Choosing a restaurant-specialist GC when you're opening multiple locations

Why a kitchen tenant improvement is not an office fit-out with appliances

Most of the cost and risk in a restaurant buildout sits in systems an office contractor rarely touches. These include a Type I hood with its exhaust duct, roof fan and makeup air; gas piping sized for the cook line; floor sinks and a grease interceptor; dedicated circuits for refrigeration and dish machines; and finishes a health department will accept. Each system needs its own trade, permit or inspection, and each one depends on the others. Duct routing changes the roof curb, the curb changes the structural review, and the makeup air unit adds both a gas load and an electrical load. A GC who has never sequenced those dependencies tends to find them one at a time, during construction.

Owners describe exactly this. In the thread Cost and Length of Time for Full Build Out?, an operator who had run four projects said one of them went badly because they hired a contractor and architect they had used for an office buildout. That team was inexperienced with restaurants ("a completely different animal"), and absorbing their mistakes cost a lot of time and money. The same operator said a later contractor who worked almost exclusively in restaurant and hospitality "made all the difference." Replies in a thread asking for the steps to build out an empty retail space give the same advice: hire an architect and GC who have designed and built restaurant kitchens. Both threads are self-reports, and we have excerpts, not every comment. They show a recurring pattern, not a measured failure rate.

Why a wrong choice costs more when you are opening several sites

A single-site owner who picks the wrong contractor pays for the mistake once. A group that is rolling out a format usually turns the first build into its template: the same kitchen layout, the same equipment package, the same hood length, the same drawings adjusted to each space. If the first GC routed duct poorly, undersized a panel or skipped a step with the fire inspector, that decision ends up in the prototype. Unless someone catches it, it gets built again at the second and third sites.

So the selection question changes. You are not only asking whether this contractor can build this space. You are asking whether they will help you find the defects in your own prototype, report what went wrong at site one before site two is priced, and deliver as-builts and closeout documents good enough to reuse. A contractor who is good at one-off jobs and treats every project as new may be fine for a single location and a poor fit for a rollout. Our related article on writing a repeatable prototype spec covers the owner's side of that loop.

Portfolio evidence that actually tests restaurant experience

Photos of finished dining rooms say little about whether the contractor can handle the back of house. Ask for a short list of restaurant projects completed in the last few years with the details that matter. For each one, you want the scope (full kitchen or front-of-house refresh), whether a new Type I hood was installed, whether a grease interceptor was added or replaced, whether the electrical service was upgraded, how long the job took from permit to final approval, and who the owner contact is. A contractor with real kitchen experience can answer these quickly. Vague answers, or a portfolio that is mostly retail and offices with one cafe, tell you something too.

Ask how they staff the work. Which trades do they keep in-house, and which mechanical, electrical, plumbing and hood subcontractors do they use repeatedly? Will the same superintendent run every one of your sites? Restaurant experience often lives in a few people, so you are partly hiring the superintendent and project manager. Get their names into the contract or proposal. A Montgomery County, Maryland owner in a thread seeking restaurant-capable trades was looking separately for an architect, electrician, plumber and deck contractor for a ventless concept in an early-1900s house. It is a reminder that in older buildings the trade bench matters as much as the GC's own staff.

• Three to five completed restaurant projects with owner contacts you can call
• For each: hood installed or reused, interceptor work, electrical upgrade, months from permit to final
• Named superintendent and project manager, with their own restaurant jobs
• Repeat MEP and hood subcontractors, and what the GC does when one is backlogged
• A sample closeout package: as-builts, equipment manuals, test reports, warranty list

A reference-check script for past restaurant clients

Generic reference questions get generic answers. "Were you happy?" tells you little. Ask about specific moments where restaurant jobs usually go wrong, and listen for whether the owner describes the contractor finding problems or creating them. Schedule twenty minutes with each reference and ask the same questions every time, so the answers can be compared.

Take notes on dates and amounts the reference offers, but treat them as one owner's account. If two references independently describe the same weakness, such as slow closeout or poor coordination with the hood installer, take it seriously. One unhappy reference among several good ones may say more about that project than about the contractor. Ask the contractor about it directly and see whether the explanation is specific.

• When the hood, makeup air and roof work happened, who coordinated the trades, and was anything redone?
• Did any inspection fail? What was the reason, and how fast did the GC fix it?
• How did change orders work? Were they priced before the work was done?
• Were invoices itemized well enough that you could check materials and labor?
• How did the final schedule compare with the one in the contract, and what explained the gap?
• Who showed up for punch list and warranty calls after opening?
• Would you hire them again for a second location, and would you change anything in the contract?

License and insurance verification, with DC as the example

Licensing is a minimum check, not proof of restaurant skill. Requirements are set by states and localities, so look them up for every jurisdiction where you plan a site. In Washington, DC, the Department of Licensing and Consumer Protection's Business Licensing Division page lists General Contractor/Construction Manager licenses in classes A, B, C, G and H, and points to BOSS and Scout for verifying an existing business license. Search the exact legal entity named on the proposal, not a trade name, and check that the license is active.

The class matters for multi-site work. DLCP's Mandatory Insurance Levels and Claims Support page ties each class to a maximum value for any single contract project. Class A has no limit; Class B is capped at $10 million, Class C at $5 million, Class G at $2 million and Class H at $500,000, each citing 17 DCMR § 3902.1. A restaurant TI will usually fall well inside these caps. But if you plan to bundle several DC sites under one contract, ask DLCP or your attorney how the cap applies before you sign, rather than assuming. The same page lists the minimum commercial general liability coverage for each class. DLCP's separate License Types Requiring Insurance page cautions that a certificate of insurance does not automatically mean coverage will be provided for all issues. Our article on checking license, insurance and workers' comp covers that verification in detail. Outside DC, including Maryland and Virginia sites a DC-area group might add, use that state's or county's own lookup.

Capacity: the pass/fail test most selection sheets leave out

A restaurant specialist with a two-person project team can be excellent on one job and overwhelmed on three. Before scoring anyone, sketch your rollout calendar and mark where construction periods overlap. Then ask each candidate how many active restaurant jobs they are running now and expect to run during your dates, who would supervise each of your sites at the same time, and how they handle a hood or electrical sub who is booked solid. Ask for current workload in writing rather than accepting "we can handle it."

If two of your builds overlap and the contractor can only staff one experienced superintendent, you have three honest options. You can stagger the openings, split the sites between two qualified contractors, or pick the contractor with real depth even if their single-site score is slightly lower. What you should not do is average capacity into a weighted score where strong portfolio points hide a team that cannot be in two places at once.

Hypothetical worked example: scoring three contractors for three sites

All names and figures here are hypothetical. A group plans three restaurants. Sites 2 and 3 overlap in construction by about four months. The owner scores three bidders from 1 to 5 on six weighted criteria, after a pass/fail gate for license and insurance that all three pass. Each weighted score is the rating divided by 5, multiplied by the weight. Contractor A is a general commercial builder with mostly office work and the lowest price. Contractor B is a small restaurant specialist. Contractor C is a regional restaurant contractor with two superintendents.

Contractor A totals 58: 10 + 12 + 8 + 12 + 10 + 6. Contractor B totals 81: 25 + 16 + 20 + 6 + 6 + 8. Contractor C totals 78: 20 + 16 + 16 + 12 + 6 + 8. On points, B wins. But B rated 2 on capacity because it can field one superintendent, and sites 2 and 3 overlap. The owner treats capacity as a gate for overlapping work, so B either gets site 1 alone, with sites 2 and 3 staggered, or C takes the rollout.

Price looks different once you account for delay. Suppose A bid $410,000 per site and C bid $452,000. C costs $42,000 more per site, about 10.2% (42,000 divided by 410,000). If the owner's own carrying cost while a site sits unopened, meaning rent, preopening payroll and loan interest, is a hypothetical $18,000 a month, that $42,000 equals about 2.3 months of delay (42,000 divided by 18,000 is 2.33). The comparison does not prove A would run late. It tells the owner how much schedule risk the price gap is buying, and that is the question to put to A's references.

Red flags, and what to keep on file once you choose

Some warning signs show up before a contract is signed. A proposal that excludes the hood, makeup air, electrical and roofing "by others" without saying who those others are pushes coordination back onto you. Watch for a contractor who cannot name the superintendent, references who only did front-of-house refreshes, reluctance to itemize, or pressure to start before permits are in hand. None of these proves bad faith. Each one is a question to resolve in writing.

Once you pick a contractor, save your selection evidence. That means the license lookup with the date, the certificate of insurance and any verification you did with the insurer, reference notes, the capacity commitment and the named personnel. If the relationship goes badly at site two, you will want to show what was promised and whether staffing changed. Our related articles cover comparing bids line by line, payments and lien waivers, and what to do when a GC is months late.

• Exclusions that list major kitchen systems as "by others" with no named party
• No restaurant references with a new hood or interceptor in scope
• Superintendent "to be determined" on a multi-site proposal
• Unwillingness to provide itemized pricing or a written change-order procedure
• License entity name that does not match the proposal or contract
• File to keep: dated license lookup, COI and insurer confirmation, reference notes, staffing commitment