What Is an IP Address? – Definition and How It Works
Learn what an IP address identifies, how IPv4 and IPv6 differ, what NAT changes, and why an address rarely identifies one person.
An IP address identifies a network interface so packets can be routed to it. It identifies a connection point, not a person, and the difference matters constantly in practice.
1. The Two Versions
| IPv4 | IPv6 | |
|---|---|---|
| Length | 32 bits | 128 bits |
| Notation | 203.0.113.5 |
2001:db8::1 |
| Address space | About 4.3 billion | Effectively unlimited |
| Typical allocation | Scarce, traded | Abundant, whole ranges per subscriber |
| NAT | Ubiquitous | Rarely necessary |
IPv4 exhaustion is the reason so many users share addresses, which shapes everything about how addresses behave as identifiers.
2. Public and Private
Certain ranges are reserved for private networks and are not routable on the internet. A device on a home or office network has a private address, and traffic leaving that network is translated to a public one.
That translation is why a whole household, office, or campus can appear as a single address to every site they visit.
3. Carrier-Grade NAT
Mobile networks and some providers translate at a larger scale, placing thousands of subscribers behind one public address.
The consequences are direct: a site blocking one address may block thousands of unrelated users, and a rate limit keyed on address is shared among all of them.
4. What an Address Reveals
- Approximate location, reliably at country level and weakly below it.
- The owning organisation, from public registries.
- The network class — hosting provider, consumer ISP, mobile carrier.
- Reputation, accumulated from behaviour attributed to it and its neighbours.
What it does not reveal is who is using it, which is why treating an address as an identity is a category error.
5. Static and Dynamic
Server addresses are usually static. Consumer connections are often dynamic and change periodically, so an address observed today may belong to someone else next week.
Reputation systems account for this by decaying scores over time, and it is why an address's history is a fading signal rather than a permanent one.
6. Why It Matters for Collection
- Rate limits and blocks are frequently keyed on address.
- Shared addresses mean inheriting other people's reputation.
- Geography changes content, so egress location is part of an observation's context.
- Hosting ranges attract more scrutiny than consumer ones.
Recording the egress address and region alongside collected data is what makes observations comparable later.
7. Addresses Are Not Anonymity Boundaries
Changing address changes one signal. Cookies, account state, protocol fingerprints, and behaviour all persist across a change, which is why address rotation alone rarely resolves friction that was never about the address.
8. Summary
An IP address identifies a network interface, in a scarce 32-bit space or an abundant 128-bit one, frequently shared through translation.
It reveals rough location, owning organisation, and network class, and it is shared often enough that treating it as an identity is unreliable — which is exactly why address-keyed blocks and limits catch so many uninvolved people.