Tutorial: Domain Tasting ICANN meeting San Juan, Puerto Rico 24 June 2007 >>JACQUELINE MORRIS: Good morning, everyone, we'll be starting in a few minutes. But in the meantime, if you haven't yet got a copy of the issues report, there are English language versions here on the podium for you. good evening, everyone, and welcome to the domain tasting workshop. I'm Jacqueline Morris, the chair of the At-Large Advisory Committee. And I will be moderating -- thank you -- this session. The domain tasting issues report came out of a request from the At-Large Advisory Committee. And that's why I'm here today. We have a panel of people who are going to go over the report and to explain how the issue of domain tasting affects different constituencies in the Internet world. So without further ado, let me introduce you to our first speaker, Tim Cole, ICANN services staff. >>TIM COLE: Thank you, Jacqueline. Thank you, everyone, for being here. And for those of you who are online, this, just by way of background, I'm just going to give some terminology and some data before the people with the substantive concepts speak. But I just wanted to mention that this is the -- I believe this is the fourth consecutive ICANN meeting that we've had some program related to domain monetization practices. And so we're trying very much to make sure that the entire ICANN community is aware of developments in this field. And this is one more step, and we're grateful to ALAC for hosting this tutorial today. First of all, I'd like to just give a few definitions. And these definitions come from the issues report. But I think it's helpful to have a common understanding before we go into the various presentations. First I want to mention the add grace period, which refers to a specific -- a specified number of calendar days following a registry operation or a grace period in ICANN terminology refers to a specified number of calendar days following a registry operation in which a domain action may be reversed and a credit may be issued to a registrar. The add grace period is typically the five-day period following the initial registration of a domain name. The add grace period appears as a contractual term in some but not all gTLD registry agreements. And it is designed to allow for the correction of typos and other errors by registrants. Domain tasting, the subject of this tutorial today, is a monetization practice employed by registrants to use the add grace period to register domain names in order to test their profitability. During the five-day period, registrants conduct a cost-benefit analysis to determine if the tested names return enough traffic to offset the registration fee paid to the registry over the course of the registration period. Another set of terms, parking pages and pay per click. The reason that these are tested, primarily, the primary reason is because the domain names will be used to create parking pages, which are a variety of tools -- or a variety of tools can be used by domain holders during the tasting period to assess the market value of a domain name and to generate revenue. During this period, the name may resolve to a parking page that contains advertising or links to determine -- determined to be relevant to the name or to certain associated search terms for which the registrant has made pay-per-click arrangements with advertisers. So one of the ways that -- we've gone into much greater detail in other workshops about pay-per-click and parking pages. But just to give you background, one of the main reasons that someone might acquire a domain name that they would not otherwise have use for is because they would use it to direct people to parking pages where they could then collect revenue from people clicking on links. A term that has been used quite often is domain kiting. And registrants could also use the add grace period for continual registration, deletion, and reregistration of the same name in order to avoid paying registration fees. This practice is sometimes referred to as domain kiting. This term, we believe, has been mistakenly used as being synonymous with domain tasting, but it refers to multiple and often consecutive tasting of the same domain name. While the ICANN staff has received anecdotal reports that this type of activity is occurring, we do not have data to demonstrate definitively that domain kiting occurs or if it does, to what extent. This diagram should start to seem familiar to many of you. It's a diagram we've been using for some time now to demonstrate the life cycle of a typical domain name. And specifically when we talk about domain tasting, we're talking about the happening at those periods where the names become available at either the very end, after a name has been deleted, or before it has ever been registered and existed. So those are the points in the domain name cycle where the domain tasting would begin. Finally, I just want to share with you some data, because there's been some interest in knowing how the magnitude of this activity has changed over time. And what I just did was, I have snapshots in time, they're not exactly one year, they're about 15 months apart. But data is from February 2006 and May 2007. During this time period, the number of registrars that have actively added names has grown from 527 to 802. So that's just of all registrars, the growth during this period was about 152%. Just to give you a baseline, that's how many registrars have grown. It's not necessarily those involved with the activity. Deletes, the names being deleted grew by 191%, or from 2.49 million to 4.76 million, during the months of February 2006 and May of 2007. Now, this next number, I think, is key. And that is that the number of registrars that are actively deleting more names than they register in the sense that their deletes are at the level of 50% or higher of the names they've registered, grew from only 11%, which is the bar on the chart, you see the bar to the left is 11%. It grew to almost 30%, or 28%, of all registrars now have deletes in excess of 50%. If you carry that to the next highest level, those that are deleting over 75%, the number of registrars drops very little. Over 90% drops very little. And then there's a -- somewhat of a significant dropoff at the 95% level. So the total number of registrars that are actually deleting a lot of names went from 58 in February 2006 to 226 registrars that are deleting in excess of 50% of the names. So this is just some background information I wanted to share with you before the other speakers proceed. >>JACQUELINE MORRIS: Great. Thank you very much, Tim. Now we have an overview of the issues report, both with two speakers, Alan Greenberg, who is an At-Large Advisory Committee member, and our liaison to the GNSO, and Maria Farrell. And they will divide that among them. So Alan first. Thank you. >>ALAN GREENBERG: Thank you, Jacqueline. I'm going to spend just a few minutes talking about the origins of why the ALAC came to request the issues report and the process we went through. A little bit of history. I'm new to the ALAC. I only joined in December. The idea of formally looking at domain tasting predates me by perhaps a year or so. So I came into this a little bit late. Clearly, from any discussions with people, it's a very emotional issue. And monetization is closely linked to it, so much so that very often people have a great amount of difficulty separating the two. Monetization hits people, certainly individual users, heavily, because every time you mistype a name, you'll likely hit a monetization page. Every time you do a search, chances are the first few entries will be monetization pages in many cases. A little bit of investigation. First of all, certainly at the time we started looking at it, very little data. There was some data, is data available on deletes. But not add grace period deletes. And the numbers we had at that point were closer to the beginning of the numbers you just showed and therefore much lower. It certainly was unclear, and perhaps still is to some extent unclear, to what extent domain tasting facilitates/increases monetization. But clearly there is a link. And on further investigation, there were a whole bunch of other issues which impact users at various levels, and certainly registrants at various levels, both from intellectual property issues to stability of the overall DNS. And as you can see from the numbers that were just presented, even then it was clear it was growing. I was amazed when I heard these numbers yesterday, how much it's grown in the last few months. So the bylaws allow an advisory committee to request an issues report, and thus, potentially, start a PDP. But at the time we looked at it, no advisory committee had ever done that, and there was an awful a lot of confusion both within ALAC and the ICANN staff of exactly how do you go about doing this kind of thing. Now, as tasting is potentially a result of the add grace period allowing domains to be available at very low cost, partly the result of increased automation which allows names to be entered into the zone file virtually instantaneously and taken out almost instantaneously. In the old days, it took so long to do it that the five-day period didn't matter very much. Now it happens in seconds. And lastly, of course, the increased opportunities for pay per click. At the time the add grace period was first made available, the concept was -- I don't want to say nonexistent, but certainly a lot less common than today. All right. Request for an issues report. There was an awful a lot of confusion about what the add grace period was and how it came about. I talked to a good half dozen people who told me that it was a result of a consensus policy and told me which one. That, of course, turned out not to be true at all. And as I think was mentioned, it was there primarily not to enable people to get free registrations, but to correct errors. I think we can see, what, now, four million errors per month? That's a lot of typos. And the decision that was made in Lisbon to request an issues report, with a little bit of a lag time to be able to formulate it properly. Although not required by the bylaws, the ALAC made the decision to work cooperatively with a number of GNSO constituencies in requesting the report. There was clearly a lot of interest in some of these constituencies, and we also knew that they had very different perspectives from us. And we felt it was important in phrasing the request that we looked at all of those issues. And the report was issued in early May. And that's all I have to say, because now we're going to hear about the report. >>MARIA FARRELL: I'm Maria Farrell, the GNSO policy officer with the ICANN staff. And I'm going to talk you through the issues report in not too much detail. The first thing to know about the issues report is that it was created largely by the ICANN services department, by our in-house experts on registry and registrar issues. And so it's very much a cooperative and collaborative report. So don't blame me. [ Laughter ] >>MARIA FARRELL: Of course not. First of all, what is an issues report? Well, an issues report is something that is a very defined moment in a GNSO policy development process. And it is something that, as Alan has described, is created, it's triggered by an official request from, in this case, the At-Large Advisory Committee. When that happens, the ICANN staff has got 15 days to produce this issues report. And in this, I think we managed it in about 17, which we all felt possibly unreasonably proud of. The issues report looks at -- first of all, it has to define what the issue itself is. And it also, as required by the bylaws, is -- tells us who raised the issue and why did they raise the issue. So what are the sorts of interests in the people who raised that. And then the other thing that the issues report does is, it gives a staff determination of two things. One, is the issue within scope for a GNSO PDP, and, ultimately, within the scope of ICANN's role. And, two, is there effectively a positive recommendation of the staff that a PDP should in fact be launched? When the GNSO Council receives an issues report, it then has -- it has a binary decision. It has to choose whether or not to launch a policy development process. I'm going to cut straight through to what was the staff recommendation. First of all, the staff, when we are making our recommendation of whether a PDP should be launched, we have to answer several questions which are laid out in the bylaws. And these are written up here in this slide. I'm just going to read them out for you. The first is really to determine, is this issue, as a policy issue, is it within ICANN's mission statement? Is it broadly applicable to multiple situations or organizations? Is it likely to have lasting value or applicability? Will it establish a guide or a framework for future decision-making? And does it implicate or affect an existing ICANN policy? These are the broad areas that we consider when we're making this staff recommendation. Well, to put you out of your misery, the ICANN staff did, in fact, find that domain tasting, specifically with reference to the add grace period, is within scope of the ICANN policy process and within scope of the Generic Names Supporting Organization policy process. So we did recommend that a PDP be launched. At the same time, we also recommended that while -- when a PDP was being launched, that there should be further fact-finding and research done, and also, there should be, at the same time, consideration of other mechanisms that could be used, non-PDP mechanisms, to look at the issue of domain tasting. So I'm just going to tell you what those other mechanisms could be. These are the ones we identified in our report. We saw that assuming the ICANN community does wish to act -- of course, which itself is not a staff determination -- there are several other ways to look at this as well as a GNSO PDP. The first is the ICANN budget process, which has something to say about registrar fees, particularly add-delete fees and whether or not these could be changed. Margie is going to touch on that briefly. Then, secondly, there are two registry-focused activities. One is contractual changes or negotiations by a registry, for example, at the time of renewal or other times. And then the other is the result of a consensus policy. And that is the process for consideration of new registry services, the RSEP process, as it's called, sometimes also known as the funnel. Actually, I'm just going to speak a little bit about the funnel just for a moment. David Maher, from PIR, is going to talk briefly about a funnel request which PIR put in with respect to the add grace period and to add-deletes. And so that is something that's, I think, also worth looking at. So these are the basic mechanisms that can be looked at. We've got the PDP, we've got the budget process, and we've got changes to registry contracts, however they arise. As to the PDP, Avri Doria, who is the new chair of the GNSO Council, is going to cover the PDP process and what we can expect from it. And that is the conclusion of what I have to say about the issues report. Thank you. >>JACQUELINE MORRIS: Thank you. Next, we'll have presentations on the impacts of domain tasting to the different constituencies. And first we have Jothan Frakes from domain sponsors, and VC.net. >>JOTHAN FRAKES; Thank you, Jacqueline, and thank you for including me on this panel. My five minutes here is largely to just put some clarity to this -- both to the definitions of the add-grace use that was presented -- I think Tim had in his slide some statistics that illustrated use of add-grace. I have a very simple slide, it's, again, a rehash of the life cycle of generic top-level domains slide. But what I've done is somewhat make an attempt to untangle some of the types of domains, tasting labels, et cetera, that have been used to define the AGP practices. The original add grace activity surrounds add tasting, where it's, in essence, a brand-new registration or a registration that hasn't existed for quite a period of time, in essence, a fresh registration of a domain name, an initial registration. The next is what I'd define as drop tasting. This would be where a domain name had been previously owned by a registrant and was allowed to delete through the drop process. And very significantly different type of activity there, because it often is a previously owned Web site, and largely the focus of registrations because, typically, tasting when you're doing PPC viability tests to determine the keep or not keep look at traffic, look at clicks, and look at the way that is a domain can monetize to meet a particular threshold. So those in drop tasting typically come with some inherent traffic from previous Web sites, links, or other activity, mentions in blogs or e-mail, and tend to be far more viable for PPC consumption. Lastly, there's another type of activity, which is a recurring type of activity, which the very strongly emotive term "kiting" gets applied to. And kiting is a situation where the domains are added and deleted and then re-added in a manner to use and leverage that five-day period to avoid actually ever paying for the domain name, but continuing to perpetually run your PPC viability. We see different versions of this where the domain registered and reregistered in add tasting. Many registrars dig through the Dumpsters of the other registrars to determine a name that's previously tasted and go and look at the domain for themselves, thinking their PPC model might hit a viability model. And so we see that happening across registrars. So there's a variety of different activity here. And it's really important if there is something to be done here about tasting that that be defined, because it really helps to remove some of the cognitive distortion that using an emotive term like "kiting" to describe the entire process causes folks to somewhat not react and untangle the mess and look at how that works. The summary of my presentation is that the -- technical difficulty here -- the main point I wanted to make is that, as Tim identified, there's a lot more registrars now today -- thank you, Marilyn, that was very kind. And the growth in the number of registrars and the growth in the activity, marketplace factors might be playing a role here to, in effect, change the registration in tasting. Because there's such a high volume, as illustrated by Tim's slide, of the growth in this, one would argue that that is in advance of the price increases that were approved by ICANN of the different registries, that people are racing to get those domains before the price increases happen. Secondarily, that price increase is a very important fact, because it raises the bar, the threshold under which some domain might hit a PPC viability. In addition, the increase in the activity that we see for PPC, all of these different domains, these many domains that are registered by these many new entities, are all, in essence, competing for those same advertised key words or revenue-generating model, and so it actually reduces the amount that a per-click might make on an individual domain, while the cost threshold raises. So I would offer that sometime around October, that we will see a significant drop in this activity just by market forces. This is an area where there's already something in motion that would affect this as a deterrent because of the cost viability model. Thank you very much. >>JACQUELINE MORRIS: Thank you, Jothan. Now, Marilyn Cade, consultant and -- will take us through a business user's perspective. >>MARILYN CADE: Love it when a plan comes together. Like others here, I really welcome the opportunity to talk to you about that, but I'm going to suggest a slight correction to a comment that was made earlier. This, in fact, is the first meeting at which we have discussed a key aspect of monetization, and that is the harmful aspect, the dark side, when it is not just a monetization scheme, but it is a scheme that is involving the abusive registration and exploitation of the rights of others. I by no means mean to imply that the secondary market is all bad. But what I'm here today to do is to show you some examples of the experiences that people are having as users, both businesses and others. First of all, I might just note, we have a problem. And I call it joy-riding. Because it's basically taking advantage of investments that are made by others. Cybersquatting is still cybersquatting, even if it's on steroids. And when it's described as a business model, that doesn't eradicate or change that it's still cybersquatting. We might just consider an analogy. If I have a leaky roof and I ignore the problem and I put pans under the drips, eventually I'm going to see a significant increased cost in maintenance and repairs, but more than that, ugly molds are going to grow in the attic and in the walls. They won't be discernible initially, but eventually there will be major structural damage. And no one's going to want to live in that home. There's a big scope to the problem that is just beginning to become apparent. And I'm going to let others give the statistics on that because of the limitation of five minutes. But I think it's a fairly compelling story. So let's look at some pictures. Later, for those of you who have more time, maybe in the Q and A, I can walk through some of these slides, one of which -- two of which are provided by CADNA. But this slide shows the linkage between when a user mistypes a domain name, many spells a domain name, they're taken to a site that is not associated with the brand holder. It's a park page. So the person who bought the key word, the brand holder, pays to -- because there's a mistyping, they pay to take you to a parked page. And then other ads are served, often ads, primarily ads that are not associated with the brand holder. Let's look at some images here that are important messages. Here's one from OECD. The OECD is a highly respected international organization. I don't think this site is associated with the OECD. How about this one? World intellectual property organization.com. Hmm. Well, you know some of the related links are links that will take you to information on WIPO's site. But the rest of the ads or the information that is served has nothing to do with WIPO, a highly respected U.N. agency. Okay, how about another one? FedralTradeCommission.com. "Federal" misspelled. Well, in the United States, the U.S. Federal Trade Commission, the FTC, does provide information to consumers. But there's a lot of information on this site not affiliated with the FTC. How about one more? Ah, Commerce Department.com. Not sure you can read across the top, but you might be shocked to know that if you live in the U.S., your taxpayer dollars bring you pornography, dating, triple X. PatrickLeahy.com, a highly respected U.S. senator, the chair of the U.S. judiciary committee. Well, I guess he won't be needing to spend money on campaign financing since he can sell tickets for air travel. Or music downloads. By the way, if you go to some of these music downloads, they're all for free. I checked a couple of them out. Not sure that the music companies know that they are for free. Here's one for Orin Hatch, another highly respected senator in the United States. How about the Senate judiciary committee. None of this information, by the way, has to do with the Senate judiciary committee. But you'll be glad to know that you can check out home addresses and personal telephone numbers from one of these links. Here's one that's pretty shocking, jonbenetramsey.com. A few years ago, a very sad murder situation in the United States in Colorado where a child was murdered in her home. This site is the exploitation of the interest that people might have in keeping track of the efforts to find out information about the murder. Here's another one that's pretty shocking, Virginiatechmassacre.com, a recent tragedy that took place in the United States. None of these links, by the way, are associated with Virginia Tech. I didn't show you business sites because some of my colleagues are going to talk more about that. But the number of sites that you go to, you can find many of them, attsbcglobal.com, attphones.com, VerizonWireless.com with any number of misspellings. You go to those sites, and you're served information that may or may not be linked to the brand holder. But a lot of the information you're served has nothing to do with the business. In many cases, it is information or sites that are pornography or are perhaps even more interesting. I suggest that there are more than one pieces to the puzzle of trying to address domain name joyriding. The abusive part of it. And, again, I don't say that all domain name parking is abusive. But when it does exploited the rights of others, i think it's important for us to take a careful look at what it's going to take to eliminate the dark side and make sure that the roof's not leaking. >>JACQUELINE MORRIS: Thank you very much, Marilyn. Now, on the topic of possible next steps and context, Avri Doria, chair of the GNSO, will take us through the PDP process overview. Avri. >>AVRI DORIA: Thank you. It will take me but a second. >>JOTHAN FRAKES: While Avri is adjusting that, Marilyn, were those park pages or were those tasting pages? >>AVRI DORIA: I will start and perhaps we can get back to the question when it gets to the question. So I'm going to quickly go through the PDP process, basically launching an issues report, launching a PDP, conduct of a PDP, some of the PDP timelines, little bit of discussion and then sort of the next steps that the GNSO will have to go through on the issues report. So the request came through. In this case it came through from the ALAC. And the staff sends the issues report to the council. Then the council has to decide whether to launch a PDP. If the staff recommendation says the issue is within the scope of GNSO policy process, then it only takes a 33% of council members present to start -- to initiate the PDP. If the staff recommendation says the issue is out of scope, then a supermajority vote of the council would be required. In this case, the report does indicate it is within scope. Just for information purposes there is weighted voting within the council and the number of the votes that the registries and the registrars have, are balanced by the votes by those in the other constituencies, with the three NomCom-appointed members being outside that balance. Within the process of a PDP, first public comments are pretty much right after the start of the PDP. The council chooses whether to create a task force or whether to operate as a committee of the whole. That's a decision that the council makes. Constituency statements are collected and they need to include clear statement of the constituency's position, how the constituency arrived at its position, for example, did they have meetings? Was it telecommunications? Was it, perhaps, teleconferences? Was it, perhaps, some other method that they used to arrive at their position? An analysis of how the -- the issue would affect the constituency. And that includes a financial impact and how long it would take to implement a new policy. There is a preliminary task force report. There is a second round of public comments. Those comments are dealt with and then there is a final that basically is discussed, integrated as appropriate and then a final task force report is produced. This is assuming we followed the task force process and then the GNSO Council votes on the recommendations. And then that is sent to the board and then there is an open review process and then the board makes its decision. When the council takes its vote, if there is a supermajority vote, then the board needs to adopt the policy unless 66% of the board determines that the policy is not in the best interest of the ICANN community or ICANN. And then there is the implementation of the policy which is done by the staff. Just wanted to get into -- and that was a quick overview and there is a similar process for when it is a committee of committee of the whole of the council doing it as opposed to the task force with a little bit of variation. In terms of the bylaws, they sort of mandate a timeline for a PDP. And from the request to the issues report is 15 days. As Maria said, they got this one done in 17 days. By and large, the staff is very good at meeting this 15-day time requirement. It is harder to say anything about the rest of them. The issues report to the initiation of a PDP is 15 days. In this case, we are already outside that 15 days it. Initiation, decision to a charter or a term of reference, another ten days. If we're following the task force route, then the charter from the time of the charter to the time of the appointment of a chair, five days. Initiation to a constituency statement, 35 days. Initiation to the preliminary task force report, 40 days, which is five days after the deadline for the constituency statements. The preliminary task force through the deliberation meeting, there is five days. Then there is the meeting to the task force report, another five days. And then the task force, sending that report to the council after a 20-day comment period. So basically you have a 90-day cycle. Some just examples of reality of timelines, the Contractual Conditions Task Force was created in February '06 and produced its report in May of '07. And we'll be starting to act on that at the open meeting in -- where are we? San Juan, thank you. [ Laughter ] I lose track. The WHOIS task force -- I remember a lot of staff but where I am. The WHOIS task force in February '05, it was completed in March '07. We have since creating a working group to try to iron out some of the details so that has not terminated quite yet. The new gTLD committee of the whole is still ongoing. That was created in December '05. Now, by and large this is because it takes that long to basically get all the constituencies to deal with all the issues, to listen to comments, to deal with the comments, to have public discussions. So, basically, the bylaws require that every time we slip the schedule, we have to ask the board and so periodically at the board meetings, there is -- the chair gets up and says, and the task force schedule is now. Okay. So in the discussion, one of the things that's often discussed is that the scope of a PDP should be a narrow scope. It is very difficult to get the scope narrow enough. Certainly, if we look at domain tasting how narrow we can get that scope so we can try to do it within the 90-day schedule, it is hard to think of it. One of the things that happens is that task force -- and lately we've started doing some work in working groups -- need clear and defined terms of reference. Those are also difficult. Those take time to create. The council workload has often meant that the creation of a task force has been preferred to a committee of the whole. The committee of the whole as the name means that the whole GNSO, perhaps, with some substitution of representatives from some of the constituencies, but essentially the council needs to deal with this issue in the PDP as well as all the other work. So, basically, task forces have been created which can have a different membership in them than the councillors themselves. As I showed, though, the time lines are unrealistic and complex PDPs, diversities of use, basically trying to reach some sort of broad consensus on an opinion before sending it on to the board takes 12 to 24 months, not 90 days. And, in fact, in the GNSO review that's happening at the moment, this is one of the issues that's being talked about. And we are moving towards a more inclusive way of working and with the working groups where we've included more people so that we can get a broader range of opinions. So our next steps. We have, I believe, it's Wednesday -- there will be a council open forum discussion where the issues report will be presented and the community will be invited to comment on the details, the recommendations, whatever. Then after that, there will be a council open meeting which will consider the issues report and at that time the council needs to work its way towards a decision. Will it initiate a PDP? As I said, it would take a vote of a third of the councillors to make that decision. If there is, would it be a task force or committee of the whole? A decision that the group needs to make. There is a possibility that a working group to study the issue further and make limitations, something with a very tight term of reference and limited time could be created and, perhaps, out of the discussions in the open forum, some other possibilities will be presented that could be considered. At this point, the GNSO Council has not discussed this yet in any detail. In fact, what we discussed was the -- basically the fact that we delayed initiating that until this meeting and until we had time for some open discussion. So I have no guess as to what the council will decide to do. That's it. Thanks. >>JACQUELINE MORRIS: Okay, thank you very much, Avri. Next we have David Maher from PIR who will give us some options for example on registry action related to the add grace period, alternative ways to deal with this issue. >>DAVID MAHER: I have no slides and I would like to give an explanation of the excess deletion fee which has unfortunately been the subject of some misunderstanding. PIR, the manager of the registry of dot org, found last year that we were, we thought, the victims of the abusive use of the five-day add grace period. It came to our attention primarily because of someone who took a respectable name and turned it into a pornographic site. In addition we found that, I think, a very small number of registrars were registering literally millions of names and then deleting well over 90% of them. So we set out to do something, and we did not set out to cure or solve what might be called the domain tasting problem. We set out simply to cure the abuse of the five-day add grace period. And our proposal was really very simple. The proposal was that we would charge five U.S. cents, a nickel, for every deletion if a registrar deleted more than 90% of the registrations it had made within a 30-day period, arbitrary time period. We started out using the funnel process, not the PDP. The funnel process is a much simpler process within ICANN, and it was very successful in our case. We submitted a request to ICANN for this new service that charged for excess deletions on the 28th of August, 2006. The request was published and ICANN asked for more information. That was provided, and on the 22nd of September, ICANN initiated its 15-day review of the proposal and gave a preliminary okay. The proposal was published for public comment and even though there had been a great deal of discussion in the media, e-mail lists and publications, there were only two comments. One was very favorable, and the other was very unfavorable but it was the comment from the International Trademark Association which said this doesn't go far enough. This is not a cure for the perceived problem of domain tasting. Well, that was true. We didn't set out to cure domain tasting. In any event, on the 22nd of November, the ICANN board approved our proposal and on the 1st of March of 2007, an agreement was signed with ICANN that amended our registry agreement. You can find it in appendix 7 and appendix 8 of the dot org agreement that's posted on our Web site. And, finally, as I'm sure you were all waiting to hear, what is the result? Well, very simple. It seems that we've solved our problem. The two registrars who were registering millions of names appear to have stopped. We're never going to collect a penny, but oh well, we've cured our problem and the funnel process worked very well for us. Thank you. >>JACQUELINE MORRIS: Thank you. >>MARGIE MILAM: Hi, I'm Margie Milam and I am vice president and general counsel of Markmonitor. We are a corporate domain name registrar that does brand protection services as well so we have a little bit of an unique perspective. I wanted to talk about the possible solutions to the domain tasting problem. But before I talk about solutions, I wanted to give you some statistics on the scope of the cybersquatting problem. At Markmonitor, we conducted a survey during April, we call it our brand Jacking index where we announced our results. And basically what we did is we looked at the world's largest brands, we looked at the top 25 brands. In addition we analyzed different industries -- top brand within different industries to see what kind of problems are out there and this slide highlights some of the results. Some of them don't relate to kiting but I want to point during April, in this small subset of companies, there were 11,000 kiting incidences. You know, this is a significant problem. From our perspective, cybersquatting has increased over 248% over the numbers in 2006. And as Marilyn aptly pointed out, it is really abusers using the brand strength to try to siphon traffic and to make money off somebody else's brand. What we found is that the volume is very difficult for brand owners to manage. It is very hard to deal with the scope of problem and deal with it quickly and efficiency. From our perspective, we try to analyze the money involved associated with kiting, and we estimate that it's $125 million industry alone in kiting. When you look a little bit deeper within the domain kiting issue, we tried to see what industries were targeted. As you can see on the slide, clearly the financial industry is the one that's targeted the most. And it certainly ranges from financial all the way down to pharma. There are other types of industries targeted. The reason the financial industry is targeted is because of the -- it is a very financially lucrative business to be used financial targets. What we found astounding was the first statistic you have here, which is when we looked at a particular financial brand, in a one-month period, they had an average of over 980 kited sites. If you are trying to protect a brand for a bank, that's really a very difficult problem to deal with. In talking about solutions, I wanted to think about solutions that don't involve the PDP process, although I do think the PDP process is good. And I have highlighted three of them here. The first one involves the ICANN budget. What we talked about earlier is that there's two components to the registrar fee. One component is a fixed fee portion in which there is this new forgiveness policy in which ICANN has made a determination that if a registrar participates in excessive add-grace abuses, that they should pay more for their ICANN fees. That's actually built into the budget as it exists today. There is also the transaction fee of 25 cents, and at the moment it's not -- it isn't charged if there is a name that's deleted. But that could be changed, and that's one of the things that might be considered here, is to actually make that 25 cents transaction fee permanent so that if -- if someone does participate in the add-grace abuses, they at least have to pay 25 cents. And so there's some reason to not participate in that process. I also think that the registry-based solution is also a possibility and David certainly highlighted what PIR did in connection with the registry service evaluation procedure. When I took a look at this, I also looked at other aspects of registry solutions including in particular in the dot com agreement, there is section 3.1A and it is a provision where the ICANN board is allowed to adopt temporary specifications or policies if it is in response to a stability issue. So if this affects the stability of the Internet, as an example, the ICANN board could certainly adopt temporary specifications to deal with this issue. And I don't know if it was clear from the prior presentation, but this whole add grace period concept is addressed in the specifications for the dot com registry and for the other registries. And so that's one way you could solve this problem. And as was mentioned earlier, a lot of this has to do with traffic. And so if the registries were actually to sell their traffic data to domainers, the companies that participate in this activity may not need to do the domain tasting. They could get that information elsewhere. The other thing that I wanted to talk about, which is a little unique, is that I think this is also a registry-accreditation issue and that ICANN should look into whether the current registrar accreditation agreement has provisions that deal with add grace period abuses. And when I looked at it, I looked at Section 3.7.4 of the accreditation agreement, and it basically states that registrars may not activate a domain name without a reasonable assurance of payment. That section also has a provision that the obligation of payment becomes final and non-revocable by the registrar name holder upon active registration. How I see this applying to the add grace period is, as was mentioned earlier, the reason that the add grace period policy was adopted was to deal with specific instances that could be clarified. The first one is when a registrant doesn't pay for the domain name and has defrauded a registrar. That could still be in effect, but when you think about what's happening in the industry, if a registrant has defrauded a registrar, there is no reason for the registrar to continue to do business with them. So if a registrar tries to get a credit, that's great, they can go ahead and apply for it but then why would they continue to do business with a company that isn't paying them for the domain name. The other issue that seems to be addressed with the add grace period is when a registrant makes a mistake, like they mistyped a domain name. Again, you know, that could certainly be carved out and that could be dealt with specifically. But that isn't what's happening with domain tasting. And as long as the rules become more clarified, you may be able to, you know, deal with this from a policy perspective without having to go through the PDP process. So in conclusion, and I think -- as I've mentioned before, I do think we should continue with the PDP process to see if any of these short-term solutions, you know -- if they don't solve the problem, we may need to deal with it from a more global perspective. If there is a way to take short-term action, I think ICANN should consider that and look into whether the contract provisions that I've cited earlier can be relied on to try to take a firm stand against add grace period abuses. This is just an example. As a registrar, we typically get from time to time advisories from ICANN on specific issues, and ICANN, in this case staff, could issue an advisory about when the add grace period should apply. It could take a position that, for example, it only applies if the registrant asserts a basis for a mistaken registration. It could deal with the typo situation we just talked about. You could also have it apply when the registrar demonstrates there has been fraud by the registrant. Again, that's a very narrow situation and I think it could continue to exist, but to deal with the add-grace abuses, you could actually make an affirmative statement that there should be no resolution to a PPC site. If the domain name has been, you know, mistakenly registered, then there is no reason it should -- it should resolve to a pay-per-click site. And I think that's one way of clarifying when it's appropriate to give a credit. The other area is in actually the registry specifications that talk about grace period. What I have here on the slide is from the dot com agreement. It basically states that a grace period is a specific number of days in which a domain action may be reversed -- I emphasize "may" -- and a credit that may be issued to a registrar. In other words, it doesn't have to be automatic, and ICANN can work with the registries to issue an advisory as to when it's appropriate to issue an add grace period credit. So that's, again, another easy fix to the solution. The other thing that the registry agreements have is they have a statement in there that says that registries have the right to charge a fee for a disproportionate number of add grace period deletes. And I think that's what PIR did in their approach. But that is certainly something that I think is a reasonable thing for the registries to do, if, for example, the dot com registry would charge the same fee to those registrars who have excessive deletes, I think you will see this problem go away quickly. And then, finally, as I mentioned before the budget issue. I think it is a good idea to think about the transactional fee that I mentioned earlier and to have it be non-refundable so that there would always be a fee associated with deletes during that five-day period. >>JACQUELINE MORRIS: Great. Thank you very much, Margie and all the panelists. Now, we have a period of questions from the floor. First, I want to call Paul and Jay Westerdal. >>JAY WESTERDAL: Hi, Jay Westerdal. >>JACQUELINE MORRIS: Great. >>JAY WESTERDAL: I will start off with David. So, David, the rate counseling case you cited was not domain tasting. That was a person who actually had a beef with them, set up a Web site, that was very disparaging. I think it is deplorable, but I don't want to confuse the issue. I want to applaud you for solving the domain tasting issue. I think you are confusing apples and oranges. That shouldn't have been the catalyst for domain tasting because it's not. >>DAVID MAHER: It happened to have been the catalyst. I am sorry if it was illogical, but it was the catalyst. >>JAY WESTERDAL: I want to say good job on solving the domain tasting. >>DAVID MAHER: Thanks. >>JAY WESTERDAL: Marilyn, all of your examples were primarily just parking pages. So I'm very confused what you are trying to accomplish. >>MARILYN CADE: If you looked at the title, Jay, of my presentation, I did link tasting and parking. The experience that we're having as business users is that very often the pages are initially tasted and then often they're converted and they remain permanently parked. And they are -- so we're basically seeing that the trademark holder who has paid for a keyword is also funding the traffic that is -- we pay to take the potential customer to a parked page and then, of course, the ads that are served may or may not be associated. In some cases, we find pages that have no association at all with the company. Most pages will have at least one or two ads that are in the industry. But there is a huge amount of conversion, it seems to me, between the issues where we see somebody benefitting from traffic that's intended to go to -- that's intended to be related to the famous and well-known brand holder and, yet, what we're doing is we're paying third-parties and they are benefitting and exploiting the brand investment, companies like Verizon, Neiman Marcus, AT&T and I can go on with the list, Dell computers, et cetera, et cetera. So I'm not exactly sure why if it's exploitation that doesn't make it good just because it is now parking and no longer just tasting. >>JAY WESTERDAL: Well, I mean, I think you may be confusing the issue and, you know, when you go to Capitol Hill and you present on this, you're really confusing the people up there on the hill that don't understand technology. And you say, hey, look, here is a parking page, tasting is bad. And you are confusing half the audience and these guys are technical. But in Washington you are confusing everybody. >>MARILYN CADE: Jay, I have a lot more faith in policy makers, including government policy makers. >>JAY WESTERDAL: I don't. >>MARILYN CADE: The presentation that Sarah. And I did on domain tasting with David Still included long lists of the pages that were just tasting. In five minutes, it is clear I didn't have time to go into that level of detail. And what I wanted to do was to really make the imaging messaging here, which I think is important to the audience at ICANN to understand that if we are not about solving problems, both in tasting and we've talked a little bit about solutions in tasting, the reality is that the cybersquatting relationship to this and the exploitation of brand names continues to make parking lucrative and that is going to drive other solutions. You will remember the puzzle slide that I showed. ICANN is not the total answer. >>JAY WESTERDAL: I think cybersquatting needs to be handled separately and I do think it needs to be handled but separately. I would like you to maybe concentrate on different issues at the same time. I don't want to attack you right now. Let's just drop the issue. [ Laughter ] But just to say, try to keep the paths clear so we can talk about one subject at a time. >>MARILYN CADE: Jay, I am really open from taking guidance from you on where I should concentrate. >>JAY WESTERDAL: Okay. And I'd like to say that domain parking or, actually, I would say this domain tasting thing, I'd like to see domain tasting end. So I applaud what PIR has done. They've actually solved the problem. If VeriSign wanted to solve the problem, they would have. But they don't see it as an operational problem, so I don't think they will. I think it's going to be up to ICANN to solve the issue. And I think that moving the 25 cents to the other side of the transaction would be fine. But let's hope that that's the outcome of the PDP process. >>MARILYN CADE: I'm a skeptic that further charges are going to do anything about the exploitation of the names that are associated with brands, because the traffic is just to attractive. >>JAY WESTERDAL: Marilyn, that's a different subject. We're talking about tasting. >>JACQUELINE MORRIS: Great. Thank you very much, Jay. [ Applause ] >>MICHAEL PALAGE: Mike Palage. >>JACQUELINE MORRIS: Hi, Mike. >>MICHAEL PALAGE: And not to put you back on the spot, Marilyn, but I will. And I think it is, first off, within the BC, which both you and I are members, I have been an advocate to end this problem for well over a year. I discuss it had on the list, provide a definition. So unlike perhaps Jay, who, shall we say, derives some revenue from the tasting, I derive none. >> No. >>MICHAEL PALAGE: No? >> Clarify that. >>MICHAEL PALAGE: I'll clarify that. Jay gets none. What happens, Marilyn, is there's a coupling. Margie, if you can go to PatrickLeahy.com, I think you're still online. >>MARGIE MILAM: No, I'm not. >>MICHAEL PALAGE: What I did, Marilyn, was I pulled up the WHOIS, something you would appreciate, and that domain name, PatrickLeahy.com, was registered in 2003, assuming that Godaddy's records are accurate, and I have not reason to not believe the accuracy of those records. So I think that really goes back to what Jay was talking about here is, you've kind of blurred tasting and cybersquatting together. >>MARILYN CADE: I linked it, Mike. I didn't blur it; I linked it. >>MICHAEL PALAGE: I think you threw a lot of red meat with senators. And, again, Patrick Leahy, great senator from the state of Vermont, someone, copyright attorney, us I.P. brethren like to stick together. So great senator from Vermont. But what you did, Marilyn, was really try to throw out a lot of red meat, as Jay sort of said, that may be attractive on Capitol Hill, but is it really solving the problem? And I think that when we look at what David and PIR has done is, they have addressed or solved the problem for that registry. And, again, I will -- we can point to some of the discussions that happened on the BC list where people were critical, oh, this is just an effort by a registrar to make more money. But as David said, the problem worked. David, are you making a lot of money off of the tasting? As he said, no. So I guess what I think is important about these forums is not for people to grand stand or throw red meat, but, really, let's try to solve the problem. And, you know, again, I think it was just important, as Jay was trying to do, let's decouple tasting from, if you will, the other issues of cybersquatting. And, again, one of the things that Mike Rodenbaugh from the business constituency is working on is, let's look at abusive registration practices, automatic take-downs, these other things that might involve an amendment or revision to the UDRP or other types of proactive actions by the registry. Again, I just wanted to put that out on the record. And one other point of clarification. You said that the businesses have never had the opportunity to voice their concern in this forum, as Tim alluded to. I believe Sara Deutsch had participated in one of the other forums. Just wanted to correct you on that. >>MARILYN CADE: Let me respond to that. One of the things I see is that the tasting period is used in order to determine how much traffic is associated, and then often the names are converted. So while a lot of names are dropping -- and I think we heard the statistics from Tim -- there is a huge amount of churn. But we're also seeing that some names are converted because they are very lucrative. So I do think it's important to learn more about the entire situation and the linkage between parking, tasting, and kiting. >>MICHAEL PALAGE: I guess, as I said, I guess we sort of disagree. We agree on eradicating the problems. I guess it's how you couple them that I think actually impedes the ability to get to the solution. So I guess you could take the -- one of us will take the high road, one of us will take the low road. And, hopefully, we'll arrive at a solution. >> DHARMA DAILEY: That's lowbrow. That's lowbrow to talk to somebody like that. That's not nice to say. If you were talking -- if that was a man up there, it wouldn't have been both off you using the kind of language that you -- >>JACQUELINE MORRIS: Okay. So we have another questioner coming up. >> JOHN KANE: My name is John KANE. I'm from eNOM. Just a couple of quick points. I do agree, Marilyn, cybersquatting is cybersquatting. There's no doubt about that. Unfortunately, what we're talking about here is not cybersquatting. So I agree that in some sense -- there is one solution to end cybersquatting. And that is to end domain name registration. That is the only solution. [ Laughter ] >> JOHN KANE: -- to ending cybersquatting. >>MARILYN CADE: I think you're being a little radical here. >> As a registrar, we're incentivized to do domain name registrations. We are not incentivized to help people register domain names that are trading on other people's good names. That is not -- that is not what people here are trying to do. So I think that sometimes we, you know -- you've had enough people throw out, you know, that you've sensationalized the issue a little bit by doing this, because I doubt that any of those instances that you put up on the board were done through domain tasting. I would virtually guarantee -- and, Marilyn, I'll check them all if you like -- I will virtually guarantee that all of those were actually domain name registrations, where someone went in and registered a domain name because they consciously knew they were trading on someone's good name. >>MARILYN CADE: And are you suggesting, just for the record, that trading on someone's good name is a good thing? >> JOHN KANE: As I stated at the beginning, do I not think that is a good thing. I think that is a very bad thing. But there's something called the UDRP, and that is in place to help protect trademark holders for specifically this situation. So we are kind of -- again, we're kind of linking a few different issues. Margie, I saw you threw out kiting a couple times on yours. And I'm not sure if you meant kiting specifically there or if you're really talking about the add delete. >>MARGIE MILAM: Kiting. I confirmed. And to address the issue you raised, I focused on the statistics with respect to brands. And so the problem that Marilyn is highlighting is very real. And it is involving cybersquatting. >> JOHN KANE: I agree. It's real. But stopping this will not stop cybersquatting. >>MARGIE MILAM: The problem with the five-day add grace, is you threw out the UDRP. But it's almost impossible to get the UDRP put together in a five-day period before the name drops. That's part of the problem with the add grace period, is the traditional means of dealing with cybersquatting are almost impossible to deal with. >> JOHN KANE: David Steele will be up next and he will tell you how to catch those guys using the WHOIS data. >>JACQUELINE MORRIS: Do we have another question? >> ANDRES PIAZZA hi, Jacqueline. Is there a problem to speak in Spanish? Do you have translation? Do we have? >>JACQUELINE MORRIS: And just before we talk in Spanish, we will have the report in Spanish for those of you who want it in Spanish instead of English. We just don't have it today. It's online already. Okay. It's just not printed out and available for us yet. >> ANDRES PIAZZA: It's not in the same way the translation. I am going to speak in English. It's a simple question, but hard to answer, I think, or we think. I am from Argentina. We were talking today, and we will be continuing talking tomorrow in the ALAC RALO space, about this issue. I was really looking forward to come here and assist in this workshop. I was expecting more information than the information we already knew. The people who read the documents and are on the discussion lists have already a perspective, a clear perspective, about the problem. I think that registrars have too many resources displayed to continue with these earnings and this practice. And ICANN doesn't put in their priorities this matter. The board of ICANN should take more to put this issue as a priority on the agenda. And I don't think it -- any of you think that having a PDP process, calling for a term of reference, naming, as always, a task force, and the traditional way to deal with the problem is the only solution. Don't you think that there is need more than this? >>JACQUELINE MORRIS: Marilyn. >>MARILYN CADE: I'd like -- >> ANDRES PIAZZA: Sorry, because I am a Spanish speaker, I did my best. >>MARILYN CADE: And you did fabulously. Let me give you my opinion on it. A couple of things that I would note. I, too, had expected to have the opportunity to have this be a tutorial. And it turned out to actually be a tutorial on the issues reports rather than a tutorial on the issue. For a number of reasons, I thought we should spend more time hearing from people who are dealing with the problem. And I still think there's a need to do more of that. So maybe some other people from the audience will come and speak. That still, to me, is not going to replace the need for further examination. I do think there's a role for a PDP. But I think that there is a role as well for other steps. I think that ICANN needs to close the five-day gap or we need to spend a huge amount of time teaching people how to type if, in fact, the real problem is typos. So we could perhaps do capacity-building or we could close the five-day grace period. But enough of that. So I think we also need to raise awareness of people about what is going on. I think that there are other steps that need to be taken outside of ICANN. I think there's a direct linkage here to the fact that people are not paying for the behavior that they're engaging in, and so they're not bearing any cost. I think there may be the need for -- it isn't just, you know -- we need to actually insist that people pay for the names that they register and have a different process for how they get a refund. I'm not sure that I agree that the restocking fee is the answer. I think that people ought to pay the full cost of the registration. But I also think we probably need to know more about the characteristics. And I understand that the ICANN staff do expect to have more data available in the near future, which might be helpful. I think this is a virus. So I don't think you can leave it up to a registry-by-registry solution, because viruses like this spread. And just because it's dominantly a dot com problem today doesn't mean that's going to be the case in the future. >>JACQUELINE MORRIS: Maria. >>MARIA FARRELL: Hi. I'm the ICANN staff person who has been, in some measure, dealing with this issue from the policy side. And I think it's important to recognize that while, of course, there's a very strong and understandable desire on the part of some people to, quote, do something, we should do something very, very quickly, but it's really important within ICANN that we follow our set multistakeholder policy development processes and/or the other processes that are set out in the issues report. So we didn't just say we can do a GNSO PDP and it will take quite some time, which it will, as Avri said, it will. But we did set out three other things that can be done, and three other mechanisms that can be used by this community as a whole. So I don't think it will be quite accurate to say that, you know, there's not a will within this community to do something. But I think it's very important that we use the processes at our disposal -- and there are several of those -- to determine what is the best way to do this rather than simply jumping in and doing something. And just briefly on the fact that we are having a tutorial on the issues report, we are. And that's because we are at a certain point in a process. We are at a binary decision where the GNSO will decide, will it launch a PDP or will it not launch a PDP. And that -- because that's quite an important decision for the GNSO, in consultation with the entire community, that's why we chose to have this tutorial about the report, so that everybody understands where we are. Because we're not just talking about domain tasting anymore. We're actually in a formal process. So to make sure everybody understood where we are and where potentially we could go from here, that's why we've been dealing with the issues report. Thank you. >> ANDRES PIAZZA: Thank you. Of course, I didn't mean to say that this aspect, this tutorial wasn't a great initiative. But for me, there is need more. And that was the point. But, okay, I congratulate all of you about your explanations. They were quite useful. I agree with that. So thank you. >>JACQUELINE MORRIS: Gracias. And we have, what, three more people in the queue? Four? Five? >> Actually, I yield for a second to Peter. He had a quick comment about translation. >>PETER DENGATE THRUSH: Peter Dengate Thrush from the board. I just wanted to say, I think we could have coped just then with the translation. I think we have translation and it was about to be transcribed. I just wanted to say, there's another $200,000 in the budget for translation, taking it to $490,000 for next year, something some of us are very concerned about. I just want to let that gentleman from Argentina know we take that seriously and to thank him on this occasion for speaking English for us. Hopefully, next time, he won't have to. >> DAVID STEELE: Thank you, Peter. Good afternoon, or good evening. My named is David Steele. I have just a few comments that I'll make. And, firstly, I guess I'd like to point out that in my experience dealing with a number of cybersquatting cases -- and, principally, the cybersquatting cases tend to involve the practice of domain name tasting -- but I am specifically talking as to domain name tasting as it ends up getting trademarks for domain names which are confusingly similar to trademarks, just for the purpose of clarification. One of the first things that I've observed in the cases that I've worked on is that the actors involved in domain name tasting and cybersquatting in the cases I've been involved in are, in fact, ICANN-accredited registrars. So when we talk about registrants as the lady from MarkMonitor did, I think it's important for us to keep in mind that many times and in most of the cases that I've dealt with, the actually parties who are cybersquatting are domain name registrars, ICANN-accredited domain name registrars, not third parties. And this gives rise for me to call on ICANN to take a look more carefully at some of these issues as it pertains to people within the fold of ICANN. Obviously, as one of the other commenters said earlier, obviously, Network Solutions -- I'm sorry, excuse me, obviously, VeriSign, the registry operator for dot com, doesn't have a problem with the practices that are taking place, and, obviously, the individuals who are registrars who are, in fact, tasting also don't have a problem with it. So, essentially, it's a different segment of the ICANN community that does have a problem with it. I believe it is an epidemic. And one of the comments that were made revolves around WHOIS data to track them down. And I guess I'll just add, in most of the cases that I've been involved in, the listed registrant for the domain name is, in fact, not the actual registrant. It's either false information, it's nonexistent information, or it's misleading information. A specific comment I'll leave for some of the members of ICANN is that on several occasions when I've contacted ICANN to report a domain name which was being tasted which was also identical or confusingly similar to a famous trademark and, in fact, had no WHOIS data whatsoever, ICANN has done whatever they've done, but I've never received any information back, and, in fact, those actors continue to taste, continue to taste infringing domain names, and continue not to list any WHOIS information or materially false information. So when we talk about compliance, obviously, that's an area where ICANN can step in where there's already existing rules in place, and ICANN can take a leadership role rather than taking a back seat. I'll leave you with this comment, which is, if ICANN continues to take a back seat on this issue, that -- and fails to address the epidemic, other legislative and legal actions are going to be taken, and it's going to surplant ICANN's role in this important area. ICANN has an opportunity to act or to not act, and if it chooses not to act, it's going to find itself being, essentially. Displaced in the process. So thank you. >>JACQUELINE MORRIS: Okay. We've got -- almost out of time. So if you guys can just keep it really short, one question, one comment kind of thing, we'd appreciate it. Thank you. >>WERNER STAUB: Yes, my name is Werner Staub from CORE. I have two comments. One of them is, I actually know where the add-grace period comes from. I was present when it was produced back in '99. It was announced unilaterally by NSI, which was then the registrar and registry. And we essentially found it neat, nice to have. But we were told at the same time by David Gaves, who announced it, that we were not to abuse it. And in no uncertain terms did the registry in those days say, "Do not abuse this or you will face consequences," which we took very seriously -- which we told very seriously. And we told CORE and our members, do not abuse the grace period, because there will be consequences, and we, CORE, will act first before the registry acts against all of CORE. Now, when we discovered that this was actually now changed, we kind of felt, you know, it's not nice, you know, suddenly, the registry has changed the rules, and we discovered that at some point this rule ended up in the contract as a purported obligation of the registry, and that the registry would actually have to beg on its knees to ICANN, oh, please let us change that. There's no need for that. The registry originally had the ability to change that. But the question is now, can we change this and do we want to change it and do we want to have the data for this? And here again comes another problem that started increasingly over the period that I have been watching this. And that is that we play with identities. I saw the statistics here about the behavior of the number of deletions by registrar and so on. And, quite frankly, I mean, it is not true that the registrars stop or actually have a lesser rate of dropping when they have 90%. It just happens to be the second registrar of the same group, or the third registrar of the same group that just happened to take this one specific case. So we have the problem of what, in financial markets is called cornering, where, in financial markets, there are rules and have been rules for a long time, and there have been, in financial markets, obligations to announce. For instance, in financial markets, people will have to announce if they do something. If they act in concert, they will have to announce that. And if they do not announce they're acting in concert, they will face penalties for not announcing that they act in concert. I think you might think about a solution like this for ICANN. >>JACQUELINE MORRIS: Thank you very much. >>JON NEVETT: Thanks. This is Jon Nevett. I'm the chair of the registrar constituency, and I just wanted to highlight one thing that Tim said in the beginning, Tim Cole, that I believe the statistic was about 25% of registrars were engaging in some kind of tasting activities. Is that correct? >>TIM COLE: I said that roughly I think it was 28% have deletes in excess of 50%. >>JON NEVETT: Okay. So that means that 72% do not. And I just wanted to make it clear to everyone here that it's not every registrar engaging in this activity and that the registrar constituency has not taken a position on tasting. Obviously, there are folks who engage in the activity that I'm sure support it. And, clearly, Margie's up on the panel coming out against it. A number of registrars have come out against it. And it's definitely a debate we're having internally. But I didn't want to leave this tutorial with anyone having the impression that every registrar was engaging in this activity. So thank you. >>JACQUELINE MORRIS: Thank you very much. >>MARILYN CADE: Can I say something, Jon? Thank you very much for doing that. Because that's something that I skipped in my points, but it is quite apparent that it is a small percentage of the registrars. I still think it's a serious problem, but I welcome the fact you made that clear. >>MICHAEL PALAGE: Mike Palage again. Since this was intended to be a fact-finding mission, a couple of facts that I did want to leave on the table. And this, as I said, notwithstanding my opposition to domain name tasting and wanting to find a solution, I want to put some facts on the table of how a solution can negatively impact two businesses. I cite to two cases. The first is in connection with a credit union in Canada. This incident was actually documented in connection with dot coop's proposal for a new registry contract that, hopefully, will be taken up by the ICANN board this Friday. In the appendix, there was a situation where a credit union in Canada went and registered a bunch of domain names that included the word "sucks" at the end of the credit union's name. The registry contacted the credit union and said, why are you registering these names that you -- why are you registering them? And they said, well, this is just part of our defensive registration process. When it was explained to them that the protocols in the dot coop registry made that not necessary, the credit union said, "Can I cancel these names?" Since it was within that five-day add grace period, the registry worked with the registrar to cancel those names and give the credit union back its money for those domain names that were registered. Case number one where the add grace period did have a positive impact for business. Case number two. A number of years ago, there was a newly ICANN-accredited registrar in south Florida who became operational. Within the first one or two months of operation, its system was hacked by a hacker, and their entire account with VeriSign was depleted. Okay? What this registrar was able to do was, after realizing that their systems had been hacked, that their entire account had been deleted with VeriSign, they worked with VeriSign, using the add grace period, to cancel all of those names, get their money back, and continue operations. So, as I said, Marilyn, you and I are in agreement that the domain name tasting is a bad thing. I don't like it. You don't like it. Where we differ is how we go about addressing the solution so that it is done in a manner that does not have unintended consequences. And that is -- just from a fact-finding standpoint -- why I wanted to get this on the record, so that, if you will, an elimination of the five-day grace period that has been proposed by some in the community may lead to unintended consequences for legitimate businesses. >>JACQUELINE MORRIS: Thank you, Mike. >> PHILIP CORWIN: Phil Corwin. I'll be succinct since I'm the last one here, which means I'm the last one standing in the audience. Three quick points. One, the ICA is firmly opposed to domain tasting. We supported the PIR proposal and are happy to hear that it is being effective. And we are writing a draft code of conduct for our members, soon to be published, which comes out squarely against tasting, says basically if you want a domain name, pay for it, don't take it for free. And we look forward to participating in the PDP process. Second, there's more than one dark side going on out there. I just came from Domain Name Industry Conference in New York, and one of the main topics of conversation in the hallways was increased incidence of what in that industry is referred to as reverse domain name hijacking cases in which hyperaggressive corporate interests, knowing the economic disparity between them and small domain name owners, are going after names and often getting them because the person who owns them cannot afford to fight the action, and now we're hearing reports of trademark lawyers taking these cases on a contingency basis, which is quite worrisome. And, third, if we're going to link tasting to trademark issues, let's talk about that comprehensively and dispassionately and have trademark law which is even handed across the board for all types of search, whether it's Google-type search or direct search, and for all types of advertising-driven models. And in that context, it's well known that the U.S. courts are consistently finding that the search engines cannot just sell typos of trademark names, but can sell the trademark names themselves to competitors of the trademarked companies, and there's no trademark infringement. The courts are holding that's not a use in commerce. And we type simple typographical variations of trademark names like Verizon, if you type "Verizen," "zen," instead of "zon," into Google, you get a search results page in which Verizon, rather than being horrified at the abuse of their name is paying for the most prominent-placed ad on that search results page. So let's deal with tasting and then trademark separately and dispassionately. >>JACQUELINE MORRIS:Thank you very much. We've had a long and -- one more? All right. One more. >> DAVID STEELE: I appreciate the opportunity to respond to some of the comments made by Phil. I'll be generally brief. Firstly, I don't agree that it's settled law. In fact, if anything, the law settled in the opposite direction. But in terms of actually discussing the law and settling the law, that's already been done. And the enactment of the ACPA is clear, it establish what is is cybersquatting and what isn't cybersquatting. This is not the forum for us to discuss appropriate sovereignty and laws in the United States, the ACPA is quite clear, and, in fact, if anything, it states quite clearly and, in fact, recently, a preliminary injunction was issued against a domain name taster who, in fact, also allegedly had registered domain names which were identical or confusingly similar to trademarks. I agree about treating tasting and cybersquatting separately. But this group needs to open up its eyes to the goings-on, that is in many cases the tasters are also cybersquatting. Thank you. >>JACQUELINE MORRIS: Thank you very much. So I'd like to thank the panel. We have a lot of information on the options that we do have. Right, Maria? The GNSO will be discussing this in the public forum, Wednesday, 8:30. This room? >> Yeah, 8:30. >>JACQUELINE MORRIS: 8:30 Wednesday morning, this room. At that time, we will be discussing the statement -- we can take statements for or against the various actions that we have that the GNSO has -- that the GNSO has to decide between when they're about to make their decision. Okay? Thank you very much for coming. >>MARILYN CADE: Could I just announce -- [ Applause ] >>JACQUELINE MORRIS: Sorry, Marilyn, go ahead. >>MARILYN CADE: Somebody left their Blackberry charger up here. And I have somebody's flash drive -- >> How many gig? >>MARILYN CADE: If somebody is worried about their flash drive, tell them to see the ICANN staff so we can get it back to them. [7:10 p.m.]