Dated: 20th May 2004 Dear Kurt, ICANN Finance Committee and Board Members and other participants, Thank you for a very detailed and quick response, I realise you have burnt midnight oil on this one, so I must upfront clarify that it is not my intention to give ICANN sleepless nights :). Please do not take my entire communication below in the wrong sense. I do criticise almost all of the points that you have brought up in your email, but that is not to mean any disrespect. I realise that we are both seeing different sides of the same issue. The important aspect in this whole exercise is the fact that we are both aiming for the same thing, namely a strong budget which allows ICANN to fulfill its goals without compromising on fairness and without being burdensome to the participants in a way which negatively impacts their business. Your assurance that even ICANN feels the same way is not even required since we are all aware that ICANN itself does not wish to see Registrars, especially deserving ones, go out of business. Having said the above, please find below the flaws in the arguments that you have sent to us. I believe we can mutually overcome these flaws, but I also believe that this will not be achieved by further patching the existing budget (for instance I consider the portion of forgiving fees of certain Registrars as a patch). Instead we will need to modify certain basic premises of this budget to overcome these fundamental flaws Assumption/Flaw 1: Batch Pool Revenue will subsidise ICANN fees The most fundamental assumption your entire email is based upon, is that of the batch pool connection revenue. Infact, out of the 50 or so paragraphs in your email almost 12-15 reference the batch pool in some way or the other (some of these paras are entirely devoted to the batch pool). A majority of your email content and argument therefore relies on the fact that the batch pool revenue derived by a set of Registrars will allow these Registrars to pay the higher ICANN fees and therefore the newer budget model makes sense. This assumption is flawed in so many respects (please do not get me wrong) that I was actually surprised this occupied so much content of the response - * Firstly and most importantly, the ICANN Board has itself approved WLS. This obviously means that in the near term future we are likely to see the implementation of the same. If a LARGE chunk of the reason for the new budget structure is the batch pool revenue, then I am surprised that the ICANN budget committee based their decision upon a revenue source that they themselves have taken steps to eradicate 3 months ago. Infact most Registrars who are participating in the batch pool themselves perceive this revenue to be a short term opportunity which could disappear any given moment. So is the ICANN Budget meant to be a short term one to be modified as business models change or as WLS gets implemented? * Many places in the email, you mention that one of the basic aspects you will use to determine whether a Registrar should pay the $19000 fee or the fee should be forgiven, will depend on whether the Registrar participates in the batch pool or not. What will you do after the WLS comes in? Forgive this fee for EVERYONE or apply this fee to EVERYONE? Your primary forgiveness of fees factor comes across as non-participation in the batch pool. So after WLS will you be modifying the parameters based on which you will forgive Registrar fees? Or will you stop forgiving them? * After WLS will you then transfer this $19000 per Registrar burden to the Verisign Registry since the income that the current Registrars make from the batch pool will directly get siphoned off to the Registry at $24 + $6 per WLS * Since you are going this granular how about taking this one step further. Some Registrars make $5k per month on their batch pool connections while some make $20k. Will you charge each Registrar differentially based on their income from the batch pool? * What about Registrars who are not participating in the batch pool today and begin to participate tomorrow. Will ICANN keep a constant HAWK watch on every Registrar on a temporary short term business model that will soon be eradicated in order to fulfill its revenues? * So why restrict yourself to the batch pool. I see some Registrars making a GOOD chunk of money in selling secondary market domain names. Some Domain Names are selling for $1 million and above and would fetch handsome brokerage to Registrars who participate in the aftermarket. Many a Registrar run secondary market places making EXCESSIVE incomes over and above their current income thanks to a creative business model. So does ICANN intend to also investigate these types of business models in the future and base their revenue upon such models? * Lets also take into account landrushes for the launch of new TLDs. As ICANN clearly knows, many a Registrar made a sizeable amount of money by auctioning off spots in the .INFO landrush and the .BIZ landrush. On the other hand many Registrars who were not aware of these landrushes made nothing. I know Registrars who made as much as $80,000 in just 3 weeks by adopting a creative business model which required no labour during the launch of a new TLD. So in the future there will be many such opportunities which a certain section of the Registrar community will avail of. Does this mean ICANN will inspect everyone of these creative opportunities and modulate its fee structure per Registrar who participates in such opportunities? * Add to this I know MANY Registrars who reserve expiring domain names for themselves and then redirect the traffic on such names to Pay Per Click programs. Some Registrars are making cool 6 figure incomes through pay per click programs by registering expired domain names. So is ICANN looking at that too? * In the email you mention that newer Registrars can subsidise this activity through the batch pool. While ICANN is up close and personal to the batch pool issue you need to realise that there are a large set of potential applicants out there looking at the ICANN website for accreditation who do not even know of the existence of the batch pool. Infact I run a separate entity called LogicBoxes where we provide Accreditation consultancy to companies. We are currently talking to several Hosting companies who do not know or care about the batch pool. I know several of these companies are potential Registrars with 5000+ domains each. They are willing to take the step forward because currently apart from the fixed $4000 fee they do not bear any fixed costs. However these potential applicants if they visit the ICANN site and find out that the annual fee is hiked up to $23000 they will not bother to even ask us or ICANN for help. So is ICANN going to state on its website that potential applicants need not be scared of the $23000 fee since the applicant can make more than that by participating in the batch pool? If not then the only applicants that ICANN will be getting will be from those who know about the batch pool and are getting accredited solely to monetize their connection threads. The genuine applicants will get discouraged from the site itself and will not apply, unaware that they can participate in the batch pool. * In the email you clearly mention - "There are a number of accreditation applications in the pipeline, including several with clear indications that the accreditation is to be used to gain access to the batch pool." and "it has been estimated by others that over 110 registrars presently derive revenue from using or selling their contractual right to access the batch pool in an effort to register deleted names". As soon as WLS kicks in a large number of these Registrars will cease to exist. They will shut their shop and leave. If you have taken a closer look at the numbers of these Registrars you will find none of them register any domain names and all the names registered through their batch pool are transferred away immediately after the first 60 days. So has ICANN given a thought to what will happen to its revenues when 20%-50% of the Registrars just walk away from the Registrar business * In your email you differentiate between Registrars who use the batch pool and those who do not. So is ICANN going to personally inform all existing Registrars who are not using their batch pool that they can do so in order to subsidise their $19000 fee? Infact as you can see from the above flaws, the direction this is taking seems to be even more precarious than what my earlier email suggested. If ICANN will base its revenue and budget model upon short term temporary creative business models of Registrars, then the consequence will be a very weak, fluctuating, subjective and seemingly unfair budget. Some participants will be advantaged and some will not and this will keep changing. A budget must be rock solid and not depend on ANY short term or creative business model of a Registrar outside of pure Domain Registration activities. I realise that many a ICANN staff as well as other key industry participants are not too happy about Accreditations which have spawned solely for taking advantage of the batch pool. I personally have voiced my concerns in the past about this practice. However the ICANN budget is not a weapon to be wielded against this issue. I am quite concerned that the ICANN budget committee sought to modify the budget to correct this anomaly or, even worse, to use this anomaly in the short term to make revenue for ICANN. Both of these objectives come out in Kurt's email and I am worried by the thought to a great extent (please don't get me wrong here). While I realise that ICANN and many others may feel that it is unfair that a set of Registrars out there are making between $5-$20k a month without significant effort, this is a completely unrelated matter which has already been addressed by approving the WLS proposal. If ICANN feels that any further steps need to be taken to address this then these steps should be taken as a separate issue and not mixed with the budget at all. For the budget to be strong, stable and long term, it is essential for it to be NOT DEPENDANT on any short term business model at all. It must only be dependant on pure domain name registration business. Assumption/Flaw 2: It makes sense to charge a fixed cost portion of $3.8 million equally to all Registrars This is another fundamental weakness in the budget. The budget assumes that since ICANN needs to spend $19000 per Registrar as a fixed cost it must charge it in the same fashion to all Registrars. ICANN forgets however as Tom Barrett from Encirca mentioned that it is a quasi-governmental body with one of its primary aim being to foster and promote competition. Here I discuss a few of my observations - * We run a website called WebHosting.Info which publishes live statistics about the entire Domain Names and Web Hosting industry. Please visit the below URL and check the Ranking column - http://www.webhosting.info/domains/country_stats/ You will find out that except for the TOP 25 countries worldwide the remaining 218 countries have all less than 50,000 Domain Names attributed to them (this calculation methodology is not based on Registrant, but is based on hosting company, however it gives a decent approximate). Also you will find that over 200 countries have less than 20,000 Domain Names. This means that with the new budget a sustainable Registrar will not be able to exist in any of these countries. Has ICANN actually conducted such research itself, considering that its goals are to promote competition? It makes more sense to adopt a budget approach which does not make life so difficult for the smaller countries to participate in the ICANN accreditation process. * By raising the fixed fee component from $4k to $23k, ICANN has straight away increased the barrier to entry. Previously a Registrar could enter the market and sustain operations with around 4000 domains and a $1 profit. Now a Registrar needs 23000 domains at a $1 profit to just cover his ICANN costs (add insurance and operations and the figures change). If the profit margin is lesser than $1 then the number of domain names required to break even further increases. In the current market, to compete with GoDaddy, Enom etc a new Registrar must sell at the same price that the larger Registrars do. This is not possible and hence will reduce applicants. * For any body like ICANN, which is responsible for creating healthy international competition, it is important to create a reasonable barrier to entry and a way for newer entrants to be able to compete with existing ones. This current budget move is counter to that. * You mention in your email - "why should larger Registrars bear a larger cost of this $3.8 million as compared to the smaller Registrars". The answer is - the fundamental way of looking at the costing by any Registrar is their per domain cost. The larger Registrars and the smaller Registrars both will always look at their per Domain Cost and determine their selling price. It is important for ICANN to ensure that the per Domain Cost for EACH Registrar remains nearly the same no matter from which country the Registrar is and no matter whether the Registrar is new or old. That is the only way a new Registrar in a small country can survive. * Infact in any government regime you will always observe that the tax component paid by the higher income brackets is higher per dollar than the lower income brackets. This allows people who are at the bottom rung of the income ladder to climb up and compete with those in the higher rung. What ICANN is doing with this budget is the EXACT reverse. It is reducing the per domain tax for larger Registrars and increasing the per domain tax for smaller Registrars. That is like a government asking poor individuals in the country to pay a higher per dollar tax than the richer individuals. How can such a government expect to sustain the budget? * Lets look at the long term view of both the models. Take two Registrars today Godaddy - whose per domain cost is $6 + $0.25 + $0.005. Godaddy can therefore sell at $6.5 and still make a profit "Small Registrar Inc" with 10000 domains in US - whose per domain cost is $6 + $0.25 + $2. This Registrar can therefore sell at $8.5 only Domain Names are commodity products and therefore a consumer can buy them from pretty much anywhere. It is obvious that GoDaddy will continue to grow while "Small Registrar Inc" will lose customers to the likes of godaddy due to its higher selling price. The more clients that "Small Registrar Inc" loses the higher it has to raise its price, thus starting a vicious circle which will not allow this company to survive. Infact "Small Registrar Inc" realises that it can buy domains cheaper from Godaddy than from ICANN itself. In the long term this effect will be felt across a large number of Registrars causing a REVERSE DRAIN of domain names. In the past 4 years ICANN has successfully created a large belt of Registrars and resulted in market share getting distributed across all of these Registrars on a gradual basis. The new budget will begin to reverse that effect by causing a consolidation of domain names towards the larger Registrars Assumption/Flaw 3: There are many new applicants in the pipeline as well as many existing Registrars who will sustain the new Budget by adding $19000 to the kitty I want to ask two questions at this stage - * How many of your new applicants (except those who have applied solely for the batch pool) have you actually spoken to and figured out as to what their reaction will be if their fee was increased from $4000 to $23000? I am of belief that a good percentage of the ones who are not aware of the batch pool will immediately question their ability to sustain. * Of the current set of Registrars that exist, how many active Registrars with less than 50,000 names have you spoken to with regards to how a increase of $19000 in their annual fee will affect their business (in a post-WLS world). I am of opinion that several of them will begin to question their ability to sustain * This new budget policy will infact not do anything to discourage the applicants who solely want to monetise their batch pool, and on the other hand completely discourage genuine applicants unaware of the batch pool, which is the opposite of the desired result. * Until now there were many mid-sized hosting companies who could apply for accreditation. These companies currently buy domains at $6.49 or so from existing Registrars. In the new budget if these Registrars get accredited their per domain cost will range from $7 to $9. There is no reason for them to get accredited, they will continue to be resellers of larger Registrars. It is no wonder that the larger Registrars find this budget a positive one. They were earlier afraid of losing some market share as some of their clients themselves became Registrars. Now that fear is gone. ICANN however will lose out on the $4000 annual accreditation fees it would have made from these companies (and there are over 2000 such companies worldwide). Moreover ICANN will lose out on being able to achieve its aim of creating international competition. Assumption/Flaw 4: Smaller Registrars will not be affected since their annual fees will be forgiven Please do not mistake my statements here as an allegation or a remark that this additional aspect of "forgiving fees" is a knee-jerk reaction. But please take a look at the budget policy yourself. Does it not appear that the "forgiving fees" process seems like a PATCH applied on the budget. Does it not seem out of the ordinary, as if a budget was drafted and then some flaw was realised and so a patch was applied to the budget to sort of workaround that flaw? The concept of forgiving fees for some Registrars and not for others initself seems so shaky as a model that the very fact that people in the committee started thinking in that direction should have been warning enough to try and rework the fundamentals - * Firstly a good chunk of the "forgiving fees for certain registrars" seems dependant on the batch pool monetisation. There are many flaws with this assumption covered already in the first section * Secondly the very concept of forgiving fees for a set of Registrars irrespective of the objectivity of criteria seems onerous. Whatever conditions you can think of will still result in certain Registrars who deserve a waiver of fees not getting it and viceversa, some Registrars who DO NOT deserve a waiver getting it. This type of treatment is actually a breeding ground for dissonance and mistrust. ICANN CANNOT have a grey area and differential charging policy for its fixed annual fee based on criteria such as batch pool monetisation or other such subjective criteria Conclusion In conclusion, once again let me request to kindly take all of my comments above in the right spirit. At some places the language maybe a little harsh, please attribute it to the fact that it is 4:00am here right now :). I do not think that ICANN has spent such a lot of time and come out with a bad budget. Infact I do think that several portions of the budget are perfect. All I think is that a certain set of base fundamentals in the budget need to be modified. I firmly believe that if the budget was passed in its current form it would be catastrophic for the industry as a whole. ICANN must revert to a simple per domain year variable fee. It does not matter what this amount is since whether it is 10 cents or a dollar, it is applicable to all Registrars and will thus enable unencumbered free competition. This model is also scalable for ICANN. The variable fee amount can be modified in any year without impacting a specific segment of Registrars. Any such modification would impact all Registrars equally. The effect would be felt by the entire industry as a whole. I would once again urge ICANN to reinvestigate its current budget and make this modification to ensure that the budget is not dependant on short term revenue models of a set of Registrars, and that the budget does not stifle competition, or discourage newer or existing participants in any way now or in the future. Yours sincerely Bhavin Turakhia Founder, CEO and Chairman DirectI -------------------------------------- http://www.directi.com Direct Line: +91 (22) 5679 7600 Direct Fax: +91 (22) 5679 7510 Board Line (USA): +1 (415) 240 4172 Board Line (India): +91 (22) 5679 7500 --------------------------------------