I have walked more than 200 potential restaurant sites over the last decade. Some became units I ended up running. Most did not. The best real estate broker I have worked with once told me that his job was to show me sites and my job was to say no to almost all of them. He was right, and the operators who say yes too fast are the ones I see rebuilding their P&L in year two.
Here is how I walk a site now, what I ask, and the deal terms I will not sign no matter how good the walk went.
The one number that decides most of it
Before any of the qualitative questions, the site has to pass one quantitative test. Total occupancy cost (base rent, CAM, taxes, insurance, any percentage rent) has to come in at or below 10 percent of the realistic year-two revenue for the concept.
Realistic year-two revenue, not the aspirational number. If similar concepts in similar neighborhoods do $2.4M in a stable year, use $2.4M. Do not use $3.2M because you have a better chef. You might. Most operators do not. The site either works at $2.4M or it does not work at all.
Do this math before you fall in love with the space. Once you have walked a beautiful room three times, your brain will start reverse-engineering revenue numbers that justify the rent. Every operator does this. Almost every operator regrets it.
The three visits
Walk the block at least three times before you make an offer, at different times of day.
Visit one: your target meal
If you are opening a dinner concept, walk the block at 7pm on a Thursday. If you are opening a lunch concept, walk at 12:30pm on a Wednesday. Count the pedestrians who pass the frontage in 15 minutes. Extrapolate to two hours. Then apply a realistic capture rate. Good restaurants in walkby locations capture 1 to 3 percent of pedestrian traffic. Excellent ones might hit 5 percent. Nobody hits 10.
Visit two: an off-peak hour
Wednesday at 3pm. This is where you see whether the neighborhood has activity between meal periods or dies. Restaurants in neighborhoods that die between meals struggle to fill the second turn of any service, because there is no ambient energy in the block.
Visit three: 11pm on a weekend
This tells you two things. Whether the block is safe at night, and whether the neighborhood has a nightlife identity that supports a late seating or a bar program. A block that empties at 9pm on a Saturday cannot support a late dinner concept, no matter what the daytime traffic looks like.
Fig. 1 · Pedestrian counts across three site visits.
The eight questions I ask on every walk
- Where does the trash go and how does it get out? If the answer is "through the dining room," walk away. If the answer is "out the back to a shared dumpster the whole strip uses," negotiate hard on shared costs.
- Where does the grease trap live? An interior grease trap that needs pumping during service hours is a nightmare. An exterior one accessible from the parking lot is a gift.
- What is the electrical service? Restaurants pull a lot of amperage. A space that used to be retail may need a full service upgrade, which is a $30K to $80K surprise you find in month three of the buildout.
- Is there gas? Some cities are restricting new gas hookups. If you need gas cooking and the site is all-electric, you have a problem you did not sign up for.
- What is the ventilation situation? The hood system is the single biggest infrastructure cost in a restaurant buildout. A second-generation restaurant space with a functioning hood is worth 20 percent more than an equivalent shell.
- Where do deliveries happen? Loading zone, alley access, freight elevator? A restaurant that has to receive through the front door blocks the guest entrance during morning deliveries.
- What is the sound situation? Adjacent to a residential building, and you will get complaints about noise starting on week two.
- How does the space feel at 60 percent capacity versus full? A room that only looks good full will look empty at 60 percent, and 60 percent is where you will spend most of your operating year.
Second-generation space is almost always better
A previous-restaurant space, if the previous tenant closed for reasons other than location, is almost always a better deal than a shell. The hood is in. The grease trap is in. The three-compartment sink is in. The electrical is sized. The permits are documented. The landlord has already accepted restaurant use.
You save $200K to $500K on buildout and 6 to 12 weeks on the timeline. You also inherit some risk. Old equipment that may not meet current code. Layout compromises the previous tenant made. Sometimes deferred maintenance the landlord has not disclosed.
Do a professional inspection of any second-generation space before you sign. Bring your chef. Bring your GM. Bring a commercial kitchen equipment specialist. Spend $3K to $6K on the inspection. Save yourself $80K in surprises.
The critical question is why the previous restaurant closed. If it was operations (bad menu, bad management, wrong concept), the location is still viable. If it was location (no traffic, hostile block, bad access), walk away no matter how good the space looks.
The lease terms that will hurt you
Every restaurant lease has terms in it that the landlord's lawyer put there and hopes you will not push back on. The ones that matter most:
Personal guarantees that do not burn off
A personal guarantee turns a bad unit into a personal bankruptcy. Push for a burn-off, usually after 24 to 36 months of on-time rent payments. Some landlords will not agree. Some will. Ask.
Percentage rent on top of base rent
Percentage rent that kicks in above a threshold ("breakpoint") is common in high-traffic locations. Sometimes acceptable. What is not acceptable is percentage rent with no breakpoint, or with a breakpoint set so low that any successful operation pays it. Model the number at year-two revenue. If it is meaningful, negotiate.
CAM caps only in year one
Common area maintenance charges can escalate 8 to 20 percent a year in a badly negotiated lease. A CAM cap that only applies in year one gives the landlord unlimited upside starting year two. Push for a year-over-year cap of 5 percent for the full term.
Narrow exclusive-use clauses
An exclusive-use clause protects you from the landlord opening a competing concept in the same center. Make sure the clause is written broadly enough to cover your actual business. "Full-service Korean BBQ restaurant" as an exclusive does not protect you from a Korean fried chicken concept opening next door.
Radius restrictions on your other units
Some landlords include a clause that prevents you from opening another unit within a certain radius of the leased premises. This blocks your growth plan before you have even opened the current unit. If the concept is one you plan to scale, negotiate this out or make it very narrow.
Bring the operator to the walk, not just the broker
The single biggest thing I have learned is that real estate scouting is an operations job as much as a real estate job. A broker sees the deal. An operator sees the shift. The GM who will run the unit needs to walk the space before you sign, and needs to be able to say no.
I have signed leases on sites that my GM was quietly worried about. All three of those units underperformed in year one. I have walked away from leases my GM did not like even when the deal looked good on paper. I have never regretted the walkaways. I have regretted every one of the ignore-the-GM signatures.
The person who is going to open the door every morning has to be able to picture opening the door every morning in this specific space. If they cannot, do not sign.
The point
Real estate scouting for a restaurant is not the same job as real estate scouting for a retail store or an office. The infrastructure requirements are heavier. The traffic assumptions are more fragile. The lease terms have more places to hide bad news. The wrong site is the one mistake in this business that a great operator cannot fully fix.
Run the math before the emotion. Walk the block three times. Ask the eight questions. Get a real inspection on any second-generation space. Push back on the lease terms that matter. Bring the operator to the walk. Then decide.
Sites that pass all of that are rare. That is fine. You do not need many. You need the right ones.