The message has been on repeat for years: IPv4 is exhausted, IPv6 is the future, migrate now. And sure, directionally that’s correct. But the push toward an IPv6-only infrastructure for public-facing services tends to gloss over the practical, financial, and technical friction that hasn’t gone away in 2025. The IPv6-only hidden costs rarely make it into vendor whitepapers. They do show up, though—in analytics dashboards, support tickets, and lost revenue. Usually all three, and usually within the first week.
The Reachability Problem: Not Everyone Can Reach You
Global IPv6 adoption sits around 45% according to Google’s statistics, with strong numbers in India, France, and Germany—and significant gaps elsewhere. The United States hovers near 50%. Meanwhile, plenty of enterprise networks, government agencies, and entire regions across Africa, the Middle East, and parts of Asia remain predominantly IPv4.
Need IPv4 addresses?
Browse clean, RIPE-verified subnets at $0.50/IP/month.
So what happens when you publish a service as IPv6-only with no translation mechanism? Every IPv4-only client gets a connection failure. Simple as that. Not a hypothetical edge case—a measurable chunk of your addressable audience. For an e-commerce platform, even a 5% reachability gap maps directly onto abandoned carts.
DNS64/NAT64 Is Not a Silver Bullet
Some operators point to NAT64 gateways in mobile networks (T-Mobile, for instance, runs large IPv6-only mobile fleets) as proof that IPv6-only works. It does—but only in one direction. NAT64 helps IPv6-only clients reach IPv4 servers. It does nothing, nothing at all, for IPv4-only clients trying to reach your IPv6-only server. If the entire internet needs to reach you, you need an IPv4 presence in some form.
Infrastructure and Tooling Gaps That Add Hidden Costs
Reachability aside, running IPv6-only infrastructure surfaces operational costs that are easy to underestimate during planning. I’ve watched teams budget carefully for the migration and still get blindsided by these:
- Legacy vendor appliances and SaaS integrations. Firewalls, load balancers, monitoring agents, third-party APIs—many still assume IPv4 connectivity. You end up with workarounds or expensive upgrades.
- Email deliverability. Plenty of receiving mail servers still deprioritize or reject IPv6-sourced mail unless PTR, SPF, and DKIM alignment are properly configured. IPv4 senders face fewer default friction points.
- DDoS mitigation and WAF coverage. Some scrubbing services and CDN plans price or provision IPv4 and IPv6 protection separately, which pushes up total security spend.
- Staff training and debugging overhead. Troubleshooting dual addressing, prefix delegation, ND/RA issues takes skills many teams haven’t fully built yet. MTTR during incidents stretches out.
SEO, Analytics, and User Experience Impact
Search engines crawl over IPv4 far more reliably than over IPv6 in many hosting environments. If your IPv6-only origin introduces timeouts or crawl errors, those signals can hurt indexing. Worse, your analytics will undercount IPv4-only users who bounce before the page even loads—so the problem looks smaller in the dashboard than it actually is.
Session-level metrics suffer too. An IPv4-only user on a CGNAT network who hits a failed connection produces no server logs whatsoever. You lose visibility into an entire class of failure, which quietly degrades both capacity planning and marketing attribution.
Security and Compliance Blind Spots
Several compliance frameworks and enterprise procurement requirements still explicitly assume IPv4 logging and filtering capabilities. PCI DSS assessors, for example, expect complete network visibility—an IPv6-only edge can complicate established logging pipelines if your SIEM rules were built around IPv4 semantics. Rebuilding them takes time nobody budgeted for.
And then there’s the address space itself. IPv6’s sheer scale changes how scanning and inventory work. Teams have to re-learn network discovery, and misconfigured prefix delegation can accidentally expose entire /64 subnets that legacy scanning tools never bothered to check.
IPv6-Only vs. Dual-Stack at a Glance
| Factor | IPv6-Only | Dual-Stack (IPv4 + IPv6) |
|---|---|---|
| Global reachability | ~45–55% of clients | Near 100% |
| IPv4 address cost | $0 upfront | ~$30–$50 per address (one-time purchase) |
| Operational complexity | High (translation workarounds) | Moderate, well-understood |
| Legacy compatibility | Frequent breakage | Seamless |
| Future-readiness | Excellent | Excellent |
Practical Recommendations for Network Teams
- Default to dual-stack for public-facing services. Enable IPv6 fully, but keep IPv4 reachability via your own addresses or a translation layer you control.
- Measure before you cut. Analyze traffic by address family, per region and per customer segment. IPv6-only may be viable for internal networks or specific CDN-fronted workloads—just not for direct-to-internet services.
- If you need IPv4, buy rather than rent long-term. The IPv4 market has matured; a /22 or /24 purchased through a verified marketplace typically pays for itself versus multi-year leasing, and the block holds resale value.
- Ensure clean, reputable address space. Pre-purchase block history checks (blacklist screening, RIR transfer compliance) prevent deliverability and reputation headaches later.
- Keep the IPv6 migration momentum. Dual-stack is a bridge, not a retreat. Keep building IPv6 capability while protecting revenue with IPv4 reachability.
When acquiring IPv4 space, the marketplace you work with matters enormously. IP4 Market, for example, provides verified sellers, clean-block due diligence, and competitive pricing on both purchases and leases—removing much of the risk that historically made IPv4 acquisition intimidating for smaller organizations.
Frequently Asked Questions
Is IPv6-only ever a good idea for public services?
Yes—for CDNs, content platforms, and services where a translation layer or partner network guarantees IPv4 reachability on your behalf. Running your own IPv6-only origin behind a dual-stack frontend can be an excellent architecture.
How much do IPv4 addresses cost in 2025?
Market prices generally range from $30–$50 per single address depending on block size and region, with /22 and larger blocks commanding efficient per-address rates. Prices have trended upward as remaining supply tightens.
Can I buy a small block instead of a /24?
Minimum transferable blocks are typically /24 in most RIR regions, though leasing smaller allocations from holders is possible through marketplaces like IP4 Market.
What’s the fastest way to add IPv4 back to an IPv6-only service?
Lease a block for immediate needs while pursuing a purchase for long-term cost efficiency—leasing can be live within days, while transfers take longer due to RIR processing.
The IPv6 transition is real and worth investing in. But the IPv6-only hidden costs of abandoning IPv4 at the edge remain substantial for most organizations. A dual-stack strategy, backed by well-sourced and cleanly vetted IPv4 address space, gives you full internet reachability today—and still positions you for the IPv6-dominant future.
