Is Go Daddy’s size a competition concern?

Go Daddy is undoubtedly the runaway success story of the domain name industry.
It may not be as big as VeriSign, but unlike VeriSign it was not simply handed a multi-billion dollar resource to manage. It was essentially scratch-built. It didn’t even have first-mover advantage – Register.com and Network Solutions had that, and Go Daddy’s been eating their lunches for years.
The company has got where it is today through, in my opinion, a combination of cheap prices, decent customer service and populist marketing. Mainly the cheap prices, but I doubt that putting a great big pair of boobs on TV during the Super Bowl can have hurt sales.
But how big is the company? And with the introduction of new gTLDs, is its size now a cause for concern?
Go Daddy’s market share of live gTLD registrations is around the 30% mark. Big, but not necessarily dominant.
However, the company caused a bit of a stir when it said in a press release last week that it holds a “near 50 percent market share of all active new domains registered in the world”.
I obviously don’t have access to all the data Go Daddy does, so I mined RegistrarStats.com to see if I could work out where this 50% number comes from.
Note that these numbers may be somewhat low. RegistrarStats only measures the 13 largest gTLDs and one ccTLD, .us. Also, it counts registrars by their accreditation ID number; some registrars have several of these. Here, I’m using only Go Daddy’s primary ID, excluding its affiliates.
According to RegistrarStats, there are 117,942,539 active domains today, compared to 109,538,023 a year ago. That’s 8,404,516 net new domains, growth of 7.67%.
Go Daddy has at present 35,186,522 domains under management, compared to 29,938,953 a year ago. That’s an increase of 5,247,569 domains, or 17.53% growth.
Ergo, Go Daddy is growing more than twice as fast as the market.
Of the net new registrations recorded over the last 12 months across all these TLDs, 62.44% of them belong to Go Daddy. If the market as a whole had grown by 100 domains, Go Daddy would have grown by 62 domains.
Of course, this might be somewhat misleading as not all of these Go Daddy domains will be new registrations, some will be transfers in from other registrars.
It’s possible that if you filter out transfer data, which I don’t have access to, and factor in ccTLDs and affiliates and so forth, you arrive at the 50% ballpark cited in Go Daddy’s press release. I don’t know.
Regardless, I think it goes some way to showing just how massive Go Daddy has become in retail domain sales.
For comparison, Xin Net was the second most successful registrar in terms of its slice of the market growth over the last year, with 7% of the pie. That’s 7% versus 62%.
It is reasonable to assume that when new gTLDs are launched, Go Daddy will get a big slice of that market too.
Which brings me to my point.
As Antony Van Couvering of Minds + Machines said during the ICANN public forum last week, referring to the Go Daddy press release, “They are not going to want to carry all new top level domains. That means if they don’t want to carry you, you’ve lost 50% of your market.”
He very reasonably pointed out that Go Daddy’s sales interface has a limited amount of real estate, and it will have trouble fitting 500 new TLD options into it. It already carries 60-odd TLD options, and that’s a squeeze. At the very least, the company will have to prioritize.
Van Couvering put forth the view that enforced vertical separation of registries and registrars will not be helpful for new gTLDs – if a startup is not allowed to retail its own domains it will be beholden to Go Daddy and other large registrars.
To play devil’s advocate, one could quite easily make the opposite case. If vertical integration is permitted, and large registrars are allowed to operate as registries, it might be in their best business interests not to carry any competitive TLDs. Might.
In either case, Go Daddy’s market power is an issue.
As of Friday’s board meeting, ICANN has opted for strict vertical separation in the absence of a consensus policy from the GNSO, which has a working group looking at the issue.
Interestingly, an economic study produced for ICANN in January, published last week, addresses the issue of market share and vertical separation head-on.
The paper basically says that vertical integration should be permitted, but that ICANN should notify competition authorities if a registrar with a certain amount of market share attempts to acquire a major interest in a registry.

ICANN must also specify the appropriate market share threshold to trigger any action. We recommend that ICANN choose a market share threshold in the 40-60% range (the market share measured would be that of the acquiring company). The lower end is the market share at which U.S. competition authorities begin to be concerned about market power. The upper end is the market share at which U.S. competition authorities typically begin to be concerned about monopoly power. EU competition authorities typically begin to be concerned about possible market dominance (a concept similar to market power) at a market share in the 40% range.

You might think that Go Daddy, with its 30ish% share of extant domains does not meet the threshold for concern. Not so. The paper goes on to recommend:

For gauging market power with respect to registrars, we believe that the percentage of newly created gTLD registrations is a more appropriate measure, because this measure is a more accurate proxy for the potential buy-side market power issues that exist at the registrar level.

This measure, as press released by Go Daddy, is close to 50%, easily meeting the European threshold. If my 62% number has any meaning at all (and I’m not convinced it does) then the threshold where the US authorities become concerned is also met.
Go Daddy may find it has to tread carefully over the next few years. Its market share gains show no signs of slowing down, and any perception of unfair play is bound to attract criticism.

Tweet

Tagged: go daddy, ICANN, new gTLDs