What the build-out costs you is the number you then have to hit.

Restaurant build-out cost is usually discussed as a budget question: what can we afford to spend on this unit. That framing hides the part that actually governs the outcome. Money spent on a build-out does not disappear — it converts into a sales number the unit has to clear every year, forever, before anyone calls it a success. You are not choosing a budget. You are choosing a hurdle.

A group that spends $2.7M to open a 200-seat restaurant and a group that spends $940K to open the same 200-seat restaurant are not running the same business. They are running the same concept with two different definitions of “working.”

The same restaurant, two capital structures

Take a 6,000 square foot space, 200 seats, and run it both ways. These are illustrative figures built on published construction benchmarks, not a project we ran — the point is the method, so redo it on your own market and your own rent.

Full gut, everything moves. Ground-up and full-gut full-service construction runs roughly $250 to $450 per square foot. At $350, that is $2,100,000 in hard cost. Design fees, permits, FF&E and smallwares commonly add 25 to 35% on top; at 30%, another $630,000. Total invested: $2,730,000, or $13,650 per seat.

Second generation, nothing moves. A second-generation space with a usable Type 1 hood, walk-in and grease interceptor already in place runs 30 to 50% below new construction, and cosmetic-only work runs below that again. At $120 per square foot, that is $720,000, plus the same 30% for FF&E and design: $216,000. Total invested: $936,000, or $4,680 per seat.

Now convert both into the hurdle. At a 20% unlevered return on invested capital and a 15% store-level cash margin, the full-gut build must generate $546,000 a year in unit cash flow, and the second-generation build must generate $187,200. Divide each by the margin and you get the sales each one needs simply to earn its cost of capital.

Same seat count. Same menu, if you like. The full-gut build has to do $2.39M more every year to earn the same return on the money, and that gap was set by a construction decision rather than an operating one. No amount of operating excellence closes it.

What a moved wall actually costs

The capital number is the visible half. The calendar is the half that gets people.

Permits are not a cost line. They are a queue. Every change that touches a permitted system — mechanical, plumbing, egress — sends the drawings back for revision and the space back into a queue you do not control. And that queue gets paid for twice.

Take the same 6,000 square foot space at $55 per square foot in annual rent, in a unit built to do $6M. Rent runs $27,500 a month whether the doors are open or not. $6M in annual sales is $16,438 a day, and at a 15% margin that is $2,466 a day in cash flow. Sixty days of permit delay therefore costs $55,000 in dark rent plus $147,960 in contribution you never get back — roughly $203,000 for two months in which nothing was wrong, nothing was over budget, and everyone did their job. The drawings were simply revised.

That number appears in no construction budget. It appears in the P&L a year later, where it reads as a slow ramp rather than as a decision made in a design meeting.

What moves, and what it triggers

This is the filter worth keeping. The question in a build-out review is never “how much does this cost.” It is whether this is money or whether this is calendar. Money you can decide about. Calendar hands your opening date to someone else.

Jurisdictions vary enormously in how long each queue runs, which is exactly why the useful column is not weeks. Sequence every build-out decision by which side of that third column it falls on, and the schedule stops being a hope.

The constraint that does the work

The strongest build-out discipline is not a budget cap. Budget caps make people cheap in the wrong places — they cut the chairs and the lighting, which the guest feels immediately, and leave the wall move, which the guest will never notice, sitting in the plan.

The constraint that works is positional: spend what the concept deserves, but nothing moves. Lights can change; they cannot relocate. The bar can be rebuilt; it cannot shift. The kitchen can be re-equipped; it cannot be re-laid-out. Restrooms stay where the plumbing already is.

That single rule does three things at once. It caps the capital without capping the quality, because finishes are cheap relative to systems. It removes almost every permit trigger, so the opening date stays yours. And it forces the design conversation onto what the guest actually experiences instead of onto a floor plan nobody will consciously register.

It also changes what you look for in a site. A space with good bones and a weak former operator is a far better asset than a raw shell in a marginally better position, because you are buying a hood, a grease trap, a walk-in and an approved egress plan for the price of the lease.

When a from-scratch build is the right call

Sometimes it is, and the discipline above is worth nothing if it talks you out of a build you should make.

• —The existing kitchen cannot produce your menu. A space with no ventilation path for the equipment your concept requires is not a saving, it is a ceiling.
• —The seat count in the existing plan cannot carry the rent. No amount of finish work fixes a restaurant that seats 90 against a rent underwritten for 160.
• —The lease term is long enough to amortize it. A $2.7M build on a seven-year term is a different decision from the same build on a twenty-year term.
• —The location is genuinely irreplaceable, and the alternative is not a worse build but no site at all.

What is almost never right is the middle: a second-generation space taken because it was cheap, then gutted anyway because the floor plan was not quite what the design wanted. That path pays the second-generation lease premium and the ground-up construction cost together, and it is the most common way a group ends up owing $3.6M in annual sales from a site that was never positioned to produce it.