Summary

  • The June 2014 .HOTEL Community Priority Evaluation was a 15-point result crossing a 14-point threshold. Under the 2012 Applicant Guidebook, that threshold did not merely improve HOTEL Top-Level-Domain S.a.r.l.’s rank. It caused the community application to prevail over six directly contending standard applications, removing those applications from the path that otherwise led towards settlement or auction.
  • The Economist Intelligence Unit panel supplied the decisive factual judgement, but the exclusionary force came from ICANN’s rules. Reconsideration tested claimed departures from established policy and procedure. The BGC also said the requesters remained free to ask the Ombudsman to review alleged unfair treatment, a fairness and alternative-dispute-resolution route that could investigate and recommend but not set aside the score. The Independent Review Process tested ICANN’s conduct against its governing documents. None of those routes was constituted as a second Community Priority Evaluation with authority to assign a replacement score.
  • ICANN’s adopted 2026 Round rules now use a 12-of-16 threshold, three-member panels, calibration and quality assurance, and a no-fee Evaluation Challenge filed within 21 days for factual or procedural error; the Challenge Panel must also address a system error if it finds one. The challenge is available when a community application fails. On the text of the current Guidebook, it does not give a standard applicant made ineligible by another application’s passing score a corresponding right to challenge that pass.
  • Later review of .HOTEL examined provider communications, cross-report consistency and research references without reopening the score. A later IRP ended in October 2023 after all claims were withdrawn and was dismissed with prejudice without an award. ICANN records a registry agreement dated June 2025, while .HOTEL remains absent from the IANA Root Zone Database at publication. Priority, contracting and delegation were therefore separate exercises of authority.

The report that changed seven positions

On 12 June 2014, ICANN published a short document with an unusually large consequence. The .HOTEL Community Priority Evaluation report, dated the previous day, awarded HOTEL Top-Level-Domain S.a.r.l. 15 points out of 16. Fourteen was the minimum required to prevail. The title of this case is therefore about the 14-point gate, not a claim that the application received only 14.

Nothing was auctioned that day. No rival announced a negotiated withdrawal. No registry agreement was executed, and no root-zone change followed from the report alone. Yet the procedural and economic position of every applicant in the seven-member contention set changed. The community applicant crossed the threshold. The six standard applications no longer had an available route to win .HOTEL through the contention-resolution sequence that would otherwise have remained open to them.

The six standard applications were submitted by Despegar Online S.R.L.; DotHotel Inc.; dot Hotel Limited; Fegistry, LLC; Spring McCook, LLC; and Top Level Domain Holdings Limited. Later accountability records use successor or group names including Travel Reservations, Donuts, Famous Four Media, Minds + Machines and Radix. Those changes in claimant identity did not enlarge the contention set. There were seven applications: one community application and six standard applications.

The decisive mechanism was neither ordinary procurement nor judicial adjudication. ICANN had written a conditional consequence into the 2012 Applicant Guidebook’s contention procedures. A qualifying community applicant could elect Community Priority Evaluation. An external provider would score it against four criteria. If the score reached 14, the community application would prevail over directly contending standard applications and non-prevailing community applications. The external report supplied the threshold fact; the Guidebook supplied the consequence; ICANN organisation administered the result.

That allocation matters because it prevents two misleading descriptions. It is too simple to say that the Economist Intelligence Unit “awarded” .HOTEL: the provider had no power to sign the registry agreement or direct a root-zone delegation. It is also too simple to say that ICANN’s Board independently chose to eliminate six applicants in June 2014: the Board had already approved a rule under which a passing score produced that result without a fresh merits vote. The practical power lay in the combination of delegated judgement and an automatic rule.

The exclusion was written before the evidence was scored

Community Priority Evaluation was one of several methods for resolving string contention in the 2012 round. ICANN’s programme description identifies the Economist Intelligence Unit as the independent evaluation provider and states the four evaluation areas: community establishment; nexus between the proposed string and the community; registration policies; and community endorsement. Each area was worth up to four points. Requiring 14 of 16 left little room for weakness, while the consequence was not a small preference. A successful community application displaced other applications that may themselves have passed ICANN’s ordinary technical, operational and financial evaluation.

Only an applicant that had self-designated as community-based could elect CPE. The standard applicants could not require the community applicant to forgo it, and they could not demand an auction merely because they preferred a price-based resolution. The Guidebook also said that standard applicants in the contention set would not participate in the CPE, although public comments and opposition material could form part of the record supplied to the evaluator. Their direct procedural leverage came later through ICANN’s accountability mechanisms.

Participation, where permitted, was not control: the community applicant chose whether to seek priority, the provider controlled the evaluation judgement, and the Guidebook controlled the immediate consequence.

The design used a high entry threshold as its principal safeguard. A community applicant needed 14 of 16 points, leaving very little room for weakness across the four criteria. That reduced the chance that a loosely defined or poorly supported claim could defeat standard applications. But a high threshold is not the same institution as an appeal. The threshold changes how difficult it is to win at first instance. An appeal changes who can correct an asserted error after the first-instance judgement has been made.

The Guidebook’s structure preferred front-loaded certainty. Applicants were told the criteria, the available points and the effect of a passing score. The provider’s report would explain the assessment. Yet the Guidebook did not establish a second specialist panel with authority to reconsider the evidential merits. The available accountability routes had different objects. Reconsideration focused on whether ICANN or its provider had departed from established policy or procedure. The Ombudsman could investigate alleged unfair treatment, use alternative-dispute-resolution techniques and recommend action within the office’s authority.

Independent review tested ICANN action against the Articles and Bylaws. None of those routes was a specialist CPE rescore.

This was a consequential institutional trade. It reduced the prospect that every losing applicant could turn CPE into a second full evaluation, with more cost and delay. It also made the quality of the initial evaluation record unusually important. When a private evaluator’s judgement is attached to an automatic exclusion rule, imperfections in the first-instance record cannot be treated as ordinary advisory defects. They determine whether six other applications remain economically alive.

The report itself recognised a boundary that is often lost in later summaries. Its disclaimer said that the CPE result was not necessarily the final outcome of the application and could change in limited circumstances. That was accurate. The prevailing applicant still had later programme steps to complete, including contracting and the requirements preceding root-zone delegation. But the disclaimer did not make the score inconsequential. It distinguished victory in the contention mechanism from completion of the entire application process.

Why .HOTEL received 15 points

The panel awarded four points for community establishment. Its analysis accepted the applicant’s delineation of a global hotel community extending across hotels, hotel chains, hotel marketing organisations, hotel associations and other organisations with a direct relationship to the sector. The application used hotel definitions and verifiable records, including industry lists, directories and registers. The panel referred to established bodies such as the International Hotel and Restaurant Association, HOTREC, the American Hotel & Lodging Association and the China Hotel Association.

It treated the community as organised, pre-existing and of considerable size rather than as a population invented for the application.

That conclusion contained two judgements that competitors could reasonably regard as contestable without their objections thereby becoming established fact. First, the panel had to decide whether a global commercial sector could possess sufficient structure, recognition and continuity to count as a community under the Guidebook. Second, it had to decide where the community’s boundary lay. A definition centred only on establishments that provide hotel accommodation is narrower than one that also includes associations, marketing organisations and related bodies. The panel found the broader definition sufficiently delineated and extensive.

The report, not the existence of hotels as such, supplied the institutional conclusion that this was the relevant CPE community.

For nexus, the panel awarded three of four points. It gave two of the three available points for the relationship between .HOTEL and the community. The string directly identified the community’s core members, but the community as defined also included bodies that were not themselves hotels. That reduced the nexus score from the maximum. The panel nevertheless awarded the separate uniqueness point because it found no other significant meaning of “hotel” beyond the community identified in the application.

This three-point result is important to the later dispute. It shows that the panel did not simply accept every element of the applicant’s case. It recognised overbreadth at the edge of the community definition. Yet the deduction was only one point. The application stood one point above the line, but the 2012 rule made that position resistant to a single-point reversal: a reduction from 15 to 14 would still have produced priority. Any merits appeal capable of changing the outcome would therefore have needed to identify at least two net points of reversible error, not merely one debatable judgement.

The panel awarded all four points for registration policies, one point for each subcriterion. Eligibility was restricted to members of the defined hotel community and verified through membership lists, directories and registers. Name selection was a separate rule: once eligible, an applicant could register any available name that was neither reserved nor already registered, while major hotel brands and sub-brands were placed on a reserved list. Content and use had to be hotel-community-related and relevant to the domain name. Enforcement included a challenge process, random checks, possible revocation and an appeal against revocation.

The CPE question was not whether these measures would later be enforced perfectly in every registration case. It was whether the application’s eligibility, name-selection, content-and-use and enforcement policies met the Guidebook’s four registration-policy subcriteria. The panel awarded a point for each.

The final four points came from community endorsement. The panel did not treat the applicant itself as the recognised institution representing the entire hotel community. Instead, it relied on support from recognised institutions and member organisations that, in its assessment, represented a majority of the relevant community. It also evaluated opposition and concluded that the opposition placed before it was either negligible in scale, not from institutions relevant to the defined community, or better characterised as obstruction than as substantial community resistance.

That last judgement was among the most power-laden in the report. Support and opposition are not counted like ballots in a universal electorate. The evaluator must decide which organisations are representative, what population they represent, how much duplication exists between memberships, and whether an opponent has standing within the defined community. A commercial rival may have a large economic interest in the string and still carry limited weight under a criterion aimed at community institutions. Conversely, an association’s support may count for many members even though those members were not individually consulted for the CPE.

The panel’s conclusion should therefore be stated as a finding, not converted into an independent fact about the motives or legitimacy of every opponent. The report found the opposition insufficient to reduce the score. Competitors alleged that the treatment of support and opposition, like the definition of the community and the uniqueness finding, was mistaken or inadequately transparent. Their allegations initiated review; they did not themselves prove that the score was wrong.

Taken together, the points produced a result with a striking internal structure: 4 for establishment, 3 for nexus, 4 for registration policies and 4 for endorsement. The only point lost was at the boundary between the core meaning of “hotel” and the broader set of related organisations included in the community. The report’s logic was therefore not that every hotel-sector participant had a single, identical interest. It was that a sufficiently organised global community existed, .HOTEL substantially identified it, the proposed registration rules were community-oriented, and recognised institutions supplied adequate support.

The power map behind the score

Five institutional functions were separated, but not equally.

ICANN’s policy and Board structures defined the criteria and attached the exclusionary consequence to a passing score through the Applicant Guidebook. That was the foundational exercise of authority. Without it, the provider’s report would have been advice.

HOTEL Top-Level-Domain supplied the application, its proposed community definition, registration commitments and supporting evidence. Supporting organisations supplied endorsements. Competitors and other participants could supply opposition and public material. Evidence production was dispersed, but the applicant bore the central burden because only its community claim was being scored.

The Economist Intelligence Unit panel classified the evidence and assigned points. Later process-review material described two evaluators, coordination and a consensus process, with ICANN organisation performing administrative exchanges and commenting on clarity and the sufficiency of stated reasoning rather than assigning the score. The provider’s formal authority came from its engagement by ICANN and the Guidebook’s recognition of its role, not from public law or a direct contractual relationship with every losing applicant.

ICANN organisation received, published and operationalised the report in the application system. It did not need to persuade the Board to make six new, applicant-specific findings. Accepting the report as produced under the programme was enough to activate the pre-existing rule.

The BGC, later the BAMC, and ultimately the Board controlled reconsideration. An IRP panel could examine whether Board action or inaction complied with ICANN’s Articles and Bylaws. Those bodies could identify procedural failure and prompt further Board consideration or a remand-like response. They were not constituted as a second CPE provider with a free-standing mandate to determine whether the hotel community deserved 13, 14 or 15 points.

This separation created the core legitimacy problem. The actor making the decisive merits judgement was external. The actors answerable under ICANN’s Bylaws did not make that judgement directly. Accountability therefore depended on whether ICANN’s own bodies had done enough to assure themselves that the provider followed the rules. The question after June 2014 was less “Which score is best?” than “How much checking must ICANN perform before it can rely on an external score with automatic exclusionary effect?”

Request 14-34: reconsideration without a merits appeal

The six standard applicants filed Reconsideration Request 14-34 on 28 June 2014. They challenged the community-establishment analysis, the nexus and uniqueness findings, the treatment of support and opposition, the panel’s independence and qualifications, and the adequacy of information about materials and communications used in the evaluation. Their requested outcome was effectively to set aside the result and return it for a proper evaluation.

The BGC’s 22 August 2014 determination drew a sharp line around reconsideration. Where ICANN had engaged a third-party expert, disagreement with the expert’s substantive conclusion was not enough. The requesters had to identify a failure to follow established policy or procedure by the panel, or a failure by ICANN staff to follow policy or procedure when accepting the result.

That boundary did not make the report unreviewable in every sense. The BGC examined the asserted Guidebook failures criterion by criterion. It considered whether the panel had used the required definitions and whether the Guidebook demanded additional inquiries of the kind the applicants proposed. It concluded that the panel had addressed the required elements and that the challengers were seeking a different substantive judgement rather than identifying a procedural departure.

The BGC also rejected the proposition that standard applicants possessed procedural rights beyond those conferred by the Guidebook. They had a direct economic stake, but CPE was an evaluation of the community application rather than an adversarial trial between seven co-equal parties. The Guidebook did not give them a right to cross-examine evaluators, obtain litigation-style discovery or insist that the BGC hear the merits afresh.

The distinction is defensible but not neutral. “Procedure” and “merits” overlap when a criterion is open-textured. Suppose an evaluator defines the relevant community too broadly. That can be described as a wrong merits judgement. It can also be described as a failure to apply the Guidebook’s required test. The answer depends on the intensity of review. If the reviewer asks only whether the report mentions the criterion, almost every interpretive error becomes insulated as merits. If the reviewer asks whether the stated reasoning rationally applies the criterion, some errors become procedural or legal failures capable of correction.

The BGC denied Request 14-34. Its determination was final for that staff-action reconsideration and required no further Board or New gTLD Program Committee consideration. It did not state that every ICANN accountability avenue had disappeared: the determination expressly said the requesters remained free to ask the Ombudsman to review whether they had been treated unfairly. That office’s authority was different from a CPE appeal. Under the ICANN Ombudsman Framework, the Ombudsman could investigate, use alternative-dispute-resolution techniques and notify or recommend action to the Board, but could not make, change or set aside an administrative or Board decision. Independent review remained a separate route for testing ICANN conduct against the Articles and Bylaws. The applicants therefore retained review avenues, but not a specialist forum authorised to replace the provider’s 15 points with another merits score.

The later Despegar .HOTEL final declaration approached the dispute through ICANN’s own conduct rather than by conducting a second CPE. The panel considered the BGC’s treatment of the asserted Guidebook failures and did not overturn the determination or order a rescore. That outcome should not be expanded into a substantive endorsement of every point awarded. It established that the accountability panel found no bylaw-level basis to disturb ICANN’s reliance on the evaluation after the BGC’s review.

Request 14-39: transparency was not discovery

The applicants also sought the records behind the report through ICANN’s Documentary Information Disclosure Policy. Their request covered material about the selection and appointment of evaluators, evaluator qualifications, communications between ICANN and the provider, and documents used to prepare the .HOTEL report. ICANN disclosed some information and withheld other material under the DIDP’s conditions for nondisclosure. The applicants then brought Reconsideration Request 14-39.

The BGC’s 11 October 2014 determination again separated compliance from preferred outcome. Its inquiry was whether staff had followed the DIDP process, not whether the BGC would have struck a different balance between disclosure and confidentiality. It also rejected treatment of the DIDP as litigation discovery. The policy offered institutional transparency subject to defined exceptions; it did not give disappointed applicants a general entitlement to the provider’s entire working file. As in Request 14-34, the BGC said the requesters remained free to approach the Ombudsman if they believed they had been treated unfairly; that possibility did not convert the Ombudsman into a disclosure tribunal or a CPE scoring panel.

The Despegar IRP later noted that staff could have explained more specifically why particular records fell within the cited nondisclosure conditions. More precise reasons would have made the disclosure decision easier to test. The panel nevertheless declared ICANN the prevailing party and ordered neither a renewed disclosure process nor a CPE rescore. Request 14-39 concerned access to the record; it did not turn transparency into authority to replace the evaluation.

The Despegar IRP reviewed ICANN, not the hotel community

Despegar Online, Donuts, Famous Four Media, Fegistry and Radix commenced independent review, challenging the handling of Requests 14-34 and 14-39 and seeking relief connected to the CPE outcome. The claimant list reflected corporate groups connected to standard applications rather than a simple repetition of all six application names. The proceeding was consolidated with a separate .ECO IRP for hearing, but each string retained its own underlying evaluation record and requested relief.

The IRP panel’s jurisdiction was anchored in the ICANN Bylaws then in force. It asked whether the Board and its committee had acted consistently with ICANN’s Articles and Bylaws, including duties of diligence, independent judgement, fairness and transparency. The provider itself was not transformed into a public tribunal directly answerable to the claimants under those instruments. The review therefore approached the score through ICANN’s reliance on it.

The panel’s jurisdiction made formal oversight relevant but bounded. It examined whether the Board and BGC had taken the challengers’ asserted Guidebook failures seriously enough to satisfy ICANN’s governing commitments. That inquiry could reach the adequacy of the committee’s reasoning; it did not turn the IRP into a replacement evaluation panel.

On .HOTEL, the panel upheld ICANN’s handling. The BGC had addressed the challengers’ assertions about community definition, self-awareness and cohesion, string nexus, uniqueness, support, opposition and claimed departures from the Guidebook. The IRP declared ICANN the prevailing party and did not order a rescore, restore the six applications or award the string to any claimant. The decisive difference was between reviewing ICANN’s fidelity to its rules and independently deciding how many points the evidence deserved.

The declaration was nevertheless more than a ritual endorsement. It identified structural concerns about a mechanism in which different evaluator teams could make high-consequence judgements without a then-visible system for comparing outcomes and assuring consistency across reports. It noted the sensitivity of outcomes to small evaluation differences and encouraged measures to improve predictability and quality control. It also questioned the weakness created when ICANN’s core commitments were not expressly flowed through to the external provider whose work ICANN would use.

Those observations were institutionally bounded. They did not establish that .HOTEL was scored incorrectly. They identified why outsourced merits work cannot be made accountable merely by calling the provider independent. Independence from staff influence is one value. Consistency, expertise, reason-giving, record preservation and susceptibility to remedy are others. A provider can be independent and still produce a judgement that is difficult to test.

The IRP also addressed allegations arising from a configuration problem in ICANN’s new gTLD applicant portal, through which credentials associated with individuals connected to the wider .HOTEL commercial network had been used to access information belonging to other applicants. The claimants argued that the access and the individuals’ relationships to the community applicant tainted the process. Because the Board was still investigating, the panel did not convert the unresolved factual issue into a merits ruling on the CPE. It left the requested institutional response to the body then holding the application-processing power.

On 10 March 2016, the Board accepted the declaration, directed that future programme reviews consider CPE consistency and predictability, encouraged more specific DIDP responses, and ordered completion of the portal investigation. The immediate remedy was investigation and institutional follow-up, not cancellation of the community application.

The portal dispute produced a Board decision, then Request 16-11

The Board considered submissions from the standard-applicant groups and the results of ICANN’s forensic inquiry. In its 9 August 2016 resolutions, it stated that access to confidential applicant information had occurred after the community application was submitted and after CPE had been elected. It further stated that ICANN had found no evidence that the accessed information was used to support the .HOTEL application or enabled it to prevail in CPE. The Board therefore concluded that cancellation was not warranted and directed ICANN organisation to continue processing the community application.

Those were ICANN’s findings on the investigation before it. The applicant groups disputed the adequacy and implications of that inquiry. Their allegations should not be rewritten as proof that the CPE result was corrupted; the Board’s absence-of-evidence conclusion should not be enlarged into proof that no misuse was possible. The operative governance fact is that the Board, which had authority over application processing, found the record insufficient for the requested cancellation remedy.

Travel Reservations SRL, Spring McCook, Minds + Machines Group Limited, Famous Four Media Limited, dot Hotel Limited, Radix FZC, dot Hotel Inc. and Fegistry brought Reconsideration Request 16-11. Unlike 14-34, which attacked staff acceptance of the CPE report, 16-11 challenged a Board action: the August resolutions refusing cancellation and directing the application to move forward. That changed the decisional route. The accountability committee would make a recommendation, but the Board itself would determine the request.

The formal test was also narrower than a reopened inquiry into every fact. Under the February 2016 Bylaws applicable when the request was filed on 25 August 2016, a challenge to Board action required identification of material information that existed when the Board acted but was not considered, or reliance on false or inaccurate material information. The requesters could present their case, but they still needed to fit it to the remedial standard. Meaningful participation did not transfer control over the standard of review.

The process review mapped machinery without reopening merits

While Request 16-11 was pending, ICANN widened its response to concerns about CPE. On 17 September 2016, the Board directed an independent review of how ICANN staff interacted with the CPE provider. The BGC later added two further scopes: consistency in applying the CPE criteria and compilation of the research used in evaluations subject to pending reconsideration requests. Request 16-11 and several other CPE-related requests were put on hold.

FTI Consulting conducted the work and ICANN published three reports on 13 December 2017. Their boundaries are as important as their conclusions.

The Scope 1 report examined communications between ICANN organisation and the provider. It reviewed 100,701 ICANN emails and attachments, interviewed relevant personnel and considered working materials supplied by the provider. It described a process in which two evaluators undertook reviews, a coordinator and core team supported consistency, and the provider reached a consensus score. ICANN organisation transmitted materials and could comment on clarity and the sufficiency of stated reasoning, but the review found no evidence that ICANN exerted undue influence over scores or engaged in impropriety.

That conclusion was expressly bounded by the record. FTI did not receive the provider’s internal emails or the provider-side copies of external email communications requested in its first two document categories, although it reviewed ICANN’s copies of the external exchanges and received provider working materials. Many communications between ICANN and the provider were verbal and not memorialised, so FTI could not evaluate them directly. FTI reported that nothing it observed indicated improper verbal influence. This is a finding of no evidence in the available material, not an omniscient proof about every interaction.

Independence of the reviewer does not expand the evidence that exists.

The Scope 2 report compared application of the criteria across 26 CPE reports. It found no deviation from the applicable guidelines and no inconsistent application. For .HOTEL, the comparison placed its nexus treatment within a broader pattern: applications receiving two points for nexus could still receive the separate uniqueness point where the string had no other significant meaning.

Yet Scope 2 expressly was not a re-evaluation of the applications. It asked whether the provider applied a common method, not whether the best reading of the evidence required a different score. Consistency can support legitimacy because like cases should be treated alike. It cannot alone establish correctness. A methodology may be consistently permissive, consistently strict or consistently contestable. The review answered the horizontal question—whether reports diverged from one another—not the appellate question—whether .HOTEL itself crossed the threshold on the strongest interpretation of the Guidebook.

The Scope 3 report compiled references found in final reports and working papers for eight pending cases. For .HOTEL, it identified 63 references: 42 connected to community establishment, three to nexus, none to registration policies, 12 to endorsement and six classified as additional. It also recorded no separate research references for opposition in the .HOTEL final report or working papers. That documentation fact does not prove that opposition was ignored—the CPE report discusses opposition and could assess submitted material without additional research—but it shows why a published reference index would have improved contemporaneous traceability.

Scope 3 also disclaimed a merits function. It did not assess the propriety, reasonableness or persuasive force of the research. The work made it easier to see what sources were present; it did not decide whether the sources justified 15 points. Transparency increased after the decisive event, while the scope of remedy remained unchanged.

On 15 March 2018, the Board accepted the three reports, declared the review complete, concluded that no overhaul or change was required for the current application round, and directed the BAMC to resume the pending reconsideration requests. The Board relied on the review as evidence of process regularity. It did not commission a fresh evaluation panel, and FTI had not been asked to act as one.

This sequence is often described as a review of CPE, but that phrase obscures three different objects. Scope 1 reviewed ICANN-provider interaction. Scope 2 reviewed consistency across reports. Scope 3 reviewed the availability of research references. None reviewed the ultimate merits in the sense that mattered to the six applicants: whether the evidence, under the Guidebook, warranted at least 14 points. Calling all three exercises “review” does not give them the same jurisdiction.

January 2019 closed Request 16-11, not every question about CPE

After the process review, the requesters made a telephonic presentation to the BAMC on 19 July 2018 and submitted additional material. The BAMC recommendation of 16 November 2018 advised denial. The committee concluded that the Board had acted on accurate and complete information in August 2016, had considered the alleged unfair advantage connected to the portal configuration, and had not discriminated against the standard applicants.

The Board’s 27 January 2019 resolutions adopted that recommendation. The Board also considered a rebuttal even though the earlier Bylaws governing the request did not provide for one. It held that the requesters had not identified material information existing at the time of the 2016 decision that the Board failed to consider, or false or inaccurate material information on which the Board relied.

That decision did not retrospectively transform the process review into a merits appeal. Nor did it declare that every institutional concern identified by the Despegar panel was imaginary. It resolved the specific request under the applicable reconsideration standard. The Board’s question was whether its 2016 cancellation decision had been made on a materially defective record, not whether a newly convened evaluator might score the hotel community differently.

The sequence produced reasoned determinations, an independent declaration, a forensic inquiry, a three-scope process review, an oral presentation and a final Board vote. Those completed proceedings did not restore the six applications.

A later IRP ended without an award

The January 2019 Board decision did not end all formal contest. Fegistry and related claimants later opened another .HOTEL IRP. That docket ran through interim-measures applications, document-production disputes, amendments and a partial summary-adjudication decision. It therefore matters to the case history even though it did not produce a final declaration rewriting the 2014 score.

The ending is unusually precise. In October 2023 the claimants withdrew all of their claims. The panel’s termination order recorded that ICANN did not object, that there was no settlement, and that the IRP was dismissed with prejudice without entry of any award. The dismissal closed that proceeding; it did not affirm the CPE merits, invalidate the score, restore an application or award .HOTEL to anyone else.

This distinction protects the chronology from two opposite errors. The later IRP cannot be omitted as though the 2019 reconsideration decision was the last accountability event. Nor can its termination be described as a substantive victory for either side. Withdrawal with prejudice ended the available claims in that proceeding. It supplied finality without a merits declaration.

Priority was not a registry agreement, and a registry agreement is not delegation

The later record confirms the distinction drawn by the 2014 report’s disclaimer. ICANN’s registry agreement page for .hotel lists HOTEL Top-Level-Domain S.a.r.l. as operator and gives an agreement date of 6 June 2025. It classifies the agreement as a base, community, non-sponsored agreement. That is a contractual milestone established by a primary ICANN record, nearly eleven years after the CPE result.

At publication on 31 July 2026, .HOTEL does not appear in the IANA Root Zone Database. The current absence establishes that the string is not presently delegated in the DNS root; it does not by itself establish every earlier application status or explain the interval after contracting. The narrower verified account is enough: CPE priority ended the six rivals’ contention route in 2014, ICANN records a registry agreement in 2025, and the current root database does not yet show delegation.

These stages belong to different holders of control. The CPE provider controlled the evaluation fact. ICANN’s contention rules controlled elimination. ICANN and the applicant controlled execution of a registry agreement subject to programme requirements. Root-zone delegation requires the later IANA naming-function process. Collapsing those stages would exaggerate what the 15-point report conferred.

The 2026 rules added a challenge—but not for applicants displaced by a pass

The next-round design is no longer hypothetical. ICANN’s authoritative 2026 Round Applicant Guidebook, dated 24 April 2026, and the Community Priority Evaluation Guide and panel procedures, dated 18 June 2026, now prescribe a different first-instance architecture. They do not reopen .HOTEL, which remains governed by the 2012 rules. They do show which defects ICANN chose to address after the earlier round and which remedial boundary it retained.

The threshold is now 12 of 16 rather than 14 of 16. The allocation of points has also changed: community establishment carries six points, nexus four, registration policies two and community endorsement four. That redistribution prevents a mechanical conversion of the 2014 .HOTEL result. Its 4-3-4-4 score was produced under four equally weighted criteria and cannot simply be translated into a 2026 total. A new panel applying the current definitions and weights would have to evaluate the record afresh. The lower numerical threshold therefore does not by itself prove that the current test is easier in every case.

The provider structure is more explicit. Analysys Mason has been selected for the 2026 round. Each initial application is to be assessed by three independent senior experts, rather than treated as the work of an opaque unit. A Secretariat manages delivery, conflict review and process consistency. A Quality Assurance Board reviews whether deliberations are sufficiently documented and whether criteria are applied consistently across panels. Panelists participate in calibration sessions, and the Quality Assurance Board may transmit comments before the panel approves its final report.

These arrangements do not guarantee a correct score, but they create named quality-control functions at the point when correction is still comparatively cheap.

The evidential record is also more deliberately constructed. The current Guidebook allows limited independent research and consultation with community experts, but requires disclosure of that material with citations or links. The community applicant receives 30 days to respond before the decision. Clarifying questions may be directed both to the applicant and to a person or entity that filed a letter of opposition, with a 21-day response period. This gives opponents a defined route into fact clarification. It remains participation in an evaluation controlled by the panel, not a right to decide how evidence will be weighted.

Most importantly, the current Guidebook creates a CPE Evaluation Challenge. If a panel determines that a community application has not met the criteria, that applicant may file within 21 days of transmission of the determination. The CPE filing grounds are factual or procedural error. The broader challenge provision and the CPE remedy also direct the Challenge Panel to address a system error if one is found. A quick-look stage screens the filing.

The same provider administers the challenge, but the published procedures require a three-member Challenge Panel from which the original evaluators are excluded; the Guidebook describes use of a different set of panelists when practicable. If the Challenge Panel finds an error, it reevaluates the application with that finding in mind. If it finds none, the failed application continues to the next contention-resolution stage. ICANN states that no additional fee is charged, and the Evaluation Guide says the result is to be communicated within 30 days of filing.

This is more than the reconsideration route available in 2014. It places a criterion-specific error inquiry inside the evaluation system and gives a separate panel power to reevaluate the application. The Quality Assurance Board reviews the Challenge Panel’s documentation and consistency in the same way it reviews an initial report. The adopted mechanism therefore answers part of the institutional problem exposed by .HOTEL: an unsuccessful community applicant is no longer limited to arguing about ICANN’s procedural reliance on the provider while lacking a specialist evaluation remedy.

The standing rule, however, is outcome-asymmetric on the published text. The CPE-specific provision begins with a determination that the community application has not met the criteria and authorises that applicant to challenge the failure. The general Evaluation Challenge provisions likewise contemplate a filing by the applicant whose own evaluation is challenged and permit dismissal at quick look when the filer is not the applicant. The current CPE provisions do not grant a directly contending standard applicant a corresponding right to challenge another applicant’s successful score after that pass makes the standard application ineligible.

Applied to the .HOTEL pattern, the difference is concrete. Had HOTEL Top-Level-Domain failed under the current rules, it would have had 21 days to allege a qualifying error and seek reevaluation by a different panel without an additional fee. But the 2014 problem ran in the opposite direction: the community application passed, and six standard applicants were eliminated by the consequence attached to that pass. On the present text, those six would not fall within the CPE-specific filing rule for challenging a failed evaluation. They might still invoke a general ICANN accountability route applicable to a later ICANN action, subject to that route’s standing rule, standard of review and available remedy. That is not the same as a specialist right to have the passing CPE reevaluated.

The distinction between opposition and challenge is therefore decisive. The 2026 rules give a standard applicant or other opponent a better opportunity to place material before the first panel and to answer clarifying questions if asked. They do not transfer control of the evaluation, and the published CPE challenge provisions do not provide a post-decision specialist remedy to the standard applicant made ineligible by a pass. An applicant can participate in building the record yet remain outside the mechanism that corrects an erroneous CPE outcome.

The 2026 mechanism is adopted, but its operation in a contested CPE has yet to be demonstrated. The practical tests are how quick look defines eligible error, whether a Challenge Panel genuinely reassesses contested findings, how quality-assurance comments are recorded and whether the displaced-standard-applicant gap becomes consequential after a passing score.

What the six applicants lost—and what the community applicant did not yet gain

The six standard applicants lost their route through the .HOTEL contention set. They retained the ability to seek information, request reconsideration, invoke independent review and challenge later Board action. Those routes produced reasoned determinations, a forensic inquiry and a programme-level process review. They did not restore the applications or confer a right to a specialist reevaluation of the passing score.

HOTEL Top-Level-Domain gained priority under the Guidebook. It did not receive a judicial declaration of ownership, an immediate registry contract or a root-zone delegation. Its advantage was powerful but conditional: the exclusive ability among the seven applicants to continue towards later programme stages, subject to ICANN’s requirements and later decisions.

The current rules sharpen the institutional conclusion: a specialist reevaluation now exists for a failed community application, while the published standing rule still appears to place applicants exposed to a passing score’s exclusionary effect outside that remedy.

The 14-point gate was neither a private opinion nor a complete registry award. It was a delegated judgement connected to an automatic institutional consequence. That connection—not the length of the report, the prestige of the provider or the later volume of review—is what eliminated six .HOTEL applicants.