Single/plural gTLD combos to be BANNED

Singular and plural versions of the same string will be banned at the top level under proposed rule changes for the next round of new gTLDs.

The final set of recommendations of ICANN’s New gTLDs Subsequent Procedures working group (SubPro), which were published after four years of development last week, state:

the Working Group recommends prohibiting plurals and singulars of the same word within the same language/script in order to reduce the risk of consumer confusion. For example, the TLDs .EXAMPLE and .EXAMPLES may not both be delegated because they are considered confusingly similar.

The 2012 round had no hard and fast rule about plurals. There were String Similarity Review and String Confusion Objection procedures, but they produced unpredictable results.

At least 15 single/plural string pairs currently exist in the root, including .fan(s), .accountant(s), .loan(s), .review(s) and .deal(s). Sometimes they’re both part of the same registry’s portfolio, other times they’re owned by competitors.

But others, including .pet and .pets and .sport and .sports, were ruled by independent panels too “confusingly similar” to be allowed to coexist.

The proposed new rule would remove much of the subjectivity from these kinds of decisions, replacing the current system of objections with a flat no-coexistence rule.

If a gTLD that was the plural of an existing gTLD were applied for, the application would be rejected. If the singular and plural variants of the same word were applied for in the same round, the applications would likely end up at auction.

But there would be some wriggle room, with the ban only applying if both applied-for strings truly are singular/plural variations of each other in the same language. The working group wrote:

.SPRING and .SPRINGS could both be allowed if one refers to the season and the other refers to elastic objects, because they are not singular and plural versions of the same word. However, if both are intended to be used in connection with the elastic object, then they will be placed into the same contention set. Similarly, if an existing TLD .SPRING is used in connection with the season and a new application for .SPRINGS is intended to be used in connection with elastic objects, the new application will not be automatically disqualified.

In such situations, both registries would have to agree to binding Public Interest Commitments to only use the gTLDs for their stated, non-conflicting purposes. Registrants would also have to commit to only use .spring to represent the season and .springs for the elastic objects, also.

The ban will substantially eliminate the problem I’ve previously referred to as “tailgating”, where a registry applies for the plural variant of a competitor’s successful, well-marketed gTLD, prices domains slightly lower, then sits back to effortlessly reap the benefits of their rival’s popularity.

One could easily imagine applicants for strings such as .clubs or .sites in the next round, with applicants content to lazily ride the coat-tails of the million-selling singular namespaces.

The rule change will also remove the need for existing registries to defensively apply for the single/plural variants of their current portfolio, and for existing registrants to be compelled to defensively registry domains in yet another TLD.

On the flipside, it means that some potentially useful strings would be forever banned from the DNS.

While it might make sense for a film producer to register a .movie domain to market a single movie, it would not make sense for a review site or movies-related blog, where a .movies domain would be more appropriate. But now that’s never going to be possible.

SubPro’s work is still subject to final approval by SubPro, the GNSO Council and ICANN board of directors before it becomes policy.

Plural gTLDs give ICANN huge credibility risk

Can .pet and .pets co-exist peacefully on the internet? Or would they create such confusion among internet users that the whole new gTLD program would look irresponsible and foolish?
That’s the question ICANN is due to face today, as constituents line up at the public forum in Beijing to question its board of directors about the problem of plurals in new gTLDs.
The Governmental Advisory Committee, the Business Constituency, the Intellectual Property Constituency and others have all openly questioned the sanity of allowing plurals in recent days.
Right now there are 59 collisions between singular and plural gTLD applications (in English, at least, according to my analysis), involving 23 unique string pairs.
These are: .accountant(s), .auto(s), career(s), .car(s), .coupon(s), .cruise(s), .deal(s), .fan(s), .game(s), .gift(s), .home(s), .hotel(s), .kid(s), .loan(s), .market(s), .new(s), .pet(s), .photo(s), .review(s), .sport(s), .tour(s), .web(s) and .work(s).
None of these singular/plural clashes are currently in contention sets with each other, meaning there’s nothing to stop them all being delegated by ICANN. We could have a .loan alongside a .loans a year from now.
It seems to be only common sense that these clashes will cause frequent confusion. I doubt many would pass the longstanding “shouted across a crowded bar” test for URL clarity.
Would you want to register a .photo domain if you knew .photos was also available, and vice versa? If you did, wouldn’t you also want to register the .photos equivalent, just in case?
That’s one of the things ICANN’s commercial stakeholders are worried about: 23 extra TLDs means 23 extra defensive registrations for every brand they want to protect.
But there’s also the risk that gTLD registries that are successful in this application round will feel obliged to apply for the plurals of their strings in future rounds for defensive purposes.
The plurals issue also highlights shortcomings in how the new gTLD program was structured.
Why is this happening?
Unless two companies applied for the exact same strings, there are only two ways they could end up in a contention set together.
The first way was if the String Similarity Panel decided that the two strings were too visually similar to be allowed to co-exist, and that didn’t happen in the case of plurals. The panel only decided two things in the end: that I and l are confusingly similar, and that rn and m are confusingly similar.
To date, nobody except the Panel and ICANN knows what the logic behind this decision was, but it appears to be based on a very narrow (though not unreasonable) interpretation of what constitutes visual similarity.
The second way to end up in a contention set was to file a successful String Confusion Objection, or to be on the receiving end of one.
But of the 33 such objections filed, only 11 were filed against plurals, covering only six new gTLD strings in total: .pets, .tours, .webs, .games, .cars, and .kids. There was also an objection to .tvs, due to a clash with the existing ccTLD .tv.
(UPDATE: it appears that only approximately half of the String Similarity Objections filed have actually been revealed to date).
The main reason there weren’t more objections is that only existing registries and new gTLD applicants had standing to file an objection. Nobody else was allowed to.
Applicants were of course disincentivized from filing objections. Winning a String Confusion Objection doesn’t kill off your rival if you’re an applicant, it merely places both applications in a contention set.
Being in a contention set means you’re going to have to pay money to get rid of your competitor, either by negotiating some kind of private deal or by punching it out at auction.
By not filing objections, applicants in singular/plural situations risk looking like they don’t care about user confusion or are blasé about forcing defensive registrations.
(And by defensive registrations, remember here we’re not only talking about trademark owners, we’re talking about every potential future registrant in those gTLDs.)
They do have the slight excuse that they were only given a week or so to file objections after the results of the String Similarity Panel’s deliberations, delayed several times, were revealed.
There’s also the possibility that some of the apparent clashes won’t be as big of a concern in the marketplace due to, for example, registration restrictions.
What happens next?
The GAC is almost certain to issue advice about plurals in the next day or two, having brought the topic up with ICANN’s board of directors earlier this week.
The Business Constituency is also expected to make a few proposals directly to the board during the Public Forum in Beijing, Thursday afternoon local time.
The BC is likely to suggest, for example, that if one String Similarity Objection decision finds that a plural and a singular are confusingly similar, then that ruling should apply to all plural clashes, even if no objection has been filed.
It’s an audacious idea: it would certainly do the trick, but it would require some severe goal-post moving by ICANN at a time when it’s already under fire for pulling last-minute stunts on applicants.
It would also risk capturing fringe cases of strings that look plural but, in the context they are used in everyday language, are not (such as .new and .news).
Without some kind of action, however, ICANN is pretty much guaranteed to attract negative publicity.
Looking like it’s the stooge of the domain name industry, forcing regular registrants to double-buy their domains to the enrichment of registries and registrars, could look bad.

After eight months, similarity review creates only TWO new gTLD contention sets

ICANN has finally delivered its String Similarity Panel’s review of all 1,930 original new gTLD applications, finding that only four applied-for strings are confusingly similar to others.
Two new contention sets have been created:

• .hotels (Booking.com B.V.) and .hoteis (Despegar Online SRL)
• .unicorn (Unicorn a.s.) and .unicom (China United Network Communications Corporation Limited)

Only one applicant in each contention set will survive; the strings may go to auction.
The list is bafflingly short given that the panel’s review was originally due in October and has been delayed several times since.
ICANN last month said it was forcing the panel to address “process” issues and heavily suggested that it was trying to make sure the review process was legally defensible, leading to speculation that the list was going to be much longer than expected.
But the panel and ICANN appear to have taken a super-strict approach to finding similarity instead.
The string similarity panel was tasked with deciding whether each applied-for string “so nearly resembles another visually that it is likely to deceive or cause confusion”.
I think the four strings included in its final report pretty conclusively pass that test.
Even the controversial Sword tool – the software algorithm ICANN commissioned to compare string similarity on objective grounds — agrees, scoring them very highly.
According to Sword, .unicorn and .unicom are 94% similar and .hotels. v .hoteis have a score of 99%. For comparison, .hotel versus .hotels produces a score of 81%.
It all seems nice and logical and uncontroversial. But.
All four affected applicants are applying for single-registrant gTLDs, in which the registry owns all of the second-level domains, drastically reducing the chance of abuse.
One of the reasons ICANN doesn’t want to create too-similar TLDs is that they could be exploited by phishers and other bad guys to rip off internet users, or worse.
But with single-registrant TLDs, that’s not really as big of an issue.
It will be interesting to see the affected applicants’ response to these latest findings. Expect complaints.
The results of the review will be a huge relief to most other applicants, which have been wondering since the list of gTLD applications was released last June what the final contention sets would look like.
The high standard that appears to have been used to find similarity may set some interesting precedents.
For example, we now know that plurals are fair game: if Donuts’ application for .dentist is approved in this round, there’s nothing stopping somebody else applying for .dentists in a future application round.
Also, brands with very short acronyms have little to fear from from being found too similar to other existing acronym dot-brands. The applicants for .ged and .gea were among those to express concern about this.
But the question of confusing similarity is not yet completely settled. There’s a second phase.
Applicants and existing TLD registries now have about a week to prepare and submit formal String Confusion Objections, kicking off an arbitration process that could create more precedent for future rounds.
Unlike the just-published visual similarity review, SCOs can take issue with similar meanings and sounds too.
Losing an SCO means your application is dumped into a contention set with the winner; if you lose against an existing TLD operator, your bid is scrapped entirely.

Anger as ICANN delays key new gTLDs milestone

New gTLD applicants could barely disguise their anger tonight, after learning that ICANN has delayed a key deliverable in the new gTLD program — originally due in October — until March.
On a webinar this evening, program manager Christine Willett told applicants that the string similarity analysis due on all 1,917 remaining bids is not expected to be ready until March 1.
The analysis, which will decide which “contention sets” applications are in — whether .hotel must fight it out with .hotels and .hoteis, for example — had already been delayed four times.
The reasons given for the latest delay were fuzzy, to put it mildly.
Willett said that ICANN has concerns about the “clarity and consistency of the process” being used by the evaluation panel — managed by InterConnect Communications and University College London.
Under some very assertive questioning by applicants — several of which branded the continued delays “unacceptable” — Willett said:

When you don’t have a consistent process, or there are questions about the process that is followed, it invariably would put into question the results that would come out of that process…
I don’t want to publish contention sets and string similarity results that I can’t stand behind, that ICANN cannot explain, and that only frustrate and potentially affect the forward progress of the program.

See? Fuzzy.
It sounded to some applicants rather like ICANN has seen some preliminary string similarity results that it wasn’t happy with, but Willett repeatedly said that this was not the case.
It’s also not clear whether the pricey yet derided Sword algorithm for determining string similarity has had any bearing on the hold-ups.
One of the reasons that applicants are so pissed at the latest delay is that it presents a very real risk of also delaying later stages of the evaluation, and thus time to market.
Willett admitted that the remaining steps of the program — such as objections and contention resolution — are reliant on the publication of string similarity results.
“I am quite confident we will have results on string similarity by March 1,” she said. “We need to publish contention sets — we need to publish string similarity results — by March 1 in order to maintain the timeline for the rest of the program.”
March 1 is worryingly close to the March 13 deadline for filing objections, including the String Confusion Objection, which can be used by applicants to attempt to pull others into their contention sets.
Just 12 days is a pretty tight deadline for drafting and filing an objection, raising the possibility that the objection deadline will be moved again — something intellectual property interests would no doubt welcome.
The IP community is already extremely irked — understandably — by the fact that they’re being asked to file objections before they even know if an application has passed its Initial Evaluation, and will no doubt jump on these latest developments as a reason to further extend the objection window. Some applicants may even agree.

Seventh new gTLD bid withdrawn

ICANN has now received seven requests to withdraw new gTLD applications, according to documentation published today.
While we learned today that Google and KSB AG are behind four of the junked bids, the identities of the other three are not yet known.
ICANN has said that it will not reveal the withdrawing applications until all the formalities, such as refunds, have been finalized.
The updated stats came in a slide deck (pdf) set to be used in an ICANN webinar scheduled for noon UTC today.
The slides also reveal the aggregate status of applications’ progress through Initial Evaluation.
As you can see from the slide below, over a quarter of applications have had their String Similarity Review already. Just 65 have had their Geographic Names Review, while 127 and 141 have had their technical and financial evaluations respectively.

ICANN also states that there have been 57 requests for changes to applications — up from 49 at the last count — and that so far nobody has filed a formal objection against any bid.

Is .city confusingly similar to .citi? UDRP says yes

In one of the more surprising twists to hit the new gTLD program, Citigroup has claimed that its proposed dot-brand gTLD, .citi, is not “confusingly similar” to the proposed generic gTLD .city.
The company appears to be trying to avoid getting into a contention set with the three commercial applicants for .city, which would likely put it into an expensive four-way auction.
It’s a surprising move because you’d expect a financial services company to want to at least try to mitigate the risk of future .city/.citi typo-based phishing attacks as much as possible.
Indeed, its .citi application states that the mission of the gTLD “is to further assist Applicant in accomplishing its mission of providing secure online banking and financial services”.
Nevertheless, the company is now arguing, in a few comments filed with ICANN today, this:

CITI and CITY are not so similar in an Internet context as to create a probability of user confusion if they are both delegated into the root zone. Thus, the .CITI application should not be placed into a contention set with the .CITY application.

The new Citigroup position is especially bewildering given that it has argued the exact opposite — and won — in at least two UDRP cases.
In the 2009 UDRP decision Citigroup Inc. v. Domain Deluxe c/o Domain Administrator, Citigroup contended that:

Respondent’s citywarrants.com domain name is confusingly similar to Complainant’s CITIWARRANTS mark.

The panelist in the case concluded that the Y variant of the name was merely a “mistyped variation” of and “substantively identical” to the Citigroup trademark.
A similar finding appears to have been handed down in Citigroup v Yongki, over the arguably generic citycard.com, but the decision is written in Korean so I can’t be certain.
The company’s current view, which I’m going to go out on a limb on and characterize as expedient, is that ICANN has delegated multiple ccTLDs that have only one character of variation in the past (it hasn’t — the ccTLDs it cites all pre-date ICANN) without causing confusion.
It also states in its comments that the meaning and proposed usage of the two strings is “very different” (which one commenter has already suggested is historically dubious).
So what’s going on here?
Is Citigroup really willing to risk potential phishing problems down the line to save a few measly bucks today? On the face of it, it looks that way.
If it is put in a contention set with the three .city applicants, it could wind up at auction against Donuts ($100m funding), TLD Registry Ltd (apparently backed by the Vision+ fund) and Directi.
Will Citigroup’s gambit pay off?
That’s down to a) the String Similarity Panel and b) whether any of the .city applicants tries to force the company into the contention set via a String Confusion Objection, which seems unlikely.

DI launches new gTLD application tracker with built-in string similarity checker

I’m excited to announce the launch of a comprehensive new gTLD application tracking service, featuring a unique built-in string similarity checker, right here on DI.
The service will provide the foundation for all of DI’s new gTLD program analysis over the coming months and years, and is designed to bring together all the best information about each application under one roof.
DI PRO subscribers can start playing with it now here.
All 1,930 applications can currently be searched and sorted by applicant, string, back-end registry provider, and status.

Users can also cross-reference applications in contention sets and read salient extracts from each application.
The gTLD application database will shortly be linked to the existing PROfile service, meaning DI PRO subscribers will have access to a database of over 3,000 domain name industry companies.
More features and bid-by-bid analysis will be added as the program progresses, but the feature I’m most excited about today is the string similarity checker, which is already built into every application profile.
This tool checks for visual and phonetic similarity with other applications, existing gTLDs and ccTLDs, as well as strings that are specially protected by the ICANN Applicant Guidebook.
Semantic similarity functionality will be added in the next few days.
Similarity is important for two reasons:
1) the String Similarity Panel, which will create new contention sets based on similar but not identical strings in a couple of months, and
2) the String Confusion Objection, which enables applicants to force rivals into the same contention set based on visual, aural or semantic similarity.
In testing, it’s already thrown up some possible future objections and contention sets that I had not previously considered, and early beta testers — applicants themselves — tell me they think it’s fantastic.
Here’s a screenshot from one of the .sex applications, to give you a taste.

Note that, unfortunately, the string similarity feature does not currently support the relatively small number of IDN string applications.
If you’re not already a DI PRO subscriber, you can sign up instantly here using PayPal. If you have any questions about the service, please email subs@domainincite.com.