I don't know that all these registrations are defensive, Antony. That's why they are estimates. But when the same names are being registered across TLDs and also on the same hosters, they point to a brand being protected. This is the Cross TLD registration domain. The domain and hoster in .com (for example) and the hoster for the same domain name in other TLDs are identical. The most common pairing is the .com and .net version of the domain being registered. This is a visual representation of the data for .com TLD: HosterStats.com: .com Brand Protection Registrations The com/net and com/net/org registrations are the commonest. But the super brands are there too. These are the brands that are registered right across the gTLDs with the same hoster. The data supports your studies concerning defensive registrations in the newer TLDs. But the important point is in what constitutes a defensive registration. Most Brand Protection centres on large businesses protecting their trademarks and brands. However that's just the tip of the iceberg. There are many small business owners and website owners who will register their domain across various TLDs where they think that it may be at risk and of course where it is still available. These small players are, to a large extent, ignored by the Brand Protection service providers like MarkMonitor and others. I'll probably have to put all this data in a few Excel spreadsheets. The other important thing to consider is the weighting of the registrations. The .com registration would be, in the current survey, the most important registration except when it comes to ccTLDs. And the problem with ccTLDs is that the ccTLD registries don't provide open access to their zones for comparison. The methodology is straightforward: take a domain and its hoster in a TLD, check if it exists in other TLDs and if it does, check if the hoster is the same. And repeat for every domain in the TLD being checked. Simple in theory but large scale in practice. :) But you've got to think in terms of the TLD being checked and the registrations in all other TLDs being relative to that TLD. Otherwise it gets very confusing.

I agree Michele, The new TLDs have to prove themselves attractive to the public. The .tel sTLD was excluded from this survey because it uses the same nameservers for all of its domains. Brand protection will play a part in these new TLDs but not to the same extent as it does with the more established ones. It will still be a gamble for brand owners and some will automatically register their brand in the new TLDs - these super brands can be seen in the com/net/org/biz/info/mobi/asia figure of 3,928 domains. Some brands will have multiple or different hosters. The regional ones are the most obvious TLDs where brand protection may not be as prominent. If a business doesn't sell into the region covered by the TLD, then it doesn't make much sense to register the domain in that TLD. This is what has hit .asia sTLD - the main TLDs (.com/net/org) are Global. It makes sense for a business to hit a Global market with one registration. If I am selling widgets in Ireland and my market is completely local, it doesn't make sense to register widgets .asia - but it does make sense to register widgets .ie. The new TLDs are not just competing with the gTLDS. They are competing in a far more diverse environment of ccTLDs and gTLDs. Many of the small business owners will have to take a gamble with some of the new TLDs - so will the backers of new TLDs.

"As for Google.food - there's a nice opportunity for a food orientated search engine that could help people find ingredients, restaurants or take-aways with the aid of Google Maps. Just because you can't see the opportunity in registering a brand in a new TLD does not mean that somebody else cannot see that opportunity. And if they register a brand name that they don't own, set up a website or service that trades upon the good name of the brand owner, then that's a problem." Google is a rare company because they are in the business of making money through "keywords." They can pretty much monetize all TLD keywords. Other example includes Google.travel. Both Microsoft Bing and Google have been very active in the "travel" keyword game. I assure you they can get their return of investment for a "keyword" TLD registration in the first hour by feeding content related to the associated TLD. I understand the basics of branding but since Google controls search, it would be very difficult. Just the mere fact that search engine results are based on relevant links coming from "credible" websites with related content, it is very difficult to be discovered. If there is a suspicious site, it will be reported to Google by the user community. Like you already mentioned, the brands will find how to monetize the opportunity. On aggregate, they will find ways to make more money than they lose. You can not just claim "costs" from the release of new TLDs as the sole determinant. Profit is revenues minus costs. I outline the word "revenues" generated because all everyone talks about is "costs." Do you believe that the big brands can not recoup their registration investment and have their marketing departments work their branding magic? Do not get me wrong, even I expect minimal cybersquatting. However, I do not buy the concept that TLDs will not bring social benefits overall. I guarantee you the big brands will change gear once new TLDs are introduced and be the first to innovate and use them towards their advantage. Does anyone actually believe that brands will not use them towards their financial advantage? "Take music.us for example. You also own music.mobi. Now if .music is successfully launched, will you drop these registrations or simply point them to the .music registry website? If you had had the chance to acquire the term 'music' in other TLDs would you have declined that chance? " I have declined many terms "music" in other TLDs. Those included ".net" and ".travel." I declined the first based on the price requested and the other one because I saw no apparent value to it. I do own music.us, music.biz, music.mobi, music.pro, music.fm, music.im and others because I see direct value creation potential in regards to music business, mobile, industry professionals, radio and messaging. Do not get me wrong here, once .music launches, I would do what is best for the extension to benefiting .music. We have not launched yet because we are waiting on .MUSIC. The delays have certainly resulted in lost income as well as not yet launching the platform we have built. So yes, there is an opportunity cost associated with everything happening at ICANN and new TLDs. "Many website owners are protecting their brand name (for which may not have a registered trademark) by simply registering the .com or .net version of it. You seem to be thinking in terms of Big Brand protection and ignore these small website owners who have to deal with protecting their brands and their businesses." Yes, I faced the same issue with the most recent .CO launch. I defensively registered a lot of my brands / company names not because I wanted to but because i had to. Also I missed out on some because I was not quick enough during the landrush. However, there is a key difference between and overly generic TLD eg. .CO, .SITE, .WEB over a specialized TLD such as .MUSIC or .HOTEL or .SPORT. I believe that overly generic TLDs with no specialized value added should not be launched. Those general TLDs pose the biggest danger for user confusion and exploitation. The issue with brands is that you have many brands with the same name but are in different industries. Who defends what TLD? The new TLDs will take to time to get into the minds of consumers, let alone have a meaningful cybersquatting impact. Constantine Roussos .music

" Like you already mentioned, the brands will find how to monetize the opportunity. On aggregate, they will find ways to make more money than they lose. You can not just claim "costs" from the release of new TLDs as the sole determinant. Profit is revenues minus costs. I outline the word "revenues" generated because all everyone talks about is "costs." "

" However, there is a key difference between and overly generic TLD eg. .CO, .SITE, .WEB over a specialized TLD such as .MUSIC or .HOTEL or .SPORT. I believe that overly generic TLDs with no specialized value added should not be launched. Those general TLDs pose the biggest danger for user confusion and exploitation. "

" The issue with brands is that you have many brands with the same name but are in different industries. Who defends what TLD? The new TLDs will take to time to get into the minds of consumers, let alone have a meaningful cybersquatting impact. "

A hoster is a webhoster grouped by its nameservers, John, Basically the zonefiles are processed by grouping all a hoster's nameservers under the domain name of the host and using the domain name to identify the hoster. It is the same kind of shorthand that NameIntelligence/Domaintools.com and Dailychanges.com use. The processing these statistics is done at zonefile level rather than WHOIS level. What emerges is a landscape far richer and more diverse than the registrar level view. Based on com/net/org/biz/info/mobi/asia/ie/us domains, there are 5075 hosters with more than 1000 domains hosted, 33530 hosters with between 100 and 999 domains hosted, 32075 hosters with between 50 and 99 domains hosted, 154109 hosters with between 10 and 49 domains hosted, 270250 hosters with between 2 and 9 domains hosted and 795010 hosters with just one domain hosted. Most people see the market for .com and the other TLDs as one simple global market. The problem is that it is actually a set of country level markets. While there are super hosters which are transnational (Godaddy, Enom and 1&1;) there is a high level of concentration of distinct country based markets with the bulk of a country's domains and websites hosted on those hosters and IP ranges. Depending on the level of maturity of a country's market, between 10 and 80% of a country's domains can be hosted outside of a country's hosters/IP ranges. As such a market matures the level of externally hosted domains will decline. When it comes to ccTLDs, that level of in-country hosting will increase massively so that over 80% of a ccTLD will be hosted on the hosters/IP ranges of the associated country. Some of this data can be used to profile a particular brand before ever having to go near WHOIS data. If a number of ccTLD versions and gTLD versions of a domain are hosted on the same hoster and the websites all have the same IP and the domain in one gTLD is not hosted on this hoster and IP then there is the possibility that this domain is not owned by the brand owner. The complicating factors are that not all ccTLD domain owners own the equivalent .com domain, the website data adds an extra layer of complexity and some domains' websites may be left parked or not set up. It is a very crude metric but it does work for the larger brands but the more fringe gTLDs that are added, the more problematic it becomes (perhaps it might require a weighting system with a .com having more importance than a .asia for example). The .eu fiasco was an almost perfect test for this simple brand profiling. The non-EU owned cybersquatted and cyberwarehoused .eu domains had a completely different characteristic to the typical EU brands and domains. In addition to many generic terms, the cybersquatters and cyberwarehousers targeted .eu versions of domains that were really small local brands in various EU countries. Thus the potential cyberwarehouser/cybersquatter hoster in .eu would have had, in most cases, a completely different profile (minimal or no Cross TLD domains) than a typical hoster dealing with an EU market. The irony is that this is exactly the basis for the kind of "inference engine" that Eurid was talking about developing in the Brussels ICANN meeting.