Mapping Service Coverage Areas

A practical guide to drawing the line around where a mobile service operates: the four common ways to define coverage, why a simple radius misleads, and how to store, test, price and review zones.

Every mobile service business answers the same question dozens of times a day: do we go there? The answer is your coverage area, and how you define it decides whether bookings are profitable, whether technicians spend their day in traffic, and whether customers just outside the line get a clear answer or a confusing one. This guide covers the main ways to define coverage, how to test an address against it at booking time, and how to keep it honest as demand changes.

Four ways to define coverage

Radius circle
Everything within n kilometers of a base, measured in a straight line. Trivial to compute: geocode the customer, calculate the great-circle distance to the base, compare with the limit.
Drive-time isochrone
The area reachable from a base within a time budget (for example 20 minutes) along the actual road network. Isochrones are generated by routing engines such as Valhalla or GraphHopper from road data, and they reflect bridges, motorways and dead ends.
Postcode or district polygons
Coverage expressed as a list of administrative units: postcode districts, municipalities, named communities. Easy to communicate ("we cover these districts"), and the boundaries already exist as data in many countries.
Hand-drawn polygons
A boundary you draw yourself, usually following roads, rivers and the edges of neighborhoods your team knows. The most flexible option and, for small operators, often the most accurate.
The same base location with three coverage shapes: a radius circle that spills into water, a drive-time isochrone stretched along a highway, and a polygon that follows district edges Radius water Drive-time highway Polygon
One base, three coverage shapes. The radius includes water and the far shore; the isochrone stretches along the fast road; the polygon follows boundaries you choose.

Pros and cons at a glance

MethodStrengthsWeaknessesBest for
RadiusInstant to compute, easy to explainIgnores water, barriers and road layoutFirst drafts and quick pre-filtering
IsochroneReflects real travel effortNeeds a routing engine; shape changes with time of day and data updatesTime-sensitive services, call-out fees
District polygonsMatches how customers describe where they liveDistricts vary wildly in size; boundaries may not match travel effortMarketing pages, regulated or franchise territories
Hand-drawn polygonEncodes local knowledge exactlyMust be maintained; easy to leave gaps between zonesDense cities, small and mid-sized teams

Why a straight-line radius misleads

A radius treats the city as an open field. Real cities have rivers, creeks and harbors with a handful of crossings; motorways that can only be crossed at interchanges; one-way systems that turn a 300-meter hop into a two-kilometer loop; and gated communities with a single entrance. Any of these can make a point that is "close" on the map far away in practice.

A concrete example from Dubai. Take a base near Dubai Marina Mall (25.0766, 55.1402) and a 7 km radius. Atlantis The Palm, on the outer crescent of Palm Jumeirah (25.1330, 55.1187), is 6.6 km away in a straight line, so it is inside the circle. Mall of the Emirates in Al Barsha (25.1180, 55.2004) is 7.6 km away, so it is outside. Yet an OpenStreetMap-based routing engine (OSRM, free-flow, checked at the time of writing) gives roughly 15.4 km by road to Atlantis, because the crescent is reached via the Palm's trunk, and about 14.2 km to Mall of the Emirates. The radius includes the harder trip and excludes the easier one. You can reproduce the straight-line figures in the distance calculator; the road figures need a routing engine.

Building a polygon and testing addresses at booking

A practical workflow for a hand-drawn or isochrone-derived boundary:

  1. Draft the boundary on a map, following features a technician would recognize: main roads, waterfronts, the edges of communities. Start from an isochrone if you have one, then simplify.
  2. Simplify to a few dozen vertices. Precision beyond about 50 meters rarely matters for a service boundary and makes the data harder to review.
  3. Geocode the customer when they enter an address, using a forward geocoder such as the address to coordinates tool. Check the result type: a rooftop or building match is reliable; a street or district centroid near the boundary is not, so ask the customer to drop a pin.
  4. Pre-filter with the polygon's bounding box (four comparisons), then run a point-in-polygon test. The standard ray-casting algorithm counts how many polygon edges a ray from the point crosses: odd means inside. Libraries such as Turf.js (booleanPointInPolygon) and PostGIS (ST_Covers) do this for you.
  5. Decide boundary behavior explicitly. A point exactly on an edge is rare with real coordinates, but pick a rule (for example "on the line counts as inside") and use a function that implements it.

The guide to bounding boxes explains the pre-filter step in more detail, including the west, south, east, north ordering.

Storing zones as GeoJSON

GeoJSON, defined in RFC 7946, is the most portable format for zones: every mapping library, PostGIS and most databases read it. Three rules catch people out:

  • Longitude first. Positions are [longitude, latitude], the reverse of how coordinates are usually spoken. A Dubai point is [55.14, 25.08], not [25.08, 55.14]; swapped values land in Lithuania.
  • Closed rings. A polygon ring needs at least four positions, and the first and last must be identical.
  • Right-hand rule. RFC 7946 says exterior rings should be counterclockwise and holes clockwise. Parsers should not reject polygons that ignore this, but some tools compute areas or fills incorrectly when winding is reversed, so normalize on save.
{
  "type": "Feature",
  "properties": { "zone": "A", "name": "Marina–JLT–Al Barsha core", "travel_fee": 0, "priority": 1 },
  "geometry": {
    "type": "Polygon",
    "coordinates": [[
      [55.130, 25.068], [55.160, 25.055], [55.215, 25.095],
      [55.210, 25.125], [55.165, 25.110], [55.135, 25.085],
      [55.130, 25.068]
    ]]
  }
}

This ring is counterclockwise (its signed area in longitude/latitude is positive), closed, and stored longitude first.

Worked example: a mobile bike and e-scooter repair service

Consider a hypothetical mobile repair service maintaining bicycles, e-bikes and e-scooters in western Dubai, with a van based near Dubai Marina Mall. Using the polygon above as Zone A and running a point-in-polygon test on a few public locations gives:

LocationCoordinates (lat, lon)Straight line from baseInside 7 km radius?Inside Zone A?
Jumeirah Lakes Towers25.0709, 55.13870.7 kmYesYes
Dubai Internet City25.1022, 55.17394.4 kmYesYes
Atlantis The Palm25.1330, 55.11876.6 kmYesNo
Mall of the Emirates25.1180, 55.20047.6 kmNoYes

The polygon reverses the radius's two mistakes. For this kind of work, a few practical refinements follow:

  • Riders often keep e-scooters in apartment towers, so the job location is a building entrance, not a street. Store the pin the customer confirms, not just the geocoded street.
  • Palm Jumeirah can become its own Zone B with a fixed travel fee rather than an outright "no".
  • Communities with controlled gates may need a visitor pass; record that as a property of the zone or the address, because it adds minutes to every visit.

For how a city's naming and road layout shape zones like these, see mapping Dubai neighbourhoods and service zones.

Handling overlaps and gaps between zones

Once you have more than one zone, two failure modes appear. Overlaps mean an address matches two zones and the price or team depends on which test ran first. Gaps are slivers between hand-drawn neighbours where an address matches nothing and the customer is wrongly told "not covered".

  • Give every zone a numeric priority and always resolve overlaps by it, so the answer is deterministic.
  • Draw shared edges once and reuse the same vertices for both zones (snapping), rather than tracing the border twice.
  • Check gaps by computing the union of all zones and comparing it with the outer boundary you intended; anything missing is a gap. GIS tools and PostGIS (ST_Union, ST_Difference) make this a one-line check.
  • Prefer nested rings for tiers: Zone A inside Zone B inside Zone C, tested from the inside out.

Tiered pricing by zone

Zones let you price travel honestly instead of refusing work. A common structure is a core zone with no travel fee, an extended zone with a fixed fee, and an outer zone where jobs are quoted individually or batched onto set days. Keep the fee tied to measured travel effort: if a zone's average round trip is twice the core zone's, its fee should reflect that. Show the zone and fee to the customer before they confirm, and store the zone ID on the booking so later analysis does not depend on re-running geometry against a polygon that may since have changed.

Reviewing coverage against actual demand

A coverage map is a hypothesis. Test it every month or quarter:

  • Log every coverage check, including declined addresses outside the area. Plotting declined requests shows where the boundary is costing you work.
  • Compare planned and actual travel time per zone. A zone whose actual travel consistently exceeds plan is drawn too generously or priced too low.
  • Look at job density: a zone with few jobs but long trips may be better served on fixed days.
  • Version your polygons with an effective date, so historical bookings can be analyzed against the boundary that applied at the time.

The broader operational picture, from geocoding job addresses to dispatch, is covered in location data for field service businesses, and the next step after coverage is planning routes for mobile repair technicians.

Key takeaways

  • Use a straight-line radius only as a quick pre-filter; water, motorways and one-way systems make it unreliable as the boundary itself.
  • Hand-drawn or isochrone-derived polygons match real travel effort; test addresses against them with a bounding-box pre-filter plus point-in-polygon.
  • Store zones as GeoJSON: longitude first, closed rings, counterclockwise exteriors per RFC 7946.
  • Resolve overlaps with explicit priorities, check for gaps with a union, and price tiers by measured travel effort.
  • Log declined addresses and actual travel times, and revise the map on a schedule.