Catchment area analysis · the area, not the dot
Where does a location's real reach begin and end?
A catchment area (also known as a trade area) is the zone from which a place draws its customers. Drawing it on purpose — by distance, by a boundary you trace, or by measured drive times — is what turns a pin on a map into an area you can reason about. Inside it you read population, competing places, density and access, and every figure arrives labeled with its source, dataset version and retrieval date.
The definition
What a catchment area (trade area) actually is
A catchment area is the geographic area from which a place draws its customers, visitors or deliveries. In retail site selection it is the zone whose population, competing businesses and road access determine whether a location can trade at all. It is drawn deliberately — as a distance ring, a boundary you trace, or a set of measured drive times — rather than assumed from a pin on a map.
That definition is worth keeping narrow. A catchment is not the area a business wishes it served, and it is not the circle a mapping tool draws around a marker for you. It is the area you have declared, measured and labeled, so that anyone checking your work knows exactly which boundary produced the numbers. Two analysts using the same site and different boundaries will get different figures, and both will be correct — which is exactly why declaring the boundary is what makes any comparison possible.
Vocabulary
Catchment, trade area, service area and radius are four different things
These words get used interchangeably in retail planning, which makes comparison harder than it needs to be. They describe four different objects, and only one of them is a fact about the world rather than a choice you make.
A catchment area is the zone a place actually serves. It exists whether or not anyone has drawn it, and it can only be observed indirectly, through where people come from.
A trade area is a boundary drawn around a site for the purpose of measuring it. It is a methodological choice — a decision about where to stop counting — and a perfectly good one, provided the method is declared alongside the result.
A service area is what a business actively commits to serving, often constrained by a delivery promise, a contractual radius or a branch policy. It may be deliberately narrower than the catchment.
A radius is a geometric measurement and nothing more. A 4 km ring around an address is a fact about circles, and on its own it says nothing about whether those four kilometres are hilly, cut off by a river, or drivable in four minutes.
LocateRight uses both the catchment and trade-area vocabulary deliberately. The area you draw is the trade area; what you then measure inside it is the evidence about the catchment. Both words appear here because both describe real steps in the same workflow, and collapsing them would hide which one you chose.
Bounding the area
Three ways to bound a catchment area, each labeled for what it is
These are not three versions of the same claim. Each answers a different question, and the product labels the difference on the map rather than leaving it to be guessed.
| Boundary | What it measures | What it explicitly is not |
|---|---|---|
| Distance rings | Dashed reference circles (800 m and 4 km) around the selected point, tooltip-labeled by distance | Never presented as travel time — a minute-based guess would be a different claim, and the rings refuse to make one |
| Radius & drawn boundary | A study area from roughly 100 m to 20 km; the places view additionally honors a polygon you trace on the map | Not a model of buying behaviour — a fixed, reviewable extent that every count below is computed within |
| Drive-time probe paid plans | Sampled points across the study area, actually routed to the site, reported in minutes | Not a smooth travel-time polygon — a labeled sample, described precisely below |
Whichever you pick, the boundary travels with the figures. Any measurement here can be reproduced by declaring the same boundary, and how each figure on this page is produced sets out the formulas as implemented.
Evidence inside the boundary
What you can read once the area is drawn
Once the extent is set, the same analysis surface fills it with measurements — each one carrying the dataset it came from and the version it was read at.
People
Population in range
Grid figures from WorldPop summed within the radius, with the nearest real settlement named as context and the boundary attached to the number.
Supply
Competing places & density
Same-category places with distance, plus business density per km² and category mix across the area, counted from Overture Places and OpenStreetMap.
Gaps & anchors
White space & co-location
Which categories are absent from inside the boundary, and which anchor categories genuinely share it — presence and absence, checked per area rather than assumed from a map view.
Ground
Land & buildings
Land-cover shares from ESA WorldCover and building-footprint counts from Google Open Buildings and the Microsoft footprint dataset, where those tiles are published.
Stated limits
What the population figure does and does not cover
Population is the figure most likely to be over-read, so its limits belong on the page rather than in a footnote. Three of them are real.
It is modelled, not counted. WorldPop distributes census and survey inputs across roughly 100-metre grid cells. The result is an estimate built from a model, and a grid-cell total is never an individual-level count. The WorldPop reference sets out the method.
The vintage is 2020. Whatever has changed in the years since — new housing, a closed factory, a road realignment — is not in these numbers.
The ingested raster covers three countries. In this deployment the population raster covers Saudi Arabia, the United Arab Emirates and Egypt. An analysis anywhere else returns no population figure: it does not substitute an estimate from a neighbouring country, and it does not report a zero that could be misread as "nobody lives there" — the field is marked as having no data. Place counts, land cover and building data have much wider coverage, so an international site is still measurable on everything except population. Every dataset's actual coverage and licence is inventoried on the data sources page. An account of what footprint counts can and cannot tell you sits on the Google Open Buildings reference.
The measured version
How the drive-time probe works, and how it reports itself
Distance assumes the road network is uniform, and it isn't. The accessibility tool answers the routing question directly: you fix a site as the destination and a study area around it, and LocateRight samples points across that area a few dozen at a time, routes each one to the site through a routes matrix by road, and reports the trip times it gets back.
What makes the result trustworthy is how it labels itself. The output states its method and retrieval time, and its coverage note reads "sampled points, not full population." Points that cannot be routed are excluded from the statistics and counted, never quietly imputed. The full sequence — how origins are sampled, which points are excluded and counted separately, and what the 10, 15 and 20 minute bands do and do not establish — is documented on the drive-time methodology page. You receive the average, fastest and longest travel minutes, the average distance, the share of successfully routed points reachable within 10, 15 and 20 minutes, and a five-minute-bucket distribution of the whole set.
A few plain consequences of that design, stated so nothing is over-read: it is car travel — walking and transit are not routed; it samples rather than exhaustively measures; and shares are computed over routed points, so a large excluded count is itself a finding about the area. The guide documents what each statistic covers.
Putting the boundary to work
What a deliberately drawn area changes
- Reading competition honestly: "eight coffee shops nearby" means different things in an 800-metre ring and a 20-minute drive. The boundary decides the answer, so the boundary has to be stated.
- Comparing catchments: two candidate sites in the same district can hold very different populations inside the same real distance. Measured areas make that difference visible rather than arguable.
- Framing territory: networks planning coverage boundaries can size what an area actually contains before assigning it to anyone.
- Sourcing over intuition: once an area's counts, density and reach are written down with their sources, a decision can survive being questioned by someone else — and be re-run months later to check what shifted.
The boundary of this tool, too
What this analysis does not claim
- No smooth travel-time polygons. Reach is reported as routed-sample statistics — minutes, shares and a distribution — not as a drawn blob whose outline nobody measured.
- No buying-behaviour model. A population total is not a count of people who will buy, and drive-time reach is not a sales curve. The analysis note shipped with results is firm: raw observations only — no ratings, no predictions, no advice.
- No live congestion. Travel times use routing preferences; they are route computations, not a real-time traffic forecast.
- No silent coverage gaps. Where a dataset has nothing for the area, the block says so with its reason — an empty area reads as empty, not as a zero wearing a number's clothes.
In the workflow, this page's unit — the area — sits between the point examined in business location analysis and the shortlist compared in site selection: analyze the location, bound and read its area, then let the areas decide between locations. The underlying data practice is location intelligence; what each source contributes is inventoried honestly on the about and data-sources page, and a whole portfolio of sites is a different workload — bulk analysis covers that.
Common questions
Catchment area questions, answered plainly
What is a catchment area?
What is the difference between a catchment area and a trade area?
How do I define a catchment area for a business location?
What can I measure inside a catchment area?
How does LocateRight calculate drive time?
What are the limits of the population figure?
Does catchment area analysis tell me which location is best?
How do I compare two candidate locations on the same basis?
Who can use the drive-time accessibility tool?
Draw the area, then read what is inside it
Pick a site, set the boundary — distance, drawn, or measured in minutes — and inspect the evidence with every source named.
Start with a location Comparing candidates? Site selection →