Summary

  • The official .SHOP auction report supports a precise but bounded reconstruction. Nine applications were in the contention set, deposits were recorded for seven of them, and the resulting seven-bidder, fourteen-round clock rose from a first-round upper price of $1 million to a final band of $36.8 million to $46 million. GMO Registry, Inc. won through application 1-890-65213 at $41,501,000. The report does not identify bidders round by round, disclose individual exits, show the runner-up, reveal GMO’s maximum valuation or report final-round aggregate demand.
  • Under the Auction Rules, a bidder’s authority was real but narrow. A qualifying deposit set its bidding limit; the bidder could continue, place a proxy instruction or submit a final exit bid. An exit was ordinarily irreversible, and only a bidder that remained eligible could enter the next round. Power Auctions operated the clock, set increments and administered payment and default procedures under the rules. ICANN retained authority over programme eligibility, final interpretation of auction-rule disputes, contracting and the later governance of the proceeds.
  • Commercial Connect could seek judicial, reconsideration and Independent Review Process relief even though no deposit was recorded for its application. None supplied an operative stay before 27 January 2016. The federal court denied preliminary relief without deciding the ultimate fairness of the auction; the emergency IRP request was withdrawn; the Board Governance Committee rejected reconsideration; and the later IRP ended for non-payment of the claimant’s required cost share. Review access therefore did not become an enforceable remedy.
  • “Winning .SHOP” describes only the first of several institutional gates. The result was conditional on payment and continuing eligibility; ICANN and GMO signed the .shop Registry Agreement on 8 April 2016; the .SHOP root record is dated 5 May; and IANA’s 9 May delegation report documented applicant matching, contact and technical checks associated with the root-zone process. The $41,501,000 payment was not distributed among unsuccessful applicants. After costs, ICANN records that $40,470,230 entered its pooled, segregated auction-proceeds fund, over which the Board retained legal and fiduciary authority.

Seven bidders enter a clock built for nine applications

The most revealing document in the .SHOP allocation is not a press release about the final price. It is the official final auction report, whose first pages reduce a high-stakes contention set to a ladder of price intervals and headcounts.

Round one opened at $1 and closed at an upper clock price of $1 million. Seven bidders were eligible, but aggregate demand at the upper price was six. Round two began at $1 million and ended at $2 million; six bidders were eligible and four continued at the upper price. Four remained through rounds three to six. Demand fell from four to three at the end of round seven, then from three to two at the end of round nine. Two bidders were still eligible through round thirteen. Round fourteen opened at $36.8 million and set an upper clock price of $46 million. The auction cleared at $41,501,000.

The report’s public sequence is exact up to a point:

Round Clock-price range Eligible bidders at start Aggregate demand at upper price
1 $1–$1,000,000 7 6
2 $1,000,000–$2,000,000 6 4
3 $2,000,000–$3,000,000 4 4
4 $3,000,000–$4,000,000 4 4
5 $4,000,000–$5,400,000 4 4
6 $5,400,000–$7,200,000 4 4
7 $7,200,000–$9,600,000 4 3
8 $9,600,000–$12,000,000 3 3
9 $12,000,000–$15,000,000 3 2
10 $15,000,000–$18,800,000 2 2
11 $18,800,000–$23,500,000 2 2
12 $23,500,000–$29,400,000 2 2
13 $29,400,000–$36,800,000 2 2
14 $36,800,000–$46,000,000 2 Not disclosed

That last blank is institutionally important, but it was not an accidental omission in this report. The Auction Rules required ICANN to publish aggregate demand for each round except the final round. Nor did they require bidder identities to be published round by round. The report therefore allows a reader to see the contention set narrow without assigning an exit to Amazon, Beijing Jingdong, Charleston Road Registry, Dot Shop, DotShop, GMO or Sugar Maple. It establishes the clock’s progression, not the bidders’ private valuations.

Nine applications, two GMO files and one result that must not be conflated

The .SHOP report lists nine applications, not seven. The seven-bidder opening count resulted from a participation gate, not a smaller contention set.

Applicant Application number Auction deposit recorded
Amazon EU S.à r.l. 1-1317-37897 Yes
Beijing Jingdong 360 Du E-Commerce Ltd. 1-889-24496 Yes
Charleston Road Registry Inc. 1-1138-5993 Yes
Commercial Connect LLC 1-1830-1672 No
Dot Shop Limited 1-1176-45062 Yes
DotShop Inc. 1-1051-32260 Yes
GMO Registry, Inc. 1-890-65213 Yes
GMO Registry, Inc. 1-890-52063 No
Sugar Maple LLC 1-1632-57390 Yes

The two GMO application numbers carry different procedural histories. Application 1-890-52063 was a community application. Its Community Priority Evaluation report, dated 13 March 2015, awarded six of sixteen possible points. The passing threshold was fourteen. It therefore “did not prevail” in the priority evaluation and did not remove the other applications from contention.

Application 1-890-65213 was GMO’s standard application. GMO submitted the required deposit for it, entered it in the auction and won. That is not a technical distinction. Community Priority Evaluation and last-resort auction were different allocation mechanisms with different decision rules. The community application asked whether it met the Guidebook’s threshold for priority over competing applications. The standard application entered a price-based mechanism after the contention set remained unresolved.

Saying simply that “GMO’s .SHOP application failed CPE and then won the auction” compresses two applications into one and obscures what legal object moved through each stage.

The primary documents do not explain GMO’s internal decision to advance 1-890-65213 into the auction while no deposit was recorded for 1-890-52063. The failed evaluation establishes the community application’s status; the auction report establishes the deposit choices. Neither establishes motive. It would be speculation to say that GMO switched strategies for a particular commercial reason, abandoned the community theory because of a specific internal assessment or used the two applications as a coordinated bidding device. The record supports procedural sequence, not corporate intent.

This application-level precision matters because membership in a contention set confers no single bundle of powers. An application may remain in the set without winning community priority. Its applicant may invoke an accountability route yet choose not to bid; it may qualify to bid and later exit below the clearing price; and even an auction winner may remain unable to contract. Each status carries a different institutional consequence.

A deposit was the gate to the auction room

The 2012 Applicant Guidebook treated auction as a mechanism of last resort. Contending applicants were first expected to clear objections, evaluation and any applicable community-priority process, and they could seek voluntary resolution among themselves. Only an unresolved contention set moved to ICANN’s auction process. The auction programme page likewise explains that contention had to be resolved before an application could proceed to a Registry Agreement.

Presence in that unresolved set did not itself make an applicant an active bidder. The Auction Rules required a bidder agreement and a qualifying deposit. The deposit was not merely security against later default. It established bidding authority on the platform.

Under the rules, a deposit below $2 million generally produced a bidding limit equal to ten times the deposit. A deposit of at least $2 million produced an “Unlimited” bidding designation for the purposes of the auction. That label did not mean an applicant had infinite resources or was excused from paying the winning price. It meant the platform did not impose the same ten-times-deposit ceiling. The actual deposit amount and corresponding limit for each .SHOP participant were not published in the final report.

Deposits were placed in an Auction Bank Account and segregated by bidder. They enabled bidding but did not transfer control over the auction. A bidder could not dictate the round schedule, choose the next increment, demand that another bidder’s identity be disclosed or unilaterally suspend the event. The auction manager controlled the platform under ICANN’s rules, while ICANN controlled eligibility and the legal consequences attached to the result.

The report records deposits for seven applications. Commercial Connect’s application and GMO’s separate community application did not. The practical consequence was immediate: the auction opened with seven eligible bidders, not nine. The official announcement of the 27 January 2016 result therefore described seven eligible participants and identified Power Auctions as the entity that conducted the proceeding.

For Commercial Connect, the absence of a deposit did not erase its application from the contention set or prevent it from seeking review. It did mean that it had not acquired the platform-specific authority to submit a continue, proxy or exit bid. The Board Governance Committee later relied on that procedural fact when considering Commercial Connect’s requests to stop or undo the auction. The distinction should not be converted into a moral judgment. Not bidding may reflect litigation strategy, an objection to the process, a financing decision or another reason. The public record establishes only the choice and its procedural consequence.

The bidder’s actual powers: continue, delegate or leave

Within the auction, a participant had three principal instruction types. Each was meaningful, but none transferred control over the system.

A Continue Bid indicated willingness to remain active through the round’s upper clock price. The bidder could also set a higher Proxy Bid, authorising the system to carry its instruction forward automatically. If the proxy amount remained at or above a later round’s upper price, it functioned as a continue instruction. If a later round moved above the proxy amount, the system treated the proxy as an exit at the authorised value. The bidder could therefore delegate execution to the platform without disclosing its maximum to the public.

An Exit Bid was a final amount between the round’s lower and upper prices. It allowed a bidder to leave at a value more precise than the clock endpoint. Subject to the rules’ tie provisions, that exit was irrevocable. A bidder that exited could not re-enter merely because the field narrowed or because the next increment looked different from what it had expected.

Eligibility for a new round depended on continuing through the preceding one. This is the most important operational consequence of the clock. A bidder did not possess a standing right to jump back into the competition. The option to bid in round ten was earned by remaining active through round nine. As price rose, the field became not simply smaller but legally closed to those who had left.

The rules also addressed inaction. A valid bid left on the auction site at round close was binding. If an eligible bidder failed to submit a valid instruction and could not correct the problem under the prescribed process, the system could place a bid at the round’s starting price. That rule prevented silence from becoming an indefinite option to wait outside the clock while preserving full participation rights. It also made access, credentials and operational readiness part of bidding risk.

The published concept of aggregate demand needs care. It was not a measure of consumer demand for .SHOP, nor a count of every applicant that still wanted the string at some lower price. It was the number of Continue Bids at the round’s upper clock price. An eligible bidder could submit an Exit Bid within the round and therefore appear in the opening eligibility count but not in aggregate demand at the upper price. That is why round one could begin with seven eligible bidders and end with aggregate demand of six.

Power Auctions used the change in aggregate demand to help set later increments. The rules gave the Auction Manager discretion to determine the next price range in light of demand and other relevant price information. The increments visible in the .SHOP report show that this was not a fixed arithmetic progression. The clock accelerated as the field stabilised at high prices.

Why the public record cannot reverse-engineer $41,501,000

The Auction Rules describe two main clearing-price branches relevant when the field collapses in a round.

If exactly one bidder submits a Continue Bid at the round’s upper price, that bidder wins. The clearing price is determined by the highest Exit Bid submitted by another bidder in the round. If no bidder submits a Continue Bid, the bidder with the highest Exit Bid wins and the clearing price is generally the second-highest Exit Bid, subject to provisions dealing with ties and related contingencies.

Round fourteen began with two eligible bidders. Its lower price was $36.8 million and its upper price was $46 million. The published result was $41,501,000. But the final report does not disclose aggregate demand for that round. It therefore does not reveal whether one bidder continued to $46 million while the other exited at $41,501,000, or whether both submitted exit instructions and the second-highest exit set the price. Nor does it disclose whether the rules’ tie machinery was engaged.

This missing field blocks several common inferences. The winning price is not proof that GMO’s maximum valuation was $41,501,000. It is not proof that the losing finalist valued the string at precisely that amount. It does not identify a runner-up by name, reveal whether the winner had authorised a proxy substantially above the clearing price or disclose the sequence and revision history of the private instructions submitted during the final round.

A defensible account should therefore distinguish mechanism transparency from valuation transparency. Mechanism transparency means the rulebook, price intervals, bidder counts, aggregate-demand figures for earlier rounds, winner and clearing price were published. Valuation transparency would require individual deposits, bidding limits, proxy amounts, exit instructions and a complete audit log. The public record provides the first and withholds most of the second.

That boundary is not a reason to reject the result. It is a reason to state exactly what the result proves. ICANN’s report establishes the officially declared winner and price. The rules explain how such a price could be produced. The missing final-round demand and bidder log prevent a more granular claim about which branch actually operated.

Power Auctions ran the clock; ICANN kept the legal keys

The .SHOP auction distributed authority across a chain rather than vesting it in one decision-maker.

Power Auctions’ role was most visible during the event. It provided the platform, administered bidder access, processed continue, proxy and exit instructions, disclosed round information to participants and selected the next increment within the discretion allowed by the rules. It also administered the payment and default timetable. Participants depended on its system to record binding bids correctly. A platform error could therefore have immediate financial consequences.

ICANN’s role was broader. It defined the contention-resolution framework in the Guidebook, decided which applications were eligible for last-resort auction, appointed the auction provider, set the rules and reserved final and binding decision power over disputes concerning their interpretation or application. It also controlled what the result meant for the New gTLD Program. The auction announcement did not say GMO became the registry operator at the final bell. It said the application could proceed toward contracting subject to payment of the winning price and continuing eligibility.

The bidders controlled valuation decisions but not institutional progression. An applicant could choose an exit price, yet it could not convert a winning bid into a Registry Agreement by private declaration. It could pay, supply information and satisfy programme requirements; ICANN still had to contract. Nor could the winner instruct IANA to enter .SHOP in the root without the delegation process.

The public’s role was weaker still. Observers could read the rules and final report, compare price bands and later inspect the contract and delegation record. They could comment through ICANN’s broader policy processes and, depending on status and timing, invoke accountability mechanisms. They had no vote over the auction manager’s increment, no access to the live platform and no power to approve the winner.

This division reflects a recurring feature of multistakeholder governance: extensive participation can coexist with concentrated executable authority. Consultation, publication and review routes may influence how power is exercised, but they do not themselves determine the outcome. In .SHOP, the operative acts were performed by identifiable holders: bidders submitted binding instructions; Power Auctions ran the clock; ICANN declared and conditioned the result; ICANN and GMO signed the contract; IANA processed delegation; and the Board later controlled use of the pooled proceeds.

Review access did not create an automatic stay

Commercial Connect’s challenge is the clearest test of whether access to a court or accountability process could stop the clock merely by being invoked.

Commercial Connect’s application 1-1830-1672 remained in the .SHOP contention set, but the auction report records no deposit. The Board Governance Committee’s 25 February 2016 determination states that auction invitations were sent in October 2015 and that Commercial Connect did not submit the bidder agreement and deposit required to participate. It later sought relief through Reconsideration Requests 16-1 and 16-2, asking ICANN to suspend, undo or prevent consequences from the auction.

At the same time, Commercial Connect went to the United States District Court for the Western District of Kentucky. On 26 January 2016—the day before the auction—the court denied its motion for a preliminary injunction. That order mattered because it left no judicial command blocking the 27 January clock. It did not amount to a final trial judgment that every aspect of ICANN’s auction design was fair, lawful or correctly applied. Preliminary-injunction analysis concerns interim relief under a different posture and evidentiary record from final merits adjudication.

Commercial Connect also invoked the Independent Review Process. The emergency IRP docket records that the emergency-relief request was withdrawn and the emergency panelist’s appointment was terminated. There was therefore no emergency-panel order in force staying or reversing the auction.

The Board Governance Committee then rejected the reconsideration requests. Its determination emphasised that no operative court or IRP order required ICANN to halt the proceeding, that the requester had not completed the steps to bid and that filing a reconsideration request did not by itself suspend staff action. The BGC’s role was institutional review within ICANN, not an independent judicial rehearing of every contention issue. Its decision supplied no auction reversal.

Commercial Connect’s federal action later ended on procedural grounds. The court’s 28 April 2016 dismissal order dismissed the claims without prejudice after failure to perfect service and failure to respond to the court’s show-cause order. That disposition cannot be cited as a merits endorsement of the auction. It shows only that the case did not proceed to a final substantive judgment in that action.

The main IRP also failed to produce a merits declaration. Commercial Connect submitted the claim on 10 February 2016. A hearing was scheduled for October, then cancelled and the proceeding suspended when the claimant did not fund its required share. The panel’s 11 April 2017 termination order ended the case after further non-payment. The order did not decide whether the .SHOP auction was substantively fair and did not direct ICANN to reverse the result, cancel the contract or remove the TLD from the root.

The procedural chain therefore supports a narrow but consequential conclusion. Commercial Connect had access to several venues. It could file in federal court, request reconsideration and initiate IRP procedures. Those filings created no automatic stay. An effective pre-auction remedy would have required an actual injunction, emergency order or ICANN decision directing suspension. None existed when the clock opened.

This distinction protects against two opposite errors. One is to say that Commercial Connect had “no review” because it did not obtain relief. It plainly used review routes. The other is to say that the existence of those routes meant the auction should have paused automatically. The governing documents and observed outcome show otherwise. Review access was available; a binding remedy was not.

The auction result, the contract and delegation were three different statuses

ICANN’s 27 January 2016 announcement described GMO’s result conditionally. The application was eligible to proceed toward contracting only after timely payment of the winning amount and confirmation that it remained eligible under the programme. The Auction Rules generally required the winning bidder to pay the balance, after credit for its deposit, within twenty business days.

No standalone payment receipt, wire confirmation or exact payment date has been identified in the public source set for this article. That is a meaningful evidence limit. The official result, later proceeds accounting, executed contract and delegation record establish that the application progressed; they should not be represented as a publicly available copy of the payment instrument.

The next legal step occurred on 8 April 2016. ICANN’s Registry Agreement record identifies GMO Registry as the .shop registry operator and gives that date for the agreement. The executed agreement designated GMO as operator subject to the agreement’s requirements and to the approvals needed for delegation and root-zone entry.

Contracting changed the legal relationship. The auction had selected an application within a contention set. The Registry Agreement created ongoing rights and duties between ICANN and GMO: technical obligations, service commitments, coordination requirements, reporting and compliance exposure. The winner could no longer be understood only as a bidder. It was now a contracted registry operator—but still one awaiting completion of the root-zone process.

The IANA functions records supply that separate operational gate within the root-zone process then in force. The .SHOP root-zone entry identifies GMO Registry as the sponsoring organisation and records 5 May 2016 as the TLD registration date. The delegation report dated 9 May 2016 records that application 1-890-65213 matched the approved and contracted party, was eligible, had completed contact confirmations, satisfied technical-conformance checks and completed the other required processing.

The dates should therefore remain separate:

Institutional status Date What the record establishes
Conditional auction result 27 January 2016 Application 1-890-65213 won at $41,501,000, subject to payment and continuing eligibility
Registry Agreement 8 April 2016 ICANN and GMO established the contractual registry relationship, still subject to delegation and root-zone approvals
Root-zone registration record 5 May 2016 IANA’s root database records .SHOP and GMO as sponsoring organisation
Delegation report 9 May 2016 IANA documented applicant matching, eligibility, contacts, technical conformance and completion of processing
IRP termination 11 April 2017 The later accountability case ended for non-payment, without a merits declaration or remedial order

The $41.5 million did not become compensation for the losing field

The financial consequence of a last-resort auction differed from the consequence of a voluntary resolution among applicants. The Guidebook allowed contending applicants to resolve their conflict before auction. In that route, the applicants rather than ICANN’s auction mechanism determined the terms on which applications were withdrawn. Once the ICANN auction proceeded, however, the winning payment followed the programme’s proceeds framework.

ICANN’s auction-proceeds accounting page records the .SHOP transaction as follows:

Item Amount
Gross proceeds $41,501,000
Auction costs $1,030,020
Escrow costs $750
Total disclosed costs $1,030,770
Net proceeds $40,470,230

The unsuccessful applicants did not receive shares of that net amount under the ICANN auction. It entered a segregated New gTLD Auction Proceeds fund controlled by ICANN. That difference affects incentives. A voluntary resolution leaves applicants in control of the withdrawal terms. A last-resort auction requires the winner to pay ICANN while the losing applicants leave without a distribution from the clearing price.

The proceeds were also pooled. The $40,470,230 net contribution was not placed in a .SHOP account over which GMO, Commercial Connect or the other applicants retained direction. Nor did the payment create a donor relationship between GMO and later grant recipients. ICANN’s accounting could identify the string-specific contribution, but the governance process treated auction proceeds as a common fund.

The Guidebook had said that auction proceeds would be reserved and earmarked until their use was determined, and that any use had to remain consistent with ICANN’s mission, core values and non-profit status. The auction programme page repeated that the funds would remain reserved pending a community-informed Board decision. Those provisions postponed spending power; they did not transfer it to the bidders or create an automatic community appropriation.

Community recommendations shaped the proceeds policy; the Board retained decision power

The governance of the proceeds took much longer than the allocation of .SHOP. The clock cleared in January 2016. A cross-community working group began designing recommendations for the pooled money, and the ICANN Board made the key mechanism choice years later.

The Board’s 12 June 2022 resolutions describe the Cross-Community Working Group on New gTLD Auction Proceeds as a process chartered across ICANN’s supporting organisations and advisory committees. The community developed recommendations about objectives, safeguards, conflicts, evaluation and administration. At that stage, participation was broad and substantively important. It created the design record on which the Board acted.

But the working group did not possess final legal authority over the money. The Board retained fiduciary responsibility for ICANN’s assets and responsibility for ensuring that grants complied with the organisation’s mission and governing documents. In 2022 it selected Mechanism A, under which ICANN would design and administer the programme itself rather than transferring complete control to an external foundation or independent body.

The Board’s revised scorecard later made that allocation more concrete. ICANN would remain responsible across the grant lifecycle, although it could use external expertise and service providers. Programme goals, eligibility, evaluation, contracting, monitoring and compliance would operate inside ICANN’s governance constraints.

The first cycle demonstrates how that authority was exercised. The Board’s 26 January 2025 resolution records that the application window ran from 25 March to 24 May 2024 and attracted 247 applications. ICANN organisation and an Independent Application Assessment Panel, supported through an external service provider, produced a final slate. The Board approved that slate within an aggregate ceiling of $10 million and authorised the organisation to enter grant negotiations, subject to continuing eligibility and execution of a grant agreement.

Board approval was not the same as payment. A project still had to satisfy the programme’s conditions and execute a grant agreement. Nor does the resolution show the Board independently rescoring all 247 proposals. It records instead that the Board accepted ICANN organisation’s representation that the prescribed evaluation process had been followed, approved the slate and authorised negotiations under its mission, risk and fiduciary responsibilities.

The framework also narrowed review of individual grant decisions. The Board approved amendments to Bylaws Sections 4.2 and 4.3 on 10 November 2024; the Empowered Community approved them on 9 January 2025; and the amended provisions entered into force. The 26 January 2025 resolution then adopted the updated recommendation that Reconsideration and the Independent Review Process could not be used to challenge a decision to approve or not approve an individual grant application. That design protects the programme from repeated internal appeals over comparative funding judgments. It also means that participation in the application process and visibility into the framework do not carry those ICANN accountability remedies against an adverse individual selection decision.

On 29 May 2025, ICANN announced the first cohort of 23 grant projects, with listed awards totalling $9,018,746.71. The announcement supplied an operational milestone for the pooled proceeds programme. It did not trace any recipient or dollar specifically to the .SHOP contribution.

That evidentiary boundary should remain firm. The .SHOP auction unquestionably added $40,470,230 net to the fund. The 2025 programme unquestionably approved and announced a first cohort. What cannot be established is a one-to-one path from GMO’s payment to a named project. Pooling was a governance choice precisely because the proceeds were to be administered as a common resource.

What the .SHOP chain actually proves

The .SHOP case supports neither the claim that bidders were powerless nor the claim that the highest bidder simply bought a root-zone label.

The seven participating bidders exercised consequential choice. Each could set its bidding authority through a deposit, use continue or proxy instructions and submit an exit bid. Those choices narrowed the field and produced the clearing price. Once a bidder exited, its loss of eligibility was generally irreversible. The auction therefore depended on private judgment under financial pressure.

Power Auctions exercised operational control. It ran the non-public platform, processed binding instructions and selected clock increments within the rules. Its discretion affected pace and the information environment, but it did not carry authority to contract, delegate or spend proceeds.

ICANN retained the broadest institutional control. It defined eligibility, approved the auction framework, declared the result, required payment, determined whether the winner could contract, executed the Registry Agreement and received the proceeds. Its accountability bodies could review certain actions, yet no filing automatically paused the auction. Later, the Board retained fiduciary authority over the pooled fund and chose the internally administered grant mechanism.

The IANA functions process supplied another distinct gate in the operational chain. Its report documented that the contracted applicant, contacts and technical readiness had been checked as part of the root-zone process. Winning and contracting were necessary but not sufficient conditions for root-zone entry.

The public received meaningful but incomplete transparency. The round ladder, price bands, eligible counts, aggregate demand through round thirteen, winner, contract and delegation records are public. Bidder identities at each exit, individual instructions, deposit amounts, proxy limits, final-round demand and the winner’s maximum valuation are not. That is enough to understand the architecture and official outcome, but not enough to reconstruct the private competition.

The counterfactuals expose the power map. If auction victory itself conferred registry control, the institutional story would end on 27 January 2016. It did not. If filing reconsideration, litigation or IRP automatically stayed an auction, the clock could not have proceeded without disposing of every filing first. It did. If community recommendations controlled the proceeds, the working group rather than the Board would have selected the administration mechanism and assumed fiduciary responsibility. It did not.

The defensible conclusion is narrower and more useful. The .SHOP auction was transparent at the level of published rules and aggregate operation, not at the level of bidder valuation. The challengers had procedural access, not an automatic remedy. GMO obtained a conditional allocation, then a contract, then delegation after separate checks. The payment became an ICANN asset subject to pooled governance, not compensation for the losing applicants or a grant fund directed by the winner. The clock allocated a contention set; it did not collapse the institutions around it.