Summary

  • ICANN's final Applicant Support Program table records 56 fully approved applications from 78 submissions, alongside 99 organizations that began the process but did not submit an application. None of those figures alone proves that outreach worked or failed.
  • The GNSO's adopted measurement guidance calls for pre-agreed conversion benchmarks, stage-specific data and surveys that distinguish informed withdrawal from friction. A privacy-safe cohort concordance could close that evidence gap without exposing applicants or controlling regional outreach.

The celebratory number is 56. On 26 August, ICANN announced that 56 entities had qualified for financial and nonfinancial assistance in the 2026 Applicant Support Program, or ASP. Seventy-eight applications had been received across all five ICANN regions. Intensified outreach and engagement, the announcement said, “seem to have paid off.”

Fifty-six is a meaningful result. It is far larger than the 2012 round, when three support applications were received and one qualified. It also means that applicants from several organizational classes can enter the next stage with fee reductions, counseling, capacity development and access to volunteer professional services.

But the public results contain another number that deserves equal attention. The final statistics page records 99 organizations under Process Initiated – No ASP Application Submitted. Those organizations began the ASP process and did not submit an application for evaluation.

That is not a failure count. Some may have learned enough to make a sensible decision not to proceed. Others may have been browsing, created duplicate or test records, lost interest, missed a deadline or encountered legal, documentary, language, system or financing constraints. The final public page does not assign those reasons, and the checked record does not justify inventing them.

The 99 are therefore not evidence against ICANN. They are the missing population for testing ICANN's claim.

The result table is not the outreach test

The final page reports 78 submitted applications. Of those, 56 were fully approved, 15 were terminated, four were withdrawn and three were found not eligible for support. Within that submitted-application denominator, the approval share is 71.8 percent.

The regional projection for the 56 approvals is also useful: 30 in Asia/Australia/Pacific, 14 in North America, seven in Africa, four in Europe and one in Latin America and the Caribbean. The entity-type table assigns 25 approvals to micro or small businesses from a less-developed economy, 24 to nonprofits or equivalent organizations, six to social-impact or public-benefit small businesses and one to an Indigenous or Tribal People's organization.

Those tables describe final status and one qualification category. They do not measure exposure to outreach. They do not show how many eligible organizations in each region saw a campaign, understood it, began the process, used help, submitted, or made an informed decision to stop.

The units also need care. The table's column says “number of applications,” yet the footnote for the 99 describes organizations. The announcement refers to 56 entities, while the GNSO's later success metric deliberately counts successfully delegated applications rather than applicants or strings. Adding 99 and 78 may look like a complete funnel, but the public definitions do not prove a one-record-per-organization mapping. An initiation-to-submission percentage would therefore claim more precision than the page supplies.

This is not a semantic quibble. A conversion rate is valid only when its numerator and denominator count the same defined unit across a stable observation window.

ICANN's own guidance asks the harder question

The GNSO Council's Applicant Support guidance anticipated this problem. Its first recommendation concerns awareness and outreach. The success indicators are not simply impressions or final approvals.

They include conversion rates proportionate to industry standards, with specific metrics agreed in advance with ICANN's communications function and relevant contractors. They also include clicks, inquiries and registrations, plus surveys about the quality and clarity of information.

Most importantly, the guidance asks whether a recipient understood the information and made an informed decision either to continue or to walk away. That distinction protects both sides of the assessment. It stops ICANN from counting every click as success, and it stops critics from counting every departure as institutional failure.

The same guidance creates two later clocks. At contracting and delegation, the floor is no fewer than ten, or 0.5 percent of all successfully delegated gTLD applications, coming from supported applicants. After delegation, ICANN is expected to examine regional results, three-year operating survival and the experience of successful and unsuccessful supported applicants.

The 56 approvals do not yet satisfy or miss either later test. Support qualification is not a main-round application, a successful evaluation, a delegated string or an operating registry three years later.

Monitoring existed, but the final join is absent

ICANN did not operate blind. Its October 2025 comprehensive applicant report described event data, social-media measures, website traffic, age-of-stage reports, email engagement, use of volunteer services and an applicant survey. It recorded a genuine operational problem: many participants were moving slowly at the beginning, and ICANN could not easily distinguish curiosity from serious intent or see where people were getting stuck.

In May 2025, a five-question survey went to everyone then in the process. Four people answered. Their responses mentioned system navigation, documentation and legal-document difficulties. ICANN contacted the respondents who asked for help, updated guidance, investigated system issues and made volunteer assistance available earlier.

That is evidence of observation and repair. It is also a very small sample. ICANN's report says so explicitly: four responses may not represent people who did not complete the process.

By October 2025, ICANN said it had added active and inactive states, regular messages, age reports and recurring surveys. In January 2026, the Board told the Governmental Advisory Committee that staff were contacting participants weekly and that it remained too early to know whether the delegation goal had been met. In July, ICANN warned that applications no longer progressing because required actions remained outstanding would close after the main-round deadline unless activity resumed.

The institutional record therefore contains more than the final table shows. What remains missing on the final public surface is the join: which outreach class led to which defined entry state, which assistance was offered or used, how long each stage lasted, and whether an exit was informed, inactive, incomplete at deadline, withdrawn, ineligible or unknown.

Without that join, “outreach seems to have paid off” is a plausible institutional interpretation. It is not yet a reproducible result.

Publish a cohort concordance, not applicant files

The answer is not a public list of the 99. It is not disclosure of proposed strings, financial evidence, counselor messages or evaluation material. Those records may be confidential for good reasons.

The minimum common layer is a privacy-safe cohort concordance. Its protected version would preserve:

  • the metric version and the unit counted at every stage;
  • entry, transition, help-offer and help-use times;
  • outreach-source and multiple-contact classes where lawfully available;
  • region and eligible-entity class at an aggregate level;
  • stage age and the last outstanding action;
  • a bounded exit class such as informed decline, inactivity, deadline-incomplete, withdrawal, ineligibility or unknown;
  • survey invitations, responses and nonresponse denominators;
  • the benchmark, its approval date and comparison method; and
  • later links to main-round application, delegation, correction and three-year survival.

The public version should contain only aggregates, definitions and small-cell suppression. No reader needs to know why a named community group withdrew. Readers do need to know whether “walked away after understanding the offer” and “could not complete required documents” are being counted as the same event.

That design leaves outreach local. ICANN's regional teams and partners can choose language, channel, event, mentor and message. The concordance standardizes only the evidence needed to compare the result with the metric ICANN already accepted.

Count the transition, not the applause

The Board allocated up to US$5 million and later up to another US$4.9 million from 2012 auction proceeds so the program could support as many as 75 qualified applicants. Fifty-six approvals do not prove that the remaining capacity reflects insufficient demand, strict eligibility, prudent savings or avoidable loss. The public evidence does not choose among those explanations.

The same discipline applies to geography. One approval in Latin America and the Caribbean is a fact. It is not a regional conversion rate without the number of eligible organizations reached, the number that began, and the number that understood the offer. Thirty approvals in Asia/Australia/Pacific do not prove proportional success without the same denominator.

An institutional claim becomes stronger when the record can survive a different narrator. The 56 should remain. So should the 99, the 78, the changing unit names, the assistance record, the exit classes and the later delegation results.

That is the difference between a milestone and a measurement. ICANN has published the milestone. Its own governance guidance already describes the measurement. The remaining task is to join them.

Sources

  1. ICANN results blog, 26 August 2026
  2. ICANN Applicant Support Program final statistics
  3. ICANN Applicant Support Program overview
  4. ICANN ASP application-process page
  5. GNSO Council Applicant Support Guidance Recommendations Report
  6. ICANN comprehensive applicant report, 3 October 2025
  7. ICANN Board comments for the 12 January 2026 Board–GAC interaction
  8. ICANN ASP update, 15 October 2025
  9. ICANN Board ASP funding resolution, 26 March 2026
  10. ICANN 2026 new-gTLD round overview
  11. ICANN SubPro IRT July 2026 ASP statistics update