No .web until 2021 after Afilias files ICANN appeal

Afilias has taken ICANN to arbitration to prevent .web being delegated to Verisign.
The company, which came second in the $135 million auction that Verisign won in 2016, filed Independent Review Process documents in late November.
The upshot of the filing is that .web, considered by many the best potential competitor for .com — Afilias describes it as “crown jewels of the New gTLD Program” — is very probably not going to hit the market for at least a couple more years.
Afilias says in in its filing that:

ICANN is enabling VeriSign to acquire the .WEB gTLD, the next closest competitor to VeriSign’s monopoly, and in so doing has eviscerated one of the central pillars of the New gTLD Program: to introduce and promote competition in the Internet namespace in order to break VeriSign’s monopoly

Its beef is that Verisign acquired the rights to .web by hiding behind a third-party proxy, Nu Dot Co, the shell corporation linked to the co-founders of .CO Internet that appears to have been set up in 2012 purely to make money by losing new gTLD auctions.
Afilias says NDC broke the rules of the new gTLD program by failing to notify ICANN that it had made an agreement with Verisign to sign over its rights to .web in advance of the auction.
The company says that NDC’s “obligation to immediately assign .WEB to VeriSign fundamentally changed the nature of NDC’s application” and that ICANN and the other .web applicants should have been told.
NDC’s application had stated that .web was going to compete with .com, and Verisign’s acquisition of the contract would make that claim false, Afilias says.
This means ICANN broke its bylaws commitment to apply its policies, “neutrally, objectively, and fairly”, Afilias claims.
Allowing Verisign to acquire its most significant potential competitor also breaks ICANN’s commitment to introduce competition to the gTLD market, the company reckons.
It will be up to a three-person panel of retired judges to decide whether these claims holds water.
The IRP filing was not unexpected. I noted that it seemed likely after a court threw out a Donuts lawsuit against ICANN which attempted to overturn the auction result for pretty much the same reasons.
The judge in that case ruled that new gTLD applicants’ covenant not to sue ICANN was valid, largely because alternatives such as IRP are available.
ICANN has a recent track record of performing poorly under IRP scrutiny, but this case is by no means a slam-dunk for Afilias.
ICANN could argue that the .web case was not unique, for starters.
The .blog contention set was won by an affiliate of WordPress maker Automattic under almost identical circumstances earlier in 2016, with Colombian-linked applicant Primer Nivel paying $19 million at private auction, secretly bankrolled by WordPress.
Nobody complained about that outcome, probably because it was a private auction so all the other .blog applicants got an even split of the winning bid.
Afilias wants the .web IRP panel to declare NDC’s bid invalid and award .web to Afilias at its final bid price.
For those champing at the bit to register .web domains, and there are some, the filing means they’ve likely got another couple years to wait.
I’ve never known an IRP to take under a year to complete, from filing to final declaration. We’re likely looking at something closer to 18 months.
Even after the declaration, we’d be looking at more months for ICANN’s board to figure out how to implement the decision, and more months still for the implementation itself.
Barring further appeals, I’d say it’s very unlikely .web will start being sold until 2021 at the very earliest, assuming the winning registry is actually motivated to bring it to market as quickly as possible.
The IRP is no skin off Verisign’s nose, of course. Its acquisition of .web was, in my opinion, more about restricting competition than expanding its revenue streams, so a delay simply plays into its hands.

ICANN urged to reject .com price increases

The Internet Commerce Association has asked ICANN to refuse to allow Verisign to raise its wholesale prices for .com domain names.
The domainer trade group wrote to ICANN last week to point out that just because the Trump administration has dropped the US government objection to controlled price increases, that doesn’t necessarily mean ICANN has to agree.
Verisign’s deal with the National Telecommunications and Information Administration “does not of course, compel ICANN to agree to any such increases. Any such decision regarding .com pricing
remains with ICANN” ICA general counsel Zak Muscovitch wrote.
The deal allows Verisign to increase the price of .com registrations, renewals and transfers by 7% per year in four of the next six years, leading to a compound 30% increase by the time it concludes.
The arguments put forth Muscovitch’s letter are pretty much the same as the arguments ICA made when it was lobbying NTIA to maintain the price freeze.
Namely: Verisign already makes a tonne of money from .com, it has a captive audience, it cannot claim credit for .com’s success, and .com is not constrained by competition.
“As NTIA makes clear, it is up to Verisign to request a fee increase and ICANN that may agree or disagree. ICANN should not agree. Indeed, it would be a dereliction of ICANN’s responsibilities to the ICANN community if Verisign were permitted to raise its fees when it is already very well paid for the services which it provides,” Muscovitch’s letter (pdf) concludes.
For many years ICANN has been reluctant to get involved in price regulation. It remains to be seen whether it will make an exception for .com.

Hebrew .com off to a slow start

The Hebrew transliteration of .com has only sold a couple hundred domains since it went into general availability.
Verisign took the new gTLD קום. (Hebrew is a right-to-left script, so the dot comes after the string) to market November 5, when it had about 3,200 domains in its zone file. It now stands around the 3,400 mark.
The pre-GA domains are a combination of a few hundred sunrise regs and a few thousand exact-match .coms that were grandfathered in during a special registration period.
It’s not a stellar performance out of the gates, but Hebrew is not a widely-spoken language and most of its speakers are also very familiar with the Latin script.
There are between seven and nine million Hebrew speakers in the world, according to Wikipedia. It doesn’t make the top 100 languages in the world.
The ccTLD for Israel, where most of these speakers live, reports that it currently has 246,795 .il domains under management. That’s a middling amount when compared to similarly sized countries such as Serbia (about 100,000 names) and Switzerland (over 2 million).
Verisign’s original application for this transliteration had to be corrected, from קום. to קוֹם. If you can tell the difference, you have better eyesight than me.
In the root, the gTLD is Punycoded as .xn--9dbq2a.

Wagner takes dig at Verisign as GoDaddy reports $310 million domain revenue

GoDaddy CEO Scott Wagner ducked a question about how the company will react to future .com price increases during its third-quarter earnings call yesterday, but used the opportunity to take a gentle swipe at Verisign.
Asked by an analyst whether the first 7% price increase, almost certainly coming in 2020, would have any effect on GoDaddy’s gross margins (ie, will they shrink as the company swallows increased costs, or swell as it increases its own prices above 7%), Wagner said:

the last time VeriSign took a price increase the industry passed that through to the end registrant.
.com and more importantly the software around bringing somebody’s .com to life is valuable and, modestly, we’re providing the value in that relationship around taking a domain name and actually turning it into something that somebody cares about.

I’m interpreting that as a pop at the idea that Verisign enjoys the fat registration margins while GoDaddy is the one that actually has to market domains, up-sell, innovate, deal with customers, and so on.
The remarks came just a few days after Verisign, in a blog post, branded GoDaddy and other secondary-market players “scalpers”, infuriating domainers.
Wagner was talking to analysts as the market-leading registrar reported revenue for Q3 of $679.5 million, up 16.7% year over year.
Revenue from domains, still the biggest of its three reporting business segments, was $309.5 million, up 14.0% compared to the year-ago quarter. GoDaddy now has 18 million customers and over 77 million domains under management.
Overall net income was down to $13.2 million from $22.4 million, as operating expenses rose over 16% to hit $642 million, after the company invested more in marketing, development and so on. Its operating income was $37.5 million.
Contrast this with Verisign’s performance for the same quarter, reported two weeks ago.
It saw revenue about the same as GoDaddy’s domains revenue — $306 million — but net income of $138 million and operating income of $195 million.
GoDaddy and Verisign could find themselves competing before long. As part of its deal with the US government to allow it to raise .com wholesale prices once more, the government also lifted its objection to Verisign becoming a registrar, just as long as it does not deal in .com names.

Will ICANN take a bigger slice of the .com pie, or will .domainers get URS?

Will ICANN try to get its paws on some of Verisign’s .com windfall? Or might domainers get a second slap in the face by seeing URS imposed in .com?
With Verisign set to receive hundreds of millions of extra dollars due to the imminent lifting of .com price caps, it’s been suggested that ICANN may also financially benefit from the arrangement.
In a couple of blog posts Friday, filthy domain scalper Andrew Allemann said that ICANN will likely demand higher fees from Verisign in the new .com registry agreement.
Will it though? I guess it’s not impossible, but I wouldn’t say it’s a certainty by any means.
Verisign currently pays ICANN $0.25 per transaction, the same as almost all other gTLDs. Technically, there’s no reason this could not be renegotiated.
Putting aside some of the legacy gTLD contracts, I can only think of two significant cases of ICANN imposing higher fees on a registry.
The first was .xxx, which was signed in 2011. That called for ICM Registry, now part of MMX, to pay $2 per transaction, eight times the norm.
The rationale for this was that ICANN thought (or at least said it thought) that .xxx was going to be a legal and compliance minefield. It said it envisaged higher costs for overseeing the then-controversial TLD.
There was a school of thought that ICANN was just interested in opportunistically boosting its own coffers, given that ICM was due to charge over $60 per domain per year — at the time a ludicrously high amount.
But risk largely failed to materialize, and the two parties last year renegotiated the fees down to $0.25.
The second instance was .sucks, another controversial TLD. In that case, ICANN charged registry Vox Populi a $100,000 upfront fee and per-transaction fees of $1 per domain for the first 900,000 transactions, four times more than the norm.
While some saw this as a repeat of the .xxx legal arse-covering tactic, ICANN said it was actually in place to recoup a bunch of money that Vox Pop owner Momentous still owed when it let a bunch of its drop-catch registrars go out of business a couple years earlier.
While the .sucks example clearly doesn’t apply to Verisign, one could make the case that the .xxx example might.
It’s possible, I guess, that ICANN could make the case that Verisign’s newly regained ability to raise prices opens it up to litigation risk — something I reckon is certainly true — and that it needs to increase its fees to cover that risk.
It might be tempting. ICANN has a bit of a budget crunch at the moment, and a bottomless cash pit like Verisign would be an easy source of funds. A transaction fee increase of four cents would have been enough to cover the $5 million budget shortfall it had to deal with earlier this year.
On the other hand, it could be argued that ICANN demanding more money from Verisign would unlevel the playing field, inviting endless litigation from Verisign itself.
ICANN’s track record with legacy gTLDs has been to reduce, rather than increase, their transaction fees.
Pre-2012 gTLDs such as .mobi, .jobs, .cat and .travel have all seen their fees reduced to the $0.25 baseline in recent years, sometimes from as high as $2.
In each of these cases, the registries concerned had to adopt many provisions of the standard 2012 new gTLD registry agreement including, controversially, the Uniform Rapid Suspension service.
Domainers hate the URS, which gives trademark owners greater powers to take away their domains, and the Internet Commerce Association (under the previous stewardship of general counsel Phil Corwin, since hired by Verisign) unsuccessfully fought against URS being added to .mobi et al over the last several years, on the basis that eventually it could worm its way into .com.
I’m not suggesting for a moment that ICANN might reduce Verisign’s fees, but what if URS is the price the registry has to pay for its massive .com windfall?
It’s not as if Verisign has any love for domainers, despite the substantial contribution they make to its top line.
Since the NTIA deal was announced, it’s already calling them “scalpers” and driving them crazy.
ICA lost the .com price freeze fight last week, could it also be about to lose the URS fight?

Trump gives Verisign almost $1 billion in free money

The Trump administration may have just handed Verisign close to $1 billion in free money.
That’s according to the back of the envelope I’m looking at right now, following the announcement that the National Telecommunications and Information Administration is reinstating Verisign’s right to increase .com registry fees.
As you may have read elsewhere already (I was off sick last week, sorry about that) a new amendment to the Verisign-NTIA Cooperative Agreement restores Verisign’s ability to raise prices by 7% per year in four of the six years of the deal.
The removal of the Obama-era price freeze still needs to be incorporated into Verisign’s ICANN contract, but it’s hard to imagine ICANN, which is generally loathe to get into pricing regulation, declining to take its lead from NTIA.
Verisign would also have to choose to exercise its option to increase prices in each of the four years. I think the probability of this happening is 1 in 1.
Layering this and a bunch of other assumptions into a spreadsheet, I’m coming up with a figure of roughly an extra $920 million that Verisign will get to add to its top line over the next six years.
Again, this isn’t an in-depth study. Just back-of-the-envelope stuff. I’ll talk you through my thinking.
Not counting its occasional promotions, Verisign currently makes $7.85 for every year that a .com domain is added or renewed, and for every inter-registrar transfer.
In 2017, .com saw 40.89 million add-years, 84.64 million renew-years and 3.79 million transfers, according to official registry reports.
This all adds up to 129,334,643 revenue events for Verisign, or just a tad over $1 billion at $7.85 a pop.
Over the four-year period of the price increases transaction fees will go up to $8.40, then $8.99, then $9.62, then $10.29. I’m rounding up to the nearest penny here, it’s possible Verisign may round down.
If we assume zero transaction growth, that’s already an extra $762.2 million into Verisign’s coffers over the period of the contract.
But the number of transactions inevitably grows each year — more new domains are added, and some percentage of them renew.
Between 2016 and 2017, transaction growth was 3.16%.
If we assume the same growth each year for the next six years, the difference between Verisign’s total revenue at $7.85 and at the new pricing comes to $920 million.
Verisign doesn’t have to do anything for this extra cash, it just gets it.
Indeed, the new NTIA deal is actually less restrictive on the company. It allows Verisign to acquire or start up an ICANN-accredited gTLD registrar, something it is currently banned from doing, just as long as that registrar does not sell .com domains.
Verisign’s .net contract also currently bans the company from owning more than 15% of a registrar, so presumably that agreement would also need to be amended in order for Verisign to get into the registrar business.
I say again that my math here is speculative; I’m a blogger, not a financial analyst. There may be some incorrect assumptions — I’ve not accounted for promotions at all, for example, and the 3.16% growth assumption might not be fair — and there are of course many variables that could move the needle.
But the financial markets know a sweetheart deal when they see one, and Verisign’s share price went up 17.2% following the news, reportedly reaching heights not seen since since the dwindling days of the dot-com bubble 18 years ago.
The reason given for the lifting of the price freeze was, for want of a better word, bullshit. From the NTIA’s amendment:

In recognition that ccTLDs, new gTLDs, and the use of social media have created a more dynamic DNS marketplace, the parties agree that the yearly price for the registration and renewal of domain names in the .com registry may be changed

Huh?
This seems to imply that Verisign has somehow been disproportionately harmed by the rise of social media, the appearance of new gTLDs and some unspecified change in the ccTLD marketplace.
While it’s almost certainly true that .net has taken a whack due to competition from new gTLDs, and that the domain marketplace overall may have been diminished by many small businesses spurning domains by choosing to set up shop on, say, Facebook, .com is still a growing money-printing machine with some of the fattest margins seen anywhere in the business world and about a 40% global market share.
If the Trump administration’s goal here is to make some kind of ideological statement about free markets, then why not just lift the price caps altogether? Give Verisign the right to price .com however it pleases?
Or maybe Trump just wants to flip the bird to Obama once more by reversing yet another of his policies?
Who knows? It doesn’t make a lot of sense to me.

New gTLDs rebound in Q2

New gTLD registration volumes reversed a long trend of decline in the second quarter, according to Verisign’s latest Domain Name Industry Brief.
The DNIB (pdf), published late last week, shows new gTLD domains up by 1.6 million sequentially to 21.8 million at the end of June, a 7.8% increase.
That’s the first time Verisign’s numbers have shown quarterly growth for new gTLDs since December 2016, five quarters of shrinkage ago.
[table id=51 /]
The best-performing new gTLD across Q2 was .top according to my zone file records, adding about 600,000 names.
.top plays almost exclusively into the sub-$1 Chinese market and is regularly singled out as a spam-friendly zone. SpamHaus currently ranks it as almost 45% “bad”.
Overall, the domain universe saw growth of six million names, or 1.8%, finishing the quarter at 339.8 million names, according to Verisign.
Verisign’s own .com ended Q2 with 135.6 million domains, up from 133.9 million at the end of March.
That’s a sequential increase of 1.7 millions, only 100,000 more than the total net increase from the new gTLD industry.
.net is still suffering, however, flat in the period with 14.1 million names.
ccTLDs saw an increase of 3.5 million names, up 2.4%, to end June at 149.7 million, the DNIB states.
But that’s mainly as a result of free TLD .tk, which never deletes names. Stripping its growth out (Verisign and partner ZookNic evidently have access to .tk data now) total ccTLD growth would only have been 1.9 million names.

Have your say on single-character .com domains

ICANN wants your opinion on its plan to allow Verisign to auction off o.com, with a potential impact on the future release of other single-character .com domain names.
The organization has published a proposed amendment to the .com registry contract and opened it for public comment.
The changes would enable Verisign to sell o.com, while keeping all other currently unallocated single-character names on its reserved list.
The company would not be able to benefit financially from the auction beyond its standard $7.85 reg fee — all funds would be held by an independent third-party entity and distributed to undisclosed non-profit causes.
The arrangement would also see the buyer pay a premium renewal fee of 5% of the initial outlay, doubling the purchase price over the course of 25 years.
They would not be able to resell the domain without selling the registrant company itself.
It’s a pretty convoluted system being proposed, given that there may well end up only being one bidder.
Overstock.com, the online retailer, has been pressuring ICANN and Verisign to release o.com for well over a decade, and the proposed auction seems to be a way to finally shut it up.
The company has a US trademark on O.com, so any other bidder for the name would probably be buying themselves a lawsuit.
The proposed auction system does not address trademark issues — there’s no sunrise period of trademark claims period.
One party already known to be upset about lack of rights protection is First Place Internet, a search engine company that has a US trademark on the number 1.
It told ICANN (pdf) back in January that the o.com deal would “set a dangerous precedent” for future single-character name releases.
The ICANN public comment period, which comes after ICANN received the all-clear from US competition regulators, closes June 20.
As a matter of disclosure, several years ago I briefly acted as a consultant to a third party in support of the Verisign and Overstock positions, but I have no current interest in the situation one way or the other.

.com adds 5.5 million names, renewals back over 70%

Verisign reported first-quarter financial results that reflected a healthier .com namespace following the spike caused by Chinese speculation in 2016.
The company Friday reported that .com was up to 133.9 million domains at the end of March, an increase of 5.5 million over the year.
The strong showing was tempered slightly by a further decline in .net, where domains were down from 15.2 million to 14.4 million.
Over the quarter, there was a net increase of 1.9 million names across both TLDs and the renewal rate was an estimated 74.9%, a pretty damn good showing.
Actual renewals for Q4, measurable only after Verisign announced its earnings, were confirmed at 72.5%, compared to a worryingly low 67.6% in Q4 2016.
In a call with analysts, CEO James Bidzos confirmed that the turnaround was due to the surge in Chinese domainer speculation that drove numbers in 2016 finally working its way out of the system.
In Q1, the cash-printing company saw net income of $134 million, compared to $116 million a year earlier, on revenue up 3.7% at $299 million.
Bidzos told analysts that it’s “possible” that the company may get to launch .web in 2018, but said Verisign has not baked any impact from the contested gTLD into its forecasts.

Child abuse becoming big problem for new gTLDs

There were 3,791 domain names used to host child sexual abuse imagery in 2017, up 57%, according to the latest annual report from the Internet Watch Foundation.
While .com was the by far the worst TLD for such material in terms of URLs, over a quarter of the domains were registered in new gTLDs.
Abuse imagery was found on 78,589 URLs on 3,791 domains in 152 TLDs, the IWF said in its report.
.com accounted for 39,937 of these URLs, a little over half of the total, with .net, .org, .ru and .co also in the top five TLDs. Together they accounted for 85% of all the abuse URLs found. The 2016 top five TLDs included .se, .io and .cc.
New gTLDs accounted for a small portion of the abuse URLs — just over 5,000, up 221% on 2016 — but a disproportionate number of domains.
The number of new gTLD domains used for abuse content was 1,063, spread over 50 new gTLDs. Equivalent numbers were not available in the 2016 report and IWF does not break down which TLDs were most-abused.
According to Verisign’s Q4 Domain Name Industry Brief (pdf), new gTLDs account for just 6.2% of all existing domain names, and yet they account for over 28% of the domains where IWF found child abuse imagery.
IWF said that the increasing number of domains registered to host abuse imagery can be linked to what it calls “disguised websites”.
These are sites “where the child sexual abuse imagery will only be revealed to someone who has followed a pre-set digital pathway — to anyone else, they will be shown legal content.”
Presumably this means that registries and registrars spot-checking domains they have under management could be unaware of their true intended use.