Restaurant location strategy is the method a brand uses to decide where its next unit goes: pick the format, size the trade area by drive time, forecast sales for that trade area, then test the asking rent against the forecast. The occupancy cost decides the deal. Most restaurant real estate mistakes get made in that order, long before anyone walks the site.
The order is the whole thing. Teams that start with an available space and work backward toward a justification sign leases that need everything to go right. Teams that start with the format and the math walk away from good-looking corners for reasons they can explain to a committee.
Restaurant site selection is not retail site selection
The general retail site selection process gets you most of the way there. Restaurants break from it in four specific places, and those four are where the money is.
The trade area is tiny. A restaurant pulls from a much smaller catchment than a big-box or a specialty retailer. That compresses the margin for error in the demographic read, because a mile in the wrong direction is a meaningful share of your customer base.
The daypart has to match the traffic. A site with excellent counts that empties at 6 p.m. is a bad dinner site and a fine breakfast one. General retail cares about volume. Restaurants care about volume arriving at the hour their concept makes money.
Direction of travel matters. Side-of-street position decides whether commuters can reach you going home. A median, a hard left turn, or a difficult ingress can take a site off the list even when the trade area reads well.
The build is expensive and format-specific. Kitchen infrastructure is the largest line in a restaurant buildout, and it does not transfer between formats. That is why the format decision comes first.
Start with the format, not the site
The format sets the capital, the timeline, and the kind of space you are allowed to consider. Decide it first and the site search becomes finite.
Freestanding with a drive-thru carries the highest sales ceiling and the highest cost. The gating items are zoning and stacking depth, meaning how many cars can queue without spilling into the street. Municipalities have gotten less willing to approve new drive-thru lanes, which makes existing approved pads more valuable than the dirt suggests.
End-cap is the compromise most growing brands live in. You get corner visibility and often a path to adding a drive-thru, at meaningfully less capital than ground-up construction. Whether the center will allow a drive-thru and whether the parking field supports it usually settles the deal.
Inline is the cheapest way into a strong center and the most dependent on neighbors. Your traffic is the center's traffic, which makes co-tenancy a real variable rather than a lease footnote.
Second-generation conversion is the format that moved this cycle. Chain closures returned well-located former restaurant boxes to the market, and operators are taking them specifically to cut construction time and cost against a ground-up build. Trade press tracking 2026 closures reported Smokey Bones winding down its remaining units by late April, after parent Twin Hospitality filed Chapter 11 on January 26, and Darden closing 14 of Bahama Breeze's 28 locations while converting the rest to other Darden brands. Those boxes have kitchens, grease traps, and hoods already in place.
The catch on second-gen is the question the discount is hiding: why did the last operator leave? Sometimes the answer is a bad concept in a good location, which is the deal you want. Sometimes it is a good concept in a location that never worked, which is the deal that gets you.
The occupancy cost rule that decides the deal
Once you have a format and a trade area, the deal comes down to one ratio. Total occupancy cost, meaning rent plus CAM, insurance, and property tax, should sit between 6 and 10 percent of gross sales. This is a long-standing operator heuristic rather than a dated study, and it holds up because it encodes something real: rent is fixed and sales are not.
Run it backward and it becomes a screen.
A landlord asking $12,000 a month is asking for $144,000 a year. At an 8 percent target that site needs roughly $1.8 million in sales. If your honest forecast for that trade area is $1.3 million, the site is out at the asking rent, and no amount of enthusiasm about the corner changes the arithmetic.
This is why sales forecasting is the step that decides the deal rather than a formality after the site is picked. The forecast is not a number you produce to satisfy the committee. It is the input that sets your maximum rent, and getting it wrong in the optimistic direction is how a brand ends up with a unit that cannot be fixed by operating it better.
The margin context makes the stakes concrete. Restaurant accounting benchmarks put operating margins in the single digits to low teens across every segment, with full-service the thinnest and quick-service the widest. Those are directional industry figures rather than audited averages, but the direction is the point: a site carrying two extra points of occupancy cost is giving up a large share of its profit before it opens.
Size the trade area by drive time, not by radius
Quick-service concepts draw most of their business from roughly a 5 to 7 minute drive time, often only 1 to 3 miles. Casual dining pulls from a wider band, more like 10 to 15 minutes. These are practitioner conventions rather than published research, but every restaurant real estate team works with numbers in this range, and the shape matters more than the exact figure.
Drive time and distance are not the same shape. A 3-mile ring drawn on a map crosses a highway, a river, and a rail line as if they were not there. The actual 6-minute polygon around a site can be a lopsided blob that reaches 5 miles in one direction and 1.5 miles in another. Teams that screen on rings and then wonder why a site underperformed usually bought a trade area half the size of the one they modeled.
Inside that polygon, the questions are:
- Who is there, and when? Residential and daytime population produce different curves. Compare them against your concept's daypart mix rather than against a generic income threshold.
- What else is there? Competing supply inside the same polygon is the shelf you are fighting for. Category density inside the drive time matters more than the single competitor across the street.
- Can people actually get in? Ingress, egress, signal position, and side of street. This is the variable that reads fine on a screen and kills a site in person.
- Does delivery extend the math? The major delivery platforms cap their standard service radius at roughly 5 miles, which can widen the addressable demand around a site well past the dine-in polygon. The ghost kitchen shakeout is the cautionary version of leaning on that too hard.
What the failure numbers actually say
The claim that 90 percent of restaurants fail in year one is repeated constantly and has no source. H.G. Parsa at Ohio State went looking for it and found no evidence of a 90 percent failure rate anywhere in the literature. The most traceable origin is an unsourced early-2000s television commercial.
Parsa's own study of Columbus restaurants put first-year failure at about 26 percent, 19 percent in year two, and 14 percent in year three, for a cumulative three-year rate near 59 percent. Franchised units failed at 57 percent over three years against 61 percent for independents. That data is from 1996 to 1999, so it corrects a myth rather than benchmarking 2026.
For the current cycle, the honest numbers are these. The National Restaurant Association's 2026 State of the Restaurant Industry, published in February 2026, projects $1.55 trillion in industry sales with 1.3 percent real growth, while reporting that 42 percent of operators were not profitable in 2025 and 60 percent saw softer customer traffic. Technomic data published in February 2026 shows independent restaurant unit count falling 2.3 percent in 2025 to 412,498 locations, a net loss of more than 9,500 units, while chains grew 1.4 percent to over 263,000 units. Black Box Intelligence projects 9 percent of full-service units are at risk of closure in 2026, defined as units that lost 30 percent or more of their peak sales.
Growth and contraction in the same industry at the same time is the actual condition. Circana's 2026 ranking found the top 50 restaurant brands take 61 percent of total industry spending from just 24 percent of locations. Chick-fil-A added 178 net new restaurants in 2025 and grew US system sales 5.2 percent to nearly $24 billion, an opening pace well ahead of the year before. The brands with a repeatable location method keep opening while the market shrinks around them.
Cannibalization is a restaurant problem first
Small trade areas mean overlapping trade areas. A new unit placed 2 miles from an existing one can sit largely inside the same 6-minute polygon, and the transfer shows up as a sales decline at a store that was performing fine.
Cannibalization analysis belongs in the site decision for that reason, not in a postmortem six months after opening. The right question is not whether the new site will do volume. It is whether the new site will do enough incremental volume to justify the capital after the transfer from your existing units. A site that opens at $1.4 million while pulling $400,000 out of a neighbor is a $1 million site wearing a disguise.
For franchise systems the same math shows up as encroachment, with a contract attached. Exclusive territories are defined in the franchise agreement, and most systems require notice to incumbent franchisees within a defined radius plus a window to raise an impact objection. Franchise real estate counsel commonly cite a projected sales-impact threshold of no more than 15 percent as the point that triggers franchisor review, though the number is set system by system rather than by any standard. Ambiguous territory boundaries are the most common source of disputes, which makes an honest overlap model useful to both sides of that conversation. The franchise site selection process has to carry this step or the system litigates it later.
Market-level saturation is the same question at a higher altitude. Before adding the fourth unit in a metro, market saturation analysis tells you whether you are capturing new demand or redistributing your own.
How restaurant teams run this now
Most restaurant real estate teams are small. One to three people is common, even at brands opening dozens of units a year, and they are choosing between real estate committee dates while the broker deck sits in an inbox. The tooling reflects that pressure.
Placer.ai measures visitation and is the common source for trade-area and competitive foot traffic reads. Buxton builds forecasting and cannibalization models as an analytics partner. SiteZeus is built around sales forecasting for franchise and multi-unit restaurant expansion. CoStar is where the available space and the lease comps live. Esri supplies the demographic and mapping layer underneath much of the category. Most teams run several of these at once, which is the actual problem: the trade area is in one tool, the forecast in another, the lease terms in a third, and the site packet gets rebuilt by hand before every committee.
GrowthFactor scores sites across configurable lenses and shows the inputs behind the score, so the trade area, the competitive set, the overlap with your existing units, and the deal status sit together. Site scoring produces a full site report in about 10 seconds, and the deal pipeline holds the LOI and lease terms next to the score instead of in a separate spreadsheet. The weights are yours to change, which matters for a restaurant brand whose format works nothing like the retailer next door.
The results our restaurant and food customers report are about pace and discipline together. Lil Sweet Treat went from 2 locations to 8 in a single year with a two-person team, cut site evaluation from 3 weeks to 2 days, reviews 120 or more sites a month, and every location has come in at or above underwriting projections. Cavender's went from 9 new stores in 2024 to 27 in 2025 after evaluating more than 2,000 sites, cutting analyst time per site in half, with every new location performing at or better than expected.
More sites reviewed is only worth something if the ones you sign perform. That is the pair worth holding onto: a method that lets a small team look at everything, and a forecast honest enough that the rent you agree to is a rent the site can carry.
Frequently Asked Questions about Restaurant Location Strategy
Here are concise answers to common questions about restaurant location strategy from multi-unit operators and franchise real estate teams.
What is restaurant location strategy?
Restaurant location strategy is the method a brand uses to decide where its next unit goes. It sets the format first, sizes the trade area by drive time rather than by a fixed radius, forecasts sales for that specific trade area, and then tests the asking rent against the forecast using an occupancy cost target. The order matters, because the format determines which sites are even eligible.
What percentage of sales should restaurant rent be?
The widely used operator rule keeps total occupancy cost between 6 and 10 percent of gross sales. Occupancy cost is rent plus CAM, insurance, and property tax, not base rent alone. Below 8 percent gives you room when sales come in soft. Above 10 percent, the site needs the forecast to be right on the first try, which is rarely how a new unit performs.
How big is a restaurant trade area?
Smaller than most operators assume. Quick-service concepts draw most of their business from roughly a 5 to 7 minute drive time, which is often only 1 to 3 miles. Casual dining pulls from a wider 10 to 15 minute band. Because the boundary is drive time and not distance, a highway, a river, or an unprotected left turn can cut a trade area in half without changing the mileage at all.
Do 90 percent of restaurants fail in the first year?
No. Researcher H.G. Parsa at Ohio State went looking for the source of that claim and found no evidence of a 90 percent failure rate anywhere in the literature. His own study of Columbus restaurants put first-year failure closer to 26 percent, with 19 percent in year two and 14 percent in year three. The data is from the late 1990s, so treat it as the correction to a myth rather than a current benchmark.
How does GrowthFactor compare to SiteZeus for restaurant site selection?
SiteZeus is built around sales forecasting for franchise and multi-unit restaurant expansion, and forecasting is the center of its product. GrowthFactor is a site selection and deal pipeline platform, so the trade-area analysis, the site score, the cannibalization check, and the lease terms sit in one place through the whole deal rather than handing off to a separate pipeline tool after the forecast. Every input behind a GrowthFactor score is visible and the weights are yours to change, which is what lets a real estate team defend a site to a committee. Lil Sweet Treat went from 2 locations to 8 in a year reviewing 120 or more sites a month with a two-person team.