Why 'Within 10 Miles' Is the Wrong Way to Define Your Service Area

Drawing a circle on a map and calling it your service area causes real problems. A look at why postcode and territory boundaries need more care than a radius check.

The US Postal Service just started publishing official ZIP code boundary polygons, the actual shapes of postal areas rather than a rough point and radius. It's a small, technical-sounding update. But it points at a problem I see in a lot of UK business software: the "service area" or "territory" logic was built as a circle on a map, and a circle is almost never the right shape.

The radius check feels simple, so it gets built first

If you ask a developer to work out whether an address is "in range", the quickest way is to take the customer's postcode, convert it to a latitude and longitude, and measure the straight-line distance from a central point. If it's under 10 miles, they're in. Over 10 miles, they're out.

It's quick to build and easy to explain. The trouble is that a straight line doesn't care about rivers, motorways, national parks, one-way systems or the fact that two postcodes 400 metres apart by road can be 11 miles apart as the crow flies and 25 minutes apart in real driving time.

I've seen this cause genuine problems:

  • A plumber's software quoting a job as "in area" for an address across a river with no bridge for three miles.
  • A delivery business excluding a customer who is nearer than several customers it already serves, because the circle cuts through their street.
  • A franchise system assigning a lead to the wrong territory because two radius circles overlap and the logic picks the first match rather than the correct one.
  • A clinic's catchment area including a postcode that's genuinely a 40 minute drive, because the motorway makes the straight-line distance misleading.

None of these are edge cases. They're what happens the first time a circle meets a real map.

What a proper boundary actually gives you

A real boundary, whether it's a postcode district, a local authority area, a drawn-on-the-map delivery zone or a drive-time isochrone, is a shape with edges that follow something real: a road, a river, an administrative line, or an actual travel time calculation. In the UK, that data mostly comes from Ordnance Survey boundary products or the ONS postcode directory, both of which map postcodes to proper geographic areas rather than a single point.

Once your software checks "is this postcode inside this shape" instead of "is this point within X miles of that point", a few things get easier:

  • Sales territories can be assigned by postcode district or sector, matching how your reps or franchisees already think about their patch, rather than a distance that ignores roads entirely.
  • Delivery zones can follow your actual coverage, including the awkward bits where you serve one postcode but not its neighbour because of a contract boundary or a driver's route.
  • Service area checks on a booking form can give a straight yes or no, instead of quietly letting through a booking that's technically outside the useful range but passed the radius test anyway.
  • Reporting can group customers by a boundary that means something to the business, not just a distance band that was convenient to calculate.

Drive time is sometimes the better measure

For some businesses, neither a radius nor a fixed boundary is right. A trade that charges a call-out fee based on distance, or a business that promises same-day delivery within a travel time, needs an actual routing calculation rather than a shape on a map. That's a different kind of integration (a mapping API that returns drive time, not just distance), and it costs more to run because you're paying per lookup. It's worth it when travel time genuinely drives your pricing or scheduling. It's wasted cost if a simple postcode boundary would have done the job.

This is an integration, not a toggle

Swapping a radius check for proper boundary data isn't a config setting you flip on. It usually means:

  • Getting hold of the right boundary dataset (postcode districts, sectors or a custom-drawn shape) and keeping it updated as boundaries occasionally change.
  • Storing those boundaries in a format your database can query efficiently, since checking "is this point inside this polygon" is a different kind of query to "what's the distance between two points".
  • Deciding what happens at the edges. A boundary still has a line somewhere, and you need a clear rule for addresses that sit right on it, not a different answer depending on rounding.
  • Handling postcodes that don't map cleanly, like new builds that haven't made it into the postcode directory yet.

None of this is exotic. It's the kind of systems integration work I do regularly, connecting a boundary dataset or mapping API into whatever system is making the "in area or not" decision. The point is that it's a proper piece of work, not a one-line fix to the radius value.

Worth checking before you expand

If your business is planning to grow its coverage, add new territories, or start charging differently by zone, it's worth looking at how the current "in area" logic actually works before you build more on top of it. A radius check that's been fine for years at a small scale tends to show its problems exactly when you expand into denser areas, cross a geographic feature, or start relying on the boundary for pricing rather than just a yes or no. Getting the shape right once is a lot cheaper than untangling it after it's been feeding your reporting and your sales targets for two years.