Nobody in 1981 thought we’d run out of internet addresses. The engineers who designed IPv4 gave us 4.3 billion unique IPs, which seemed absurd for a world that barely had personal computers. Fast forward to now, and that number looks almost quaint.
Every smart doorbell, cloud server, and corporate laptop needs its own address. IPv6 was supposed to fix this years ago, but IPv4 stubbornly remains the default for most commercial networking.
IPv6 Keeps Waiting in the Wings
IPv6 has been “just around the corner” since 1998. Google’s transparency reports put global adoption somewhere near 45%, but that number is misleading. Mobile carriers in a handful of countries (India, especially) skew it dramatically. Pull those out, and corporate adoption looks much thinner.
The holdup isn’t ignorance. Migrating a large enterprise network to IPv6 costs serious money and introduces compatibility headaches that nobody wants to own. Dual-stack setups, where both protocols run side by side, just add complexity most IT teams would rather avoid.
There’s also the application layer to worry about. Thousands of firewalls, monitoring dashboards, and homegrown scripts have IPv4 logic hardcoded into them. Rewriting all of that is a multi-year project with almost no visible upside to the people approving budgets.
Scarcity Created a Real Market
A single IPv4 address now trades between $30 and $60, depending on block size and who’s buying. For companies building proxy infrastructure or spinning up cloud deployments, those per-IP costs get painful quickly. Many organizations that need to buy IPv4 proxies at IPRoyal find it cheaper to work with a dedicated provider than to acquire and manage raw address blocks on their own.
How did we get here? ARIN, which manages IP allocations for North America, ran out of free IPv4 addresses in 2015. Europe’s registry, RIPE NCC, followed in 2019. A secondary market popped up where IPv4 blocks trade hands like commercial property, and prices haven’t come back down.
Smaller hosting companies get hit hardest. A startup trying to launch a CDN can’t just grab a /16 block (that’s 65,536 addresses) the way the big cloud players did ten years ago. The math simply doesn’t work anymore.
The Sectors Still Running on IPv4
Proxy networks burn through IPv4 space faster than almost anyone. Datacenter proxies, residential gateways, ISP proxy services: they all need massive pools of routable addresses. Clients expect coverage across dozens of countries, and every single connection requires a valid IP.
Web scraping is a good example. A mid-size e-commerce intelligence company might rotate through 50,000 IPv4 addresses in a single day just to pull pricing data without getting blocked. Research published through IEEE estimates that automated traffic now makes up roughly 40% of all internet activity. Almost all of it runs over IPv4.
Gaming holds on tight to IPv4 too. Multiplayer servers, anti-cheat tools, matchmaking systems: they’re built around stable IPv4 connections. No major publisher wants to be the one who switches to IPv6 right before a big title launch and watches the bug reports roll in.
The Incentive Problem
Here’s what rarely gets discussed honestly. IPv6 migration stopped being a technical challenge a while ago. It’s now an incentive problem, plain and simple. Companies sitting on large IPv4 blocks have zero reason to change anything. Their addresses work everywhere, with everything, today.
Gartner’s own documentation reflects this reality: most organizations treat IPv6 planning as a “someday” line item, not an active initiative. Budget committees would rather fund things that show returns this quarter.
Carrier-grade NAT (CGNAT) has made the situation even stickier. ISPs can now share a single public IPv4 address across hundreds of customers, which effectively kicks the can down the road. It’s a workaround, not a solution, but it buys everyone more time than anyone expected.
Where Things Stand
IPv4 isn’t going anywhere soon. Secondary market prices will stay elevated as cloud providers, proxy operators, and the ever-growing IoT sector keep competing for limited supply. Any business planning new infrastructure should treat IPv4 costs like they treat domain registration: a fixed line item, not an afterthought.
The realistic play for most organizations is dual-stack: adopt IPv6 where it’s easy, keep IPv4 where it’s necessary. Based on how slowly things have moved so far, that “necessary” column won’t be shrinking for quite a while. This protocol has outlived every prediction of its death, and the obituary writers keep getting it wrong.