The Circle That Lies to You
Draw a 3-mile radius around any retail location. It looks clean. It looks scientific. It gives you a population count, a demographic profile, and a sense of who lives nearby.
It also ignores how people actually get there.
I grew up in a retail family and have been evaluating sites since I was 15. For most of that time, the 3-mile radius was the standard tool. Grab ESRI data, run the rings, compare locations. Every site got the same circle. Easy to produce, easy to compare.
The problem is that customers don't travel in circles. They travel on roads. Roads that bend around rivers. Roads that dead-end at highway interchanges. Roads that turn a two-mile trip into a twelve-minute frustration or a five-mile trip into a seven-minute cruise.
A drive-time polygon accounts for that. Instead of drawing a circle based on distance, it traces the actual road network and maps every point reachable within a given drive time. The resulting shape is irregular, messy, and far more honest about who can actually get to your site.
This distinction matters more than most expansion teams realize. The method you use to define your trade area determines which demographics you analyze, which competitors you evaluate, and which demand forecast you take to committee. Get the trade area wrong, and everything downstream is wrong too.
What a 3-Mile Radius Actually Shows You
A radius ring measures straight-line distance from a center point. Three miles in every direction. It's geometry.
That geometry is useful for one thing: standardized comparisons. Population within 3 miles in Omaha is directly comparable to population within 3 miles in Tampa. The circle doesn't change based on local conditions. It's the same shape everywhere.
For initial screening, that consistency matters. If you're triaging 500 potential markets and need a rough sort, radius rings let you compare apples to apples quickly. Population density, median income, household count: all standardized by the same boundary definition.
The trouble starts when you use radius data for actual site decisions.
Consider a location on the east bank of a river. A 3-mile radius includes everything within that circle, including a large residential area on the west bank. On a map, those residents are 1.5 miles away. In reality, the nearest bridge is 4 miles south. The actual drive from that neighborhood to your site is 20 minutes. They're not your customers. But the radius says they are, and your demographic profile includes them.
Or consider a suburban site near a highway interchange. A 3-mile radius includes a neighborhood on the other side of the highway. It's close as the crow flies. But there's no exit for two miles in either direction. The neighborhood might as well be in another town for the purposes of your store's accessibility. The radius doesn't know that.
These aren't edge cases. Any site near a body of water, a highway, a railroad, a mountain ridge, or an abrupt land-use change (industrial zone between residential and commercial) will produce misleading radius-based trade areas. That covers most of the country.
What a Drive-Time Polygon Shows Instead
A drive-time polygon starts from your site and traces outward along the actual road network. It follows the roads your customers would drive, at the speeds those roads typically allow. The boundary of the polygon is every point reachable within a given time: 5 minutes, 10 minutes, 15 minutes.
The resulting shape looks nothing like a circle. It stretches along highways (fast travel) and compresses near intersections with congestion. It wraps around barriers instead of pretending they don't exist. A river that a radius ring ignores becomes a clear boundary in the drive-time polygon, because nobody's driving across a river without a bridge.
Three things change when you switch from radius to drive-time.
Your demographics shift. The people included in a drive-time polygon are different from the people included in a radius ring. Some neighborhoods that fell inside the radius drop out (separated by barriers). Others that fell outside the radius appear (accessible via highway). The demographic profile of your trade area, the one you're using to decide whether this location matches your customer, changes. Sometimes it changes a lot.
Your competitor set changes. A competitor 4 miles away but accessible in 6 minutes via highway is more relevant than one 2 miles away but separated by a 15-minute detour. Drive-time analysis reorders your competitive picture based on how customers actually move between options.
Your demand forecast improves. Trade area definition is the foundation of demand modeling. When the trade area is wrong, the demand estimate is wrong. When the demand estimate is wrong, the revenue forecast is wrong. When the revenue forecast is wrong, you're going to committee with a number that doesn't reflect reality. Everything compounds from the trade area boundary outward.
A Real Example: 16 Minutes vs. 23 Minutes
A frozen custard chain assumed their typical customer drove about 16 minutes to reach a store. That's the number they'd used for years. It felt right. It aligned with their intuition about the brand's draw.
When they ran actual drive-time analysis on customer visit data, the real number was 23 minutes.
Seven minutes doesn't sound like much. But expand that seven-minute difference into a polygon on a map, and the trade area grows by roughly 40%. That's 40% more population, 40% different demographic composition, and a fundamentally different picture of who the brand serves and how far its reach extends.
Every site decision this chain had made using the 16-minute assumption underestimated their trade area. Cannibalization modeling was wrong (stores they thought competed were actually serving different populations). White space analysis was wrong (areas they dismissed as "too far" were within the real trade area). Market saturation calculations were wrong (true density was lower than estimated because the catchment was larger).
One number. Seven minutes off. Every downstream analysis affected.
This isn't unusual. Most brands we work with at GrowthFactor discover their assumed trade area is off by 15-40% when they switch from radius or assumption-based boundaries to drive-time analysis built on actual visit data. The direction of the error varies: urban sites tend to have smaller real trade areas than the radius suggests (barriers compress them), while suburban and rural sites tend to have larger ones (highway access expands them).
The correction matters most when the brand has 20+ locations and is using trade area assumptions to model cannibalization, white space, and market saturation across the portfolio. A 7-minute difference at one site is a data point. A systematic 7-minute error across 50 sites is a strategic problem.
When a Radius Still Makes Sense
Drive-time analysis is more accurate, but it's not always the right tool.
Mass screening. If you're evaluating millions of potential sites to narrow a list, radius rings are faster to compute. Running drive-time polygons at scale requires more processing time and data. Use the radius for the first cut, then switch to drive-time for the shortlist.
Standardized benchmarks. Some internal metrics are defined by radius. "Average population within 3 miles" is a benchmark that only works if every site uses the same boundary. Switching to drive-time changes the denominator. If your organization has years of radius-based benchmarks, you'll need both methods during the transition.
Complementary analysis. The best approach often uses both. A 1-mile radius captures the immediate neighborhood (walkable trade area). A 10-minute drive-time polygon captures the realistic customer base. Layering them gives you two useful frames: who lives right here, and who will actually visit.
Resource constraints. Not every team has access to drive-time tools. If you're working with basic GIS or spreadsheet-based analysis, a radius is still better than guessing. The hierarchy, as SiteSeer's analysis frames it, is: good (radius), better (drive-time), best (customer-derived trade areas modeled from actual visit data).
What to Do With This
If your team is still making site decisions based primarily on radius analysis, here's a practical path forward.
Step 1: Run both methods on your top 5 existing stores. Pull the 3-mile radius demographics and the 10-minute drive-time demographics for your five best-performing locations. Compare the profiles. If they're meaningfully different, you know the radius is misleading your analysis.
Step 2: Check for barrier sites. Look at your existing portfolio for locations near rivers, highways, or other barriers. These are the sites where radius analysis is most likely to have overestimated (or underestimated) your actual trade area. Run drive-time polygons on these first.
Step 3: Update your cannibalization models. If your brand has multiple locations in the same metro area, cannibalization modeling depends entirely on trade area overlap. Switching from radius to drive-time may show that stores you thought were competing are actually serving distinct populations, or that stores you thought were independent are sharing more customers than expected.
Step 4: Align your scoring. If you're using a multi-lens scoring framework, make sure the demographic lens and competition lens are pulling data from drive-time trade areas rather than radius rings. The scoring is only as good as the boundary it's built on.
The location analytics market is moving in this direction. According to Fortune Business Insights, the market was valued at $21.15 billion in 2024 and is projected to reach $63.71 billion by 2032 with a 14.7% CAGR, with the retail segment holding the largest share. As the tools get more accessible, the teams still relying on radius-only analysis will find themselves at a growing disadvantage.
FAQ
What is a drive-time polygon in site selection?
A drive-time polygon (also called an isochrone) maps every point reachable from a location within a given drive time by tracing the actual road network. Unlike a radius ring that draws a circle based on straight-line distance, a drive-time polygon accounts for road speeds, barriers, and access points. The resulting shape shows the real area from which customers can reach your site.
Why is a 3-mile radius inaccurate for trade area analysis?
A radius measures distance as the crow flies, ignoring how people actually travel. Physical barriers (rivers, highways, railroads), road network gaps, and varying travel speeds mean that straight-line distance often has little relationship to actual accessibility. A neighborhood 1.5 miles away may take 20 minutes to reach if separated by a river with a distant bridge. The radius includes that population in your trade area when they're effectively unreachable.
When should I use radius analysis instead of drive-time?
Radius analysis is faster for mass screening (triaging hundreds or thousands of sites), useful for maintaining standardized benchmarks across markets, and appropriate when drive-time tools aren't available. Many teams use both methods: radius for initial screening, drive-time for detailed evaluation of shortlisted sites.
How does trade area definition affect site scoring?
Trade area boundaries determine which population is analyzed for demographics, which competitors are included in the competitive assessment, and what demand estimates feed into revenue forecasts. Using the wrong trade area method can shift your demographic profile, reorder your competitive picture, and produce forecasts that don't match reality. Getting the boundary right is the first step in any scoring framework.
The Shape of Reality
A circle is a simplification. Sometimes simplifications are useful. But when you're committing capital to a location based on who lives nearby and how they'll get there, the simplification needs to match reality closely enough to trust.
Drive-time polygons aren't perfect either. Traffic patterns change. New roads open. A development three years from now will alter the drive-time map in ways today's data can't predict. But a polygon built on actual road networks is a better foundation than a circle that pretends geography doesn't matter.
Your trade area definition is the input to every analysis that follows: scoring, committee preparation, demand forecasting, cannibalization modeling. If the boundary is wrong, the analysis is wrong. And the boundary is almost certainly wrong if it's a circle.
See how GrowthFactor models trade areas using drive-time analysis and real visit data.