Geofencing Advertising: How It Actually Works

A homeowner walks into a competitor’s showroom on Saturday morning. By Saturday night, your ad is on their phone while they scroll the news. They never searched for you. They never saw a billboard. They stood in a place you drew a line around, and that was enough. That is geofencing advertising, and it is the most precise targeting a local or regional business can buy.

I am the CEO of Killerspots, and our team builds geofence campaigns for dealers, remodelers, venues, service businesses and multi-location brands across the country, along with the display and video creative that runs inside them. This guide explains what geofencing is in plain terms, how a device ends up in your audience, where the fences should go, how a store visit gets tied back to the ad, and the handful of situations where I will tell a client to keep the money in their pocket.

What is geofencing advertising?

Think of it as a radius with a memory. Every smartphone that enters the boundary while the campaign is live gets added to an audience, and the ads follow that audience around in the apps people already use: weather, news, games, sports scores. The person does not have to open your website or search for your category. They had to be somewhere.

That somewhere is the whole product. Search advertising waits for a person to tell Google what they want. Social advertising guesses from what they like. Geofencing starts from what they did with their feet, which for a local business is the strongest intent signal there is. Someone standing in a competitor’s parking lot on a Saturday is shopping. Someone walking the floor of a home show is planning a project. Someone who has already been inside your store is worth more than any cold prospect in the ZIP code.

Our geofencing advertising campaigns run alongside programmatic display and streaming TV in our digital ads work, and the honest way to describe the product is this: it is a display campaign with a far better audience. The ad formats are familiar. The targeting is not.

How does geofencing advertising actually work?

Here is the mechanical version. Mobile apps that ask for location, which includes most weather, navigation, retail and sports apps, collect device location as part of running. That data is tied to an anonymous device identifier and made available to advertising platforms through data partners, with the user’s permission captured inside the app. Nothing in that stream is a name or a phone number. It is a device, a set of coordinates and a timestamp.

We draw the fences on a map, and not only as circles. A fence can be a precise polygon traced around a building footprint, a parking lot, a trade show hall or a subdivision. When a device’s reported location falls inside the polygon during the campaign window, it joins the audience. From then on, every time that device loads an app or mobile site that sells ad space through the exchanges, our campaign has the chance to bid on that impression and serve the creative. The same thing happens on the spot, while the person is still standing in the fence, which is where competitor conquesting earns its keep.

Two things people get wrong about the mechanics. First, the fence is not where the ads run. The fence is where the audience is captured. The ads run wherever those people go afterward. Second, the retargeting has a clock. Our standard window is 30 days, and we extend it to 90 for long sales cycles like a kitchen remodel or a vehicle purchase. After the window, the device rolls out of the audience, which is exactly how it should work. Nobody wants to see a flooring ad in March for a store they walked past in December.

Privacy changes on phones shrank the pool of devices that share location, and anyone who tells you otherwise is selling. The campaigns still work, on the devices that opted in, which is a smaller and cleaner set than it was a few years ago, and the reporting is honest about it.

Where should you draw the fences?

The first fence most clients want is around a competitor. That instinct is right. A device in a rival’s lot or showroom is in the market, and a well-timed ad with a reason to look at you is the most direct conquest play in local advertising. We fence the building and the parking, not the block, because the device in the sandwich shop next door is not shopping for a hot tub.

The second fence is the event. Home and garden shows, bridal expos, county fairs, trade shows, sporting events, concerts at the venue your customers favor. An event fence runs for the dates of the event and captures a crowd with a shared interest, which you then reach for weeks afterward while the enthusiasm is still warm. A kitchen remodeler who fences the spring home show gets an audience that just spent an afternoon looking at kitchens.

The third is the neighborhood. ZIP codes, a custom polygon around a cluster of subdivisions built in the same decade, a radius around your best customers’ addresses. This is where geofencing starts to overlap with programmatic display, and the question we ask is whether the place itself is the signal or just a container. A subdivision of 25-year-old homes is a signal for a roofer. A whole suburb is a container.

The fourth fence is your own location, and it is the one people forget. Fencing your store or office does two jobs. It builds a retargeting audience of people who already walked in, which is the warmest audience you own. And it becomes the conversion zone that makes attribution possible, which I will get to.

A word on what we will not fence. Hospitals, clinics, schools, places of worship, and anywhere that would reveal something sensitive about the person who walked in. Some states now prohibit fencing around healthcare facilities outright, and we treat the whole category as off limits in every state. A campaign that needs to know where someone got medical care in order to work is not a campaign we run.

What do geofencing ads look like, and where do they show up?

The format is nothing exotic: banners, native ads inside a news feed, and fifteen or thirty second video in apps that run video breaks. It is the same inventory any display campaign buys, which is why the creative, not the technology, decides whether the campaign works.

Here is the test I use on every geofence creative before it ships. Would the person understand why this ad found them? Someone who walked the home show on Saturday and sees “Saw you at the show? Here is what a real quote looks like” on Monday gets it. Someone who gets a generic logo banner with a founding year on it does not, and the targeting was wasted on a message that could have run anywhere. The offer should match the fence. Competitor conquest gets a comparison or a reason to get a second opinion. Event fences get a follow-up that references the event. Your own store’s fence gets a reason to come back.

We produce the creative in-house, the banners and the video, and we build it per fence, not one set for the whole campaign. That sounds like more work. It is. It is also the difference between a campaign that gets visits and one that gets impressions.

If your spot already runs on streaming TV, the cutdown belongs in the geofence video too. A household that saw the commercial on the big screen and then sees the same spot on a phone after visiting a competitor gets the message twice from two directions, and that is how streaming TV ads and geofencing compound each other. We covered how the TV side is bought in our guide to CTV advertising.

How do you know geofencing worked?

This is the part that separates geofencing from almost every other local ad buy. A billboard cannot tell you who drove past it and then came in. A radio schedule reports an estimated audience. A geofence campaign with a conversion zone reports the count of devices that saw the ad and then physically showed up at your door. That is foot-traffic attribution, and it is the number we lead with in every report.

We draw the conversion zone around your location, and the platform matches devices that were served an impression against devices that later entered the zone within the attribution window. The report shows impressions, clicks, visits and the visit rate per fence, so you can see that the competitor fence drove more walk-ins than the event fence and shift the budget mid-flight.

Three honest notes on the numbers. The visit count is a floor, not a ceiling, because it only counts devices sharing location. It does not count the spouse who came in without a phone or the person who called instead. Clicks are a weak signal on mobile display, because thumbs are clumsy, and a geofence campaign with a low click rate and a strong visit rate is a success. And attribution needs a flight long enough for people to act. Two weeks is not a test. Six to eight weeks with the fences live and the window running is the shortest campaign we will stand behind.

For a business that sells over the phone rather than through a door, we swap the conversion zone for call tracking and a landing page, and we judge the campaign the way we judge our PPC management services: what did it cost to get a conversation, and did the conversation turn into a job.

What mistakes waste a geofencing budget?

I have audited plenty of geofence campaigns that “did not work,” and the cause is almost never the channel. It is one of six setup mistakes.

The big fence. A fence around an entire ZIP code is not geofencing, it is a display buy with a location filter. The audience has no shared intent, the creative cannot reference anything specific, and the budget spreads across people who never did anything that suggested they were buying. Tight fences around real targets win, even when the audience is smaller.

Generic creative. Covered above, and it is the most common failure, so it gets listed twice. The targeting is precise. The message has to be, too.

The short flight. A 30-day retargeting window inside a 14-day campaign means the window never finishes. The people captured on day ten get four days of ads and then silence. Run the fences for the full flight and let the window run past it.

No conversion zone. If you did not draw a zone around your own location before launch, you cannot measure visits after it. Then the only numbers are impressions and clicks, and you are back to guessing.

Nothing behind it. Geofencing creates awareness and pulls people toward your name. When they then search for you, a search campaign has to be there to catch them, and your website has to convert them. Our post on media buying mistakes goes through what happens when channels are bought as if they were separate.

Wrong expectation. It is not a lead form on day one. The device enters a fence, sees ads for weeks, and then visits, calls or searches. A dealer who fences three competitors in April should expect to see the lift across May and June, and a campaign judged in week two will always be judged a failure.

When is geofencing the wrong buy?

Geofencing has a clear profile of where it wins, and it is not every business.

It is the wrong buy for a pure emergency service. Nobody with a burst pipe at midnight visits three plumbers’ offices first, so there is no fence to draw, and the money belongs in Local Service Ads and search. It is the wrong buy when the budget cannot sustain six weeks, because the window needs time. It is the wrong buy when you cannot or will not put an offer in the creative. And it is the wrong buy in any category where the place a person visited is itself private: health, legal trouble, finances, faith.

It is the right buy for dealers, remodelers, furniture and flooring stores, restaurants and bars, fitness, auto service, universities and private schools, venues, and any multi-location brand that wants to defend its own stores and conquest the ones across the street. If your customers go somewhere to shop, compare or gather before they choose, there is a fence worth drawing.

The businesses that get the most out of it run it as the front of a plan: geofencing to capture intent, display and streaming to build the name, search to catch the demand, and a CRM to answer the phone. Whether Google Ads or Meta ads belongs beside it depends on the business, and that is a planning call, not a product pitch. Companies in our home market can see how we build the full paid plan on our Cincinnati PPC page; the method is the same in every market we serve.

Where to start

Make three lists: the competitors you would most like to win customers from, the events your best customers attend in the next six months, and the address of every location you own. That is a fence map. From there we size the audience, build the creative per fence, set the conversion zones, and show you reach and cost in writing before anything runs.

If you want to see what a geofencing campaign would look like around your competitors, send us the addresses and we will come back with a fence map and a plan. And if the honest answer is that a different channel fits better, we will say that instead.