Home / Blogs

Why New TLDs Don’t Matter

Lost amid the furor about ICANN’s rule change that may (or may not) lead to a flood of TLDs is the uncomfortable fact that almost without exception, the new TLDs created since 2000 have been utter failures. Other than perhaps .cat and .mobi, they’ve missed their estimates of the number of registrations by orders of magnitude, and they haven’t gotten mindshare in the target community. So what went wrong? Users stopped caring about TLDs.

TLDs stopped being important about the time ICANN started. Back when I wrote the first few editions of Internet for Dummies in the early 1990s, once we got past the mechanics of getting online, most of the rest of the book was about how to find stuff, Gopher, Archie, Veronica, WAIS, with only one chapter on this newfangled WWW thing. At the time, it seemed reasonable to hope that one of the ways to find stuff would be that the DNS would grow into a directory, along the lines of what .MUSEUM has tried to do.

I expected WAIS, a full text search system, to be the next big thing, since like nearly everyone else I didn’t anticipate the way the web would absorb every other application. But in a sense I was right, because the killer app for the web was and is search engines. These days I know a lot of people whose home page is Google, and who have no idea what the difference is between the Google search box and the browser address box. Domains and even URLs don’t matter. They type some words into one of the boxes, and if they get to a place they like, they bookmark it.

One of the oft cited examples of groups who would benefit from a new TLD is small language communities, that it would somehow make an online community possible. In reality, while it’s important that their writing system is included in Unicode, and that there be display fonts and input methods available for browsers and MUAs, the domains don’t matter because nobody’s going to type a domain more than once. After that, the sites are going to be bookmarked and the e-mail addresses will be in the address books.

At this point, the only reasonable argument I can see for a new TLD is branding, and there only in areas where brand vs not-brand is interesting. Among the reasons that .COOP and .AERO are flops is that nobody cares about real vs fake co-ops or real vs fake whatever it is you have to be to get into .AERO these days. It looks like .MOBI will work because .MOBI sites work on phones with tiny screens, mobile users care about that, and .MOBI has a compliance process to check that the sites work like they’re supposed to. Something like .BANK might be useful to help distinguish actual banks from phishes. Other than that, the main motivation for new TLDs seems to be wishful thinking combined with faith-based budgeting.

One of my correspondents commented out that the existence of typo-squatting as a profitable endeavour provides some evidence that some significant component of the Internet still uses the address bar.

Ah, but type a word or two into the IE7 or Firefox address bar and if they don’t resolve as a domain name, guess what happens—the browser passes them to your favorite search engine. People really don’t know the difference, and the browsers encourage that confusion. Hence the typosquatting basically steals a name out from what should have been a search where the search engine would offer spelling corrections.

By John Levine, Author, Consultant & Speaker

Filed Under


If one gets a non-resolving Gary Osbourne  –  Jul 8, 2008 10:58 PM

If one gets a non-resolving name one doesn’t necessarily get one’s ‘favorite’ search engine by default. SFAIK with IE7 one not surprisingly gets Microsoft’s own live.com search engine, with FireFox one gets Google. The rest of your article is bang-on. I suspect at least part of the push for new TLD’s comes from potential and existing registries, registrars and resellers, and ICANN itself, for monetary gain. I don’t hear the average net-user asking for it (they probably want less), and the IP folks don’t want to pay even more protection money.

Default search John Levine  –  Jul 8, 2008 11:34 PM

It’s true, by default you get the browser vendor’s favorite search engine. Changing IE to a different search is pretty easy, a couple of clicks on the icon in the search box, while changing Google is just barely possible (albeit documented) by editing about:config.

But I doubt that most users care. Both Google and Live Search are adequate for most purposes. As far as faking out search when people type a keyword, if you can’t get your web site to the top of a Google listing for under $100,000, something is seriously wrong.

URLs aren't going away, users are just now learning. Jeremy Armer  –  Jul 9, 2008 4:09 PM

I have to disagree with the general tone of your article, even though your facts are correct.

I’ve heard this argument a thousand times and it just doesn’t hold up. It’s true that most users today don’t know the difference between the search and address bars, but are you really sure that will be true tomorrow? The web is still very young, people catch on eventually.

In Japan, there is a trend in ads to advertise a search term instead instead of a URL. If my ad company did that I’d fire them. Search is dynamic, URLs are not. Sure, it might work for an industry group (search: “Milk is healthy!”) but not for specific sites. Maybe you’re on top of the results for “dating” today, but where does that get you when a competitor takes the lead? I’m sure they’ll thank you for bankrolling their advertising…

I don’t think the “flood” of new TLDs will be as big as most think. The expenses are high and it doesn’t make sense for every purpose to operate their own registry. Coke and eBay are the most cited examples. For eBay it makes sense, for Coca-cola, not so much. eBay will probably get their own TLD because people will very quickly get used to typing “cameras.ebay, cars.ebay and toys.ebay” into the address bar.

Who’s going to type “coke.coke” in? No one. Even if they (Coke) did start a TLD it would only be to protect their brand. Not that ICANN would allow anyone else to start a .coke anyway.

The reason that most new TLDs have failed up until now is because the registries have bungled their opportunity. With .mobi, we have a shining example of how a new TLD can be successful even when there is absolutely no need for it. With .info, we have a shining example of how effectively a registry can squander what should have been a home run.

If the new TLD operators make the same mistakes, they’ll fail. OTOH, if they come up with some solid business plans, target the right market and BRAND BRAND BRAND I think we might have some big winners.

Will we see a semantic web? Maybe not. But that will have nothing to do with TLDs. Search is great when you’re looking for something, URLs are imperative when you are looking for someone.

search terms John Levine  –  Jul 10, 2008 6:03 AM

Your example of companies advertising search terms in Japan just reinforces the point. You may not like it, but it's the way that the Net is going. By the way, I do agree that the only possible hope to make a success of a new TLD is to make it a brand with semantic content, like .mobi is doing. I don't see much market in defensive brand protection, since ICANN's process has too many hoops to make squatting very likely.

People learn. Jeremy Armer  –  Jul 10, 2008 3:21 PM

Look, I understand where this line of reasoning comes from. I too have seen many users using teh google as the address bar. I have also seen those very same users, the "digerati illiterati", learn in reasonably short order how this thing we call the internet really works. My parents, co-workers, friends and other random people have all realized (all by themselves) how URLs work, eventually. My example of the Japanese ads does not reinforce this point. All it reinforces is that you should not (completely) cede your online marketing to clueless ad execs. How do you reconcile the fact that search is dynamic? Would YOU advertise a search term for YOUR business site? Even if that search term is your exact company name it does not guarantee the people looking for you will find you. Google and others change their algorithms all the time, competitors want to be on top of the search results too, other sites that mention yours may have a higher search relevancy and weighting. To advertise like this is foolhardy. It has nothing to do with what I like, it's simple logic. Search is dynamic. Search is powerful, it's great, it is a killer app. But it is NOT the end-all, be-all of internet navigation and it won't ever be. To say so is akin to saying we should get rid of postal codes and rely on google maps to send packages. Good luck with THAT! Not everyone's minds work in the same way. When I'm looking for a physical place and some tries to give me "directions" I always listen politely and then ask "What's the address?" Sometimes when you're looking for something specific, search is not much help. I'll give you an example. I love music. One of my all time favorite bands is "Big Black". Try typing THAT into google and see where it gets you... Yes, you can be more specific and find a relevant page but that negates your argument regarding end-users. End users aren't search experts. Isn't that your point? People can just "type in what they're looking for" and google magically finds it for them? Not in all cases. It won't ever work like that. Not 100% of the time.

Bad example... Jeremy Armer  –  Jul 10, 2008 4:19 PM

I wrote: "One of my all time favorite bands is "Big Black". Try typing THAT into google and see where it gets you…" Bad example, 1st non-sponsored result is about the band. But that's the power of wikipedia for you. I still stand by my argument.

I'm curious - by what Randall Zaia  –  Jul 10, 2008 2:47 AM

I’m curious - by what standard have these TLDs been “utter failures?” Because they didn’t challenge .com for supremacy in the market?

If you look at the number of registrations for even the “failed” TLDs I don’t think their operators have lost money.

Failed TLDs John Levine  –  Jul 10, 2008 6:00 AM

I believe that .info and .biz make money, but the rest of them are pretty dubious.  Do you even know how many registrations they have?  There are, for example, about 6000 names in .pro, and 14,000 in .coop. There are 5600 in .aero, not counting 25,000 reserved and presumably unpaid airline and airport codes. And then there’s .travel, whose size I can’t tell because they have (in egregious violation of their ICANN agreement) refused to provide zone file access, but their SEC filings, on which I’ve commented in the past, say it all.

I agree with you entirely Chris Linfoot  –  Jul 10, 2008 11:24 AM

I guess I should have posted this blog post of mine here at CircleID…

Language communities propser from having Karaitiana Taiuru  –  Jul 10, 2008 11:35 PM

Language communities propser from having thier language in use. Having a TLD “does” assist in the revitalisation process and the general perceptions of the the younger community members.
In NZ, Maori have prospered by having thier own domain within the .nz hierachy.
More importantly for the younger generations who look to the predominatley Euopean structured Internet and think thier language is too old fashioned as they are not represented in a TLD while all the countires in the world are. There a copious more examples of the benefits.

Language domains John Levine  –  Jul 11, 2008 3:40 AM

If you say so, I can believe that it is of some use to have .maori.nz, although I expect that maori.org would have been as good.

So what about .maori would make it worth NZD $130,000 as opposed to the $500 that .maori.nz cost or the $15 that maori.org would cost?

One of the advantages of Ram Mohan  –  Jul 11, 2008 12:08 PM

One of the advantages of a new TLD is that you get to set new policies and a clean start.  This was one of the prime motivators behind the .mobi effort (which I helped get started).  Branding was important for .mobi - and with a small screen being bought in millions worldwide and most web sites building only for the “wired” internet, there was a good opportunity to capitalize on.

Having a lot of TLDs ultimately benefits .com—it remain meaningfuls and easy to understand compared to the confusing choices in the marketplace.  And - it’s quite hard to get a new TLD off the ground and to be successful - I’ve helped with a few, and it’s a lot of work, persistence and money to even have a chance.

Witness the fact that .com still sells more names every day than all the other gTLDs combined.


Fresh starts Jeremy Armer  –  Jul 11, 2008 5:41 PM

I agree Ram. The biggest advantage new TLDs will have is a fresh start in regards to policies. If they stick to their guns, new TLDs might actually mean something to the public. Witness .org, even after a chaotic start most of the public realizes that .org means non-profits in most cases. While this isn't a rule, it was the intention and it is the way it has played out for the most part. While this hasn't happened in the .net landscape to as great an extent, a trend is emerging for .net to be used for social networks. Not the original types of networks intended, but maybe a more fitting (and numerous) one. More importantly it lifts the stigma of being just a "second-class .com". I'm not a big fan of .mobi (no personal or professional offense intended) but I do admire the specialty niche aspect and the branding that has been done. If .info would have insisted on quality control from the beginning it wouldn't be known today as a phisher's haven. New TLDs have an opportunity to create meaning. Many will fail this test and become irrelevant. Those that take advantage will succeed. TLDs matter.

Jeremy,We agree on many things.On Ram Mohan  –  Jul 11, 2008 7:11 PM

We agree on many things.

On INFO - yes, I realize that it has acquired a reputation that needs to be burnished.  I’ve setup a team inside my company to work on cleaning things up - for example, we’re setting up anti-abuse policies that will result in quicker and more certain take-downs or mitigation of phishing or mal-ware or drive-by sites for .info.

Long term, a registry that takes itself seriously and executes with precision and persistence will matter.  TLDs do matter, but it’s not the “get rick quick” scheme that many seem to think it is.

New TLD's have big shoes to fill Holly Kolman  –  Jul 12, 2008 9:36 AM

Thank you for recognizing the value proposition of .mobi. I don’t work for mTLD (the registry) but have had the opportunity to see what kind of effort and energy is regularly invested by their top management in making .mobi a success, as well as a number of people who have crossed the line from domain name investor to developer as a result of the demands of owning a .mobi name.

New tld’s will certainly have an impact, and in the case you mentioned of ebay I think some will be quite useful. Others will be a financial drain on companies trying to protect trademarks and run the risk of turning the internet into alphabet soup. Remember when Nokia first applied for the rights to .mobi and was turned down? They came back with some big name investors (Google and Microsoft) and were told yes.

What kind of oversight will the new tld’s have I wonder? What laws will be followed regarding trademarks and copyrights?

Comment Title:

  Notify me of follow-up comments

We encourage you to post comments and engage in discussions that advance this post through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can report it using the link at the end of each comment. Views expressed in the comments do not represent those of CircleID. For more information on our comment policy, see Codes of Conduct.

CircleID Newsletter The Weekly Wrap

More and more professionals are choosing to publish critical posts on CircleID from all corners of the Internet industry. If you find it hard to keep up daily, consider subscribing to our weekly digest. We will provide you a convenient summary report once a week sent directly to your inbox. It's a quick and easy read.

I make a point of reading CircleID. There is no getting around the utility of knowing what thoughtful people are thinking and saying about our industry.

Co-designer of the TCP/IP Protocols & the Architecture of the Internet



Domain Names

Sponsored byVerisign

Threat Intelligence

Sponsored byWhoisXML API


Sponsored byVerisign

Brand Protection

Sponsored byCSC

IPv4 Markets

Sponsored byIPv4.Global


Sponsored byDNIB.com

New TLDs

Sponsored byRadix