Almost every local campaign starts the same way: drop a pin on the store, pick a distance, press go. It is quick, it is what the ad platforms offer first, and for a business with a real delivery area it quietly sends a large share of the budget to people the store cannot serve. This page is about the difference between the circle and the zone, and what it costs.
Radius targeting shows ads to everyone within a fixed distance of a point, usually the store's pin. Delivery zone targeting fits the ads to the polygon the store actually serves. The first is a guess with a compass. The second is the store's own operational truth, the same boundary that decides whether a customer can order at all.
The gap between them is not academic. Real zones are drawn by delivery time, main roads, rivers, competitor territories and where the customers actually are. None of those follow a circle, so the circle always covers ground the zone does not, and misses ground it should have covered.
A store on one side of a river, a rail corridor or a motorway can be 900 metres from homes it takes twenty minutes to reach. The circle counts them as neighbours. The delivery zone was drawn by someone who knew about the bridge.
Real zones follow drive time, staffing and suburb boundaries. In one Sydney cluster the six zones measure between 5 and 23 square kilometres, and none of them is remotely round. A single radius setting cannot describe any of them, let alone all six.
In a dense network, circles overlap. Two partners end up paying to reach the same households, bidding against their own brand. Zones tile against each other by design, because operations drew them not to overlap.
Three kilometres is too wide for an inner city store and too narrow for a regional one. Networks solve this by letting every store pick its own radius, which replaces one wrong number with hundreds of separately wrong numbers.
For a delivery business this is literal. A person outside the delivery zone who taps the ad reaches an order page that refuses them. The impression was paid for, the click was paid for, and the sale was never possible.
Platform dashboards report reach and clicks wherever they happen. They do not report "share of spend outside your service area", so the leak never shows up as a line item anyone can challenge. See the return on local marketing.
Lay a real zone over the smallest circle that contains it and the difference is the waste. A compact 7 square kilometre inner suburban zone can sit inside a circle covering roughly 20 square kilometres, which means about two thirds of that circle is area the store does not serve. A long thin zone along a coast or an arterial road does worse. The exact share differs per store, and that is the point: it is knowable, per store, the moment the zone is on screen next to the circle. We also measured it properly: across 450 realistic zones, the median covering circle puts 66% of itself outside the zone, and the fitted plans put 11 to 12% outside. The numbers are on the zone targeting benchmark.
| Radius around the pin | Fitted to the delivery zone | |
|---|---|---|
| Matches the real trade area | By accident, occasionally | By construction |
| Budget outside the service area | Routinely a large share | Small, only the fitting error |
| Overlap between neighbouring stores | Common in dense networks | Zones tile, overlap is by exception |
| Setup per store | Someone picks a number per store | Automatic once zones are loaded |
| Keeps up with change | Nobody revisits the number | Re-fit when operations redraw the zone |
Swipe the table sideways to see the rest.
The ad platforms do not accept polygons, so zone targeting is an engineering job rather than a checkbox. The approach that works: take each store's real zone, then fit a set of smaller circles inside it so the covered area follows the shape as closely as the platform allows. It is an approximation, and an honest one: far less spend lands outside the area than with one big circle, and the fit is visible on a map instead of assumed.
This is how delivery area targeting works in Amplaro. Head office loads the zones once, and from then on every campaign a partner launches is fitted to their store's own polygon automatically, across a whole network, with nobody picking radius numbers store by store. The step by step version is on turning a delivery zone into an ad audience.
No. If a store has no defined service area, a sensible radius is a fine start, and far better than targeting a whole city. The waste starts when a business that has a real polygon keeps advertising to a circle anyway.
Not as a polygon. Meta targets circles around points, Google works with radii and named localities. Fitting smaller circles inside the zone is the practical translation, and it puts far less spend outside the area than one big circle.
It depends on how uncircular the zone is. The less the area looks like a circle, the worse the circle performs. Compare the zone's area with the circle that contains it and the share is right there.
Operations. Delivery businesses maintain exact zones in the ordering system because they decide who can order. Service networks have contracted territories. That file is the best targeting data the network owns.
Yes, whenever there is any real boundary: a gym's catchment, a home services territory, a childcare enrolment area. Budget outside the area a customer would actually come from converts at a far lower rate in every category. See how catchments differ by sector.
Local. National campaigns are supposed to cover everyone. It is the local layer, the store's own budget in its own area, where a circle quietly spends outside the fence. See local vs national.
The step by step version: from the operational polygon to what Meta and Google actually accept.
How Amplaro fits every store's campaigns to its own zone automatically.
Why targeting is usually the biggest lever on the return, and how to measure it honestly.
The fastest way to see the difference is your own store on screen: the zone, the circle a radius would buy, and the share of it your store cannot serve.