Catchment area analysis, step by step: reach, residents, rivals and your own customers
A catchment area analysis in six steps: choose how customers travel, draw the area on the roads, count residents and rivals inside, and check it against where customers live.
Part of the guide Location intelligence automation
A catchment area analysis answers one question: how many potential customers can reach this place easily, and how many other businesses compete for them? You draw the area people can reach within a travel time or road distance, count the residents inside it, count the rivals inside it, and divide one by the other. Then you check the drawing against where your real customers come from, because the best model is the one your own data confirms.
This walk-through is for owners, franchise developers and analysts who need to judge a site or compare branches. Other articles explain what a catchment is; this one is the method, in order, with a made-up worked example from Vancouver, the errors that most often distort the answer, and the AI employee that redoes it every month.
The six steps
- Decide how your customers travel. A corner shop is a walk, a clinic or a gym is usually a short drive, a furniture store a longer one. Pick one way of travelling for the main analysis.
- Choose the reach. A travel time (10 minutes by car, 15 minutes on foot) or a distance by road. Pick what matches how far people will go for what you sell.
- Draw the area on the roads. Not a circle: rivers, motorways and one-way systems make the real area lopsided.
- Count the residents inside, and the age groups that matter to you, in countries where population data exists.
- Count the rivals inside, the businesses a customer would choose instead of you, and note which are chains.
- Check against your customers. Put the postcodes or addresses of real customers on the same map and see how many fall inside.
Steps one to three: reach before numbers
Most bad analyses go wrong before any counting starts, by choosing the wrong reach. Two habits help. First, use the travel your customers actually use: on-foot areas for a pharmacy in a dense centre, car areas for a garden centre on the edge of town. Second, keep the reach the same when comparing sites, so the comparison is fair.
Drawn on the roads, a 10-minute car area around a site next to a river or a ring road can be twice as long in one direction as in the other. That shape decides which neighbourhoods are in, and so which residents and rivals you count. Drive and walk areas describe usual traffic, not the traffic of a given hour.
Steps four and five: residents per rival
With the area drawn, count who lives in it and what already serves them. The simplest single figure is residents per rival: the number of people in the area divided by the number of competing businesses. A high number suggests room; a low one suggests a crowded market where you would need to win customers from others. Narrow the residents to the ages you serve when that matters: children for a tutoring centre, people over 60 for a hearing clinic.
A worked example in Vancouver, made up for illustration
Say a physiotherapy group in Vancouver is choosing between two sites for a fourth clinic. Its patients mostly drive or take a short bus ride, so it uses a 10-minute car area around each. It counts residents, adults over 40 (its main patients), and physiotherapy clinics already inside each area.
| Site | Residents in 10 min by car | Adults over 40 | Clinics inside | Adults over 40 per clinic |
|---|---|---|---|---|
| Site 1, near the centre | 210,000 | 92,000 | 46 | 2,000 |
| Site 2, inner suburb | 145,000 | 71,000 | 18 | 3,900 |
Made-up figures for illustration only.
Site 1 reaches more people, but the area is crowded with clinics. Site 2 reaches fewer, yet each clinic there shares almost twice as many likely patients. The group then puts the home postcodes of last year's patients at its three clinics on the map: 70% fall inside each clinic's 10-minute area, which tells it the 10-minute reach is a fair model for its business. It shortlists Site 2 and goes to look at parking and visibility, which no map can judge.
Errors that distort a catchment
- Circles instead of roads. A circle across a river counts people who cannot get to you.
- The wrong reach for the business. Nobody drives 20 minutes for a coffee; a 5-minute car area for a car dealer misses most of its buyers.
- Counting every business of a type as a rival. A hospital physiotherapy ward is not the same choice as a private clinic. Filter to what customers compare.
- Skipping the check. Without placing real customers on the map, the reach is a guess. A few hundred addresses are enough to test it.
- Counting residents for a workday business. A lunch spot in an office district serves people who live elsewhere.
Running the analysis as an AI employee
In xMap Flows the six steps become one flow: an AI employee that redraws each site's area and recounts residents and rivals every month, so the picture keeps up as rivals open and close. The location intelligence automation page shows how such flows fit together.
- Bring in your sites or candidates with My data, after a monthly Schedule.
- Add Travel area with a travel time and a way of travelling, around each incoming place (up to 15 minutes or 10 km by road for now).
- Add Demographics on the Area of interest the Travel area hands on, for residents and age groups inside each area.
- Add Nearby places searching the same areas for your rivals.
- Add a Calculated field for residents per rival, and send the table by Email with a map.
What lands each month is one row per site with its area, residents, the age groups you chose, the rivals inside and residents per rival, plus a map of the areas. For step six, upload your customers' addresses as a table and use Find nearest to see which site each customer is closest to. Templates that follow the same path: Rivals within 10 minutes' drive of each store and Who my warehouse reaches in 15 minutes.
Questions
What is the difference between a catchment and a radius?
A radius is a circle of equal distance in every direction. A catchment drawn on the roads follows how people really travel, so it bends around rivers, motorways and barriers.
Which travel time should I use?
The one your customers accept for what you sell: a few minutes on foot for daily needs, around 10 minutes by car for clinics and gyms, longer for big-ticket stores. Test it against where your customers live.
Can I do a catchment analysis in the United States?
You can draw the areas and count rivals anywhere. Resident counts are not available there, so use homes, schools and supermarkets inside the area as a stand-in.
How often should I redo it?
Monthly is enough for most. Residents change slowly, but rivals open and close, which moves residents per rival.
Templates that do this
Read next
More on population and demographics · Every article · Templates