The 2026 new gTLD application window has closed. Hundreds of applicants have submitted business plans describing the namespaces they intend to build. Those plans include target audiences, pricing strategies, registrar partnerships, launch sequences, and marketing approaches. Most describe what the namespace should look like once it reaches maturity.

No business plan can fully determine the composition that will emerge. That is not a criticism of the applicants or their planning. It is an observation about how namespaces appear to form in practice. Business plans describe intentions. Namespaces are shaped by the interaction of those intentions with registrant behavior, registrar dynamics, promotional structures, and speculative demand. The resulting composition is shaped by factors beyond the applicant’s original plan.

What Existing Namespaces Suggest

Over the past year, I have run the same structural analysis pipeline across four large, established nTLD namespaces. The scale ranged from roughly 600,000 to over 9 million unique domains. The operators, positioning strategies, and target audiences differed.

Across all four, the same broad structural categories appeared. A structural noise layer accounted for 15—28% of the zone. A large neutral middle—structurally clean names with no observable commercial signal—made up 55—70%. A speculative or investor-grade layer accounted for 5—12%. And domains classified as operator-grade on structural characteristics—pronounceable, commercially plausible names—sat in the mid-single digits.

An important distinction: the classification identifies structural characteristics. It does not establish that a domain is owned by an operating business, that it is being used commercially, or that its registrant intends to renew it. A structurally neutral domain could belong to an active small business. The categories describe observable zone-file patterns, not verified registrant behavior.

What the data does show is that the proportions varied substantially across namespaces. The operator-to-investor ratio ranged nearly fourfold. In this sample, larger namespace size did not consistently correspond to a higher operator-grade ratio.

The Gap Between Intended and Observed Composition

Suppose an applicant intends to attract small businesses to a commerce-focused TLD. Its registration targets may be clear, but the eventual namespace could contain several structurally different populations. A zone-file analysis would not establish how many registrants are operating storefronts. It could, however, provide an additional view of the registration mix and identify changes worth investigating alongside operational data.

In the namespaces I analyzed, domains classified as operator-grade consistently sat in the single digits as a percentage of total registrations. The largest population was always the structurally neutral middle—names that might belong to businesses, or might be defensive registrations, or might be placeholder inventory.

This does not mean the applicant’s plan is wrong. It means the plan describes one intended population, while the existing evidence suggests that applicants should allow for populations beyond their initial target audience. Speculative and structurally noisy registrations may form part of the resulting namespace. The namespace that takes shape will likely be more heterogeneous than the one described in the business plan.

What the Operator May Be Able to Influence

If these structural categories recur across namespaces, the question is whether an operator has any influence over the proportions.

There is early evidence that they might. Across the four namespaces I analyzed, the operator-to-investor ratio varied nearly fourfold—from namespaces where speculative inventory outnumbered operator-grade inventory by roughly three to one, to namespaces where the balance was closer to even.

What might explain that difference? Pricing dynamics, registrar relationships, and category specificity are obvious candidates, but the four-namespace sample does not establish whether any of them actually drive composition. The honest answer is that we do not yet know which operator decisions most influence composition outcomes.

Three Questions Worth Testing

For applicants moving from the application phase into evaluation and eventually operation, three questions may be worth considering. These are questions to test, not claims that the existing analysis has already answered them.

First: which characteristics of your intended namespace would registration volume alone fail to reveal? If the namespace develops populations you did not plan for, deciding early what additional composition signals to track could provide useful context alongside aggregate metrics.

Second: which changes in the registration mix would warrant investigation against your launch objectives? A higher operator-grade percentage is not automatically the correct target for every TLD. A closed brand TLD, a community TLD, and a commercially distributed generic TLD may have very different objectives. Defining what a meaningful structural change looks like for your specific model would give you a way to evaluate whether the namespace is forming as intended.

Third: would a comparison after launch promotions, registrar onboarding, or the first renewal cycle reveal anything that aggregate reporting does not? Composition may shift most during specific phases. Knowing when to examine the structural profile could help operators identify emerging patterns worth investigating.

An Open Question

Every new TLD applicant in the 2026 round has described the namespace they want to build. The structural evidence from existing TLDs suggests that the eventual composition will include populations and patterns that no business plan fully anticipates. Applicants already plan for uncertainty. The question is whether an additional measurement perspective—one that examines registration composition alongside operational data—would reveal anything useful.

The next question is not simply whether these structural differences can be measured. It is whether tracking them alongside a registry’s operational data reveals something that changes an operator’s decisions. That question remains open, and answering it will require working with operating registries, not just analyzing zone files from the outside.

The application window has closed. For the applicants who filed, the transition from plan to operation lies ahead. How closely the namespace that forms matches the one they planned will depend on the decisions they make and the dynamics they encounter. Observing the difference may be worth the effort.