Case Study 2: The Restaurant That Opened On Time

This case is a clearly labeled composite. It is assembled from patterns that recur in almost every restaurant opening — a slipped certificate of occupancy, a crew already hired, a landlord's abatement clock running, and a decision made under cash pressure with incomplete information. No real business, landlord, contractor, or authority is depicted, and every figure is illustrative. It is built as a composite deliberately: the decision at the center of it is made privately, under stress, by people who do not publish their reasoning, and inventing details about a real opening would be worse than useless.

The complementary case to the first one. There, an industry built a structure to protect an opening. Here, an operator who understood the structure perfectly well ran out of the one thing the structure requires: money to wait.

Background

A 64-seat neighborhood restaurant with a full bar, in a mid-size market. Two partners, both experienced operators, neither a first-timer. Their project budget was \$540,000 and their planning was, by the standards of this book, good.

Read that table before you read the rest. These operators did the thing this chapter asks for. They built the pre-opening budget from the bottom up, they found the gap, they named it, and they decided in advance which pots would absorb it. They knew their reserve after the gap was about nineteen days of fixed cost and they accepted it, because the alternative was not opening.

Then the schedule slipped.

The operating issue

The certificate of occupancy was targeted for week −4. A failed final electrical inspection, a corrected panel schedule, a re-inspection queued behind a holiday, and a fire-marshal walkthrough that required additional exit signage put it three weeks late.

Here is the position on the day the certificate was finally issued.

THE POSITION AT THE CERTIFICATE OF OCCUPANCY            [constructed teaching example]

  Sous chef on payroll                          9 weeks   (planned: 6)
  Bar manager, cooks, prep on payroll           5 weeks   (planned: 4 at C of O)
  Servers, hosts, dish — offers signed,
    start dates deferred twice                  —         two have taken other jobs
  Cash spent on pre-opening to date                       $ 51,900
  Cash remaining, reserve + unspent pre-opening           $ 40,900
  Rent clock                                              starts in 30 days
  Abatement remaining                                     $ 14,800 (intact — the grace held)
  Announced opening (told to guests, press,
    the neighborhood, and the reservation
    platform)                                             11 days away
  Days of training the plan required after C of O         28

The partners had eleven days of announced runway and a plan that needed twenty-eight. Moving the opening date a second time meant telling a neighborhood, a press list, and two hundred soft-open invitees that the restaurant had slipped again — and, more materially, carrying the crew for another seventeen days at roughly \$6,100 a week.

They ran the arithmetic, and it is worth running with them.

Option A — hold the announced date. Compress twenty-eight days of training into eleven. Cut the soft open from four services to two. Open with the full menu because it had already been printed and posted.

• Cash cost of the compression: roughly zero. It saves \$14,800 of payroll against Option B.
• Abatement: entirely intact, all \$14,800 landing on months one and two of operation.
• Reserve on opening day: \$40,900.

Option B — move seventeen days. Run the full training program, all four soft-open services, open with a crew that has served real tables.

• Payroll carried: 17 days ≈ 2.4 weeks × \$6,100 = **\$14,800**.
• Abatement: still intact — the delay is inside the thirty-day grace by three days.
• Reserve on opening day: \$26,100.
• Plus the reputational cost of a second announced slip, and the two hourly staff already lost.

They took Option A. On the arithmetic in front of them, it is defensible: fourteen thousand eight hundred dollars of reserve, against a training program they believed they could compress because both of them had opened restaurants before. This is the contested decision, and reasonable operators make it in both directions.

What it shows

What Option A actually bought, over the following ninety days.

FIGURE — The first ninety days, planned vs. what compression produced
                                                        [constructed teaching example]

                              PLANNED           ACTUAL          VARIANCE
  ────────────────────────────────────────────────────────────────────────────
  Quarter revenue             $ 318,000        $ 302,600        −  15,400
  Prime cost %                    65.0%            71.8%
  Prime cost $                $ 206,700        $ 217,267        +  10,567
  Comps and remakes,
    % of sales                     1.8%             4.6%
  Comps and remakes $         $   5,724        $  13,920        +   8,196
  Excess prime cost vs. the
    61% annual target         $  12,720        $  32,681        +  19,961
  ────────────────────────────────────────────────────────────────────────────
  Reserve on opening day      $  26,100        $  40,900        +  14,800
  Reserve at day 90           $  21,400        $  14,700        −   6,700

Every figure above is illustrative and constructed. The shape is the point.

The \$14,800 saved on payroll came back as \$19,961 of excess prime cost inside ninety days, and it came back through three specific channels that are entirely predictable in hindsight:

Comps and remakes at 4.6% instead of 1.8%. A floor that had served two rehearsal services instead of four rang the wrong modifiers, fired tables out of sequence, and sent food back. Every comp is a plate cooked twice and sold zero times.

Food cost carried the unverified yields. The compressed rehearsal cooked every item once instead of three times. Two dishes had yields nobody had confirmed and portions nobody had weighed. Neither was discovered until the week-nine inventory, which is seven weeks of a two-point overrun.

Labor never got cut on time. A manager who does not yet know what the floor can absorb keeps everyone on, because sending someone home and then drowning is the more visible failure. Six extra labor hours a night across thirteen weeks is a real number and nobody ever writes it down.

And there is a fourth channel that does not appear on the table at all. The restaurant's first thirty days produced its first thirty days of reviews, and those reviews are the permanent record from which the following nine months of demand were partly built. Chapter 23 puts a number on the second visit. Case Study 1 explains why no convention protects you from this.

Outcome

The restaurant did not close. This is important, because the useful version of this case is not a catastrophe.

It reached day 90 with \$14,700 of reserve — about ten days of fixed cost — instead of the \$21,400 the slower path would have produced. It spent months four through seven repairing habits established in month one: re-training a floor that had learned the wrong sequence, re-costing two dishes, re-establishing a labor discipline that had never existed. By month nine the prime cost was where the plan had always said it should be.

The partners' own assessment, offered to anybody who asks, is that the compression did not sink them and did cost them roughly a quarter of a year. Their sentence about it: "We saved fifteen thousand dollars and spent six months getting it back."

The lesson

Three, and the third is the one that matters.

One: the arithmetic in front of you at the decision point is real, and it is incomplete. Option A genuinely saved \$14,800 of cash on the day it was chosen. The costs it created were larger, arrived later, and appeared on a different line under a different name — which is precisely the mechanism Chapter 1 called cost drift and precisely why it is so hard to argue against compression in the room where the decision is made.

Two: the compressible parts of an opening are not the parts you think. You can compress the deep clean, the signage, the final merchandising of the room, the marketing calendar, the wine list from forty bottles to thirty. You cannot compress repetition, because repetition is the only mechanism by which a group of people becomes able to do a thing the same way twice. Everything a restaurant does well on a Saturday, it does well because it has done it before. Cut the number of befores and you have cut the only variable that matters.

Three, and this is the actual lesson: the decision was lost before it was made. By the time the certificate of occupancy arrived three weeks late with a crew nine weeks on payroll and eleven days of announced runway, both options were bad. What produced that position was not the failed inspection — inspections fail routinely, which is why §9.2 calls plan review and the certificate gates. What produced it was committing the payroll date and the announced opening date against a certificate that had not been issued.

The countermeasure is entirely upstream and it is free. Do not announce an opening date until the certificate is in hand and the licenses are issued. Do not put the crew on payroll against a forecast. Build float into the hiring date, not the opening date, because the hiring date is the only one of the two you control. Chapter 9's thirty-day rule and this case are the same idea seen from opposite ends: the operator who protects the gap between the certificate of occupancy and opening day never has to make the choice these partners made.

Discussion questions

1. Rebuild the decision as the partners faced it, but change one fact: they had \$70,000 of reserve instead of \$40,000. Does Option B become obviously correct? Show the arithmetic and then say what non-financial factor might still push toward Option A.

2. The composite claims \$14,800 of saved payroll returned as \$19,961 of excess prime cost. Identify which of the three channels named (comps, unverified yields, late labor cuts) you would attack first with a fixed \$5,000 of remedial spend in month two, and justify it with a number.

3. The partners announced their opening date before the certificate of occupancy was issued. List everything that becomes hard to reverse the moment that announcement is made — reservations, press, staff start dates, marketing spend, guest expectation — and rank them by reversal cost.

4. §9.6 advises capping covers in opening week. If these partners had done that on top of Option A, what would it have changed and what would it not have changed? Be specific about which of the three cost channels a cover cap does and does not address.

5. The hardest one. The partners' own summary is "we saved fifteen thousand dollars and spent six months getting it back." Construct the strongest possible argument that Option A was nonetheless the correct decision given what they knew — then say what single piece of information, obtainable on the day, would have settled it.