.org back-end deal will come up for re-bid, PIR says as it acquires four new gTLDs

The industry’s most lucrative back-end registry services contract will be rebid, Public Interest Registry said today.

The deal, which sees PIR pay Afilias $18.3 million a year to run .org, according to tax records, will see a request for proposals issued in the back half of 2023, according to PIR.

Given that’s two years away, it’s strange timing for the announcement, which came at the bottom of a press release and blog post announcing that the company is acquiring four new gTLDs, three of which belong to Afilias’ new owner, Donuts.

PIR said Donuts is to transfer control of .charity, .foundation and .gives, which will be “reintroduced” to the market. .foundation currently has about 20,000 registered domains; the other two have a few thousand each.

It’s also acquiring the unlaunched gTLD .giving from a company called Giving Ltd.

All four are on-message for PIR’s not-for-profit portfolio, which also includes the barely-used .ngo and .ong for non-governmental organizations.

Those two gTLDs are getting decoupled, allowing registrants to register one without having to buy the other, PIR also said today.

The last time the PIR back-end contract came up for renewal, in 2015, Afilias was also the incumbent but increased competition — it was up against 20 rivals — meant that its slice of .org revenue was cut in half.

ICANN budget: staff bloat making a comeback

ICANN plans to ramp up its headcount starting next year to support the development of the new gTLD program.

Newly published budgeting documents show that average headcount is expected to rise to 406 for the year ending June 30, 2022, from 395 at the end of this June, with an even steeper increase to 448 a year later.

That’s after several years in which staffing levels have been fairly stable, even sometimes declining a little.

The main culprit is the Operational Design Phase for the next new gTLD round(s), which is expected to kick off soon.

ICANN expects to hire or assign nine people to manage the ODP before the end of June 2022, ramping that up to an average of 22 over the following year. The amount of non-ODP operational staff is expected to rise by 28 over the same period.

ICANN currently advertises 31 open positions on its web site, having added eight listings just this week.

This chart shows the expected growth:

At the time of the last new gTLD application round, in 2012, ICANN had 152 staffers, nine of whom were assigned to new gTLD project — and that was after the programs rules had already been developed, implemented and the application window opened and closed.

ICANN budget: no more new gTLDs before 2028

ICANN is not accounting for any revenue from a future round of new gTLDs in its just-published budget, which plots out the Org’s finances all the way through 2028.

The budget, which I gave a high-level summary of here, even predicts that dozens of 2012-round new gTLDs will disappear over the next six years.

The Org is predicting that there will be 1,091 gTLDs on the internet by the end of its fiscal 2027 (that is, June 30, 2028) down by 58 or 5% from July 2022.

Given that it’s only expecting to lose four gTLDs in FY23, this projection implies a speeding up of the rate at which gTLDs start cancelling their contracts or going out of business in the later part of the five-year budget.

The forecast comes with a big asterisk, however. A footnote reads:

These scenarios do not assume any further TLD delegations arising from the resumption of the New gTLD Program. While there is ongoing work and an intent to launch a subsequent round, the timing of its release remains unclear and potential impact(s) on funding indeterminate. Given this, ICANN org has deemed it prudent not to assume any prospective impacts from a subsequent round across the described scenarios.

In other words, ICANN is not yet ready to commit to a runway for the next application round, subsequent delegations and eventual revenue.

As I reported Monday, the next round is unlikely to be approved until the fourth quarter of next year at the earliest, and my view is that 2024 is the soonest the next application window could open.

I don’t think we can read too much into the fact that ICANN isn’t budgeting for any next-round impact on funding until after 2027.

If you’re pessimistic, you could infer that ICANN believes it’s at least a possibility that the next round could take that long, or not be approved at all, but the safer bet is probably that it merely lacks visibility and is acting in its usual risk-averse manner.

ICANN budget: mild optimism amid maturing industry

ICANN thinks the domain industry, including the new gTLD industry, is maturing and will continue to grow, in its just-published draft budget for fiscal 2023.

The Org is predicting growing transactions across the board, as well as an increase in the number of accredited registrars and a slowing decline in the number of contracted gTLDs.

ICANN is expecting funding of $152 million for FY23, which includes the $4 million bung it negotiated with Verisign as part of the deal to allow the company to raise .com prices.

That’s up from the $149.1 million is expects to receive in the current fiscal year.

As usual, the bulk of the funding comes from gTLD transaction fees — the taxes registrants pay through their registrars and registries whenever they register, renew or transfer a domain name.

Legacy gTLD transaction fees are expected to amount to $93.1 million, up 3% on a forecast of $90.1 million in the current year, while new gTLD transaction fees are expected to rise modestly from $9.5 million to $9.9 million, a 4% increase.

Transactions in legacy gTLDs are expected to be 201.2 million, versus 193.6 million in the current year.

New, post-2012 gTLDs are expected to process 25.8 million transactions, up from 24.8 million, of which 21.1 million will be billable, up from 20.3 million. New gTLDs only pay transaction fees after 50,000 domains under management.

ICANN is expecting to lose four registries in FY23 — this almost always means dot-brands that cancel their contracts — with the total declining from a June 2022 total of 1,149 to 1,145 a year later. This will have a modest impact on fixed registry fees.

But the Org is once again expecting to see an increase in the number of registrars paying fixed accreditation fees, up by 28 to 2,447 at the end of FY23.

Accompanying the budget, ICANN has published some industry trend analysis (pdf) outlining some of the assumptions behind the budget forecasts.

Basically, the document describes what regular readers already know — many domain companies benefited from pandemic-related lockdowns driving small businesses online, but overall industry volumes were driven down by low-cost new gTLDs experiencing huge junk drops.

For ICANN’s purposes, factors such as customer quality and pricing are irrelevant. A spammer registering 1,000 domains in bulk pays ICANN the same amount in fees as 1,000 small businesses building their first web sites.

The document reads:

Taken as a whole, DUMs failed to expand in the past twelve months ending in mid-2021. While this decline is at least partly attributable to lower promotional activity among some of the largest new gTLDs which could be reinitiated in the future, it nonetheless points to an industry that has shifted from a period of rapid expansion to one that is now witnessing steady maturation.

The draft ICANN budget covers the 12 months beginning July 1, 2023, and is now open for public comment before possible revisions and final approval.

ICANN says ODP will speed up new gTLDs in the long run

A time-consuming process to spec-out the next new gTLD application round before it is even formally approved will actually speed up the program over the long run, an ICANN veep has said.

The so-called Operational Design Phase is a bunch of planning, or issues such as cost and feasibility, that ICANN says it needs to do before the community’s policy recommendations can be put before the board of directors.

The board approved the ODP in September, giving the Org a $9 million budget and a 10-month deadline to complete the project, but the clock doesn’t start ticking until CEO Göran Marby formally starts the process.

Three months later, that still hasn’t happened. ICANN is still “organizing the resources needed and developing the roadmap for the work ahead”, according to a blog post from Karen Lentz, VP of policy research and stakeholder programs.

The Org is doing the preparation for the preparation for the preparation for the next round, in other words.

But Lentz says this will speed up the new gTLD program over the longer term.

We believe the ODP will actually streamline future work. It will have a positive impact on the duration of the implementation process by making the assumptions explicit, answering key questions, and considering how the recommendations on different topics work together in addition to providing a detailed timeline and visibility to the timing of implementation activities. If the Board approves the recommendations, the org and the Implementation Review team would be able to leverage a good amount of work already completed during the ODP. Future rounds would not be possible without the foundational work of an ODP. It’s important to note that without an ODP, this work would still be taking place, but without the structure and transparency that the ODP provides.

Another important consideration to note here is that we are not simply organizing only the next round. We are building a foundational structure for all of the work that the org, the community, and the Board will do over the coming years to continue to evolve the namespace along with the necessary procedures and tools. So the work from this ODP is not only for a single round — this is targeting a long-term plan and for multiple rounds.

If Marby were to start the ODP tomorrow, and ICANN managed to hit its deadline, October 2022 would be the absolute earliest the ICANN board would get the chance to approve the next round.

It’s possible, though not very likely given how intrinsic to ICANN’s mission the opening up of gTLD competition is, that the board could instead decide not to approve the next round.

After the next round gets the thumbs-up, there’s still a whole lot of extra work to do — the aforementioned Implementation Review, hiring contractors, a months-long marketing campaign — before companies would actually get to file their applications.

We’re still looking at 2024 at the earliest for that, in my view, but if there’s one thing we can rely on from ICANN, it’d delay.

ICANN takes the lamest swipe at Namecheap et al over blockchain domains

ICANN has come out swinging against blockchain domains and the registrars that sell them. And by “come out” I mean it’s published a blog post. And by “swinging” I mean “offered the weakest criticism imaginable”.

The post starts off well enough, observing that services marketed as “domain names” that are not automatically compatible with the global DNS are probably not a great purchase, because they don’t work like regular domains.

Using these alternatives requires something like a browser plug-in or to reconfigure your device to use a specialist DNS resolver network, the post notes, before concluding with a brief caveat emptor message.

All good stuff. ICANN has been opposed to alt-root domain efforts for at least 20 years, and the policy is even enshrined in so-called ICP-3, which nobody really talks about any more but appears to still be the law of ICANN Land.

So, which domain-alternatives is ICANN referring to here, and which registrars are selling them? The post states:

Name resolution systems outside the DNS have existed for a long time. One could mention the Sun Microsystem Network Information Service (NIS), the Digital Object Architecture (DOA), or even the Ethereum Name Service (ENS)…

With some ICANN-accredited registrars now selling NIS, DOA, or other similar domains alongside standard domain names, the potential for confusion among unsuspecting customers seems high.

You may be asking: what the heck (or, if you’re like me, fuck) are NIS and DOA domains, and which registrars are selling them?

Great questions.

NIS is an authentication protocol (a bit like LDAP) for Unix networks developed in 1985 (the same year the original DNS standard was finalized) by Sun Microsystems, a company that hasn’t existed in over a decade.

To the best of my knowledge they’ve never been marketed as an alternative to regular domain names. Nobody’s ever used them to address a publicly available web site. Nobody sells them.

DOA, also known as the Handle System, is a more recent idea, first implemented in 1994, before some of you were born. Handles are mostly numeric strings used to address digital objects such as documents. Libraries use them.

The main thing to know about Handles for the purposes of this article is that they’re specifically designed to convey no semantic information whatsoever. They’re not designed to look like domain names and they’re not used that way.

So how many registrars are selling NIS/DOA domains? I haven’t checked them all, but I’m going to go out on a pretty sturdy limb and guess the answer is “none”, which is a lot less than the “some” that ICANN asserts.

But ICANN also mentions the Ethereum Name Service, a much newer and sexier way of cybersquatting, based on the Ethereum cryptocurrency blockchain.

ENS allows people to buy .eth domain names (which do not function in the consensus DNS) for the Ethereum equivalent of about $5. As far as I can tell, you can only buy them through ens.domains, and no ICANN-accredited registrar is functionally capable of selling them.

The ICANN post also contains a brief mention of “Handshake”, and this appears to be what ICANN is actually worried about.

Handshake domains, also known as HNS, look like regular domain names and a handful of ICANN-accredited registrars are actually selling them.

Handshake is also based on blockchain technology, but unlike ENS it also allows people to create their own TLDs (which, again, do not function without special adaptations). Registrars including Namecheap, 101domain and EnCirca sell them.

It’s Namecheap’s storefront hover text, warning that HNS domains don’t work in the regular DNS, that ICANN appears to be paraphrasing in its blog post.

The registrar has a lengthy support article explaining some of the ways you can try to make a Handshake domain work, including an interactive comment thread in which a Namecheap employee suggests that DNS resolvers may choose to resolve HNS TLDs instead of conflicting TLDs that ICANN approves in future.

That’s the kind of thing that should worry ICANN, but it’s got a funny way of expressing that concern. Sun Microsystems? Digital Object Architecture? What’s the message here?

Twenty years ago, I interviewed an ICANN bigwig about New.net, one of the companies attempting to sell alt-root domains at the time. He told me bluntly the company was “breaking the internet” and “selling snake oil”, earning ICANN a snotty lawyer’s letter.

Today’s ICANN post was ostensibly authored by principal technologist Alain Durand, but I’m going to give him the benefit of the doubt and assume comms and legal took their knives to it before it was published.

While some things haven’t changed in the last two decades, others have.

Google to release another new gTLD next month

Google Registry is gearing up to unleash another gTLD from its stockpile of unreleased strings next month.

The gTLD is .day, one of over 100 that Google applied for in 2012 after a reported brainstorming session at the company.

According to its application:

The specialization goal of the proposed gTLD is to offer a new Internet environment that allows users to create and organize events that have or will occur on a particular day. The proposed gTLD will provide a single domain name hierarchy for Internet users globally to promote celebrations, such as a holi.day, wedding.day, or birth.day.

With that in mind, it’s difficult to see .day being a high-volume TLD along the lines of Google’s popular .app or .dev gTLDs.

While the company itself doesn’t seem to have addressed the launch publicly, it has given details to registrars and informed ICANN about its start-up dates.

It started a Qualified Launch Program program earlier this week. That’s where it gets to hand out a limited number of domains to hand-picked anchor tenants.

The sunrise period, restricted of course to trademarks, begins December 14 and ends January 24.

General availability starts January 25, according to registrars and ICANN records, with a seven-day Early Access Period during which domains can be purchased at daily-decreasing premium prices.

Full regular-price general availability begins February 1.

CentralNic takes over a dead dot-brand

CentralNic has become the latest company to pounce on a dot-brand gTLD that was on its way to the dustbin of history.

The ICANN contract for .case was transferred to a London company called Helium TLDs, a CentralNic subsidiary, last week.

That company was previously called FANS TLD, and was the vehicle CentralNic used to acquire .fans from Asiamix Digital in 2018 before later passing it on to Hong Kong-based ZDNS International.

I believe something similar is happening here.

.case was a dot-brand owned, but never used, by CNH Industrial, which Wikipedia tells me is an American-Dutch-British-Italian company that makes about $28 billion a year making and selling agricultural and construction machinery. Diggers and forklifts and such.

CNH also managed .caseih, .newholland, and .iveco for some of its other brands, but these contracts were terminated earlier in the year.

The company had also asked ICANN to cancel its .case agreement, but that seems to have attracted acquisitive registry operators, and the termination request was withdrawn as I noted in September.

While terminating a dot-brand can often be seen as a lack of confidence in the dot-brand concept, selling off the gTLD to a third party rules out reapplying for the same string in future and can be seen as an even deeper disdain.

Now, .case is in CentralNic’s hands. I believe it’s the first dot-brand the company has taken over.

Rival registries including Donuts, XYZ and ShortDot have also swept up unwanted dot-brand gTLDs, stripped them of their restrictions, and repurposed them as general-purpose or niche spaces.

.music goes live, plots 2022 launch

.music has become the latest new gTLD to join the internet, but it seems unlikely to hit the market before the 10th anniversary of the 2012 ICANN application period.

The TLD was added to the DNS root at the end of October, with the first domain, the obligatory nic.music, going live a few days later.

The registry, Cyprus-based DotMusic, said in a press release that it plans to launch the gTLD next year.

.music was one of the most heavily contested gTLDs from the 2012 application round, with eight total applicants.

It was one of two Community applications, which promised a more controlled, restricted namespace in exchange for a smoother ride through the ICANN approval and contention resolution processes.

But it failed to win its Community Priority Evaluation, leading to years of appeals and ICANN reviews.

The contention set was finally resolved in 2019, apparently via auction, with DotMusic prevailing against heavy-hitters including Amazon and Google.

But the victorious registry was slow out of the blocks after that, taking almost two years to negotiate its registry agreement with ICANN.

It’s still going to be a restricted-community space when it finally launches, which makes its success anything but assured, regardless of the unquestionable strength of its string.

Sadly, while DotMusic CEO Constantine Roussos looked every bit the part of the hip young rocker when the .music application was first filed, showing up everywhere in a sports car, cool haircut, and designer skinny jeans, today he lives in a senior-care home, drives a mobility scooter, and needs to be changed hourly.

Virgin territory as GoDaddy pushes $30 million porn domain renewals

Brand owners big and small are in for a potential surprise December 1, as their 10-year-old .xxx domain blocks expire and registrars bill their customers to convert them into a new annually-renewing GoDaddy service.

GoDaddy confirmed to DI today that it will “auto-convert” the old Sunrise B blocks, first sold by ICM Registry in 2011, to its new AdultBlock service, which provides essentially the same functionality but across four TLDs rather than one.

Tony Kirsch, head of professional services at GoDaddy Registry, said:

Registrars have been contacting all the Sunrise B owners and advising them that as of December 1 they will be grandfathered and automatically converted into an AdultBlock service, but they have a choice to expire that or stop that happening prior to December 1.

And if it is that they don’t do that before December 1, we’ll still give them a grace period of at least 45 days. If that happens they can then, as you’d normally do, just turn around to the registrar and say “We don’t want that” and we will of course refund the money.

This means that GoDaddy, which acquired .xxx and ICM from MMX earlier this year, is billing its .xxx registrar partners to convert and renew what could be as many as 81,000 Sunrise B blocks.

While the registry fee for AdultBlock has not been published, retail registrars I checked have priced the service at $370 to $400 per year, which we can probably assume is low-end pricing. Most .xxx domains are sold via the specialist brand-protection registrars like CSC and Markmonitor, which sometimes have more complex pricing.

So that’s something in the ballpark of $30 million worth of renewal invoices being sent out in the coming weeks, for something in many cases brand owners may have institutionally forgot about.

Kirsch said that AdultBlock was introduced by MMX about 18 months ago and that registrars have been preparing their customers for the Sunrise B expiration for some time.

Sunrise B was a program, unprecedented in the industry at the time, whereby trademark owners could pay a one-off fee — ICM charged its registrars about $160 wholesale — to have their brands removed from the available pool.

The domains exist in the .xxx zone file and resolve to a black page bearing the words “This domain has been reserved from registration”, but they’re not registered and usable like normal defensive or sunrise registrations would be.

Companies got to avoid not only the potential embarrassment of being porn-squatted, but also the hassle of having to explain to a tabloid reporter why they “owned” the .xxx domain in question.

The term of the Sunrise B block was 10 years. ICM told me at the time that this was because the company’s initial registry contract with ICANN only lasted for 10 years, so it was legally unable to sell longer-term blocks, but I’ve never been sure how much I buy that explanation.

Regardless, that 10 year period comes to an end in two weeks.

Because Sunrise B was unprecedented, this first renewal phase is also unprecedented. We’re in virgin territory (pun, of course, very much intended) here.

Will we see the industry’s first public “block junk drop”?

There are a number of reasons to believe trademark owners, assuming they don’t just blindly pay their registrar’s invoices, would choose to allow their blocks to expire or to ask for a refund after the fact.

First, the price has gone up — a lot.

While ICM charged $160 for a 10-year Sunrise B block (maybe marked up by registrars to a few hundred bucks) brand owners can expect to pay something like $3,000 retail for a single string blocked for 10 years.

But buyers do get a bit more bang for their buck. Unlike Sunrise B, AdultBlock also blocks the trademark in three additional GoDaddy-owned TLDs — .porn, .sex and .adult — as standard.

Kirsch said he expects buyers to see a 40% to 50% saving compared to the cost of defensively registering each domain individually.

Second, the appetite for defensive registrations has waned over the past 10 years, with trademark owners employing more nuanced approaches to brand protection, largely due to the flood of new gTLDs since 2013.

When .adult, .sex and .porn launched, without the possibility of Sunrise B blocks, they got about 2,000 regular sunrise registrations each. And that’s extraordinarily high — for most new gTLDs a couple hundred was a good turnout.

Third, the .xxx launch attracted a whole lot of controversy and overreaction, and the .xxx zone file today contains a lot of Sunrise B crap.

When I scrolled a little through the zone, cherry-picking silly-looking blocks in 2019, I found these examples:

100percentwholewheatthatkidslovetoeat.xxx, 101waystoleaveagameshow.xxx, 1firstnationalmergersandacquisitions.xxx, 1stchoiceliquorsuperstore.xxx, 2bupushingalltherightbuttons.xxx, 247claimsservicethesupportyouneed30minutesguaranteed.xxx, 3pathpowerdeliverysystembypioneermagneticsinc.xxx

Is it worth $400 a year to block the trademark “100 Percent Whole Wheat That Kids Love To Eat”? Is there any real danger of a cybersquatter going after that particular brand (apart from the fact that I’ve now written about it twice)?

Kirsch said a “small percentage” of Sunrise B owners have already said they don’t want to convert, but given that the rest will auto-convert, and that the registrars are doing all the customer-facing stuff, the company has limited visibility into likely uptake.

Brian King, director of policy at MarkMonitor, told us: “We generally encourage our clients to consider blocks. They can be cost effective and a lot of times clients would rather have their brand be unavailable without having to register in TLDs where they don’t want to own domain registrations for any number of reasons.”

One reason brand owners may want to consider converting to AdultBlock — it’s rumored that GoDaddy will be relaxing its eligibility criteria for .xxx next year, removing the requirement for registrants to have a nexus to the porn industry.

It’s always been kind of a bullshit rule, basically a hack to allow ICM to run a “sponsored” TLD under ICANN’s rules from the 2003 application round, but doing away with it would potentially make it easier for cybersquatters to get their hands on .xxx domains.

CSC told customers in a recent webinar that the rules are likely to be changed next year, increasing the risk of cybersquatting.

There’s some circumstantial evidence to suggest that CSC might be on to something — pretty much every “sponsored” gTLD from the same 2003 application round as .xxx has relaxed their reg rules to some extent, sometimes when their contracts come up for renewal and ICANN tries to normalize them with the text of the standard 2012-round agreement.

And GoDaddy’s .xxx contract with ICANN is being renegotiated right now. It was due to expire in March, but it was extended in February until December 15, a little under a month from now. We may soon see ICANN open up the new text for public comment.

Kirsch, who’s not part of the negotiations, could not confirm that the eligibility relaxation is going to happen or that it’s something GoDaddy is pushing for.

If it were to happen, it wouldn’t be for some time, and it wouldn’t necessarily impact on the December 1 deadline for Sunrise B conversions, which is going to be interesting to watch in its own right.

“There are registrations that are protecting people’s trademarks that are expiring and our primary objective here is to ensure that that protection continues, and that’s what we’ll do,” GoDaddy’s Kirsch said.

“If we just let them expire, it would create a lot of opportunity for brand infringement. Faced with that choice, our primary objective is to protect trademark owners,” he said.