Geofencing is sold as if it were a technology you buy, when in practice it is a targeting option inside programmatic advertising that almost anyone can resell. An agency draws a boundary around a place, a competitor's car park, a hospital campus, a trade show, and serves display or video advertising to devices seen inside it, sometimes for months afterwards under the name of geoconquesting or addressable targeting. The mechanics are real and for some businesses they work. What varies enormously is the quality of the location data underneath, the margin taken between what you pay and what the media costs, and whether anyone is measuring an outcome rather than an impression. This page is how to tell the difference before signing.
What is actually happening under the label
In most cases the agency is a reseller. Media is bought through a demand side platform, the location targeting comes from a third-party data provider, and the agency's contribution is setup, creative and reporting. That is a legitimate service, and small advertisers often cannot access those platforms directly, but it means the questions that matter are about the layers you cannot see. Ask which platform the media runs through and which location data provider supplies the audience. Ask whether the boundary is drawn in real time as devices enter, or built from historical observations, because the two behave very differently for a short event. Ask what the minimum viable boundary size is, since a very small polygon around a single building often yields too few devices to spend a budget against, and the difference gets absorbed by widening the target quietly.
The margin question, asked plainly
The single most useful question is what proportion of your budget reaches the media. In a resold arrangement the agency buys inventory at one price and bills you at another, and unless the invoice separates media cost from management fee, that spread is invisible. There is nothing improper about a markup that is disclosed and agreed. There is a great deal wrong with one that is not, because it means you cannot compare two quotes at all, and it means the agency's incentive is to buy cheaper inventory rather than better inventory. Ask for an invoice format that separates media, data fees and management, and ask whether the agency receives any rebate or incentive from the platform. Get the answer before the contract, in writing, and treat a refusal as your answer.
Privacy, disclosure and where the rules bite
Location advertising is one of the more scrutinised corners of digital media, and precise location data around sensitive places such as health facilities, places of worship and schools carries particular exposure. The general advertising rules still apply throughout: claims made in the creative need a reasonable basis, and any disclosure necessary to prevent an advertisement being misleading has to be clear and conspicuous in the ad itself, which the FTC's guidance on digital disclosures spells out in terms of proximity, prominence and whether the audience will actually notice it. Small display formats make that harder, not optional. Ask a candidate agency what boundaries they refuse to draw and how they handle consumer opt-out signals. A provider with no answer has not thought about the question and is exporting the risk to you.
Measuring it without deceiving yourself
Impressions and click rates on geofenced display tell you almost nothing, because display click rates are dominated by accidental taps and the audiences here are small. The measures worth agreeing before launch are walk-in or store visit measurement where the platform genuinely supports it, a holdout region that receives no advertising so you have something to compare against, and a lift test rather than a before and after. Insist on the holdout, because it is the only part of this that resists wishful attribution. If the agency also sells you search, keep the budgets and the reporting separate, since geofencing is often used to justify a broader retainer where the actual returns come from geofencing advertising campaigns that are far narrower than the invoice implies. Review at ninety days against the lift test you agreed, and be willing to conclude it did not work.
Questions people ask about geofencing agency
How accurate is the targeting really?
It varies with the data source and the environment, and honest providers will say so. Dense urban areas, multi-storey buildings and indoor locations degrade accuracy considerably, so a polygon around one floor of an office tower is optimistic. Ask the provider what accuracy they assume, whether the data is observed or modelled, and how they handle devices seen at the edge of a boundary. Vague answers usually mean modelled data.
Is geofencing worth it for a small local business?
Sometimes, and rarely as a first purchase. A single location with a modest budget usually gets more from the free map listing, reviews and search than from display advertising to a small device pool. Geofencing earns its place for events, competitor conquesting at scale, or a business whose customers are genuinely defined by where they physically go, such as a venue or a clinic drawing from a specific campus.
What is a reasonable minimum spend?
Ask the agency, in writing, and ask what happens if the boundary does not deliver enough devices to spend it. That second question is more revealing than the first. Budgets that cannot be spent inside the intended polygon get quietly redistributed to broader targeting, which is no longer the product you bought, so agree in advance what the agency does when supply runs short.
Can we run this ourselves?
Larger advertisers often can, by buying platform access directly and paying a smaller managed fee or hiring the skill in house. Below a certain spend the platform minimums make that impractical and a reseller is the only route. The decision hinges on the disclosed margin: once you know what proportion of the budget is fee, you can price the alternative honestly rather than guessing.