Geofencing Radius Best Practices for Mobile and OOH Campaigns

There is no single correct geofence radius. The right size depends on how accurately phones report location in that environment, how people move through it, and what you want to count as a “visit” or an “exposure.” A practical starting point for walk-in retail is the area within a four- to five-minute walk of the location. For app-based geofences, treat roughly 100 to 150 meters (about 330 to 500 feet) as the floor. For ad platforms, expect a minimum of about 1 km (0.6 miles). From there, you tune.

This guide explains where those numbers come from, why a circle often fails along a road, and how we set boundaries when we pair geofencing with LED mobile billboards and wrapped rideshare vehicles.

TL;DR

• Start from the use case: a walkable trade area for stores, a tight boundary around the venue for events, a corridor for roads.
• Do not go below about 100 to 150 meters for app-based geofences. Android’s guidance ties this floor to location accuracy and battery limits.
• Many ad platforms enforce a minimum radius. Google Ads requires at least 1 km for proximity targeting, so very small fences need a different buying path.
• For anything on a road, use a directional or polygon boundary. A circle counts passersby who never see the screen.
• Document your method (radius or polygon, size, validation rules) so results can be audited.

Table of Contents

• What a geofence radius actually does
• Why location accuracy sets your minimum
• Timing and latency also change what “inside” means
• Radius ranges by use case
• When a circle fails: roads, screens and direction
• Platform minimums and how to work around them
• A step-by-step method for setting your radius
• Privacy and disclosure
• Our perspective
• Measuring whether your radius worked
• Putting it into a campaign
• Common mistakes to avoid
• FAQ
• Sources

What a geofence radius actually does

A point-radius geofence is a center point plus a circle. The Media Rating Council’s location-based advertising guidelines describe geofencing as assigning a device’s reported location to a specific place, using either a point-radius or a boundary method. The radius is not just a targeting setting. It defines who gets counted, which feeds every number you report afterward.

That is why the choice is a measurement decision as much as a media one. Too small, and real visitors fall outside the fence because their phones reported a position 40 meters off. Too large, and you count people in the next building, on the highway, or in the parking structure across the street.

Why location accuracy sets your minimum

The radius has to be bigger than the typical error of the signal you rely on. A study of mobile geolocation data reported median accuracy of roughly 11 to 12 meters for GPS, 20 to 21 meters for carrier network location, and 30 to 36 meters for Wi-Fi. When a position relied on cell tower databases, the median error ran to roughly 930 to 967 meters.

Those are medians, so half of readings are worse. Indoors, accuracy falls further, because the study notes that reception depends on walls, wall material and closeness to them.

Android’s developer documentation recommends a minimum radius of 100 to 150 meters for reliable geofences. It also notes that with Wi-Fi enabled, even without a connection, accuracy can reach about 20 to 50 meters, but it can drop to several hundred meters or more in rural areas or on cellular alone.

Two takeaways:

1. A 30-meter fence around a storefront will miss visitors and produce noisy data. Do not design for precision your signal cannot deliver.
2. Radius should vary by environment. A dense downtown with strong Wi-Fi tolerates a tighter fence than a rural highway exit.

Timing and latency also change what “inside” means

Even a well-sized fence does not trigger instantly. Research reviewed in a methodology paper on geofencing notes that iOS may require a device to travel a minimum distance and stay on one side of the boundary for at least 20 seconds before an event fires. Android apps may check for entries only every 1 to 6 minutes, depending on device activity.

For a campaign promising “real-time” messages, this matters. If a person walks past a venue in two minutes, a coarse check interval can miss them entirely. The practical response is to size the fence a little larger than the area you care about and to plan creative that still works a few minutes after entry. It is also a reason to avoid promising exact-second triggers to a client.

Radius ranges by use case

The table below pulls together the sourced figures above with the common rules of thumb. Treat it as a starting range for testing, not a standard.

A caution on the first row: the four-to-five-minute walk guideline appears in several marketing articles but I would treat it as a heuristic. Test it against your own conversion data.

When a circle fails: roads, screens and direction

This is where out-of-home differs from ordinary mobile marketing. A person 80 meters away on a parallel street is inside your circle but cannot see a billboard truck. A driver heading away from the screen is also inside the circle.

The OAAA’s DOOH exposure methodology standard addresses this. It says that for some inventory, screen facing is a necessary input to validate the direction of travel. Best practice is to validate travel direction, screen orientation and dwell time, rather than counting every nearby device as an exposed viewer.

In practice, that means:

• Use polygons or corridors that follow the road and the sight line, not a circle that cuts across blocks.
• Count a device only if it is moving toward the display, or is within viewing distance while the ad is live.
• Match ad length to how long people actually stay in view. We cover that in our guide to matching ad length to venue dwell time.

If you are unsure whether you need a geofence or simple location targeting at all, our comparison of geofencing vs geotargeting walks through the decision.

Platform minimums and how to work around them

Many buyers draw a 200-meter fence in a plan, then find the platform will not run it. Google Ads states a minimum proximity radius of 1 km around a location, and says location targeting is a best-effort match based on a variety of signals. A fence that small simply is not offered for standard bidding.

Options when your ideal boundary is smaller than the platform allows:

1. Accept the larger radius and filter later. Use the wider fence for reach, then rely on attribution to separate true visitors.
2. Use a custom boundary tool. Some data and mobile ad providers support polygons built from venue footprints.
3. Move the precision into the physical media. A truck parked or routed at the exact spot does the narrowing for you; the digital layer follows.

Choosing audience layers on top of location also helps. For background on how to combine them, see this overview of audience targeting fundamentals.

A step-by-step method for setting your radius

1. Define the behavior you want to capture. Walk-in visit, event attendance, or road exposure each call for a different shape.
2. Map the real environment. Look at building footprints, entrances, parking, adjacent businesses and major roads.
3. Pick the shape. Circle for open, isolated sites. Polygon for dense blocks, venues or roads.
4. Set the size above the signal error. Start at or above 100 to 150 meters for app-triggered fences, and widen in rural or indoor-heavy areas.
5. Add validation rules. Dwell time minimum, direction of travel and exclusion zones (neighboring stores, the highway).
6. Test two sizes. Run a tight and a wide version and compare visit rates and cost per visit.
7. Write down the method. Keep the shape, size and rules in your campaign report.

Privacy and disclosure

Radius settings affect privacy risk. The MRC’s out-of-home measurement standards point to correlating audience metrics at the home block group level and avoiding high-risk identification methods, such as personally identifiable collection, unless disclosed and supported by consented opt-ins. The location-based advertising guidelines also expect marketers to disclose the geofencing technique and the proximity ranges they use.

Sensitive locations such as clinics, places of worship and schools deserve extra care, and your legal counsel should review any fence around them. Our guide to U.S. geofencing legal gates covers the main issues for marketers.

Our perspective

We run LED mobile billboards and wrapped rideshare vehicles, so our geofences have to describe something physical: a street, a venue entrance, a route. That changes how we think about radius.

First, we treat the fence as a measurement definition. A bigger circle makes a campaign look busier and a smaller one makes it look sharper, but neither says whether anyone saw the screen. That is why we pair geofencing with route customization and proof-of-posting, with GPS and photo tracking, so you can see where the media actually ran.

Second, we prefer shapes that follow movement. For a conference, we think about the approach roads, the hotel blocks and the convention center entrance, not a circle centered on the building. Our piece on targeting conferences and trade shows with LED mobile billboards shows how that planning works.

Third, we would rather you start with a modest, documented fence and test it than buy a huge radius for the sake of reach. Smart QR codes add a second signal: a scan is a deliberate action, so it confirms interest independent of a phone’s location accuracy.

— Scott

Measuring whether your radius worked

To judge a radius, compare outcomes, not impressions. Useful measures include:

• Visit rate among devices exposed inside the fence versus a matched control group outside it.
• Cost per incremental visit at each radius size.
• QR scan volume by route segment or zone.
• Share of devices that pass your direction and dwell filters.

If the wide fence delivers many more “visits” but the same number of scans and conversions, the extra area is mostly noise. Our explainer on foot traffic attribution describes what that kind of measurement can and cannot show, and the geofencing campaign ideas page has examples of pairing fences with mobile OOH to test visit lift.

Putting it into a campaign

If you are planning a localized campaign and want the fence, the route and the reporting to line up, we build them together: LED mobile billboards, wrapped Uber and Lyft vehicles, geofencing, real-time retargeting, audience filtering, smart QR codes, and attribution reporting across all 50 states. Talk to Beacon about a mobile OOH campaign and bring your venue or market, and we will help you define the boundary and the proof you want at the end.

Common mistakes to avoid

• Going too small. A 25-meter fence relies on accuracy most phones cannot provide.
• One radius for every location. Urban and rural sites behave differently.
• Counting nearby as exposed. Without direction and dwell checks, you are measuring proximity, not viewing.
• Overlapping fences. Neighboring venues can double count the same device.
• Ignoring latency. A device may enter and leave between checks.
• Skipping disclosure. Report the method and range you used.

FAQ

What is a good geofence radius for a retail store?

A common starting point is the area within a four- to five-minute walk of the store. In practice, that is often a few hundred meters in a city and larger in car-oriented suburbs. Test a tighter and a wider version and compare cost per visit.

What is the smallest geofence radius that works?

For app-based geofences, Android’s guidance suggests a minimum of 100 to 150 meters for reliable behavior. Ad platforms can be stricter: Google Ads, for example, requires at least 1 km for proximity targeting. Smaller fences produce missed entries and false triggers.

Should I use a radius or a polygon?

Use a radius for isolated sites with open surroundings. Use a polygon when neighbors, roads or a venue footprint matter. For out-of-home on roads, a directional or polygon boundary is better because it follows the viewing path.

Does geofencing work in real time?

Not exactly. iOS may wait for sustained movement of at least 20 seconds across a boundary, and Android may check for entry every 1 to 6 minutes. Plan creative and expectations around that delay.

Sources

• Create and monitor geofences, Android Developers
• Missing Data and Other Measurement Quality Issues in Mobile Geolocation Sensor Data
• Target ads to geographic locations, Google Ads Help
• MRC Location-Based Advertising Measurement Guidelines
• MRC Out-of-Home Measurement Standards
• DOOH Exposure Methodology Standardization Guidelines and Best Practices, OAAA
• Geofencing in location-based behavioral research: Methodology, challenges, and implementation

Recommended

• Geofencing Campaign Ideas Paired With Mobile OOH to Prove Visit Lift
• Geofencing vs Geotargeting: What Marketers Should Use When
• Geolocation strategies that maximize out-of-home ad impact
• U.S. Marketers: 4 Geofencing Legal Gates Aligned With FTC Orders