a certificate of occupancy is in hand. Thirty to sixty days after that package is accepted is common.

Which means you pay for \$75,000 of improvements out of your own account in T-6 through T-2 and get it back sometime after you open — possibly in month two of operation.

Two things follow. Do not count the allowance as opening cash; count it as a receivable with an uncertain date. And read the reimbursement conditions in the lease before you sign, because an allowance you cannot document is an allowance you do not get. Lien waivers collected as you go are trivially easy. Lien waivers chased six months later from a subcontractor who has moved on are not.

T-1 · Thirty days

The compressed month: certificate of occupancy, health permit, hourly hiring, training, commissioning, POS validation, first inventory, friends and family, soft open — and only now, with the CO and health permit physically in hand, the announcement. This month gets its own section; see §J.4.

Money that moves. Training payroll, opening inventory, marketing, remaining deposits, and retainage release: about **\$34,000** — most of the \$35,000 pre-opening budget in §J.9, landing in a single month.

FIGURE J.2 — Cash out of the project account, by month     [the Bellwether plan; illustrative]

   T-12  ██                                     $22,000   deposit, retainers, formation
   T-11  ██                                     $16,000   schematic design
   T-10  ██                                     $19,000   design development + MEP redesign
   T-9   ███                                    $32,000   permits, CDs, GC contract
   T-8   ████████                               $80,000   long-lead equipment deposits
   T-7   ███                                    $26,000   mobilization, insurance
   T-6   █████                                  $54,000   construction draw 1
   T-5   ███████                                $68,000   construction draw 2
   T-4   ███████                                $66,000   construction draw 3
   T-3   ███████████                           $112,000   draw 4 + equipment balance + FF&E
   T-2   █████                                  $46,000   draw 5 + smallwares
   T-1   ███                                    $34,000   retainage, pre-opening, inventory
         ───────────────────────────────────────────────
         TOTAL OUT BEFORE THE FIRST DOLLAR OF SALES   $575,000

   That is the $620,000 project cost less the $45,000 working-capital reserve.
   Soft costs — design fees, permit fees, deposits — sit inside the construction
   line, which is where a contractor-priced budget usually puts them.

   Not shown, because it has not arrived: the $75,000 tenant-improvement
   allowance. It is a reimbursement, paid after the work, after the lien
   waivers, commonly 30 to 60 days after substantial completion.

   The $45,000 reserve is deliberately absent from this column. It is what
   you live on during the ninety days in section J.6.

Read the shape rather than the numbers. Two-thirds of the cash leaves in the last five months, and the single largest month is T-3 — the month the equipment lands, which is also the month a slip is most expensive, because you have paid for everything and still cannot open.

J.4 The final thirty days

Everything collides. Four workstreams that were sequential for eleven months become simultaneous for four weeks: inspections and the CO, hiring and training, equipment commissioning, and systems — POS, inventory, scheduling, and the menu as an actual printed object.

The roster you are staffing toward:

Twenty-four scheduled positions on the weekly template, filled by thirty-one people because servers, hosts, and server assistants are part-time. Both numbers matter: the 24 drives your labor model, the 31 drives your training cost, your onboarding paperwork, and your certification count.

Week 4 before opening (days 30 to 24)

The certificate of occupancy is the whole week. If it does not issue, nothing below happens on schedule and you push everything right — which is fine, because you have not announced anything.

• Close the punch list, prioritizing items that block inspection.
• Final inspections completed; certificate of occupancy and health permit issued.
• Hourly offers accepted; start dates set; onboarding paperwork prepared.
• POS build signed off against the final menu: every item, modifier, price, and seat map.
• Menu costing reconfirmed against actual delivered invoice prices, not quotes.
• First food order placed for training production. Order small; you will throw some away.
• Utilities transferred into the operating entity; deposits paid.

Week 3 before opening (days 23 to 17)

• All-staff orientation, paperwork first and completely: Form I-9 employment eligibility verification, tax withholding, direct deposit, handbook acknowledgment, and whatever harassment and safety training your jurisdiction requires. Nobody works an unpaid minute of it.
• Food-handler certification for everyone who touches food; certified food protection manager per local rule.
• Equipment commissioning: hearth burn-in, range and fryer calibration, refrigeration temperature logs started, warewasher sanitizer checked and documented.
• Back-of-house training: station setup, mise en place lists, prep sheets, the first ten recipes produced to spec.
• Front-of-house product knowledge: menu, allergen map, wine list, cocktail list, the story of every dish.
• Standard operating procedures posted where the work happens, not in a binder in the office.

Week 2 before opening (days 16 to 10)

• Every dish produced daily, plated to the photograph, timed.
• Service simulation: staff eat, staff serve each other, tickets ring through the live POS, expo calls real times. Run it three times. It will be bad the first time, and that is the point.
• Beverage training: pours measured, cocktails specced and timed, wine service drilled.
• POS validated against real tickets — voids, comps, splits, transfers, discounts, gift cards, tip declaration.
• Reservation platform live but closed to the public.
• Health inspection and permit, if not already issued; liquor license confirmed in hand.
• Announce the opening date. With the CO issued, the health permit issued, and the license in hand, the date is now a fact rather than an aspiration.

Week 1 before opening (days 9 to 1)

• Two friends-and-family services, a dark day of corrections between them, one soft open. See §J.5.
• Full physical inventory counted the day before you open. This is the beginning inventory that makes week one's food cost real, and there is no substitute for counting it yourself.
• Opening par levels set for every product.
• Final schedule published, deliberately overstaffed — §J.6 says what that costs and why you pay it.
• Reservation book opens to the public at reduced capacity.

J.5 Opening week, day by day

Bellwether serves dinner Tuesday through Saturday plus weekend brunch, so the week is built backward from a Tuesday open.

FIGURE J.3 — Opening week at a glance                             [the Bellwether plan]

  MON  D-8   Load in. Final deliveries. All-staff meeting. Walk the room.
  TUE  D-7   Mock service. Staff serve staff. Full menu, live POS, real timing.
  WED  D-6   FRIENDS & FAMILY #1     ~40 covers   full menu, food comped
  THU  D-5   FRIENDS & FAMILY #2     ~55 covers   full menu, beverage charged
  FRI  D-4   DARK. Debrief, retrain the two stations that failed, deep clean.
  SAT  D-3   SOFT OPEN               ~45 covers   invite list, limited hours
  SUN  D-2   DARK. Full physical inventory. Reset pars. Publish the schedule.
  MON  D-1   Prep. Final walkthrough. Staff meal. Everyone home by 9:00 p.m.
  TUE  D+1   OPEN TO THE PUBLIC      capped at ~65 covers, reservations at 60%
  WED  D+2   Open. Raise the cap only if last night was clean.
  THU  D+3   Open. First weekly numbers assembled.

  Brunch does not launch in week one. It launches in week three or four.

Monday, D-8. The last deliveries land and everything gets a home. Walk the room with the general manager and the chef and fix the twenty small things you have stopped seeing: the wobbling table, the light aimed at a guest's eyes, the station that is eleven inches from where a cook can reach it. Hold the all-staff meeting now rather than on opening day, when nobody will hear you.

Tuesday, D-7. Mock service. Staff serve staff. Full menu, live POS, real ticket times, expo calling. Half eat, half work, then swap. It will be slow and disorganized, and you will discover four things about your kitchen's flow you could not have found any other way. That is the point, and it costs nothing but food.

Wednesday, D-6. Friends and family, service one. Roughly 40 covers, invited, spread across the full window. Food comped; beverage comped or charged at cost — check your license and your jurisdiction's rules on complimentary alcohol first, because some states restrict it. Run the full menu. Ask for honest feedback and mean it. Track ticket times.

Thursday, D-5. Friends and family, service two. Roughly 55 covers. Slightly larger, slightly faster, beverage charged so the POS and the bar workflow get exercised properly. You are looking for whether last night's corrections held.

Friday, D-4. Dark. Do not serve. This day exists to fix what the last two nights exposed, and operators who skip it because "we were nearly there" open with the same two broken stations.

Saturday, D-3. Soft open. Roughly 45 covers, invite list or a quietly announced limited service, shortened hours, full menu, everything charged. The first service where money changes hands and the first honest read on your timing.

Sunday, D-2. Dark, and count. Full physical inventory — food and beverage, every shelf, every walk-in, every bottle. Set opening pars, publish the week's schedule, and send everyone home.

Monday, D-1. Prep, final walkthrough, staff meal, and a hard stop. Everyone leaves by nine. An exhausted staff on opening night is a service problem you built yourself.

Tuesday, D+1. Open. Cap it. The plan's steady-state Tuesday is around 95 covers; open at roughly 65, with the book at about 60% of capacity and one section held back. You are not maximizing revenue on night one. You are producing a room full of people who tell other people it was good.

When to accept reservations from the public

The rule is simple and operators break it constantly: open the book only when the certificate of occupancy, the health permit, and the liquor license are physically in your possession, and only for dates after your soft opens have run.

Practically, open the book 10 to 14 days ahead of the public opening, at 60% of capacity, with a deliberately wide spread of seating times. Then raise the cap by roughly 15% a week for the first month — and only after a clean night, never after one where tickets ran long or the kitchen went down.

🤝 Hospitality

Tell them you are new.

There is a persistent belief that a soft open should feel like a normal restaurant and that admitting you are new is admitting weakness. It is the reverse. A guest told at the door — warmly, briefly, without apology — that this is the restaurant's fourth service ever and that the team is grateful they came will forgive a fourteen-minute appetizer. A guest told nothing will judge you against a restaurant that has been open three years, because that is the only comparison they have.

This is not spin. It is setting an accurate expectation, which is what hospitality mostly is. And it does something commercially specific: it converts a slow night from a bad review into a story the guest tells about being there at the beginning. The second visit is where the business actually lives, and an invitation to come back and see the difference is a far better offer than a comped dessert.

One caution. This works exactly once, and only in the first two or three weeks. An operator still explaining in month three that they are new has stopped setting expectations and started making excuses.

J.6 The first ninety days

A new restaurant does not open at its planned volume, and it does not open at its planned cost structure. Both take a quarter to arrive, and the gap between them is a real, budgetable amount of money that has to exist before you open.

The revenue ramp

The Bellwether plan does not assume a flat year. It assumes the year builds:

Inside Q1 the shape matters more than the total. Weeks 1 to 4 run an opening spike — curiosity traffic, friends, neighbors, the people who walked past the paper on the windows for five months. Weeks 5 to 9 dip, sometimes sharply, as the curiosity clears and before any habit has formed. Weeks 10 to 13 give the first honest read on what your restaurant actually is.

Operators who do not know this happens panic in week six. Operators who do know use weeks 5 through 9 to fix things, because a slow room is the only room in which you can actually retrain a station.

The cost ramp, which is the expensive half

Read the first row and the last row together, because that is the whole argument of this section. The plan targets a 60.0% full-year prime cost — \$930,000 on \$1,550,000 of sales. It also plans a first quarter at 66.6%, which consumes \$239,760 of that \$930,000 in thirteen weeks. Which leaves \$690,240 of prime cost to cover \$1,190,000 of sales across weeks 14 through 52.

$$\frac{\$690{,}240}{\$1{,}190{,}000} = 58.0\%$$

Weeks 14 through 52 must average 58.0% prime cost for the year to land on plan. Not 60%. Fifty-eight. The 60% target you have carried since Chapter 1 is a full-year average that already includes a quarter you have not earned yet, and the moment Q1 closes at 66.6% the number you are actually managing to for the remaining thirty-nine weeks is two full points tighter than the benchmark.

That gap is the ramp. It is normal. It is survivable. It is survivable only if it was planned for, because the alternative is discovering in month four that the target you thought you were chasing has moved away from you.

🧮 Run the Numbers

What the ramp costs, in dollars.

Suppose Q1 had run at the full-year target of 60.0% rather than 66.6%.

• Prime cost at 60.0% on \$360,000 = **\$216,000**
• Prime cost at 66.6% on \$360,000 = **\$239,760**
• Difference = \$23,760

Twenty-three thousand seven hundred sixty dollars of cost, in thirteen weeks, that produces nothing you can point to. It is not waste, exactly — it is the price of a kitchen learning its own hearth, a floor learning its own tables, and a manager learning what Tuesday looks like here.

Set that against the **\$45,000 working-capital reserve**. The ramp alone consumes **52.8%** of it, before a single slow February, a single broken compressor, or a single week that comes in \$4,000 under.

That is what a reserve is for, and it is also why \$45,000 on a \$620,000 project is defensible but not comfortable. An operator who spends the reserve on a nicer bar top has funded the wrong thing.

The 6.6 points do not distribute evenly. Most of the overage is labor: you deliberately overstaff opening week and keep more people on the floor than the covers justify for a month, because a green team at correct staffing produces bad service, and bad service in week two costs more than the labor does. Second is food cost — waste from over-production against a forecast nobody can make yet, comps and remakes from a kitchen still learning timing, portioning that has not tightened, a walk-in over-ordered because nobody knows the pars. Third and smallest is beverage: over-pouring, breakage, and a wine list nobody on the floor can sell yet.

🧾 Read the Numbers

```text FIGURE J.4 — "Week one" [the Bellwether plan] THE ARTIFACT The first weekly flash report. Five dinner services, no brunch, covers capped, week one of operation. THE CONTEXT 68-seat full-service restaurant, wood-fired hearth, staff of 31. Reservation book held at 60% of capacity. Every ticket rung on a POS the team has used for eleven days.

                 Covers    Tue 62 · Wed 71 · Thu 78 · Fri 96 · Sat 104 =  411
                 Average check                                        $44.00
                 Sales                                              $18,084
                   food (76%)              $13,744
                   beverage (24%)           $4,340
                 Food cost      (36.0% of food sales)                 $4,948
                 Beverage cost  (26.0% of bev sales)                  $1,128
                 TOTAL COGS                                           $6,076    33.6%
                 Labor: 580 hourly hours @ $14.70          $8,526
                        salaried allocation                $2,900
                        payroll taxes and benefits (10%)   $1,143
                 TOTAL LABOR                                         $12,569    69.5%
                 ──────────────────────────────────────────────────────────────
                 PRIME COST                                         $18,645   103.1%

WHAT IT SHOWS In week one you spent $103 for every $100 that came through the register. That is not a disaster; it is what a capped opening week with 580 hourly hours on 411 covers arithmetically has to produce. The average check landed at $44 against a $46 plan — servers who cannot yet sell the wine list and guests ordering carefully. Food cost at 36% against a 30% target is six points of waste, remakes, and over-production. WHAT IT DOESN'T It does not tell you which of those six food points is waste, which is over-portioning, and which is comps, because week one has no theoretical usage to compare against. It does not separate training hours from service hours. And with one week of history it cannot distinguish a problem from a start. THE DECISION Do not cut labor this week. Cut it in week three, against a forecast, one position at a time, and watch ticket times. Do start tracking waste on a sheet by the trash can on Monday, and do count inventory again on Sunday — the second count is the first one that means anything. THE LESSON Week one is not a measurement. It is a baseline you are taking so that week four has something to be compared to. The number that matters is not 103.1%; it is the slope from here to week thirteen. ```

The ninety-day discipline

1. Count inventory weekly. Food and beverage, both, counted by you for the first six months, and compute prime cost weekly against sales. Not monthly — you cannot afford to learn about week two in week nine.
2. Publish a forecast and staff to it. By week three you have three data points per weekday. The staffing guide is the mechanism by which the 6.6 points come down.
3. Read the comp and void report daily. In a new restaurant it is the fastest diagnostic you own: which dish is failing, which station is behind, which server is struggling — at a cost you have already paid.
4. Fix the menu at week six, not week two. Cut the two items nobody orders and the one that blows every ticket. Two weeks of data is noise; six weeks is a signal.
5. Do the 30-, 60-, and 90-day conversations with every hire. In an industry running roughly 75% annual turnover, the people who opened your restaurant are the cheapest labor you will ever have, and most who leave in the first quarter leave for reasons that were fixable and never surfaced.
6. Delay brunch. The plan needs it; week one does not. Launch it in week three or four, when dinner is stable, and treat it as a second opening with its own training and its own soft launch.

J.7 The permit and licensing gauntlet

Read this caveat before the table and again after it: every item below is jurisdiction-specific. Names, sequences, agencies, fees, and lead times vary enormously by state, county, and city, and they change. Some jurisdictions combine approvals that others separate; some issue in days what others take months to issue. Nothing here substitutes for calling your own authority having jurisdiction and asking directly — which is free, takes an afternoon, and is the highest-return hour in this appendix. The ranges below are directional: a checklist of what to ask about, not a schedule.

Two more that operators forget until a letter arrives: music licensing through the performing-rights organizations if you play recorded or live music, and an outdoor seating or sidewalk-café permit for the 16-seat patio, which in many cities is a separate application with a separate season and a separate fee.

⚖️ Code and Compliance

The liquor license is the one you cannot rush, so start it first.

Alcohol licensing is the most variable item in this appendix and the most damaging when it runs late, because a full-bar concept with a roughly 40-bottle wine list has budgeted beverage at 28% of sales. Opening without a license means opening with 72% of your planned revenue and a bar staff you cannot deploy.

The structures vary enormously. Some jurisdictions issue on a rolling administrative basis in a couple of months. Some run a statutory notice period, a public posting on the premises, and a hearing where neighbors may object. Some are quota jurisdictions, where the number of licenses is capped and a new operator must buy an existing license on a secondary market — at a price that can rival the equipment package, with a transfer process of its own. There are also constraints nobody warns you about: distance requirements from schools or places of worship, food-to-alcohol sales ratios you must maintain, mandatory server training, and hours-of-service limits that may not match the hours you planned.

What to do. In the first week of T-12, find out which structure you are in, what the filing window is, what total elapsed time realistically runs, and whether anything in your lease or location disqualifies you. File at the earliest legally permitted moment. Then decide, at T-3, what you will actually do if the license is not in hand at T-1 — open dry with a clearly communicated plan, or hold the date. Both are defensible; discovering the choice at T-1 is not.

All of this varies by state, county, and city. Verify locally, with an attorney who practices in your jurisdiction, before you commit money to a concept that depends on a license.

J.8 The ten things that slip

In rough order of how often they are the actual cause. At Bellwether's occupancy, every week of slip after rent commencement costs \$1,830 in rent alone, before any payroll.

1. Permit review cycles. Typical slip: 3 to 12 weeks. You submit, you get comments, you correct, you resubmit, you get more comments. First-time applicants and incomplete submissions generate the most cycles. What to do: hire an architect who has permitted restaurants in your jurisdiction, attend the pre-application meeting, and respond to comments in days rather than weeks — your own turnaround is the one variable you control.

2. A failed initial inspection. Typical slip: 1 to 3 weeks per failure. A missing hand sink, a hood clearance, a floor-to-wall coving detail, a backflow preventer, an unshielded light. Each failure costs a correction plus a scheduling cycle. What to do: request a pre-inspection walkthrough if offered, and walk the space yourself against the inspection checklist the day before — most departments publish theirs.

3. Long-lead equipment. Typical slip: 2 to 10 weeks. The hearth, the hood and makeup-air unit, the walk-in, custom millwork, anything imported. What to do: order at T-8 with ship dates in writing, in the purchase order rather than in an email, and re-confirm at T-5 and T-4. A vendor who has gone quiet has slipped.

4. Utility service upgrades. Typical slip: 2 to 16 weeks. A larger gas meter, a service upgrade, three-phase power, or a new water tap is scheduled on the utility's calendar. What to do: confirm existing capacity at T-12, not T-6. If an upgrade is needed, apply the day the design is stable and treat the utility's date as immovable, because it is.

5. Hood and grease-trap capacity discovered mid-build. Typical slip: 4 to 10 weeks. Bellwether's frozen plot point, caught in design at T-10 for a four-week cost. Caught in construction instead, it is a redesign, a permit revision, and a stopped job with a mobilized crew. What to do: pay for a mechanical and plumbing survey of existing conditions before the lease is signed, or make the lease contingent on one.

6. Liquor license timelines. Typical slip: 4 weeks to open-ended. See §J.7. What to do: file at the earliest legal moment and have a written dry-open contingency by T-3.

7. Contractor change orders. Typical slip: 1 to 6 weeks cumulative. Each is a decision, a price, an approval, and sometimes a permit revision — and delay in your approval is the most common cause, not the work itself. What to do: decide within 48 hours, in writing, at the weekly meeting. A change order sitting on your desk for two weeks is two weeks.

8. Landlord delivery of the space. Typical slip: 2 to 12 weeks. The landlord owes you a defined condition on a defined date — demising walls, a roof, a service, a demolished prior tenant. Landlords slip. What to do: negotiate delivery-condition specificity, tie rent commencement to actual delivery rather than the scheduled one, and get a delay remedy in writing.

9. The certificate of occupancy. Typical slip: 1 to 6 weeks. It issues only after every other final passes, so it inherits every other delay on this list and is where they all arrive at once. What to do: track the finals as a checklist, chase the trades that are slow to call for their own inspections, and never assume a passed inspection has been recorded — verify.

10. Hiring a chef de cuisine or sous late. Typical slip: 2 to 8 weeks of readiness, not of construction. This one does not delay the building; it delays your ability to open it safely. A sous hired at T-1 has not built a prep system, written the station setups, or learned the hearth. What to do: open the search at T-7 and target a T-3 start. Pay for those three months. It is the best-spent salary in the project.

⚠️ Where the Money Leaks

What an announced date actually costs when it slips.

You announced for the 14th. On the 2nd, the certificate of occupancy is held for a fire-alarm device that was never installed. Four weeks.

The payroll choice. You have hired 26 hourly staff to a start date. You can keep them — at, say, 20 hours a week each while they train and wait — which is $26 \times 20 \times 4 = 2{,}080$ hours at the \$14.70 blended rate, or roughly **\$30,600**, nearly the entire \$35,000 pre-opening budget, spent on a restaurant that is not open. Or you release them and rehire in four weeks, in which case your two best cooks and your best bartender take jobs elsewhere and do not come back.

The rent. Four more weeks at \$1,830 is **\$7,320** if rent has commenced.

The rest. A press cycle spent. Cancelled reservations. Perishable inventory ordered for the 14th. And a neighborhood that now carries a small, permanent belief that you are the place that kept saying it was opening.

What the disciplined operator does: announce nothing until the CO and the health permit are in hand, then announce a date at least ten days out. Ten days is enough to fill a first week. Ten weeks is enough to lose one.

J.9 The pre-opening budget

The Bellwether plan carries **\$35,000** of pre-opening cost inside its \$620,000 project — the money spent between the end of construction and the first paying guest, on things that will never appear again.

⚠️ Where the Money Leaks

Look at the last line and be honest about it.

A \$700 contingency inside a \$620,000 project is a rounding error. It covers one compressor service call, or one emergency plumber, or one replacement mixer bowl and a delivery fee. It does not cover a slow first month, an extra week of training, or the second round of smallwares you did not anticipate.

That is the standing problem with the pre-opening line in nearly every restaurant budget: it is the last number written, it is written by someone with construction fatigue, and it is sized by what is left rather than by what is needed. On a project of this size, an operator with the option should carry \$45,000 to \$55,000 of pre-opening cost, not \$35,000.

The Bellwether plan does not have that option, which means its real contingency is not the \$700 on that line. It is the **\$45,000 working-capital reserve** — the same \$45,000 that §J.6 already showed is 52.8% committed to the first-quarter cost ramp before anything goes wrong.

What the disciplined operator does: build the pre-opening budget from the bottom up — count the training hours, price the opening inventory off real cost cards, call the utilities for actual deposit amounts — and then present the honest number, even when it makes the project total uncomfortable. A pre-opening budget that is too small does not save money. It moves the shortfall into the reserve, where it becomes invisible until the reserve is gone.

Three lines deserve a note. Training payroll is the one people cut first and should cut last — every hour removed from training is an hour of learning that happens in front of a paying guest instead, at a much higher price, and 650 hours across 31 people is already modest. Opening inventory is a stocking purchase, not a usage figure, which is why the beverage line is larger than the food line and surprises operators every time: a 40-bottle list at opening depth, a back bar that has to be complete on night one, and beer stock will sit on your shelf for months, while food is bought against pars and turns over weekly. And the marketing line is small on purpose — in a neighborhood restaurant the opening's demand comes from five months of paper on the windows and the first two hundred guests. Spending your way into a full opening week is easy and mostly wasted. What you want is a full sixth week, and that is bought with service, not advertising.

J.10 What a timeline cannot do

A timeline is a way of being wrong on purpose, early, in writing, where you can see it.

It cannot tell you when your permit will issue, because that is a queue you are not in charge of. It cannot tell you what is behind the wall, because nobody knows until the wall is open. It cannot tell you whether your hearth will hold temperature on a 96-cover Friday, whether the sous you hire will still be there in March, or whether a neighborhood that supported a café will pay \$46 for dinner. The calendar is silent on every one of those.

What it can do is more useful than prediction. It tells you, on any given Tuesday, which single thing is holding up your restaurant. That is the whole trick. Most first-time operators experience an opening as an undifferentiated wall of urgent tasks and respond by working on whichever one is loudest — which is almost always the logo, the website, or the chairs, because those answer when you touch them. The critical path in §J.2 exists so you can look at a week and say: none of this matters until the mechanical review clears, so what I am actually doing this week is answering the reviewer's comments in eighteen hours instead of nine days.

It tells you what money leaves when, which is a different question from how much you need. A project can be adequately capitalized and still fail on timing — \$80,000 of equipment deposits in T-8, \$112,000 in T-3, a \$75,000 tenant-improvement reimbursement that does not arrive until after you open. Cash is not profit, and in a build it is not even budget. It is a schedule.

And it tells you what you owe yourself when you are wrong, which is the honest purpose of the document. Every one of the ten items in §J.8 will happen to somebody, several will happen to you, and the plan that survives is not the one with the fewest slips. It is the one that carried a later announced date and more working capital, so that the slips were expensive rather than fatal.

Build the schedule. Update it every Monday. Argue with your contractor about it in a meeting that keeps minutes. Then hold two things loosely: the date, which you will change, and your certainty about the date, which you should never have had.

You will not open on the day you first wrote down. Nobody does. The operators still open in year three are not the ones who guessed right. They are the ones who never told the neighborhood a date they had not yet earned, and who still had money in the account on the Tuesday in February when 41 people came in.