What is geofencing?

How it works

A boundary is defined on a map — a radius around a store, a venue, a competitor's location. When a device crosses it, an action fires: an advertisement becomes eligible, a notification is sent, or the visit is recorded.

Location comes from GPS, wifi, cell towers or IP address depending on what the device permits, and the accuracy varies enormously between them.

The accuracy problem

Geofencing is sold with a precision it rarely has. GPS is accurate outdoors and degrades badly indoors, in cities and among tall buildings. Wifi and cell positioning are coarser again, and IP-based location is unreliable below city level.

The practical consequence is that small fences produce large errors. A fence around a single shop will capture people across the street and miss people inside, and a campaign built on the assumption that it identified store visitors is measuring something considerably fuzzier.

Larger fences are more reliable and less differentiated, which is the trade-off nobody selling the capability leads with.

The privacy exposure

• Location is sensitive data in most jurisdictions, with consent requirements that go beyond general advertising permissions.
• Inference risk. Places visited reveal health, religion, political activity and relationships — categories with heightened protection almost everywhere.
• Sensitive locations — clinics, places of worship, shelters — should be excluded outright, and several jurisdictions now require it.
• Platform permissions are separate from legal consent, and having the first does not confer the second.