How to tell if a new TLD will succeed (and most won’t).

By , June 16, 2011

I just read Rick Spence’s “What’s in a name? It could be a fortune on the Internet“, (HT to HybridDomainer), which seemingly dials the needle back to (pick one), 2005 or 2007, or maybe 2010 and drinks the Kool-Aid every domainer has on an IV drip: the notion that fortune and success can be achieved simply by virtue of “owning the right names” (i.e How can we be bankrupt? Don’t we own

Only in this case, he’s talking about the seemingly impending new top-level-domains. “New domain names will rank above the current crop”, he gushes. Really? You mean now I get to defend my trademark in .www, .website, .bank, .toledo, .xxx and .FML ? Great! I think what we will see happen is a buyers’ strike in defensive names. It will become impossible to defend one’s names in all possible TLDs, brand owners will simply stop doing it, defensive registrations in new TLDs collapse, kicking out one of the two legs holding up most new TLDs (the other being speculative registrations).

But what do I know? I say myself that nobody can really predict the future, so when I thought tonight about coming out here and saying I think most new TLDs will stiff, that nobody can really say that without having some kind of criteria for what constitutes success. So I thought a bit about that and came up with the following points (I’ve put in numbers where I feel confident stating some. Where I put X% I think the number to sub in there is what the current number is for .COM, .NET or your own ccTLD, like .CA in my case)

  1. A renewal rate north of 60%, or even 70%. I am very interested to see where .CO’s renewal rate comes in, especially 5 or 10 years from now.
  2. Percentage of “parked” domains is under X%.
  3. Percentage of “defensive” domains is under Y% (meaning, the names are not simply redirecting to the .com or choice domain in another TLD)
  4. Percentage typosquats is under Z%
  5. Percentage of primary-branded domains (those domains whose primary online existence is within this given TLD) is above 10% – I deliberately set the bar low here, and I do not think very many TLDs will meet it.

It is likely that there will be two kinds of new TLDs in the future: those that exist for a reason, and those that don’t. The former will have a design philosophy baked into the cake which will enable them to exceed those 5 points above.

The latter, will be the same old, tired, worn out  “brand yourself here” crap like .neato, .hot, .money, .finance, .sports, .sporks, .wtf . Basically a loose premise on why .this is “gonna be bigger than .com” and they will launch using the typical registry cash-grab playbook:

  • implicitly stretch everybody over a barrel to “defend their mark” in the new namespace.
  • impose multi-year initial term minimums.
  • whip the speculators into a frenzy for the generics via forced auctions with minimum bid increments.

None of these TLDs will offer anything special in comparison to anything else. They will multiply like weeds, and they will for the most part, fail. If a few succeed here or there they will likely track the trajectory of a fad more than the rise of a sustainable namespace. One-time celebrity VC and blogger Rick Segal used to say: choice = confusion = inaction. And I think that assumption will apply to all these new TLDs. The majority of the buying public will be daunted and confused by the new TLDs, and turn instead back to the familiar .CNO(BI) and their own ccTLD.

I will also throw in here another observation/prediction: a lot of the enthusiasm around new TLDs touts spaces like .finance, .toronto are you f-ing kidding me? Any TLD over 3-characters makes an already uphill battle even harder.

Then there is the other type of TLD

…ones that I hope somebody somewhere gets off their ass and launches and those are TLDs that change the game, think outside the box and exist for a reason. I don’t even know what those are, but just to indicate the direction of my thinking I’m talking about namespaces along the lines of

  • An IPv6 TLD which only accepts domains running over IPv6 transport.
  • Apple doing the obvious and just implementing .mac as an extension to their entire iCloud infrastructure. There wouldn’t even be that tedious “sunrise/landrush” craze I talk about above – you’d wake up one morning and suddenly all of your Apple ID’s work under a .mac tld
  • .KEY which was just a method to exchange cryptographic KEYs and other authentication data or metadata via DNS. I’m talking about entire namespaces that don’t even have meaning within the context of a web browser.

These are my lackluster attempts to point in the direction of the other kind of TLD. Ones that will step up to fulfill an actual need, solve a real problem, do something useful other than whip everybody into a frenzy over owning this name or that name. The truly successful new TLDs will do so by looking at this advice, from the end of Rick Spence’s article:

If Canadians want to be top global players in natural resources, agriculture, technology, communications, professional services, or doughnuts and coffee, business leaders have to get moving now to own these top-level domain names – and the chance to build exciting new businesses around them. There’s gold in URLs.


…and doing the opposite. Nobody should be going into the new TLD business “because they are going to be hot and there’s money to be made”. That is the wrong reason to start any business.

What should happen instead is some people are bouncing off of an intractable problem of some sort. They’re noodling it, grappling with it, they have some workable approaches, nothing really elegant that just nails it. Suddenly somebody almost quips as an aside “too bad we couldn’t just register a domain name in some root for every single one of these [barrmp]’s” and suddenly the lightbulb goes on! A TLD with such and such characteristics is the solution!

I don’t know what those problems are, and I can’t even fathom from here what the design characteristics of the resultant TLD would be, but my guess is that when it happens, it will be barely recognizable as a traditional “TLD play”, and it will blast-off like a rocket.


Google Analytics Alternative