Location, Location, Location: How to Choose a Retail Site

Retail still lives and dies on a handful of unromantic questions. Can a customer see you from the road? Can they turn in without crossing three lanes of traffic? Is there a place to park when they get there? Does the trade area actually contain the households you sell to, or just a pretty average household income on a broker flyer? Window displays still matter. They do not rescue a site that fails those tests.

I used to treat location, location, location as a slogan. After watching operators open too fast, and after watching national chains spend real money on drive-time models instead of vibes, I treat it as a discipline. The work is not picking a cute corner. It is matching a format to a catchment, then proving the parking, the ingress, the co-tenants, and the traffic pattern will still work on a rainy Tuesday in February.

Start With the Trade Area, Not the Rent

A three-mile ring on a map is a drawing, not a market. People do not shop in perfect circles. Rivers, interstates without exits, one-way downtown grids, and school-pickup congestion all change who can actually reach the door. Sophisticated retailers draw drive-time isochrones: the set of addresses from which a shopper can reach the site in 5, 10, or 15 minutes under real road speeds.

Grocery and convenience formats live in tight catchments. A regional mall or a home-improvement warehouse pulls from much farther. The International Council of Shopping Centers still publishes typical primary trade-area sizes by center type, from about three miles for a neighborhood center to 5–25 miles for a super-regional mall.

The idea that the probability of a given consumer visiting and purchasing at a given site is some function of the distance to that site, the site’s attractiveness, and the distance and attractiveness of competing sites.

Esri GIS Dictionary, gravity model

That gravity logic is why a 12,000-square-foot specialty store and a 180,000-square-foot Target do not compete for the same drive. Attractiveness (size, assortment, price authority) pulls farther. Distance and competing options push back. David Huff’s 1964 model, still taught in Geographic Information Systems (GIS) site work, turns that into probabilities instead of hard rings. If you skip it, you will overcount rooftops on the wrong side of a freeway and call it demand.

Walmart’s own teams have said the same thing, just in different language. The company moved delivery catchment areas off fixed-mile radii and onto a hexagonal grid that mixes historic order data, Census inputs, driver capacity, and drive time.

This is helping us to adapt how we service our customers, by allowing us to go from a fixed-mile radius into a much more dynamic catchment area that caters to the needs of the customers that a particular store will serve.

Parthibban Raja, senior director of engineering, Walmart Global Tech, in Fast Company

You do not need Walmart’s engineering bench to steal the principle. Score every candidate on who can get there, not who lives inside a compass circle.

Parking, Ingress, and Traffic Patterns Are Not Soft Skills

Parking is capacity. If the lot fills before the dinner rush, the rest of your demand never becomes a ticket. Industry parking studies used by planners still cluster around four to five spaces per 1,000 square feet of gross leasable area for shopping centers, with higher ratios for grocery and restaurants.

A ULI and ICSC-backed study summarized in municipal parking analyses recommended 4.0 spaces per 1,000 square feet of GLA for centers under 400,000 square feet, rising toward 4.5 as centers get larger, with extra supply when restaurants, entertainment, or cinemas take a bigger share of the mix. A supermarket often needs closer to 10–15 spaces per 1,000 square feet because dwell time and basket size are longer.

Those numbers are planning tools, not decorations. Count the stalls. Count how many employees of other tenants take them. Watch the lot at 12:15 p.m. and at 6:10 p.m., not during the broker tour at 10 a.m.

The key recommendations of this study provide a set of base recommendations for parking supply based on center size and makeup.

ULI / ICSC parking ratios, as summarized in the Town of Hempstead parking analysis

Traffic counts without direction are another trap. A 30,000-car average daily traffic number is not useful until you know which way those cars are going at the daypart you sell. Quick-service restaurants hunt the going-home side of a commuter corridor and a clean right-in, right-out.

A furniture warehouse can live on a harder left turn because the trip is planned. A college-town lunch counter with parking only on the far side of a four-lane arterial will train students to eat somewhere else. Walk the approach the way a first-time customer would: signal, turn, park, walk, see the sign. If any of those steps feel like work, the site is already expensive.

Convenience is not a slogan. The National Retail Federation’s Consumer View work found that shoppers will abandon a purchase when the path gets hard, and that they are more likely to choose a retailer when convenience is real.

The report found 97 percent of respondents have backed out of a purchase because it was inconvenient for them. … 93 percent say they are more likely to choose to shop at a specific retailer based on convenience.

National Retail Federation Consumer View, as reported by Convenience Store News and Chain Store Age

Numerator’s convenience survey put store location at the top of the list of convenience factors, with proximity to home as the most important definition of a convenient site. If the lot, the turn, or the walk from the stall fights that instinct, you are arguing with the customer’s calendar.

How Large Retailers Actually Pick Dirt

Big-box and small-box chains do not feel a location. They score it. The stack is usually the same even when the weights change: demand (households, income, category spend), access (drive time, traffic, parking), competition and cannibalization, co-tenancy, and deal math (rent, tenant improvement, exclusives).

Walmart and other mass merchants still care about highway access, population density, and spacing from their own stores so they do not steal their own volume. Delivery has made the trade area even more operational: the store is a warehouse with a parking lot. Target and other general-merchandise chains layer lifestyle segmentation and co-tenancy (who else is already drawing the trip) on top of the same drive-time logic. Home improvement warehouses care about housing stock, renovation permits, and whether a contractor can get a truck in and out. None of that is a coffee-shop problem. It is a network problem.

Dollar General is the clearest small-box counterexample to just find a busy corner. The company has spent decades putting 7,500–10,000 square-foot boxes into towns that cannot support a supercenter. Management’s public real-estate story is consistent: about 75 percent of Americans live within five miles of a store, and roughly 80 percent of the base serves towns of 20,000 people or fewer. That is a deliberate catchment, not a failure to reach the city.

Site selection follows a consistent framework, unchanged for decades: accessibility for underserved customers, alignment with target demographics, and real estate economics that support attractive returns.

Quartr, “Dollar General: The Rural Retailer Playbook”

The lesson for an independent is not become Dollar General. It is that format and catchment have to match. A 6,000-square-foot dining room in a town that only produces lunch traffic is the inverse error of putting a tiny box where customers expect a warehouse.

ICSC’s shopping-center definitions still force that honesty: neighborhood centers are convenience-oriented with a supermarket pull of about three miles; community centers stretch to 3–6 miles with a discount or home-improvement anchor; regional malls are a different animal entirely.

If your concept needs the mall’s trade area but you signed a neighborhood strip, the rent will not be the thing that kills you. The missing rooftops will.

Format Has to Match How People Actually Use the Store

Location analysis fails when the building is a time capsule. A site that was perfect for a 1998 behavior can be wrong for a 2026 trip.

Buffalo Wild Wings built the sports-bar version of wings: large dining rooms, walls of televisions, beer, linger time. That made sense when the living room screen was small, and the game was a night out. Households now have the game at home on a screen the size of a small sail. Take-out and delivery took a third of the chain’s sales, and the company had to invent a second format to chase that trip. Traditional BWW sports bars are about 6,000 square feet. BWW Go locations are about 1,500 square feet, limited to take-out and delivery, and cheaper to build.

While a BWW sports bar is around 6,000 square feet on average, a Go location is roughly 1,500 square feet. That means cheaper real estate that’s faster to build and easier to operate.

Abigail Bowie, Buffalo Wild Wings, in CNBC

The franchise filings behind that shift are blunt. Restaurant Business, working from Buffalo Wild Wings’ 2026 FDD, reported 1,178 traditional sports bars at year-end 2025, five fewer than the year before, while BWW Go jumped nearly 60 percent to 219 locations. Average franchised sports-bar sales were about $3.6 million, barely up. Average franchised Go sales were about $928,700. Growth was in the small box. The big box was treading water.

There is the restaurant company’s core sports bar business, specializing in chicken wings, beer and televised sports, and then there is its scaled-down offshoot, Buffalo Wild Wings Go, with its small storefronts focused on takeout and delivery.

Restaurant Business, “To-go locations fueled Buffalo Wild Wings’ growth last year”

Wingstop never built the sports bar. Most of the business was take-out even a decade ago, which let the chain keep restaurants around 1,700 square feet, skip the liquor-and-TV cost structure, and ride online ordering. Business Insider reported that 75 percent of Wingstop’s mix was take-out in 2015, and that the small footprint was the point.

This allows Wingstop to keep restaurants small (about 1,700 square feet), minimize costs, and prepare for two of the biggest trends in the industry: online ordering and delivery.

Kate Taylor, Business Insider, “This chicken chain is beating Buffalo Wild Wings because customers don’t want to hang out there”

That is the Wild Wings versus Buffalo Wild Wings problem in public filings: one format assumed people would keep coming in to watch. The other assumed they would take the food to the couch. The take-out specialist kept opening small rooms near where people already were. The sports-bar specialist is now retrofitting a Go box to catch up. If you are signing a 15-year lease on a dining room whose only thesis is “big screens,” you are betting against the living room.

A Local Operator Who Outran the Lot

I watched a local restaurant operator make the same mistake without a Wall Street ticker. He had a formula that worked in college towns: hungry students, walkable blocks, late hours, not much need for a family minivan or a 40-stall lot. Success made him aggressive. He started opening in markets that looked like more of the same on a demographic printout and ignored parking, traffic patterns, and how a suburban dinner trip actually happens. There was no clean turn. There were not enough stalls when the dining room was full. Staff parked in the spaces customers needed. The college-town muscle memory did not transfer. He had to close and sell the expansion stores at a loss.

That is not a story about food. It is a story about catchment. A campus location can live on pedestrian counts and a two-hour dwell. A family location lives on the right-in from the going-home lane, and a lot that still has spaces at 6:30. Copying the menu without copying the trip is how you light money on fire. The National Retail Federation numbers above are the polite version of what his guests already knew: if it is inconvenient, they leave, even when they wanted the product.

Windows Still Sell. They Cannot Fix the Intersection.

Once the site works, the glass still does a job. Passersby decide in seconds whether the store is for them. Signage, lighting, and a display that shows the actual offer (not a lifestyle poster) remain cheap media you own. I would rather a retailer spend a week on the approach and the lot than a week arguing about a mannequin. The mannequin only gets a vote if the car can stop.

Visual merchandising inside the box matters too, after the customer has already committed to the trip. Do not confuse the two. A brilliant window on a site with a brutal left turn across a high-speed arterial is a billboard for a store people cannot use. A plain window on a site with parking, visibility from both directions, and the right co-tenants will still take money.