John: 

In your opinion, did ICANN act relatively quickly or take too long to act?

I think they may have seen the beginning of registrars building products on top of the AGP. Once that ball got rolling, it might be hard to get registrars to buy into a policy change. Did the NSI front running issue push them to act more quickly?

Do you think that there is any chance that the ICANN Board or registrars will fail to approve this resolution?

I have also posted on this issue here, and set forth the entire resolution and committee vote.

John,

Unfortunately, I don’t believe the change will go into effect that quickly.  Based on both the resolution and information the staff is providing to the GNSO Council, the fee will be included in the budget now being discussed.  The fee will then (as part of the budget) be subject to public comment, registrar approval (required if the registrars are to collect the fee), and then Board approval.  In other words, July 1, 2008 (assuming the fee makes it into the final budget).  Given the 3-month freeze on registry monthly reports, we won’t know until November 1 if the fee had the desired effect.  Too long in my opinion.  I hope the ICANN community will continue to push ICANN to take more immediate, certain action to end domain tasting.

Kristina

This is excellent news.  Oddly enough we may have Network Solutions to thank for this change.  I think it will improve alot of aspects of domaining.

  * The typo-squatting business will plummet as domain tasting is somewhat essential for profitable typo choices.
  * Expired domain after-market will thrive as lists begin to be “cleansed” from these bogus tasted names.
  * Pay-Per-Click parking services may become more reputable as domains will be long standing and garner targeted traffic.
  * Registrars will become more credible down the road, because in many cases they themselves were the biggest players in domain tasting.
  * Domainers will not have to worry about companies stealing WHOIS search data.

Via: Domain Tasting Goes Sour: ICANN Will No Longer Issue Registration Refunds

Mark:  I wonder how much of an impact the fee will really have on typosquatting.  Many typosquatted domains are intentionally contrived like www.wwwmicrosoftcom.com
There is no doubt some of these typo domains are contrived, tasted (smart squatters) and either deleted or retained based on typo traffic and PPC revenue.

But I am convinced that many typosquatters specifically and thoughtfully target trademarked domains and register without tasting at all. Many typosquatted domains are junk.,  They generate almost no traffic, and never could have generated any significant traffic. It is doubtful that the typosquatting registrant in those instances did any analytics or tasting before committing their dollars to the domain. 

There are of course typosquatted domains that have been around for more than 5 days and clearly generate enough PPC traffic to support the annual registration fee. And there are other clever ways to gauge potential traffic than tasting (although I certainly won’t share them here). Profitable typo domains aren’t going to be transfered to the trademark holder absent a UDRP or ACPA proceeding. I fear it will take more than a 20 cent delete fee to have a serious impact on typosquatting. i hope I am wrong.

For the trademark typosquatters, even those that are shell companies in the carribbean and costa rica .. there’s stuff like the dell lawsuit.

And GOOG cracking down on kited domains is more fun.

Nice, two nails (Dell, GOOG) and now a damned great stake in the heart of the business.

Days inn and Wendys time for TRAFFIC?

I fail to see what all the positive fuss is about. Ok, so there’s no more domain tasting (or less anyway, see PIR’s Org. attempt) and presumably no NSI gumming up the works (though they do know how to file or threaten lawsuits to make ICANN more pliable). SFAIK neither practice caused me any grief.

But as Bret Fausett points out, we’re talking about a multi-million dollar industry here, if the cost is $.20 extra to do business some will still do it. And some will just use the likes of Alexa (one of the best known, if not the best of many metric tools) to gather top site metrics (at $.065 or less each), compare that list to Whois shelf-life lists, and then hammer servers on the deletion date. This requires only a slightly less trivial scriptthan domain tasting, plus you wind up with a few choice domains instead of a lot of drek, and it’s cheaper.

Dropped domains have gone both ways in the UDRP, but again, just the cost of doing business. If you’re a bit more of a bad actor, you disable the site’s WHOIS contacts (through various means that aren’t unknown) so they don’t get a renewal notice, it only has to work a few times. And if you’re a bad registrar (and ICANN has them and won’t de-certify them), all the better.

I’m not giving up any secrets here, I wrote  about it way back in 2001, including the then un-named domain tasting in a followup comment. I suppose ICANN taking action 7 years after it bubbled to the surface is their version of moving at blinding speed. If a domain name is deleted what’s wrong with washing it for six months on the shelf unless the original owner retrieves it (or explicitly drops or sells it)? Even more problems would go away. As black hats were doing all this in 2001, what are they capable of now?  Perhaps we’ll deal with it by 2015. -g

Gary Osbourne said:

I fail to see what all the positive fuss is about. Ok, so there’s no more domain tasting (or less anyway, see PIR’s Org. attempt)

This changes the economics by orders of magnitude. The papers in the Dell suit reveal that high-volume tasters register and delete a million domains a day, so they are certainly churning through at least thousands of them to find the rare domain that will provide $10/yr in clickthroughs. That will all stop.

Yeah, they’ll keep squatting on deleted domains as long as Google and Yahoo keep paying for clicks on pages with no content, but that’s much harder to fix.

By the way, PIR’s change had the same effect, AGP deletes went down to almost nothing.

Gary,

You seem to have conflated domain tasting and drop catching.

“Hammer the servers” drop catching of expired names is pretty much dinosaur technology at this point.  Most major registrars do not delete names at the registry level upon expiration anymore, but shuffle them off into various auction systems.  This is the basis of such “backorder” systems as operated by SnapNames, NameJet, and others.  In the expiring and deleted name context, yes you are correct, there are numerous ways to determine probable traffic volume.

“Domain Tasting” refers to a much larger scale registration of previously unregistered domain names, and as practiced thus far, is more like using a drift trawler net to catch a few fish and throw the rest away.  The scale of domain registrations is in the range of six figures per week, and we have empirical evidence from .org that imposition of even a $.05 fee at the registry level is a sufficient deterrent to domain tasting on this scale.

The $.25 fee proposed by the ICANN board will kill “blind” domain tasting.  Of course, there will be retail speculation by ordinary folks, but the tasters would have to ratchet up their technology a bit - in particular by buying .com zone error data from Verisign, or by striking deals with ISPs, such as Verizon, which have their own monetization-of-non-existent-domain systems in place.

@ Jeffrey A Williams

I am sure you realize that what you describe would be too clever by half. You would find yourself in very very deep doo-doo with your credit card issuer if you dispute charges falsely. Especially if you dispute several hundred thousand 20 cent charges.  Or rotate credit cards and then dispute charges.

Your credit rating would be so far down the toilet, if you do that, you’d have to winch it out with a crane.  In fact you’d have to pay for your Days inn room and Wendys burger at TRAFFIC with cash. If you have any left from all the penalty charges you’d be dinged with.

Even “cleverer” tasters might want to hire some russian botnet operators, and drop / catch domains with various stolen cards and signups from a botnet of virus infected PCs. 

Spammers and botnet operators already do this on a fairly large scale. So registrars tend to have a lot of fraud management in place to spot this kind of stuff. 

Even then some stuff slips through, as all the bot domains being registered every day, at dumb or complaisant registrars, and random ccTLDs (not just some of the gTLDs) shows.

So, for “cleverer than usual” tasters, using botnets and stolen cards for this would be a very natural extension of your “suggestions”, once the legitimate players are driven from the market and the various shady types involved in this decide to get even shadier. 

In which case, well, law enforcement and the IP lobby will love that. And crack down on all domainers.  And probably use DoC oversight to

* Force ICANN to kill the AGP (or restrict AGP to registrars)

* Crack down on and shut down “registrars” that are shell companies in costa rica and the carribbean, and exist solely to kite domains.

* Increase the 20 cents to something like 3 to 4 bucks. Or figure that given you’re paying what, 6 to 10 bucks for a domain, retail, if you screw up and register the wrong domain, you can afford to lose all that. Not just 20 cents of it.

Your suggestions remind me of Wile E Coyote trying to catch the roadrunner by wiring TNT under two sides of a bridge, and then standing in the middle to blow the charge.  Wile E recovers within seconds, the domaining industry wont.

Suresh, if you don’t know of one Jeffrey A Williams please read the outdated but still damning FAQ on him here, I guess someone better update that if he’s going to spam the invaluable circleid.com. In what must be close to a decade now I’ve never managed to figure out whether he’s just another netkook, a randomized annoybot, or a disinformation specialist (that is, he’s Kent Crispin in disguise, which KC did do at least one famous time so my money is resting on the latter so far). I particularily liked Ross Rader of TuCows roadtrip to his supposed address and posting a picture of an empty field (I can’t find that on Ross’s blog anymore, Jeff’s INEGGroup must have hacked him, they’re worse than the Silentologists).

Other than that I’m on the road and my laptop grenaded but I’ll try to respond to the other excellent commentary ASAP, particularily from the two Johns. Hey, on the road, working the streets, it’s not that much of a stretch. -g

Ah. A kook. If this were usenet I’d have concluded that faster, I guess. Thanks, entertaining read, that faq.

In the unlikely but possible category:

With ICANN getting $0.20 (or more when that tax goes higher) on every name registration, if a name is tossed about frequently by tasters and drop catchers it could end up that the money flow could be much in ICANN’s favor.

Consider the unlikely event that a name is tossed about using the still existing 5-day add grace period (AGP) once a day for a year.  ICANN would then derive 365*$.20, i.e. $73, while the underlying registries that actually process these 365 adds and 365 drop transactions will get paid nothing.

As I have said several times - this entire discussion is being held in an information vacuum - until we know how much it actually costs at the registry level to handle a registration transaction all we can do is make stabs in the dark and hope that we’ve made the right policy choice.

(It may be that one reason that a deep and believable audit of registry transaction costs is not being pursued is that such an audit is likely to show that transaction costs are far, far below the registry fees that ICANN has created and thus reveal the degree to which ICANN has turned the registry business, at least for certain big registries, into a money pump that sucks money out of registrant pockets and into registry bank accounts.)

It seems extremely unlikely. When PIR added their restocking fee, the tasting dried up.  It’s pretty obvious that the gross margins for registries are gross indeed, since we’ve never heard a peep out of Verisign about the several million daily unpaid adds and drops that they’ve been handling in recent years.

I’ve been looking at some of the court documents in the Dell suit that give an interesting insight into the economics of tasting. Blog entry coming soon.

@John Levine

While I agree that .org frontrunning fell precipitously, I don’t think we can just generalize this across the board. First, it was .org after all, not as desirable as .com/.net. Second, a lot of runners probably just moved over, even tepid water runs downhill, trying to dam it entirely could be a different matter. Third, I don’t know how the ccTLD space factors into this. Fourth, regarding rare domains bringing in $10/year, that might barely cover registration and hosting costs. Not quite so rare domains can bring in $100+/day. See below for a sense of the money involved.

I don’t blame Google or Yahoo (at least unless it’s assimilated by MS :) ) in particular, although they could be doing a better job of policing, partly as many of the paid links are useless or worse, which from my experience is causing a consumer backlash. If the big two didn’t exist or changed policing/policies then squatters (I am quoting and somewhat misusing the term to mean those sites which are free of unique content) can, and some do, use any one or more of tens of thousands of affiliate programs which often pay better for very little work. Following my -g .sig is a bit (too much, hence its footnote placement) of my own history with this.

My reasons for the apparent bragging are as follows. First, the techniques still work, though with some modifications, and new ones added. It works less well given the drek and bad actors out there but it’s almost impossible not to turn at least a modest profit. Second, it takes little work or skill or cost, it’s all legal and not against any rules (only once did I get a cease and desist letter, I reasoned with the megacorp, put a prominent disclaimer I wasn’t them and linked to them and they went away, perhaps a rare case of sanity, or perhaps appreciating the traffic. It was that incident which got me into paying closer attention to the then-new ICANN). Third, others with names of equal or greater apparent value would just put up a domain for sale page or not even go live, they’d get listed on afternic or greatdomains, and wait for the money to pour in, which it almost never did. For years Yun Ye was making $millions/year by putting his ULTSearch content on countless sites, which was basic value add (not to say most of what he did was basic, it was brilliant). He sold his name porfolio for about $180 million (apparently including $9 mill in stock, various figures have been bandied about), $180 mill.equals about 900 million domain tastings at 20 cents each, a million tastings per day would take well over two years, so I’m not yet convinced tasting will just go away quietly. I would like to add, as with John Berryhill, while I don’t always agree with your opinions I always greatly respect and value them. -g
__________

How I monetized domain names: Beginning over ten years ago I had sites with names like (but not exactly, I’m almost as secretive as master Yun Ye) i-books.com. I’d fill a page with book related affiliate links (EG: Amazon.com which paid me 15% of a customer’s first purchase and 5% for returnees, what ever happened to their affiliate patent application BTW). There were affiliate aggregate sites with thousands of listings and it was a simple point and click for a few lines of HTML and graphics for numerous affiliate book sites. Additionally, there was free or inexpensive software and/or sites which would pipe in the latest book/author news, rotate sidebar featured books, etc. (hence at least daily unique content), submit me to over a dozen of the top search engines, and report and help increase my position on them. With rule-following search engine techniques (incl. metatags, tagged graphics, reciprocal links, keyword in URL, etc.) I was good to go (I was never banned from a search engine for breaking rules). I could do a site in less than an evening as could anyone who could RTFMs and knew even rudimentary HTML. Once a site was up I could just forget it, although I did tweak and value add on many of them. Most of these sites were a single page, only one was more that five pages. I also used a few legit tricks I won’t pass on but it wasn’t rocket science. One of my sites was #1 on Yahoo for a common word for over a month, others were on its first page, I did even better on some other search engines. The first site I took live brought in $30 its first day, and it was then only listed on one small search engine (Yahoo commonly took about 3 months). I covered its yearly cost in less than two days. BINGO, epiphany time. I then put up a second like i-music, and a third like i-games, and repeated as necessary :) . I was doing this in my spare time and soon automated much of the cookiecutter process. I never made any secret of most of what I was doing and would proselitize value adding to domain sale fora like afternic. In response I was treated more or less like the three contrarians in Lord of the Flies. Oh well, less competition. I never felt the need to buy a second hand name or bulk register so I never did. What I made is between me and the tax department, honest.

@John Berryhill, who wrote:

You seem to have conflated domain tasting and drop catching.

I didn’t intend to conflate the two. I would argue that there is some overlap, but as your defininition of domain tasting as only going after new names that would be pointless. I don’t know if many tasters check whether the name was previously registered, to use your analogy does a drift trawler check or care whether a fish has been caught before? It might be to their benefit to have dropped names as some would come with existing traffic, prior links in, and search engine placement (at least for a time, which might artificially inflate the potential hit numbers via just tasting). Some players have certainly used both tasting and catching, as I think you’d agree. Other than that and whether tasting will entirely die (I’m not arguing it won’t drop considerably, but see my recent response to John Levine), I’m largely in agreement with you.

On a somewhat related note, I recently read a fascinating article. from last year in which you are quoted about Ken Ham et al, detailing what the domain cognoscenti and other big money boys (Starbucks CEO? GAAKK!) have recently been up to in the namespace (note some were already talking about no longer tasting, and the value of value adding). But I am skeptical of this idea that many users type athleticshoes.com in the browser URL line without knowing whether such a site exists. I’ve watched over hundreds of shoulders, from newbies to experienced (though perhaps not too technically aware) users over the years, and I have never once seen anyone do that. On the contrary, particularily with the growth of Google/iGoogle as a preferred home page, I have often seen users type a web address in the search line rather than the browser URL line. They almost always get to where they’re going, and there’s the value add of additional links which may say more about that site, perhaps not all positive. Years ago Yahoo listed their top ten search terms and ‘yahoo.com’ was sometimes on the list. I never got my head around this, why would one search for a site one was already on, it was before the days of the Yahoo browser bar add-on which might have otherwise partly explained it. I think in general we have a more sophisticated user base now, particularily in long-wired countries. But this claim that 15% of users type example.com in the browser URL line without knowing whether it exists strikes me as suspect (plus a hyphenated athletic-shoes.com or .net or… would rank higher on various search engines, so defensive purchases could considerably increase expenses). I’d like to hear your opinion (or anyone’s, almost) on this, although I realize professional considerations may preclude that. BTW, again aware of the previous proviso, what’s the current buzz on .cm? I would like to add, as with John Levine, while I don’t always agree with your opinions I always greatly respect and value them. But you already knew that ;) . -g

Gary is quite definitely confusing domain tasting and domain squatting.  They are not the same thing at all.

Domain tasting is registering huge numbers of more or less random names of at best marginal value, typically variants on famous names, measuring the traffic for five days, and then abusing the AGP to refund the vast majority that don’t have enough traffic to be worth keeping.  Maybe someone sold a fabulous domain portfolio for $180M (keeping in mind that there’s a lot of puffery in this biz), but I doubt anyone could find those kinds of domains available today.

Have you ever refunded any domains during the AGP? Doesn’t sound like it.

Incidentally, as far as I’m concerned, registering names that resemble other people’s and putting mechanically generated content on them to trick people into clicking on links to the sites they wanted to visit in the first place is just as abusive, albeit nominally legal.  It’s sort of like going into the condiment business by visiting all your local restaurants and scooping up the free ketchup packets they set out for their real customers.

Maybe someone sold a fabulous domain portfolio for $180M (keeping in mind that there’s a lot of puffery in this biz)

To be clear, the purchase was by a public company, and if there is “puffery” in an SEC disclosure statement, someone is going to jail. 

http://www.secinfo.com/d14D5a.1658z.htm

Gary, it is simple not the case that domain names with inherent traffic are as available as they used to be.  Domain tasting is the final gasp of the evolution of extracting value from the name space.  It used to be that petroleum could be collected, even in the State of Pennsylvania, from surface pools.  After those were exploited, drilling and the use of derricks for pressurized deposits became common.  After that phase, it became necessary to pump the oil out.  Now, in many parts of the US, petroleum is extracted by injecting steam and detergent in the vicinity of those pumps. 

Perhaps gold mining is a better analogy.  Long ago, one could find surface nuggest… and then panning… and then mining… until today, gold is extracted from gold-bearing ore in a process that requires tons of ore to produce a single ounce of gold.  And that is precisely what large scale tasting is about - using a large sample space to identify a fractional percent of names worth keeping.  The process necessarily requires unbelievably large scale testing, and that is why a very small per-name fee renders it utterly uneconomical.

Frankly if I were of the belief that “not quite so rare” domain names worth $100 per day were readily available, I wouldn’t be wasting my time here, and I doubt you would as well.

As Mr. Levine demonstrates, the elimination of tasting, which I agree is a good idea, is only a step on a longer road for those of a certain mindset.  There are those who believe that the use of such domain names as lasvegasvacations.com for the purpose of displaying paid links to various providers of Las Vegas Vacations, which is indeed what the domain name does, is an inherent evil of some kind.  Or, to use another example of a PPC page, this mindset holds that it is of greater social utility for a domain name like cameras.com to be the exclusive province of some lucky camera shop (which is nowhere near you), instead of to have accessibility via such domain names be determined by competitive keyword bidding.

Those of this mindset are not above lying:

http://www.arb-forum.com/domains/decisions/434268.htm
(“For these reasons, the Panel is obliged to and does find that the Complaint was brought in bad faith, and that it constitutes an abuse of the administrative process.  See Rules, supra, §15(e); see also Deutsche Welle v. DiamondWare Ltd., D2000-1202 (WIPO Jan. 2, 2001).”)

forging evidence:

http://www.wipo.int/amc/en/domains/decisions/html/2007/d2007-0823.html
(“The Complainant has not provided any explanation for the insertion of 2003 and 2004 dates on a form of certificate which, on its face, would not have been in existence in those years.”)

and hiding facts:
(“The Panel is in no doubt that it was incumbent upon the Complainant in the circumstances of this case, noting in particular the nature of the domain name in question, to provide the Panel with sufficient detail of the Respondent’s website to enable the Panel to make a fair assessment.”)

...in the advancement of this notion that the use of domain names in connection with paid advertising search systems is some sort of unspeakable evil.

Absolutely, the parasitic registration of domain names with the intention to capitalize upon trade or service mark rights known to belong to another is an abusive practice, and is effectively addressed under the UDRP and national laws.  The automated nature of domain tasting has the incidental effect of sweeping in trademark typo variants by a mechanism similar to the way that the Google search engine is able to effectively “spell check” without “knowing” the correct spelling of anything (i.e. by comparing the search string to similar and more common search strings).  But precision in identifying the various species of domain registration practices is important, because at its root there are a number of quite subjective “I don’t like it” opinions about practices that are qualitatively distinguishable in intent, effect, and legality.

Missing reference from third citation:

http://www.wipo.int/amc/en/domains/decisions/html/2007/d2007-0770.html

John Berryhill said:

To be clear, the [ $180M domain name portfolio ] purchase was by a public company, and if there is “puffery” in an SEC disclosure statement, someone is going to jail. 

http://www.secinfo.com/d14D5a.1658z.htm

Oh, right, I’d forgotten about that bizarro 2004 transaction.  While we’re looking at SEC disclosure statements, I’m surprised you didn’t also notice this one:

http://www.secinfo.com/d14D5a.u6977.d.htm

In 2007 the company that paid $180M sold itself for $11M, suggesting that in the former transaction, it overpaid by at least $170 million.

We certainly agree that whether or not it ever was possible to find fantastically valuable domains available for the taking, it isn’t now.

As far as the rest of the rant, it just confirms my point that since we are stuck with people so ethically challenged that they can’t see any problem with value-subtracted domain squats, that abuse (which is different from domain tasting) won’t go away without specific action to get rid of it.

If you have a higher-value use for cameras.com, I’m sure it’s owners would be happy to work out a deal with you, Tovarisch.

In 2007 the company that paid $180M sold itself for $11M, suggesting that in the former transaction, it overpaid by at least $170 million.

That’s not even what that filing says.

http://finance.yahoo.com/q?s=mchx

Market Cap:  328.72M

It wasn’t sold to anyone.  Decadent capitalist counter-revolutionaries are still trading shares in it.

I realize that capitalism is sometimes difficult to understand, but you appear to have read the plan of merger in that SEC documentbackwards.

Yeah. And he’d still have to bequeath that domain to several generations of your descendants, have them monetize it with ppc traffic, before they get the money back that they paid for the thing.

Unless they can find another sucker^Wdomainer to sell to.

Well, there’s the problem.  No doubt the current link farm on cameras.com brings in lots of revenue from the suckers who foolishly hoped that cameras.com might have some actual information about cameras.  As I said before, this won’t stop without some rule changes.

Re Tovarisch, you might want to take a few minutes out from value-subtracting and read the papers now and then.  Russia hasn’t been communist for a long time. Now it’s run by rapacious capitalists who think nothing of grabbing public assets on the flimsiest of excuses and using them for individual short term profit. Take a look, could be your kind of place.

Oops, you’re quite right. I should stop blogging at 1:00 AM.

But Suresh’s point is well taken.  Marchex has yet to be more than marginally profitable, and has lost money in the first three quarters of 2007, so it’ll be a long, long, long, time before they show a return on that $180M.