Crackdown looms for new gTLD auction gaming

ICANN will be urged to consider taking a stronger position against companies who apply for new gTLDs simply to lose them at auction or immediately flip them to others.

A community working group, known as SubPro and tasked with developing rules for the next new gTLD round, delivered its final Final Report this week, and the one area that failed to gain a designation of “consensus” or stronger was private auctions.

In the 2012 application round, several companies applied for large portfolios of strings that look — in hindsight at least — like efforts to game the system by forcing rivals to auctions they planned to deliberately use.

Companies such as MMX made millions losing auctions during the round, some of which was reinvested in winning auctions for other TLDs.

Applicant Nu Dot Co was notable for losing every private auction it participated in, then quickly flipping its successful .web application when Verisign stepped up with a $135 million bankroll.

While it’s difficult to know the extent to which this was all planned in advance, it proved the business model — filing spurious applications for new gTLDs you have no intention of launching — could be lucrative in future rounds.

But SubPro has put forward a slew of recommendations that, should they pass the remaining hurdles of the policy development track, could bring in substantial sanctions for those applicants and registries found to be gaming the system.

The SubPro recommendations are heavily buttressed with square parentheses, indicating disputed text, and supplemented by some minority statements from members of the working group who think that private auctions should be banned outright in future application rounds.

But the headline recommendation, numbered 35.3, is this:

Applications must be submitted with a bona fide (“good faith”) intention to operate the gTLD. Applicants must affirmatively attest to a bona fide intention to operate the gTLD clause for all applications that they submit.

Far from merely providing a check-box assertion that they’re legit, which would itself be easily gamed, applicants would also find their applications scrutinized by ICANN and its external evaluators to check for signs of a lack of bona fides.

Factors used to determine shadiness could include how many applications for contested strings are applied for, how many private auctions are lost, whether the successful applicant has not launched its gTLD within two years, and whether contracts are flipped within the first year.

SubPro discussed penalties for gaming could include the loss of registry contracts, a ban from future rounds or straight-up monetary fines. But the group did not put forward any recommendations.

SubPro couldn’t seem to come to agreement on most of this. The recommendations were determined to have “strong support but significant opposition” during the group’s recent consensus call.

One strong objection came from a somewhat diverse group of SubPro participants comprising Alan Greenberg (At-Large), Christopher Wilkinson (At-Large), Elaine Pruis (Verisign), George Sadowsky (Afilias/ISOC), Jessica Hooper (Verisign), Jim Prendergast (consultant), Jorge Cancio (Swiss government, but signed in a personal capacity) and Kathryn Kleiman (non-commercial users). They said:

The recommendations in the final report are a mix of overly complex disclosures and attestations that needlessly complicate the program to allow for private auctions. And they will not work. The only way to prevent a repeat of the activity from the 2012 round is to ban private auctions

They also claimed that allowing private auctions would putter smaller, niche and community applicants at a disadvantage, and that ICANN’s reputation would be harmed if it was seen to be overseeing gaming.

The At-Large Advisory Committee also issued a strong objection to private auctions along the same lines:

We remain concerned about attempts to “game” the application process through use of private auction and share the ICANN Board’s concerns on the consequences of shuffling of funds between private auctions. The ability for a loser to apply proceeds from one private auction to fund their other private auctions only really benefits incumbent registry operators or multiple-string applicants and clearly disadvantages single-TLD/niche applicants. We believe there should be a ban on private auctions, and that by mandating ICANN only auctions, the proceeds of ICANN auction can be directed for uses in public interest

The assumption there of course is that an ICANN “last resort” auction, in which the winning bid is funneled into ICANN’s cash pile, would be spent on stuff genuinely in the public interest, rather than frittered away on secretly settling employee lawsuits or indulging in more expensive, self-important navel-gazing.

Perhaps unsurprisingly, the ICANN board of directors has indicated that it prefers the idea of last resort auctions to private auctions.

But SubPro has also made some recommendations that could potentially keep the price of last-resort bids down, completely redesigning the auction process compared to the 2012 round.

If the recommendations are implemented, applicants would have to submit bids towards the start of the application process, when they don’t even know who they’re bidding against.

After all the applications have been submitted, ICANN evaluators would group them all according to whether they’re identical or confusingly similar to each other, then inform each applicant in a contention set how many bidders — but not their identities — they’re up against.

Applicants would then have to submit a sealed bid stating the maximum price they’d be willing to pay for the gTLD in question. It would be only after “reveal day”, when ICANN publishes the applications themselves, that everyone would learn who they’re bidding against.

They’d then be able to engage in private resolutions (auctions could come into play at this point), but it would only be after contention resolution phases such as objections and Community Priority Evaluations were complete that applicants would find out who’d submitted the highest bid.

The winning bidder would pay the amount of the second-highest bid to ICANN.

The 400-page final report (pdf), along with the minority statements, will now be sent to the GNSO Council for approval, before it makes its way to the ICANN board.

Given how much work remains to be done on private auctions and other issues that I’ll get to in later coverage, it seems that a lot of the mechanics of how contention resolution will work will have to be devised by ICANN and the community during the Implementation Review Team and Operation Design Phase phases, along with at least one round of commentary on at least one edition of the next Applicant Guidebook.

The next round of new gTLDs has moved a step closer, but it’s still going to be well over a decade after the last application window before we see the next one.

Tweet

Tagged: auctions, contention, gaming, ICANN, last resort, new gTLDs, private auctions