Summary
- The Customer Standing Committee's finding of 18 November 2026 rated Public Technical Identifiers' October 2026 performance "Excellent", at 100 per cent, with all 64 currently defined thresholds met.
- The table underneath that figure sets a threshold for each individual request but measures compliance as an aggregate across all requests in a category, so 64 of 64 counts thresholds, not tickets served inside their target.
Start with where the yardstick lives
The ICANN Bylaws, as amended 3 July 2026, list Article 17 for the Customer Standing Committee and Article 18 for IANA Naming Function Reviews. Their Annex A is something else entirely: the GNSO policy development process (Bylaws). The service levels are not a bylaw annex. They sit in Annex A, Section 2(c) of the IANA Naming Function Contract, which fixes the structure of each service level — process, metric, threshold, type, compliance percentage, measurement period — incorporates the levels maintained on PTI's own page, and requires a monthly performance report no later than fifteen calendar days after month-end (contract, integrated through Amendment 2). That document says on its own face that it is informational and that the executed instruments control; the executed contract was not obtained in this review.
The committee itself is described on ICANN's page in four verbs: monitor PTI against those service level expectations, analyse the reports PTI supplies, publish findings, and take remedial action before escalating an unremedied issue to the ccNSO and GNSO (CSC page).
The table tells you which unit is being counted
Open the yardstick and the argument is over before it starts. The columns read Process Category, Metric, Threshold, Type, Compliance and Period. Threshold is defined as the target for each individual request. Compliance is defined as "The percentage that the target goal in aggregate must be met or exceeded within the specified time period for all requests in the specified category" (PTI service levels).
Read those two sentences together. The target is written for one request; the test is applied to a category. A month can therefore satisfy every defined threshold and still contain tickets that arrived after their target, provided the category's aggregate held. That is what the table says — not a suspicion about anyone's arithmetic.
The numbers themselves are modest and specific. Ticket confirmation within 60 seconds, 95 per cent. Lodgment of an emailed change request within three days, 95 per cent. First technical checks within 50 minutes, 95 per cent; retests within 10 minutes, 95 per cent. Contractor review within five days at 90 per cent for routine root zone changes, ten days at 90 per cent for gTLD creation or transfer, fourteen days at 100 per cent for ccTLD creation or transfer. Root zone publication within 72 hours at 99 per cent. Label Generation Ruleset validation within five days and implementation within seven, both 90 per cent.
Accuracy at 100 per cent as a minimum; availability of the root zone management system, the website and the directory service at 99.0 per cent or more. Enquiry acknowledgement within 60 seconds at 95 per cent; initial response within five days at 90 per cent. Extraction elided Category II and parts of Categories III and IV, so that list is representative, not certified complete.
What October printed
PTI's October 2026 report reached the committee on 9 November 2026, with a covering note from Selina Harrington, Senior Manager, IANA Operations: "During the month of October, we met 100% of the SLA thresholds" (transmittal). The report states its own denominator plainly — "Percentage of SLA thresholds met of those defined 64 of 64 (100.0%)" — and adds that "There were no exceptions for the reporting period" (October report).
The measured actuals are not close calls. Availability at 100.00 per cent for the IANA website, the root zone management system, the WHOIS directory service and the service level dashboard. Root zone file accuracy at 100.00 per cent. Root zone publication at 21.45 hours against a 72-hour expectation at 99 per cent. First technical check at 2.59 minutes against 50 minutes at 95 per cent. Validation and reviews at 0.56 days against five days at 90 per cent. Nobody reading that page needs to be talked out of thinking October went well.
The rows that print N/A and an em dash
The same summary carries rows whose expected value is N/A and whose actual value is an em dash: validation and reviews for ccTLD creation and transfer, validation and reviews for other changes, third-party approval for ccTLD creation and transfer, requester clarification, root zone publication for other changes, the time-per-actor rows, the request volumes. The headline percentage is computed over the 64 defined thresholds, not over every printed row. Nothing in the text obtained says how those rows relate to the denominator, and that is the point at which the figure becomes a statement about a set rather than about a service.
The word the committee published
The finding is one page, dated 18 November 2026, and its spacing collapses when the text layer is extracted; the wording is not retyped here. It records that the committee completed its review of the October 2026 report and finds PTI's performance for the month was "Excellent", that performance over October 2026 was 100 per cent, and that PTI met all 64 of the currently defined thresholds. It adds that no metrics currently require close tracking, that there are no SLAs currently in the process of being changed, and that no escalations were received during the reporting period (CSC findings).
Two days earlier, on 17 November, a cover letter sent on behalf of the chair by Claudia Ruiz reported Excellent for September and October both, and recorded that the committee met the PTI Board at ICANN87 in Bali on the ongoing SLA review. It states that the committee and PTI "expect to formally agree on the changes at the November CSC meeting", after which the proposals go to public comment (CSC-Announce). Read beside the finding's sentence that no SLAs are currently in the process of being changed, those two documents pull in different directions. They may be reconcilable — nothing has yet been agreed, and nothing is yet in force — but the text does not say so, and this review reports the tension rather than dissolving it.
The reviewer that cannot start the review
The committee's amended charter was adopted 3 May 2026 and took effect 3 July 2026. It records monthly monitoring, monthly publication, the authority to invoke the Remedial Action Procedures, and the authority to escalate an unresolved deficiency to the ccNSO and GNSO Councils. It also records a limit: the committee is not authorised to initiate a change in the IANA functions operator through a Special IANA Function Review. Composition follows Bylaws Section 17.2 (amended charter).
The Bylaws amendments that produced that text were approved by the Board on 3 May 2026; ICANN's General Counsel told the Empowered Community on 11 May that the approved Article 17 text is identical to the February 2026 public-comment draft, that the Article 18 change was confined to Section 18.2(b) on periodic-review timing, and that the package then went through Empowered Community approval under Section 25.2 (General Counsel notice). The operative Article 17 and Article 18 bodies were not obtained in this review: the Bylaws page text truncated inside Article 1.
So read the ladder as the supporting guidelines describe it. Section 18.12(a), quoted in the ccNSO and GNSO joint consultation guidance of 23 January 2020, conditions a Special IFR outside the periodic cycle on the committee's remedial procedures having been followed and failed, the IANA Problem Resolution Process having been followed and failed, review by both councils, meaningful consultation with other Supporting Organizations and Advisory Committees, a public comment period, and a ccNSO Council supermajority plus a GNSO Supermajority. Escalation to the ICANN Board is the last step of the remedial procedures, and the process formally starts when the committee notifies both councils that an issue remains unresolved (joint guidelines). That quotation is six years old and the current wording of Section 18.12(a) was not obtained here.
A body that grades performance and cannot initiate the review that could replace the performer is not a broken instrument. It is a designed one. Its published verdict tells you the aggregate held; the lever sits elsewhere, behind two council votes.
One dated posting observation
On 19 November 2026 the IANA monthly reports index still labelled September 2026 as "Latest", and its 2026 row ran from January to September only — nine days after the committee's own finding had turned October's report into a rating (reports index). The month-specific address for October returned no readable text at observation time. That is a note about one index page on one afternoon. It is not a claim that the October report is absent from the site, and the unreadable address is an access limitation, not a finding about its contents.
What would make the verdict arguable
A registry operator holding a ticket filed on 3 October cannot learn from these documents whether its ticket met its own sixty-second target. It can learn that the category aggregate held, and that the committee accepted the operator's figure.
Four things would close that gap, and none of them requires a new institution: per-request distributions, so a category prints how many requests met the threshold and how many missed; the definition and population of "those defined", including how rows printed at N/A with em-dash actuals are treated; raw monthly actuals at the same resolution as the thresholds; and the current status of the escalation path, so a reader can see whether the ladder has been climbed or only built.
The committee and PTI say they expect to agree changes at their November meeting and take them to public comment. That is the moment the yardstick is editable, and the four items above are the questions worth asking of it. A scoreboard that counts thresholds answers a governance question well. It has never been a customer-service report, and calling it one asks a number to do work the instrument never assigned it.
Member Briefing
Read the full trend briefing
Members read the full briefing with the source notes behind every finding.
Strategic Circle
A daily briefing in your language, built from changes we have verified.
Join for USD 20/monthCancel anytime.
Members also get the full research library, a daily briefing on the networks they follow and the private forum.






