.xyz sold just shy of 15,000 domain names during its first hours of general availability.
That’s according to zone files released today.
The registry added 14,829 names yesterday, ending the period with 14,924 in its zone, making it the 15th-largest new gTLD by volume, behind the likes of .center, .directory and .solutions.
(UPDATE: see the bottom of this post for important updates on these numbers.)
While it’s a respectable first-day performance — logging more sales that I expected it to — it’s doesn’t look like the registry is going to hit CEO Daniel Negari’s target of a million names in year one.
Earlier this week, Negari enthusiastically blogged, in a post tagged simply “believe”:
There is no doubt in my mind that the growth rate of .xyz will surpass the growth rate of .com. It won’t take us 20 years to reach one million registrations. It will only take 1 year. The growing number of internet users online today, the increasing demand for domain name registrations, and the way we have positioned .xyz all validate this projection. Soon we will see that there is only one competitor to .com, and that is .xyz.
Averaged out over a year, a million registrations is over 2,700 per day, a rate that no other new gTLD registry has managed to sustain to date.
If .xyz follows the path taken by its predecessors, it will probably rack up a thousand or so more names per day for the next few days, likely making it a top-ten new gTLD within a week, before settling down to a comfortable rate of 200 to 300 per day.
So far, .guru (four months old) and .club (one month old) are the only gTLDs with significant volumes that have managed to double their launch-day figures.
.xyz is going to be in desperate need of some more marketing if its wants to get to one million by June 2, 2015.
As I’ve always said, I think .xyz is a tough sell.
As Negari says, the string has “little-to-no meaning”, and the game has always been to brand .xyz as being even more generic than .com.
The registry has attempted to shoehorn it into meanings such as “It is the ending of the alphabet, and it is the ending of domain names” and “Generations X, Y and Z”.
But it seems to be one of those sales pitches you either get or you don’t. I’ve heard it many times, and I still don’t.
So 15,000 names in half a day, by my expectations, is not bad.
However, despite the 15,000 names, we’re talking about a TLD with a sub-.com registry fee, being retailed by some registrars for under $10.
With a million names, that’s a nice little business. Sub-100,000, not so much.
There are plenty of people out there who will immediately say the “success” of new gTLDs in general is a combination of many factors, of which volume is only one, and I agree.
It’s only day one for .xyz, but for a cheap-as-chips gTLD that has explicitly made itself a volume play, I think registration count is a fair way to measure its success going forward.
UPDATE: According to Gavin Brown, CTO of .xyz back-end CentralNic, the number of names in the zone file almost 24 hours after GA began is actually 31,119, over double what the registry sold in the first 10 hours.
If there’s one thing we’ve learned from the last six months of new gTLDs, it’s that predictions about massive levels of defensive registrations were way off the mark.
New gTLDs are not seeing anywhere near the same numbers of sales during sunrise periods as their predecessors.
I have managed to collate some data that I think gives a pretty accurate picture of how many sunrise registrations are being made and therefore how much new gTLDs are costing trademark owners.
About 128 gTLDs have finished their sunrise periods to date, and I have the sunrise sales figures for 101 of them. All of these numbers were provided by the respective registry operator.
The biggest sunrise, per these numbers, was for .clothing, which had 675 registrations. That’s 5.97% of the 11,301 overall names in the .clothing zone file today, over three months after launch.
At the other end of the scale is شبكة. (“.shabaka” or “.web” in Arabic), which sold just five names during its sunrise, the first of the program, which was restricted to Arabic trademarks.
The total number of sunrise sales across across all 101 gTLDs is 14,567, making for an average of 144.2 domains per new gTLD sunrise.
Sunrise currently accounts for 1.87% of all names in these 101 gTLDs, but that’s an artificially high number because some of the gTLDs I have sunrise numbers for are not yet in general availability.
But compare the real numbers to .co, which sold over 11,000 names at sunrise when it launched in summer 2010, or .xxx, which took 80,000 sunrise applications in late 2011.
Trademark owners are not defensively registering with anywhere near the same fervor as they once did.
If that 144.2 average names holds true for all 128 gTLDs that have completed sunrise, we can approximate that 18,461 names have been sold during sunrise periods to date.
I should point out that I’m assuming in these calculations that all sunrise registrations are “defensive” and that brand owners are not defensively registering during general availability.
Neither of those assumptions will be fully true.
Not all sunrise sales are made to genuine brand owners, of course. Some number of generic dictionary domains have been registered by people who obtained trademarks just in order to get the matching domain.
And only a psychic could know whether a GA registration is “defensive” or not at this stage.
But let’s assume that every sunrise reg went to a genuine brand owner. How much have they had to pay for these names?
It’s difficult to calculate a precise dollar value because each registry has a different pricing scheme and sometimes the price of a name can vary even within a specific given TLD.
I looked to the prices listed at 101domain, which has pretty exhaustive coverage of new gTLDs, for a guide.
The average first-year cost for a sunrise registration in the 75 or so new gTLDs currently being sold to trademark owners at 101domain is a little shy of $165.
Assuming that’s a good guide for pricing in sunrise periods that have already closed, we can calculate that 18,461 names at $165 a pop equals $3,046,089 out of the pockets of trademark owners in the first year.
But the sunrise fees are not the only costs, of course. In order to participate in a sunrise you must first register your mark in the Trademark Clearinghouse.
There are 30,251 marks registered in the TMCH, according to the TMCH itself. At $150 a pop — the minimum you can pay for a TMCH registration — that’s $4,537,650 spent on defensive measures.
Add in the cost of the sunrise registrations and a generous $100,000 to cover the cost of the 50 Uniform Rapid Suspension cases that have been filed to date and the total cost to brand owners so far over the first 128 new gTLDs comes to $7,683,739.
Whether this is “a lot” or not probably depends on your perspective.
It’s certainly not the billions of dollars that were being predicted by some as recently as last year.
In September the Better Business Bureau and the Coalition Against Domain Name Abuse speculated that 600 “open” new gTLDs could lead to $10 billion being spent on defensive registrations.
That statement was made in a press release calling for stronger cybersquatting legislation in the US.
But if 101 open gTLDs leads to $3,046,089 being spent, 600 such gTLDs should lead to a total cost of about $18 million, not including the fixed TMCH costs (which probably won’t grow very fast in future).
That’s not the same ballpark, not the same league, not even the same sport.
.club has overtaken .guru to become the top-selling new gTLD on the market.
According to today’s zone file report, .club now has 59,120 domains, having grown by 2,504 yesterday. That’s compared to .guru, which grew by 181 names to 58,791.
It’s been 12 days since .CLUB Domains predicted it would be at the top of the league table “within days”.
It’s taken 20 days for the gTLD to beat .guru, something the registry at first reckoned would happen within its first week of full general availability.
I don’t think yesterday’s spike has anything to do with the 50 Cent endorsement deal.
As far as I can tell, Fiddy has only been linking to his .com shop on Twitter since his .club domain went live and the deal does not seem to have generated any media coverage outside of domain blogs.
Taking the number one spot is only a relative success, of course. What’s important is that .club is still managing to move 2,500 names in a day three weeks after launch, something no other new gTLD has managed to date.
Curtis ’50 Cent’ Jackson accidentally supported “.com” while he was endorsing da new .club gTLD at a launch party at a New York nightclub last night.
As previously reported, Fiddy is da first significant celebrity endorser of a new gTLD. He’s being paid to use 50inda.club, a web site developed by .CLUB Domains, as his new social media hub.
“In Da Club” was of course his breakthrough hit, in 2003.
He showed up at da Tao nightclub in New York — which had been rebranded “.CLUB” for one night only — last night for about 90 minutes in order to meet fans and pose for selfies, etc.
I was there. As disclosure, .CLUB had paid for my airfare from London, a night in a hotel, and copious amounts of alcohol.
I didn’t attempt to get into da roped-off VIP area where Fiddy was being held, but I gather that da bouncers guarding it were somewhat selective in who he got to meet.
He also publicly spoke, alongside .CLUB’s CEO Colin Campbell and CMO Jeff Sass, for about 30 seconds, in order to provide his official endorsement of da new gTLD.
Da problem was that during his brief address he referred to his support for “.com”, which is a little bit like a celebrity being paid to endorse Pepsi referring to how much he loves Coke.
Probably just a Freudian slip. We’ve all done it.
Unfortunately, I can’t give you da full quote just yet. It was quite noisy in there, and I’d consumed quite a bit of Cristal with diamond flakes floating in it. But a lot of people who videoed da address on their phones tell me it will be on YouTube shortly.
Fortunately for .CLUB, I don’t think what he said matters that much.
What matters is how frequently his people link to his new .club domain on his social media channels, how much mainstream media coverage his endorsement generates, and how many people register .club domains as a result.
Getting Fiddy as an anchor tenant will not have come cheap — my guess, and it is just a guess, is that da deal is costing .CLUB high six figures at least — so da company will have to sell a lot of domains to make it pay off.
UPDATE May 26: Here’s .CLUB’s video of da event. Fiddy says he’s very excited to launch his “50 in da .com club”. He later gets it right, referring to “.club” more than once.
ICANN has rejected demands by the Belgian government by giving Donuts the go-ahead to proceed with its application for .spa, which Belgium says infringes on a geographic name.
Noting that the Governmental Advisory Committee had submitted no consensus advice that Donuts .spa bid should be rejected, the ICANN board’s New gTLD Program Committee said last week “the applications will proceed through the normal process.”
That means the two-way contention set is presumably going to auction.
The English dictionary word “spa” derives from Spa, a small Belgian town with some springs.
The other applicant is Asia Spa and Wellness Promotion Council, which has made a deal with Spa to donate some of its profits to local projects and give the city some control over the registry.
Donuts refused to sign a similar deal, leading to Belgium last month asking ICANN to delegate the gTLD to ASWPC and not Donuts.
The GAC’s last word on .spa was this, from the recent Singapore meeting:
Regarding the applications for .spa, the GAC understands that the relevant parties in these discussions are the city of Spa and the applicants. The GAC has finalised its consideration of the .spa string and welcomes the report that an agreement has been reached between the city of Spa and one of the applicants.
There’s no ICANN fudging here; if the GAC wanted to issue a consensus objection it could have.
The question is: why didn’t it?
Why does the string “amazon”, which does not exactly match the name of a place in its local languages, qualify for a GAC objection, while “spa”, which exactly matches the name of a city, does not?