Nobody can source the 6-10% restaurant rent rule

Every guide to restaurant leases opens with the same benchmark: rent should be 6–10% of gross sales. I went looking for where that number comes from and could not find an origin. It appears on vendor blog after vendor blog, each citing another blog. No survey, no dataset, no trade association study, no filing.

What is available, free and dated, is what public restaurant companies tell the SEC. Here are three, all fiscal year 2025:

A nearly five-fold spread, in one industry, in one year, among three profitable operators.

Before anyone builds an argument on that table, the honest caveat: those are not the same measurement. Texas Roadhouse reports rent only and books CAM, insurance and property taxes elsewhere. Chipotle and Shake Shack report occupancy, which sweeps in CAM and certain local taxes. Comparing 1.6% to 7.7% without saying so is the exact error the benchmark articles make. But the spread survives the caveat, and it points at something more useful than a ratio.

The ratio is a function of your sales, not your negotiating

Rent is a fixed dollar amount. Sales are not. A $9,000 monthly rent is 15% of sales at $60,000 a month and 6% at $150,000. The identical lease, negotiated by the identical operator, sits at either end of the "benchmark" depending on a variable that gets decided after you sign.

Texas Roadhouse's 1.6% is partly a very high average unit volume and partly the fact that it owns the land under 158 of its 714 company restaurants. Shake Shack's 7.7% is small urban boxes on expensive streets. Neither number tells you whether either company negotiates well.

This is why the benchmark is worse than useless: it invites you to evaluate a lease against a sales figure you have not achieved yet. The question that actually matters is the opposite one. If sales come in 30% under plan, which clauses in this document decide whether you survive it? Everything below is sorted by that test.

Rent commencement is the first place you lose money

The lease ties rent commencement to delivery of the premises plus a fixed outfitting period — commonly 90 to 150 days. The permit authority does not care about that date. Chipotle, a company with a real estate department and standing relationships with municipalities, disclosed lengthening construction timelines in its FY2025 filing "due to backlogs and long wait times for us to obtain required permits and utility hookups."

If a public operator with that much leverage is flagging it as a risk, an independent signing one lease should assume it. The clause to fight for conditions rent commencement on delivery of the premises and receipt of the permits required to open — not delivery alone. Failing that, negotiate the outfitting period long enough to absorb a realistic permit delay in your specific municipality, which your contractor can estimate and your landlord's broker cannot.

Paying full rent on a dark box for four months is, for most first restaurants, a larger dollar amount than everything you will win on the rent rate.

The tenant improvement allowance may be a loan

This is the clause I would most want an independent operator to read before signing, and I have not seen it explained anywhere in the consumer-facing literature.

From The Cheesecake Factory's FY2025 10-K, verbatim: "A portion of our tenant allowances at certain premises may be subject to recoupment against percentage rent otherwise payable for such sites. When we are unable to achieve sales in a sufficient amount to generate percentage rent obligations, we are not able to fully recoup available allowances at affected sites."

Read it twice. The "allowance" is structured as a contingent advance recovered out of percentage rent. Hit your sales targets and the landlord effectively gets the money back. Miss them and you never collect the balance of the allowance in the first place — which is precisely the scenario in which you needed it.

A multi-billion-dollar tenant with real estate counsel did not negotiate this mechanic away everywhere. Assume the version in your lease is worse, and read the allowance section against the percentage rent section as one clause, not two.

For scale: Shake Shack's FY2025 average investment per location ran about $2.3M gross and $1.9M net of landlord allowances — roughly $400K per unit, about 17% of build cost. That is what a large tenant with a proven prototype extracts. Every "restaurants get $100 per square foot" figure I found came from contractor and broker lead-generation sites with no sourcing and numbers contradicting each other by a factor of two, so I'm not passing any of them along.

Percentage rent: understand the breakpoint before you concede it

The natural breakpoint is base annual rent divided by the percentage rate. At $200,000 base rent and 7%, you pay percentage rent only on sales above $2,857,143.

Cheesecake Factory's filings put its percentage rent between 2% and 10% of revenues across its portfolio — a five-fold spread inside one tenant's own leases, which tells you the rate is more negotiable than operators assume.

Two practical points:

An artificial breakpoint is the version to resist. A breakpoint set below the natural one means you pay percentage rent before the base rent has been earned out. A natural breakpoint, paired with a lower base rent, is genuinely risk-sharing — it moves downside onto the landlord.

Percentage rent opens your books. The lease defines "gross sales," grants audit rights over your POS data, and sets reporting obligations. Whether that definition includes third-party delivery at gross or net, whether it includes comps, and whether gift cards count at sale or at redemption can push you over a breakpoint on revenue you never kept. Negotiate the definition, not just the rate.

The personal guarantee, and the New York solution

A guarantee typically survives assignment by its own terms. Texas Roadhouse's own risk disclosures describe remaining obligated for base rent and real estate taxes on sites it transferred to other operators. If that exposure follows a public company, it follows you.

The mechanism most worth knowing is the good guy guarantee, largely a New York City convention: rather than guaranteeing the full term, you personally guarantee payment only while you remain in possession. Give proper notice, surrender the space clean, free of subtenants and liens, with all rent current, and the personal exposure ends there.

Two cautions. It is not an exit option — you still owe everything through surrender, and you remain liable for pre-surrender defaults. And outside New York, many landlords will not recognize the term, so you will need to describe the mechanic from scratch: guaranty of payment during occupancy, not guaranty of term. The alternatives worth asking for are a dollar cap or a burn-down that reduces the guarantee each year you perform.

Assignment is your exit value

Restaurant leases nearly always bar assignment without landlord consent. The three words that decide whether you can ever sell the business are whether consent may be withheld in the landlord's sole discretion or may not be unreasonably withheld.

Watch for two related clauses: a recapture right, letting the landlord take the space back rather than approve your buyer — which converts your sale into a surrender — and profit-sharing on assignment premium, giving the landlord a cut of what you sell for.

Negotiate this on day one, when it costs nothing. On the day you have a buyer, the landlord knows exactly what it is worth.

CAM, exclusives, and what most operators get wrong about both

CAM. Estimated monthly, trued up annually, and the true-up arrives as one bill. Ask for a cap on controllable expenses, exclusion of capital replacements that should be amortized rather than expensed, exclusion of costs benefiting a single other tenant, a ceiling on the landlord's administrative fee, and — most importantly — an audit right with a real window. Practitioners commonly cite 3–5% annual caps on controllables, a 12-month audit window, and 10–15% admin fees. I'd treat those as negotiating conventions rather than market data; I found no survey establishing them as medians.

Exclusive use. Courts enforce exclusives strictly against landlords, but only where the protected category is precisely defined — vague language is frequently litigated and frequently fails. The larger problem is the carve-outs. Pre-existing leases in the center are almost always excluded, and well-drafted landlord language extends that exclusion to renewals of those leases, relocation of a tenant within the property, and replacement of a defaulting tenant. Your exclusive can be legally defeated on day one by a lease signed before you arrived. And securing the clause without negotiating a remedy for breach leaves you with a right and no leverage.

Where not to spend your leverage

Negotiating capital is finite. Three places it commonly gets wasted:

Breakpoint math you will never reach. At $200K base rent and 7%, you need $2.86M in sales before percentage rent costs you a dollar. If you're underwriting $900K, that clause is theoretical. Spend the leverage on the allowance or the guarantee cap instead.

Co-tenancy in the wrong format. Co-tenancy protects you when an anchor goes dark. In street retail or on a freestanding pad there is no anchor to fail, and asking for it signals inexperience. In an in-line center it is worth real money — Cheesecake Factory carries co-tenancy protection in a majority of its leases, and independents usually carry none, because nobody asked.

Second-generation space. If you're taking a former restaurant, the allowance conversation inverts — landlords offer less precisely because prior improvements cut your hard costs. Shift the negotiation to condition of delivery and to who owns the aging hood, grease trap, and HVAC. On a shared or communal grease trap, a clause assigning "the grease trap serving the premises" to the tenant can mean funding repairs for the entire line, and municipal violations attach to the operator, not the owner.

One more thing about who you're bidding against

Food and beverage accounted for 39.4% of Manhattan's year-to-date retail leasing volume in the first half of 2026 — nearly double the next category, apparel at 21% — while overall Manhattan asking rents hit $710 per square foot and premier-corridor availability fell to its lowest recorded level. The national average retail asking rent, for contrast, was $24.79 per square foot.

Restaurants are the marginal bidder setting street retail rents in the most expensive market in the country. That is worth knowing when a broker tells you the rate is non-negotiable: in a lot of markets, the reason rents are where they are is that operators like you keep meeting them.

Disclosure: I work at Katalyst, and clean sales reporting — which is what a percentage-rent audit will demand from you — is something we build. But nothing in this post requires buying software. It requires reading the allowance clause against the percentage rent clause, tying rent commencement to permits rather than to delivery, and asking what happens to the assignment consent standard on the day you want out. Those three edits cost nothing and outlast every ratio anyone quotes you.