Home / Blogs

New TLD Vertical Integration, Market Forces and the Path of Least Compliance

The ICANN Board will soon make many decisions, one of which is to decide whether to continue or reverse ICANN’s longstanding policy of vertical separation of registries and registrars. Since new Top-Level Domains (TLDs) are supposed to benefit registrants with lower prices, choice and what we trust will be a decision for ‘market-differentiated’ competition, the Board will no doubt consider market forces as well as compliance and enforcement issues in choosing the path that can maximize these goals.

Market Forces

Arguments are well-made that allowing 100% vertical integration (VI) would require little or no compliance oversight of the industry and that “market forces” will effectively police any abusive behaviors by new TLD registries. Unfortunately, I’ve lived long enough to know that the opposite is more the truth, and mistakes that are made on large scale activities leave marks for a long time, if not irrevocably.

Those of us who oppose allowing 100% vertical integration do so for a number of reasons, but one of the most salient points for opposing VI comes in a statement attributed to Register.com that [paraphrased] notes: An affiliated registrar would have de facto better access to registry systems, deleted names and operation support. Similar to the above this would allow the affiliated registrar to receive additional benefits that are not available to competitors which the affiliated registrar could pass on to its customers. Registrars in a 100% integrated structure will have unfettered access to registry data that would enable anyone with such access the ability to engage in front running and other gaming of the domain name registration process with ruthless efficiency. An integrated registry will also have visibility into lookups through any one within its ‘family of registrars’ (hundreds of registrars controlled by one entity which in and of themselves are an opaque world only known to, and deciphered by, those who manage them).

For non-technophiles of the ICANN community, like me, it was very unsettling when we learned about the impact of front running. And previous forms of front running were based on registrant lookups visible only by a single registrar. A vertically integrated registry/registrar or backend registry service provider/registrar will have visibility to all lookups for all names across the entire TLD in question. As metaphors go, here is one I understand: Instead of having access to a single tip on one stock based on insider knowledge, it’s like having all insider data on all the stocks traded in an exchange.

Practically speaking, the recent S-1 statement filed by Demand Media as reported in DomainNameWire.com gives one a good sense of what is to come if 100% vertical integration is allowed:

“Demand Media CEO Richard Rosenblatt has referred to the data from eNom and its value to the company before, but its newly filed S-1 explains the value succinctly.

It uses the data not only for its content business, but also to determine which domain names to acquire for its own portfolio

[to whit]:

‘Proprietary Data. In providing registration services for over 10 million domain names, our Registrar resolves an average of over 2 billion domain name system queries per day. Our Registrar also serviced, on average, more than 3 million domain name look-ups per day from potential customers seeking to register new websites or purchase existing domains during the first six months of 2010. These queries and look-ups provide insight into what consumers may be seeking online and represent a proprietary and valuable source of relevant information for our platform’s title generation algorithms and the algorithms we use to acquire undeveloped websites for our portfolio.’” —DomainNameWire.com, August 6, 2010 (emphasis added)

To get a glimpse of the gathering storm I fear, simply ‘find and replace’ the word “Registrar” with the words “integrated Registry/Registrar” or “integrated registrar/backend registry” in the above quote.

The eye-popping assertions and allegations in the KnujOn Registrar Audit, despite withering critique of the study from certain quarters, was a ‘wake up’ call for many in the community to the serious fraudulent activity that occurs in our current, small, generic name space. So it is no surprise that the VI Working Group is now focused on addressing ‘harms’ in an effort to ensure that a fair and level playing field is maintained for all who are concerned with expanding the DNS. Best explained by our co-chair, Mikey O’Connor,

“The working group is compiling a list of possible harms that in some way bear on the issue of Vertical Integration. We are developing an approach to analyze the harms on this list in a variety of ways with the hope that we can uncover new ideas that can become part of consensus recommendations. At this stage, the list of harms [noted on the Working Group Wiki] is an unfiltered compilation of every harm that anybody has raised—we are planning to substantially refine and flesh out the list before using any of it as a basis for recommendations.”

Through comments readers might add to this article, I am sure that the VI WG will be pleased to receive your contributions to the list of harms with any examples you cite. With the prospect of hundreds of new TLDs and given ICANN’s track record of reactive and tardy compliance, we would all be whistling past the graveyard to assume that allowing 100% vertical integration will mean less compliance challenges for ICANN.

Front running recently came to the fore in early 2008 because of a lawsuit that alleged Network Solutions had engaged in the practice. Reaction within the ICANN community was swift and harsh. While the lawsuit was eventually settled, the implications of potential front running were stark. By the simple act of looking up the availability of a domain name, a registrant could see that name taken out of the system by the registrar and be forced to purchase that name from only that registrar and at a higher price than originally offered. By taking the domain name for itself, the registrar strips the registrant of the ability to shop for the name at other competitive registrars for a potentially cheaper price. The ICANN community quickly realized that front running raises prices for registrants and destroys competition between registrars.

The question the Board has to ask itself is: How much abuse will ICANN permit to take place between the launch of new TLDs and the point in time when we realize there is a problem? The second question will be: How can ICANN remediate the abuse that occurs during that time? 100% integrated registries/registrars and backend registries/registrars will—without question—have a clear shot at raising prices and taking high value names off the market for themselves. These are issues that the “hidden hand” market forces can’t address. Reliance on market forces implies that (1) there will be a true TLD market (and no one will know the answer to that question until we gain the perspective of time), and (2) registrants have a choice to take their business elsewhere. The fact is that front running destroys consumer choice by forcing the sale of the name through one registrar only. Where there is no consumer choice, there is no ‘punishment.’ Specifically registrants who don’t view other TLDs as substitutes for the TLD they want can’t dole out punishment to the abusive TLD operator. They are, plain and simple, out of luck. And if current and prospective registrants are not being served through all of ICANN’s multi-lingual debate and dialogue, what is the point?

The Path of Least Compliance

When trying to resolve complex issues like vertical integration, there will be a natural tendency to seek the ‘path of least compliance’ burden for ICANN. Arguments about cost, methodology, effective sanctioning and follow-up all arise and they all have some validity. But such arguments hold no water in an open and transparent ICANN. I, and others, argue that without police on the streets and laws to abide by, chaos will ultimately ensue. A sad commentary on our times, but human nature in the 21st century is unfortunately not what it should be.

Those seeking the path of least compliance argue that 100% vertical integration can be allowed and use of registry data can be prevented through the imposition of Chinese walls between the integrated registry/registrar. Monitoring Chinese walls for potentially hundreds of new, vertically integrated TLD operators would significantly increase ICANN’s compliance burden and market forces cannot be relied on to keep abusive practices in check. Moreover, the critical aspect to preventing abusive practices in the longer term is to take a proactive (not reactive) approach to compliance and enforcement. But, compliance—lack thereof—has been ICANN’s poorest capabilities demonstration to date. One must ask oneself what has suddenly changed that the community should expect ICANN to create, deploy and enforce new compliance testing systems that will provide alerts whenever ‘a wall’ is transgressed? While the Great Wall of China may have been effective in keeping out the Huns, corporate Chinese walls are notoriously ineffective and policing them is not an easy task. Neither, will the task be simple for the small ICANN compliance team that will be operating under new leadership while ICANN moves to the application phase.

There are some proposals within the Vertical Integration Working Group, one of which I am a proponent of (RACK+), that call for continuation of the vertical separation policy and the implementation of bilateral ownership caps between registries and registrars, as well as between backend registry services and registrars. Ownership caps already exist in the ICANN contractual framework and they provide a deterrent against easy registrar access to registry data. Ownership caps, in fact, require less compliance effort and can be effectively checked through an annual declaration by the registry and registrar and validated by a third party audit firm at the registries’ and registrars’ expense. Ownership caps of 15% have a track record of vigorous competition between registrars and lower prices for registrants. Importantly, ownership caps reinforce ICANN’s policy of non-discriminatory treatment of registrars by registries—a policy that will be continued in the new TLD round.

While there are no perfect solutions to complex issues such as vertical integration, ICANN’s Board should reflect on the respective roles that registries, backend registries and registrars play in the market. Registry operators and backend registry services providers (who are the functional equivalent of a registry) operate a TLD as a trust to the community. Registries are expected to operate the TLD in a responsible manner consistent with the values of the community and with a constant eye toward the security and stability of the DNS. And Registrars are expected to provide value with honest services in their retail capacity. Each of these components manages a part of a global, public good. These important roles should not be diminished or easily discarded in the new TLD round in the name of unfettered market forces that no one has yet seen.

In my view—particularly with all of the brand new moving parts that the ‘open’ TLD application process is going to test—repealing vertical separation policy at this time would likely result in significantly more consumer complaints and more, not fewer, compliance headaches for ICANN.

By Ronald N. Andruff, President at ONR Consulting, Inc.

Filed Under


Ron, I remember our talks from Brussels Volker Greimann  –  Sep 22, 2010 2:10 PM

Ron, I remember our talks from Brussels very well, and if I remember correctly, the main reason for you to favor vertical separation in the new gTLD market was your perception of ICANN being unable to monitor and enforce compliance on the scale required. I do agree that compliance monitoring and enforcement are needed, I do not however agree that vertical separation is an adequate means to prevent abuse.

You cite frontrunning as an example for abusive behavior in the domain world. However, this occurred in a separated environment, as you correctly pointed out. If we took frontrunning to the extreme, a scenario as you describe is imaginable. Is this scenario however a direct result from vertical integration? Is it even likely?

The scenario you describe could just as likely occur with full vertical separation. The registry could act in collusion with one of its registrars and make the same data available to the registrar, who could then engage in the abusive behavior, while feeding a revenue split to the registry. The same forms of abuse are possible and not even substantially complicated with full separation as compared to partial or full integration. But they are hidden a lot better and harder to detect with separation, thus making compliance monitoring and enforcement that much harder. With integration, the level of deniability is greatly reduced.

Even in current gTLDs, there is ample examples of vertical integration: gTLD registries or backend providers such as AFILIAS are partially owned by registrars. The registry for one of the smaller gTLDs is closely affiliated with one of its registrars. CORE is a collective set up by registrars.

Regarding the likelyhood of abusive behavior, one should also bear in mind that none of the new gTLDs launched in the coming year will come even close to the numbers of registrations that the large gTLDs enjoy. Their value and success will to 90% rely on the trust of the communities they serve or the audience they target. Abusive behavior that hurts the consumer will hurt the TLD. A failed TLD that is not picked up by the consumers or search engines is a total loss of investment. While sex.com may be worth millions, sex.fail could be worth nothing.

Finally, vertical integration has been proven to work in many ccTLDs where registries also act as registrars without any claims of abusive behavior. Could those ccTLD registries abuse their position of power? I imagine so. Will they do that though? Not likely, as it will harm the success and value of the TLD in the long run. Many ccTLD registries are also cooperating with registrars as registry backend providers; other state authorities are even chosing registrars as their registry operators (see: .so offered by an affiliate of GMO) as they know the value their experience can offer to their TLD. The registrar, which technically able to abuse their access is still unlikely to do so as such actions will hurt their reputation and business in the end.

Effectively, a harm is a harm only if there is actual abusive behavior. A potential harm is like a potential crime. While one could at any time engage in illegal behavior, the majority usually does not and those that do are punished. Just as there are laws against illegal behavior in “the real world”, there need to be rules and penalties in “the domain world” as well. We may not be likely to detect all abusive behavior, but that is absolutely separate from the issue of vertical integration. Abuse can happen regardless of the outcome of the VI debate.

Potential harms can and will still occur if there is a will to abuse the system to its own advantage. All potential harms can and will still occur if there is a will to abuse the system to its own advantage. All you have essentially done with the prohibition of vertical integration is to limit competition in the registry market to the advantage of encumbent providers.

What we need instead is what we agreed upon during our talks in Brussels: Strong and effective compliance monitoring by ICANN, swift and painful penalties for abusive behavior.

We need to take the those ICANN board members by their words when they stated in Brussels that whatever the PDP WG decided would be required for compliance monitoring, they would put in place.

Look at it from the consumer side Antony Van Couvering  –  Sep 22, 2010 5:57 PM

We are going to get to an unrestricted marketplace at some point.  Absent an abuse of market power, efficiencies drive down costs and promote competition—which means better service and lower prices for consumers.  ICANN is supposed to act in the public interest, not the interest of a particular group within ICANN.

While I generally agree with Volker’s points, and disagree with Ron that vertical separation or integration will substantially affect fraud and bad behavior among registries and registrars, they are both biting the wrong end of the sandwich.  ICANN is a not-for-profit California corporation, which is supposed to act in the public interest.  We need to look at this issue from the point of view of the public.


If you favor competition, innovation and fair Constantine Roussos  –  Sep 22, 2010 10:29 PM

If you favor competition, innovation and fair business then 100% vertical integration is a no brainer for new entrants. Abuse of power and shady business practices is not a viable excuse for not allowing new entrants to choose how they will distribute their TLD. The market will dictate everything in the end. The ones with the better product or bundle will win. The launch of the new TLDs will make scarcity of domains a thing of the past. Every business should be allowed to choose who they want to work with, especially if they are adding value beyond what the business can achieve.

The concept of Vertical Separation is an antiquated concept without any economic value to new entrants who want to innovate. If new TLDs are going to be launched to increase competition and bring innovation in the space, then the strategy of ICANN should coincide with that and be aligned with that. Consistency is key. If vertical separation is voted for new entrants then I am afraid that ICANN should scrap the whole new TLD program. It will be another indication that new TLDs are about applicants filling the pockets of ICANN, irrespective of the goal of the new TLD initiative.

I have presented strong examples how the registry can add value and be abusive. Pricing is a market demand and supply component. In a marketplace where scarcity decreases, pricing will take its natural course if you want to stay in business. In regards to high prices and unavailability of premium domains due to abuse, I believe this is a moot point in many cases. All premium domains for .mobi, .asia, .me and I am expecting new TLDs will be auctioned out. Why isn’t that considered abuse? furthermore, in our case, premium domains will be reserved for registrant-generated content and marketing i.e for the benefit of the community.

The votes were in and the most popular proposal was vertical integration. Any economist would agree with this. I have discussed this numerous times with many organizations who are baffled why ICANN might be imposing such a rule to new entrants while on the other hand pushing for innovation and competition. Why should registrars get a free pass to sell to their domainer clientele if some registries believe it might hurt their brand because of the propensity of domainers not developing domains and parking domains with junk?

It is ICANN’s responsibility to be consistent with its policies in regards to launching new TLDs and not punishing new entrants because of some bad apples. The web is not dead yet. Let us hope ICANN votes for innovation and competition as opposed to pleasing a select few big players. At least, applicants should be given the opportunity to be allowed to vertically integrate in a case-by-case scenario. Applicants who will be competing for registrar shelf space for distribution would not care about integration because all they will care about is selling domains in volume. What happens to the limited few who will be adding value to their communities through bundling and service integration and are focusing on quality registrants rather than volume registrations?

I think ICANN will add exceptions for Vertical Integration for the select few. There is no ways around it if ICANN wants to be viewed as fair and consistent with the nature of increasing competition and innovation in the domain industry by launching new TLDs. I repeat, there is no innovation in launching new TLDs just for their novelty name. All it does is increase the bargaining power of registrars for distribution (shelf space) and the pockets of the big registries that will offer their backend services to the new applicants.

Constantine Roussos

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



Threat Intelligence

Sponsored byWhoisXML API


Sponsored byVerisign

New TLDs

Sponsored byRadix

IPv4 Markets

Sponsored byIPv4.Global

Brand Protection

Sponsored byCSC

Domain Names

Sponsored byVerisign