Companies planning to apply for “.brand” top-level domains still have concerns that ICANN’s new gTLD program does not adequately cater to their unique requirements.
ICANN has so far resisted calls from the likes of the Coalition for Online Accountability to create clearly delineated categories of gTLD, instead favoring the one-size-fits-all approach.
But one type of gTLD where the Applicant Guidebook has started to introduce exceptions to the rules is the so-called “.brand”.
In its latest draft, for example, the Guidebook’s Code of Conduct for vertically integrated registries/registrars does not apply to single-registrant TLDs such as .brands.
The Guidebook also makes it mostly clear that ICANN does not intend to re-assign .brands to different registry operators in the event that the brand decides to discontinue the TLD.
But those who are working with potential .brand applicants still have concerns.
The Guidebook currently calls for TLDs that are potentially confusing in meaning or appearance to be lumped into the same “contention sets” from which only one winner will emerge.
The worry is that this will capture companies with similar sounding brands. ECTA called for a mechanism to exclude .brands from these requirements:
The Draft Applicant Guidebook 6 does not take into account either co‐existence agreements or natural co‐existence. Currently a successful application from NBC in round one would preclude ABC or BBC or NBA in future years. Equally, should both EMI, the music company and ENI, the energy company apply, they would be placed in a Contention Set and could in theory face each other in an auction. In the real world these companies co‐exist.
It’s an interesting point, and not one that’s received a great deal of airplay in recent discussions.
There’s also the problem that companies with two-letter brands, such as HP or BP, are essentially banned from getting their .brand, because there’s a three-letter minimum on new TLDs.
Geographic name protections
The ICANN Governmental Advisory Committee has pushed hard for the protection of geographical terms at the second level in new gTLDs, and has won significant concessions.
One of the results of this is that if Canon, say, has .canon approved, it will be unable to immediately use usa.canon or japan.canon domains names – one of the most logical uses of a .brand.
ICANN plans to enable registries to loosen up these restrictions, but the Guidebook does not currently spell out how this will happen, which leaves a significant question mark over the value of a .brand.
ECTA wrote in its comments to ICANN:
This prohibition severely limits brand owners unnecessarily. On the contrary a .brand domain should provide clients with an intuitive replacement for ccTLDs. It would seem to be more logical if Internet users could replace www.mycompany.de with www.de.mycompany rather than having to type www.mycompany/de.
The BC has called for the Guidebook to be rephrased to made it clear that .brand TLDs should not have to offer their domains through a multitude of registrars on “non-discriminatory” terms.
The BC wants this language adding to the rules: “Single-Registrant TLDs may establish discriminatory criteria for registrars qualified to register names in the TLD.”
Given .brands will have essentially one customer, it would be a pretty crazy situation if more than one registrar was approved to sell them. It may be a hypothetical risk, but this is a strange industry.
All new gTLD registries will have to abide by the Uniform Dispute Resolution Process. The problem is that successful UDRP cases generally result in a domain name being transferred to the complainant.
This could result in a situation where a third-party trademark holder manages to win control a domain name in a competitor’s .brand TLD, which would be intolerable for any brand owner.
The BC suggests that domains won in this way should be allowed to be set to “reserved and non-resolving” instead of changing hands.
Demand Media and the Internet Commerce Association have called for ICANN to drop the “three strikes and you’re out” ban on applying for new top-level domains.
In the current version of ICANN’s Applicant Guidebook, if you’ve lost three UDRP cases in the last four years you’re considered a cybersquatter and effectively barred from applying for a new TLD.
It’s not entirely clear, but it is quite possible that this provision may capture Demand Media and Go Daddy, which, via subsidiary companies, have lost several UDRP complaints.
In comments filed with ICANN yesterday, Demand senior vice president Jeff Eckhaus said that a simple “three strikes” benchmark does not prove a pattern of cybersquatting:
losing a few contested UDRP cases in what amounts to a tiny percentage of their total domain name portfolio certainly doesn’t seem to constitute a “pattern” as most people would define the term
by all reasonable standards, it is difficult to conclude that an entity or an individual has engaged in a history/pattern of cybersquatting when they own hundreds or thousands of domain names and have lost a few UDRP or similar proceedings.
The ICA, which represents high-volume registrants, also has a problem with the rule. Principal Phil Corwin wrote ICANN:
We continue to believe that the “three strikes” criteria is too inflexible and that applicant evaluation criteria should take into account the total size of an applicant’s domain portfolio as well as the percentage of adverse UDRP decisions rendered against them in comparison to all UDRP proceedings they have been involved with.
Demand also argues that three strikes is “extremely broad standard that we believe will unintentionally disqualify otherwise qualified applicants.”
That strikes me as quite a weak argument, which could be equally applied to any of the background checks in the Guidebook. A murder conviction will also “disqualify otherwise qualified applicants”.
I’m not sure it’s “unintentional” in either case. If you work from the assumption that ICANN expects Demand and other speculators to successfully apply for new TLDs, it is. If you assume it’s designed to make their lives more difficult, it isn’t.
But Corwin noted in his comments that ICANN can waive the ban in “exceptional circumstances”, and said he suspects this could be used to allow large registrars to pass the background checks.
In any event, as Andrew Allemann has pointed out at Domain Name Wire, the way the Guidebook is phrased there may well be a loophole that would allow Demand and others to slip through.
Go Daddy, which DNW also reports could be affected by the rule, does not appear to have filed any comments on the latest Applicant Guidebook yet.
ICANN has threatened to terminate the domain name registrar EuroDNS for failing to transfer a typo domain lost in a UDRP case to Facebook.
But EuroDNS says it is subject to a court case in its home country, Luxembourg, which has prevented it handing over the name.
The original registrant of facebok.com lost a slam-dunk UDRP case back in September 2010. He didn’t even bother defending the case.
But over half a year later, he’s still in control of the domain, and he’s using it to recruit folk into a shady-looking (but probably legal) subscription text messaging service.
EuroDNS is the registrar of record for the domain, and like all registrars is responsible for transferring domains lost under the UDRP to the winning party, in this case Facebook.
ICANN’s compliance department – my guess is under pressure from Facebook – has therefore threatened EuroDNS with termination unless it hands over the domain in the next three weeks.
This is noteworthy because EuroDNS isn’t the kind of tiny, fringe outfit ICANN usually files compliance notices against. It’s a generally respectable business. It even shows up to ICANN meetings.
EuroDNS deputy general counsel Luc Seufer tells me that the company was fully prepared to transfer the domain – it had even sent the authorization codes to Facebook – but it found itself on the receiving end of a lawsuit claiming that the domain had been “stolen”.
Somebody in Luxembourg, it seems, has sued to reclaim an obvious typo domain that’s probably going to be transferred to Facebook anyway.
“We are therefore in an incredible position where if we transfer the name before the judge’s ruling we will be accountable in our own country and if we don’t transfer the name we are in breach of the [Registrar Accreditation Agreement],” said Seufer.
The Luxembourg case has not yet made it to court, hence EuroDNS’s delay, he said. ICANN is aware of the action, and has seen the court papers, he said.
According to ICANN’s breach notice (pdf), the only way for EuroDNS to avoid its obligation under the UDRP is to show proof that the original registrant has sued Facebook to keep the domain.
But the case in question was filed by a third party claiming to be the rightful owner of the domain, not the original registrant. EuroDNS seems to be trapped between a rock and a hard place.
Seufer said the company is prepared to hand over the domain, adding:
Should we simply ignore a judiciary court case against us in our own country – that could prevent us from operating the transfer since it is was asked of the judge – because of our RAA’s obligations?
The domain in question, facebok.com, currently redirects to a series of sites asking visitors to fill in a survey to win a Mac.
Those who are duped by it are actually signing up to a text service that costs, in the UK, £4.50 ($7.40) per week.
The National Arbitration Forum saw a steep increase in the number of cybersquatting complaints filed under the Uniform Dispute Resolution Policy last year.
According to a NAF announcement, 2,177 cases were filed in 2010, up 24% on the previous year.
That seems to be roughly in line with the experience of the World Intellectual Property Organization, which recently reported a 28% increase in UDRP complaints to 2,696 last year.
On that basis, it appears that WIPO has ever so slightly widened the market share gap between itself and NAF.
Between 1999 and the end of last year, NAF had handled 15,763 domain disputes, compared to WIPO’s over 20,000.
A basic UDRP filing covering a few domain names with a single panelist presiding costs about $1,500 with both providers, not including lawyers’ fees and other expenses.
With roughly 35,000 complaints filed to date, we can estimate that the revenue from UDRP flowing to WIPO and NAF together has been in the ball park of $50 million in slightly over 11 years.
Cybersquatting cases filed under the Uniform Dispute Resolution Policy have become less predictable, judging from complex new guidelines for adjudication panels.
The World Intellectual Property Organization has just published WIPO Overview 2.0, which sets out over 10 years of UDRP precedent for panelists to consider when deciding future cases.
The document is a must-read for domain investors and trademark holders.
Updated for the first time since 2005, it contains new sections covering developments such as registrar parking, automatically generated advertising and proxy/privacy services.
The Overview has quadrupled in length, from 5,000 to 20,000 words. With that, has come increased complexity. WIPO notes:
While predictability remains a key element of dispute resolution systems, neither this WIPO Overview nor prior panel decisions are binding on panelists, who will make their judgments in the particular circumstances of each individual proceeding.
The document reflects decisions already made, rather than creating new law, but as such it also reflects the tilting balance of the UDRP in favor of complainants.
For example, while the 2005 guidelines presented majority and minority views on whether [trademark]sucks.com domains meet the “confusing similarity” criterion, Overview 2.0 presents only a “consensus view” that they do, suggesting that it is now settled law.
On whether parking a domain with PPC ads meets the “legitimate interests” criterion, the guidelines refer to precedent saying that the ads must not capitalize on a trademark:
As an example of such permissible use, where domain names consisting of dictionary or common words or phrases support posted PPC links genuinely related to the generic meaning of the domain name at issue, this may be permissible and indeed consistent with recognized sources of rights or legitimate interests under the UDRP, provided there is no capitalization on trademark value
Supporting this view, the Overview states that “bad faith” can be shown even if the domain owner does not control the content of their parked pages and makes no money from the ads:
Panels have found that a domain name registrant will normally be deemed responsible for content appearing on a website at its domain name, even if such registrant may not be exercising direct control over such content – for example, in the case of advertising links appearing on an “automatically” generated basis… It may not be necessary for the registrant itself to have profited directly under such arrangement
There is a defense to this, if the respondent can show they had no knowledge of the complainant’s trademark and made no effort to control or profit from the ads.
Because the UDRP calls for “registration and use in bad faith”, the guidelines also ask: “Can bad faith be found if the disputed domain name was registered before the trademark was registered or before unregistered trademark rights were acquired?”
The original guidelines said no, with a carve-out for cases where the squatter anticipated, for example, a future corporate merger (microsoftgoogle.com) or product release (ipad4.com).
The new guidelines are a lot less clear, calling it a “developing area of UDRP jurisprudence”. The document lists several cases where panelists have chosen to essentially set aside the registration date and concentrate instead just on bad faith usage.
The question of whether a renewed domain counts as a new registration is also addressed, and also has a couple of exceptions to give panelists more flexibility in the decisions.
The Overview covers a lot of ground – 46 bullet points compared to 26 in the first version – and will no doubt prove invaluable reading for people filing or fighting UDRP cases.
The guidelines are not of course set in stone. The 2005 version read:
The UDRP does not operate on a strict doctrine of precedent. However, panels consider it desirable that their decisions are consistent with prior panel decisions dealing with similar fact situations. This ensures that the UDRP system operates in a fair, effective and predictable manner for all parties
But the new version adds a caveat to the end of the sentence: “while responding to the continuing evolution of the domain name system.”