Summary

  • On 20 July 2026, the ICANN Board approved the Ombuds Office Objectives for FY27 and said the Ombuds would be evaluated by the Board against specific objectives. The published resolution does not reproduce the objectives or link an objective document.
  • This is not an empty-planning problem. ICANN’s adopted FY27 Operating Plan already publishes three Ombuds priorities, a detailed activity list and six KPI families. The missing public object is the crosswalk showing which Board-approved objective uses which activity, measure, evidence interval and final disposition.
  • A public evaluation map need not expose cases, complainants, self-assessments, compensation amounts or confidential Board deliberation. It can identify stable objective classes, authority, public measures, protected-evidence boundaries, progress states and year-end outcomes.
  • The map matters because the Ombuds must be independent enough to examine unfair treatment involving the Board while remaining formally accountable to that same Board. Legible evaluation protects both independence and oversight; silence invites speculation about each.

A resolution approved an instrument the reader cannot inspect

ICANN’s Board resolution of 20 July is unusually clear about the chain of authority. The Ombuds developed the FY27 Ombuds Office Objectives. The Board Compensation Committee discussed them with the Ombuds and recommended approval. The Board evaluated the recommendation, found the objectives appropriate and consistent with ICANN’s Strategic Plan, and approved them as resolution 2026.07.20.06.

The rationale adds the reason the objectives exist: the Ombuds reports to the Board and is to be evaluated by the Board against specific objectives.

The public page then ends. It does not list the objectives. It does not name an attachment containing them. Its related documents are the agenda and preliminary report, which repeat the decision but do not supply the instrument. The reader can verify that approval occurred, who participated in the approval chain and what broad strategic test the Board applied. The reader cannot see the exact standards that will turn performance evidence into a Board evaluation.

That bounded fact should not be inflated. The objectives may exist in a complete internal record. They may closely track the public operating plan. They may later be published. Their detailed wording may include material that deserves protection. Nothing in the reviewed record shows that an evaluation was skipped or that any objective is improper.

Nor should the item be described as a redaction. The adjacent officer-compensation action expressly invokes confidentiality for employment information. A May Board decision on FY27 organization and chief-executive objectives also contains an express confidentiality resolution. The Ombuds-objectives item contains no corresponding confidentiality clause. That does not create a legal duty to publish; it simply leaves the public disposition of the objective text unexplained.

The governance defect is therefore precise. An authoritative public act points to a specific evaluation instrument without making that instrument, or a safe public representation of it, inspectable.

The workplan is not missing

It would be wrong to say the public does not know what the Ombuds Office intends to do. The Board-adopted ICANN FY27–31 Operating and Financial Plan contains a substantial Ombuds section. It describes the Office as a neutral dispute-resolution practitioner, records its role in unfair-treatment and harassment complaints, and identifies three strategic priorities.

The first is operational excellence and integrity. The second is accessibility, trust and stakeholder collaboration. The third is systemic improvement and preventive intervention.

Under those priorities, the plan names concrete FY27 activities. The Office will maintain a presence at ICANN Public Meetings, handle cases, refer submissions that are outside its scope, facilitate community members and groups, review processes, introduce a new case-management system, conduct surveys and report what it learns. It will maintain an accessible website, publish nine to twelve community blogs, issue community and annual reports, report to the Board, promote the Community Anti-Harassment Policy, develop resource packs and support two to four community groups with tailored interventions.

The plan also identifies six KPI families: satisfaction; budget compliance and cost efficiency; presence and perception; trust; engagement; and prevention and promotion. It says those measures will combine quantitative methods such as surveys with qualitative methods such as focus groups and semi-structured interviews.

This is more than symbolic planning. A range of nine to twelve blogs is testable. A range of two to four group interventions is testable. A new case-management system can be implemented or not. An annual report can appear or fail to appear. Survey baselines can be compared across years.

But an operating plan and a Board evaluation instrument are not automatically the same object. The July resolution calls the approved set “specific objectives.” It does not say that the objectives are identical to the activities and KPI families in the adopted plan. The public plan does not assign weights, identify a Board scoring method or connect every activity to a specific approved objective. The public record reviewed here therefore supports neither of two confident claims: that the Board evaluates something entirely different, or that the public plan already is the complete evaluation sheet.

The missing information is the relationship between them.

Four records expose the chain but not the join

ICANN already publishes most of the architecture needed to reconstruct how evaluation should work.

The Compensation Committee charter assigns the committee responsibility for recommending appropriate base compensation, performance metrics and performance-based compensation levels for the Ombudsman. The public position description says the Ombuds reports directly to the Board, reports regularly to the Chair and reports progress against semi-annual goals to the Compensation Committee. The July resolution shows the Ombuds preparing objectives, the committee discussing them and the full Board approving them. The annual report then provides public evidence about activities and outcomes.

Read separately, each record is intelligible:

  • the charter says who handles performance metrics and compensation recommendations;
  • the position description says who receives progress against semi-annual goals;
  • the resolution says who drafts, recommends and approves annual objectives;
  • the operating plan says what the Office is scheduled to do and what KPI families will be tracked; and
  • the annual report says what happened in the previous year.

What no reviewed public object does is preserve the handoff. Which approved objective corresponds to the “trust” KPI? Which evidence will the Board consider at the first semi-annual review? Is publishing twelve blogs success in itself, or is readership and changed understanding the result? Is a target revised when demand changes? When an objective depends on confidential case evidence, what safe public status records that evidence was considered? What Board disposition closes the objective at year end?

Without a common key, the reader must guess. Institutional staff may know how the pieces connect; a new director, community participant or future Ombuds may not. Accountability that depends on institutional memory is fragile even when every participant acts in good faith.

Activity is not outcome

The FY25 Ombuds Annual Report demonstrates both the value and the limit of current public reporting.

It is a serious report. It says in-scope cases rose 50 percent from FY24 and 200 percent from FY22. It reports that 78 percent of cases involved advice, support or collaborative conflict resolution rather than formal complaints. Three cases remained open at the end of the period, none beyond the Office’s maximum response times. Ninety-four percent of cases opened during the period had closed. Seventy percent of outcomes were self-managed, meaning the person receiving support took action without reporting the result back to the Office.

The report also sets out six service commitments—confidentiality, fair treatment, impartiality, independence, dignity and respect, and role clarity. The Ombuds writes that she will hold herself accountable to them and expects the community to do the same.

That sentence is important. It invites an accountability relationship, not merely passive readership. Yet the report is authored by the Office and fulfils a Bylaw duty to publish consolidated complaint and resolution analysis. It is not presented as the Board’s scoring instrument or as the Board’s year-end disposition of each annual objective.

The distinction protects the report. A rising case count may mean confidence increased, conflict increased, awareness improved or categories changed. A high closure rate says little by itself about fairness or durability. Nine to twelve blogs measure production, not comprehension. A self-managed outcome may represent useful empowerment, but the Office often cannot know the ultimate result because confidentiality and voluntary engagement limit follow-up.

The purpose of an evaluation map is not to convert all these facts into a single score. It is to show which facts answer which institutional question—and which do not.

Independence makes the interface material

Every ombuds institution contains a structural tension. It needs independence from the people and bodies whose conduct it may examine. It also needs a lawful appointment, budget, performance review and removal structure. Absolute insulation would eliminate accountability; ordinary managerial control could destroy trust.

ICANN places that tension in public view. The Ombuds operates independently from ICANN org. The Board is the formal accountability point. The Bylaws protect the Office’s budget submission and require a three-quarters Board vote for dismissal. The Framework says the Ombuds is independent in structure, function, appearance, purpose, practice and decision-making. At the same time, the Board appoints, manages and evaluates the Ombuds, and its Compensation Committee handles performance metrics and performance-based compensation recommendations.

No reviewed evidence shows that this structure produced interference. But the design creates two symmetrical risks of misunderstanding.

If evaluation remains opaque, a critic can imagine that the Board rewards institutional comfort or discourages scrutiny. If oversight is described only in broad terms, a defender can treat “independence” as a reason not to ask whether public commitments were met. Neither inference is justified by the record. Both flourish when the interface is invisible.

ICANN’s 2025 baseline survey shows why perception cannot be dismissed. Eighty-one people responded voluntarily; the sample was not a representative poll and must not be treated as one. Within that limited group, 40 percent expressed confidence that the Ombuds Office operates independently and 19 percent disagreed. Confidence was higher for confidentiality, at 58 percent, while 42 percent expressed confidence that robust complaint and dispute processes were in place.

Those numbers do not grade the Ombuds. They identify a field of uncertainty. A transparent evaluation map cannot manufacture trust, but it can prevent the Board’s oversight role from remaining an undefined shadow behind the Office’s declared independence.

Publish the map, not the personnel file

A proportionate solution is one public evaluation map with a stable row for each approved objective. If the objective wording itself contains protected personnel material, the row can use a public objective class. The map should be issued when objectives are approved, updated at the semi-annual checkpoint and closed after the Board’s year-end evaluation.

Each row needs:

  • a stable objective identifier and version;
  • a public objective statement or safe objective class;
  • the approving authority and date;
  • links to the relevant strategic priority and operating-plan activities;
  • the KPI family and public measure definition;
  • a baseline and target, or a stated reason a target cannot be public;
  • the evidence source and reporting interval;
  • a defined progress state such as not started, on track, at risk, revised, completed or not assessable;
  • the year-end Board disposition;
  • a variance explanation or supersession link;
  • the confidentiality boundary and institutional custodian; and
  • a correction history.

The map must not contain complainant names, case narratives, identifying details, confidential Board reports, private self-assessments, compensation amounts or sensitive qualitative evidence. It need not reveal how an individual director scored an item. It need not give the community a veto over an employment decision.

A protected-evidence field can say that the Board reviewed confidential case-quality evidence, with the evidence class and date, while leaving its content private. A target field can say “not public—risk of distorting complaint intake” instead of inventing a quota. A revised objective can retain its original version and explain the changed operating condition without exposing personnel discussion.

This is accountability metadata. It proves that a test exists, that relevant evidence reached the authorized evaluator and that the test moved to a known state. It does not publish the personnel file.

The strongest objections improve the design

The first objection is gaming. If the Board publishes targets, an Ombuds might be pushed toward easy closures, high volumes or visible communications rather than difficult, confidential work. That is a real risk. Complaint counts should not become sales quotas. The answer is to distinguish activity, quality, trust and systemic learning; to mark measures that must remain contextual; and to publish explicit anti-gaming notes.

The second objection is candour. The Ombuds and directors need room for honest self-assessment, critical feedback and compensation discussion. A public map should preserve that room. It can publish the objective architecture and final state while protecting narrative evaluation, individual views and compensation reasoning.

The third objection is duplication. The operating plan and annual report already carry much of the data. That is precisely why the additional burden can be small. The map should link existing records rather than demand a second report. If it cannot be generated from the objective approval and semi-annual review that already occur, the internal record may itself be too disconnected.

The fourth objection is that Board evaluation is not community adjudication. Correct. Open participation does not make every observer the principal. The Board remains the legal evaluator. The public map allows outsiders to verify the shape and completion of the authorized process without converting commentary into decision power.

Evidence limits

This analysis establishes a public-record gap, not an internal failure.

It does not establish that ICANN lacks a complete internal objective set, a private crosswalk or documented semi-annual review. It does not establish that the approved objectives differ from the operating plan, that publication is legally mandatory, or that the Board acted improperly by approving them without public comment. It does not establish that the Ombuds missed a target, that a director influenced a case or that compensation affected a decision.

The FY25 annual report describes a prior fiscal year. The 2025 survey had 81 voluntary respondents and does not represent the whole community. Activity totals cannot be converted into quality findings without definitions and context. Confidentiality is fundamental to an Ombuds service, and some evidence should never appear in a public record.

The demonstrated condition is narrower: ICANN has published a rich FY27 Ombuds workplan and has separately announced approval of specific FY27 objectives for Board evaluation, but the reviewed public surface does not join them. The repair is a safe crosswalk, not maximum disclosure.

Sources

  1. ICANN Board — Approved Resolutions, Special Meeting, 20 July 2026
  2. ICANN — Adopted FY27–31 Operating and Financial Plan and FY27 Operating Plan
  3. ICANN — Board Compensation Committee Charter
  4. ICANN Office of Ombuds — FY25 Annual Report
  5. ICANN — Ombuds Position Description
  6. ICANN Office of Ombuds — 2025 Baseline Survey Report and Action Plan
  7. ICANN — Ombuds Office Framework and Process
  8. ICANN — Current Bylaws
  9. ICANN Board — Approved Resolutions, 3 May 2026