Why renewals matter more than launch curves

The domain name industry has always had a weakness for big numbers, big zone files, big launch days, big growth charts. Fair enough. We all like a chart that goes up and to the right. Some of us have even built PowerPoint decks around them. But the longer you work in this business, the more obvious one thing becomes: registrations are only the beginning of the story. They tell you that someone tried a domain. They do not tell you whether that name became useful, trusted, renewed or built into someone’s digital identity.

The web itself gives us a useful reminder. In its May 2026 Web Server Survey, Netcraft reported responses from 1.5 billion sites across 305 million domains and 216 million active sites (Netcraft).  

Those are huge numbers, but they also show why counting things on the web is never as simple as it looks. A domain, a hostname, a website and an active web presence are related, but they are not the same thing.

The same is true in our market. A domain can sit in the domain name base without necessarily representing a developed website, a functioning service or a long-term user. That is not a flaw in the data but simply a reminder that registration numbers are a starting point and not a full measure of value.

And as we head towards the next ICANN round reveal, that distinction matters.

Launches are exciting but patience is the business model

The first year of a TLD can tell you many things: whether the marketing worked, whether registrars were engaged, whether pricing was aggressive, whether speculators showed up, whether a string had curiosity value. But it cannot say if it is going to be used. That takes time. Domain names are a slow business. They compound through habit, trust, use and distribution. A name that is printed on a shop window, embedded in email, used in campaigns, indexed by search engines and remembered by customers is not easily replaced. But that kind of attachment does not happen in a launch week. It happens over years and years.

This is why renewal rates deserve more attention than they often get. New registrations create movement, renewals create durability. And for registries, registrars and back-end operators, renewals are not a soft metric. They are the recurring revenue base.

The problem with portfolio size as a scoreboard

For years, parts of the industry have treated portfolio size as the scoreboard. Bigger must mean better. More domains under management must mean more market success. More launch-year registrations must mean stronger demand. Sometimes that is true. Often it is not.

A large portfolio built on very low first-year pricing can look impressive until the renewal cycle arrives. Names that came in for a promotion may leave just as quickly when standard pricing applies. That does not make the strategy illegitimate. Promotions have their place. They can seed awareness, test demand and bring a TLD into the registrar channel.

But if the business model depends on constantly replacing churned names with new discounted names, then the size of the zone can hide as much as it reveals. The better question is not only: how many names are in the zone? It is also: how many come back?

Where GeoTLDs become interesting

This is where GeoTLDs deserve a more careful look. Not because they are huge. They are not. Not because they can be compared directly with .com, .net or the largest ccTLDs. They cannot. Most GeoTLDs are around a decade old. .cat is 20 years old now (congratulations!), but even there we are not talking about the same historical arc as .com or .net.

That is exactly why the renewal data is interesting.

GeoTLDs are small, relatively young and often focused on a specific city, region, culture or language community. They were never going to win the global volume race. But they can tell us something about another kind of value: whether a smaller namespace can build enough relevance for registrants to keep names over time.

Using DNIB renewal data for a sample of GeoTLDs, the picture is mixed, but meaningful:

The total across these eight GeoTLDs is just over 409,000 names. In .com terms, that is not even a rounding error with a nice view. But again, volume is not the question here, especially for new TLDs.

Seven of the eight renewal rates in this sample sit at 80% or above. Several are above 90%. .tokyo is the exception at 50.8%, and it should stay in the table because the point is not to pretend GeoTLDs are magic. Different markets, pricing strategies, registrar mixes and registrant behaviours produce different outcomes.

Still, the pattern is worth noticing. A number of GeoTLDs appear to behave less like short-term launch products and more like small, stable namespaces. That means the names that are registered can be sticky.

A few individual TLDs make the point even more clearly: size and retention do not always tell the same story.

A zone can be large and still depend heavily on replacing names that do not come back. A smaller zone can be less spectacular on paper but healthier in the renewal cycle. Portfolio size gets the attention but renewal rate tells you how much of that attention is likely to become recurring revenue.

GeoTLD is not a promotion

One reason may be simple: geo domain is not a temporary keyword.

A city, region, culture or language is not a seasonal campaign. A business in New York does not stop being in New York after twelve months. A cultural organisation using .cat is not likely to decide that Catalan identity was a first-year experiment. A local initiative using .quebec or .amsterdam is not choosing only a string; it is choosing a signal.

That kind of signal is not universal. It will not work for every registrant, every registrar or every market. But where it works, it can create a kind of attachment that is different from curiosity-driven registration.

This is also why GeoTLDs should not be judged only by the standards of mass-market gTLDs. If the question is “Can this TLD become the next .com?”, the answer is no, and everyone can go home early. If the question is “Can this TLD become a useful, trusted namespace for a defined community?”, then renewal rates become a much more relevant measure.

What the next round should learn

The next round will bring new applicants, new business plans and, inevitably, new forecasts. Some will be serious. Some will be optimistic. Some will be written in the traditional registry dialect of hockey-stick growth.

We should ask harder questions this time.

Not only: how many names do you expect to sell in year one?

But also:

• Who are the registrants you expect to keep?
• What will make the name useful after the first year?
• How will you support real usage, not just initial registration?
• What role do registrars play after the launch campaign is over?
• What does success look like in year five, not just month five?

These questions matter for all TLDs, but especially for geographic, cultural and community-based strings. A GeoTLD is not a quick-win product. It is closer to local infrastructure. It needs time, patience, distribution, partnerships and a clear sense of who it is for.

That may sound less exciting than a launch-day spike. It is also much closer to how durable domain businesses are actually built.

The real metric is staying power

Registration numbers still matter. Of course they do. A TLD with no registrations is not a noble act of patience; it is a problem. But registration numbers alone are too blunt. They do not tell us enough about use, commitment or revenue quality. Netcraft’s web data reminds us that the internet is full of things that can be counted, but counting is not the same as understanding. DNIB’s renewal data reminds us that the domain business has its own retention story, and that some categories depend far more heavily on constant new inflows than others (DNIB Q1 2026).

For GeoTLDs, the fair argument is not that they are bigger than they are. They are small. They are young. They are uneven. They are sometimes hard to sell through a registrar channel built for scale. But in several cases, they show something the next round should take seriously: a namespace does not need to be massive to be durable.

In the domain business, the launch gets the applause. The renewal gets the invoice paid.

And after enough years, that is usually the part that matters.