What is an IP address in advertising?

What it is used for

• Geographic targeting, generally accurate to country and city, unreliable below that.
• Fraud detection. Many installs from one address, or addresses belonging to data centres rather than consumer networks, are strong signals — see mobile ad fraud.
• Household inference, since devices on one home network share an address.
• Probabilistic matching, which is increasingly restricted by both policy and regulation.
• Rate limiting and abuse prevention, which is the least contentious use.

Why it is less useful than it looks

IP-based geolocation is accurate at country level and steadily less so below it. Mobile carrier addresses can place a user in a different city from where they are; corporate VPNs place them wherever the gateway is. Precision claims below city level should be treated sceptically.

Addresses are also unstable — most consumer connections rotate them — and shared, since carrier-grade NAT puts many subscribers behind a single address. Both properties break naive assumptions about one address meaning one household.

Its legal status

In the EU and UK, an IP address is generally treated as personal data, because it can identify an individual when combined with other information. Processing it therefore requires a lawful basis like any other personal data.

This surprises people because the address feels technical rather than personal. It is why 'we only use IP' is not an exemption from data protection obligations, and why fraud-prevention uses are typically justified under legitimate interests with a documented assessment rather than treated as out of scope.