Domains are inefficiently priced -- small domain owners are the problem | Hacker News Reader

Domains are inefficiently priced -- small domain owners are the problem

semmyfun.blogspot.com

38 pointsnerfhammer61 comments

This is all well and good unless the distribution of domain name ownership is that 1% of domain owners own 99% of registered names. If you want a domain that happens to be owned by a small owner, I'll agree you could face much bigger problems. If, however, there were only small domain owners (and if my ratio above applies), then 99% of the time, you wouldn't face any problems at all because 99% of the time, the domain simply wouldn't be registered yet.

Despite being wrong about the problem, the author's solution is not bad. For a serious commercial endeavor, $100/year is insignificant. For a personal site, it's a fortune. The author's solution does seem to solve this problem.

There are still problems though.

1. The .com TLD is so prolific that using a different TLD is a problem even for personal sites. The author's solution would've been great if it was implemented from the start. Right now, it might be good at best. Even if you just show your site to a few of your friends, some of them are likely to get the TLD confused if it's not .com. If this was implemented from the start, though, that wouldn't be a problem because everyone would know that personal sites don't use .com.

2. Domain squatters, the real problem, will still exist. They'll just be more selective. If someone can get beer.com, they'll still do it if it's $100/year. That's still likely to be a good investment. If you raise the price enough to make that a bad investment, you've probably priced out a lot of serious commercial sites that just don't have the money for an expensive domain name.

Perhaps a better solution from the beginning would have been to have a bunch of cheap TLDs that anyone could use for anything at $1/year, and then .com which would have varying prices depending on the value of the domain. True, you can't predict with certainty what will be valuable, but you can predict with reasonable certainty that any word in the top 1000 English words is likely to be valuable. You can similarly predict that shorter names are likely to be more valuable than longer ones.

Of course, we don't have the option of changing what was done, so that's probably all moot.

I'm not sure that domain registries attempting to extract maximum value from .com would have made the market more efficient...

Why not? You could also auction being the registry for .com domains: if it's worth so much they will spend more money, which the ICANN can use to subsidised as needed.

As an example look at the auctions for the mobile phone frequency licences: that worked pretty well.

The same problem with evaluating the value of domains exists; no effective price signalling and temporal fluctuations in the value of a domain to any individual.

Moreover, by capturing much of the value of domains currently realised by the domain aftermarket, the registrars would remove much of the little amount of signalling in the market that currently exists.

It's like suggesting that dysfunction in the labour market would be reduced if wages were set by a really big organisation, like the government.

But I didn't say ICANN would set the prices, I said it would auction them. Not sure I understand how your point applies.

Maybe it's a different about inside the USA vs. outside of it, but I don't feel everything ends with .com: a lot, yes, but not a massive absolute majority.

I think it would be nice to have different prices according to what they are _supposed_ to be for. Just like .gov is reserved, so .com could be priced at 100$/year since if you are a commercial company, i.e. you make money, then it would be a fair price to pay.

Others like .org should be almost free, but only for companies which are legally non-profit. A domain for personal websites on the other hand should be cheap and have no limitations. And so on.

So basically I agree with what you are saying but I'd just make it slightly more fine grained.

As for the problem you mention, I think people would change pretty quickly to be honest. Maybe there could be some governmental ads to drive the message home, but com-> commercial is pretty easy to remember, and similarly if it's not .com then it's not commercial. :)

The point of having a hierarchical URL, with most people buying the second level, is to increase the number of choices. If everyone ends up buying .com it's just a waste. :)

I agree 100% on your 2nd point. It's tough to price out a domainer but not price out the small business or startup who probably has smaller pockets. By the time the small business has the money to buy the domain, they have already built a following & links on a lesser quality name.

Cheap pricing like Godaddy's $.99/year .info just pollutes the web with spammy sites - think of it as tons of little Mahalos.

Tiered pricing hasn't worked out that well for other extensions. Originally .tv had a tiered pricing where you paid a premium yearly rate, however, they abandoned that practice this Spring. [1]

[1] http://domainnamewire.com/2010/03/18/verisign-drops-prices-o...

The function of the registrars is as a mechanism for homesteading, much like establishing initial private ownership of some newly discovered land. As such, the price charged by the registrars should be largely irrelevant and held as low as possible to not inadvertently set a price floor (this is after all a commodity with effectively zero marginal supply cost).

Once some individual comes up with and homesteads a domain, it can be used by them toward some ends, or traded/auctioned in the market whereby a price will emerge (see Coase theorem). To be clear, there is no "right" price, just whatever the market will bear, even if the market is constrained to a single domain name. Externally applied "fixes" would only serve to distort the market in unexpected and harmful ways. As Hayek noted, "The curious task of economics is to demonstrate to men how little they really know about what they imagine they can design."

If there is a problem here it is either that the transaction costs (including non-monetary costs) of such exchanges are high (which provides an entrepreneurial opportunity), or the value of the domain to its current owner just happens to be higher than what others want to pay and those others are upset. Not much can be done about the latter, short of coercion.

Instead of selling domains, they could lease them, and auction off the new lease once the old lease has expired.

Most auctions would probably not be contested, and you had the same situation as before. But for the interesting domains, ICANN could thus capture the value of the domain.

Neither this or the current solution (what you describe) is preferable from the outside. It's just that different parties capture the values of the domains. (And if I were to make the rules, you can bet that I would ensure that I capture the value. Or at least some of it.)

we do lease domains now. You don't actually own the domain...it can be taken away at any time.

Maybe. But ICANN isn't capturing the value, or is it?

Fine, except some of the big owners will not sell under almost any circumstances. There is a domain which I am prepared to pay five figures for, but the owner (one of the largest domain squatting firms) has a policy of not entering into correspondence about the sale of their domains. Frustratingly, the domain has virtually no commercial value to anyone but myself, but the owner has obviously decided that it is not worth their time.

Personally, I think that the current domain system is a nonsense. In a system with a finite number of short, pronounceable names, it seems manifestly unjust that most names are not being put to good use. I'm of the opinion that ICANN should set a blanket ban on squatting, and that failure to make good use of a domain should be automatic grounds for revocation under the ADRP.

But how would you define what constitutes 'good use' of a domain?

Something more substantial than simply displaying automatically-generated sponsored links.

But why does a domain need to host a webpage? I get that we mostly want snazzy domains for webapps these days, but if I grab a domain, point it towards a server that only serves SSH externally, what's wrong with that (Actually, I am considering doing just that for a domain I have been eying).

Defining 'substantial' is difficult, especially if we start pulling lawyers out to sue for the lost of virtual 'real estate.'

Domains with significant type-in traffic should be very easy for the owner to price. Assuming the domain owner is rational.

Refusing to negotiate may also just be a hardball negotiating tactic. For instance reflex.com says "Our domains are not for sale.... and here's our contact info we're leaving here just to be friendly."

Immensely frustrating though

Sorry to hear about wanting a name that a seller refuses to sell. I'm in a similar situation with a neighbor who has several acres that adjoin my ranch. I've offered way above market value for the land that doesn't even have access to a public road, but he still won't sell. It's definitely frustrating.

Defining "legitimate use" is such a slippery slope. With automatic content generation becoming so advanced today, it's become increasingly difficult for both search engines and humans to tell spam from legitimate content. It's even worse when some of the spam and scraping is more useful that someone's 5 year old blog with pictures of their pets.

How about instead of increasing the cost, we decrease the value of squatting on a domain? Browsers (via a plug-in or built-in mechanism) could show a little warning page before loading the actual site if the site is known to be a squat page (the way Google does with malware sites). If enough browsers had this feature, it might cut the revenue generated by squatters to be not worth it for the vast majority of domains.

It's a great idea, except Google funds 1/2 the squatted domains and browsers.

One answer is to use an analog to a land value tax. Call it a "domain value tax". The idea is to have the most productive user of that domain (judged by the one who has the most money) pay the efficient rate. This leads to perfect efficiency, however it means your company's domain might be highly bid up by someone else - not fun. We might not want ::perfect:: market efficiency here.

Shameless plug - My article on this:

http://www.nickpinkston.com/2010/04/domain-value-fees-cybers...

Fortunately, there's an almost infinite number of domains, even within the .com tld. It just requires a little more creativity than it used to to think of a catchy one that hasn't been taken.

Now, if we were actually running out of domain names, the way we're slowly but surely running out of IPv4 addresses, then we'd have a serious problem.

     Dot-com names should cost $1000 and there should be 
     another TLD that costs $1/year for all of our hobby 
     projects

Hobby my ass, just because you can afford $1000 and I cannot, that doesn't make my project a hobby.

Livable land on earth is more limited than the number of decent domain names (yeah, you can make a living without owning sex.com ... sexandbutter.com would do just fine ;)).

And yet I don't see too much bitching about land-owners that don't want to sell. You want to get that piece of property, then offer more money (everything has a price).

Yes, I have one bitching domain name that I haven't used yet (owning it for the last 2 years), but I will and when or how is none of your business.

Get off my lawn!

technically it does.

if you can't afford $1,000 that means you aren't even making $1,000 a year. If your project only makes you $83/mo...it's just a hobby.

Technically you can make 1 million $ in revenue and not have $1000 in profits.

yes...but it doesn't work the other way...you can't have $1,000 in revenue and $1 million in profits.

I don't see the relevance.

Spending $1000 a year for a domain is $1000 that gets cut from profits and $990 more than I'm willing to spend on a domain that's on standby until I have the time to work on my idea.

On the other hand I should have the first-mover advantage, because yes, I thought about that name before somebody else did.

Properties in general are cheap for first-movers ... I can think of dozens of places where I can buy land from uninhabited areas with a breath-taking mountains panorama ... for small prices like $5-$20 per square meter.

Prices get high sooner or later due to infrastructure developments, or increased demand, but thought luck, it's my property now.

And yes, for land you pay taxes ... but you're receiving services in return, like water, electricity, public roads, hospitals, etc...

So you want that domain name so badly? Pay me a sum I cannot refuse.

Sure it will cost you more than $1000, but that's life for you, and you shouldn't have a problem if your project is not a hobby :)

So you're proposing 80 total websites with $500/yr profit each should actually be costing their fictional single owner $40k?

Plenty of people blow $1000 per year on a hobby, though it's usually cars and parts, camera lenses, ski gear, or greens fees instead of domain names.

Here on a startup site, how many of us have purchased domains to start a company? A lot of people here bootstrap their own businesses, which often can't make any money before they own a domain and put their site up. In that case, 1000 dollars is a lot.

"I don't see too much bitching about land-owners that don't want to sell"

Actually, there is, and with similar arguments to the domain name problem, only on a larger scale. Just imagine you manage some large-scale development project, but there's an old couple sitting in the middle of your projected building site that refuses to sell at any price. It happens a lot more often than you realize. (the large scale developer usually bullies its way in, which is not as easy with domains I guess).

One local example to illustrate my point:

http://maps.google.com/maps?f=q&source=s_q&hl=en&...

You see the green patch and the house in the lower left of the parking lot? The people there didn't settle with the supermarket owners, so they simply routed the parking lot around it.

There are lots of cases of inefficient land use, and land/house owning incurs in much higher fixed costs than U$ 7.

I meant that nobody disputes the protections of property the constitution of a democratic country gives you.

That family that didn't want to sell may be a PITA to deal with, but it's their right, and I would be really offended if some law passed that favors the supermarket.

As for taxes for land ... yes, they are much higher, but you get lots of stuff in return, as I've said, public roads, public hospitals, public schools, infrastructure for water / electricity, etc...

So raise the tax of owning a domain to $1000 / year.

What will that accomplish? Sooner or later all the good domain names will still be taken and we're back to being creative or to putting our money where our mouth is.

How about the cost per domain increases the more you own. For example, the first one costs $7/yr, the second, $10.50, the third, $14, $17.50, etc. in a gradual linear curve, (or maybe $7, $10.50, $15.75, $23.63, etc. for an exponentially increasing curve) to discourage hoarding.

It would be very difficult to check though: a bit too easy for people to find loopholes or cheat somehow.

True, but at least the framework would be there, and it could gradually be made more difficult to circumvent without significant effort, potentially making a good dent in the problem.

I think this is one of the most well reasoned pieces of the domain name issues I've read in a long time. I don't quite agree with the conclusion as a real solution. It's too late in my opinion. We can't start over and have such dramatic pricing changes. In fact, I am appalled to see prices going up (VeriSign increasing prices 7% annually for no reason in a contract that wasn't competed for). Prices should be going down.

As far as negotiations and tracking down buyers... it varies. I've done domain investing for a living for many years. Most people big or small are quite easy to contact. Some big guys are a real pain to deal with (it's not worth their time to sell even for a few grand). Some small guys are more than happy to sell a name they no longer use. I cannot generalize from my experience which type is better/worse/harder/easier/etc. It's as unique as the domain you want.

Finally, most of these people aren't squatters. Squatters are violating someone's trademark, property owned by someone else. Using Beer.com and saying someone is squatting on such a name is complete and utter bullshit. It's a generic word, it's not being cybersquatter unless someone has a TM for beer on its non-generic purpose and that is what they are using the domain for. Thanks.

Snapping up domains names to sit on them and resell them at a SIGNIFICANTLY higher price when someone finally wants them is commonly referred to as squatting. It doesn't matter what the official definition is. People will use words as they see fit.

People will be corrected when they use words improperly. There is a legal definition, and it's an important distinction.

At the time of this posting, there are 12 references to "squatting". While it may be arguing over semantics, it is an important distinction as there is a difference between "squatting" and "investing".

An individual who saw the foresight and bought a few acres of land that later became highly desirable for a subdivision is heralded as a real estate investor while someone doing the same with domain names is lumped with the shadier individuals in society.

According to the United States federal law known as the Anticybersquatting Consumer Protection Act, is registering, trafficking in, or using a domain name with bad faith intent to profit from the goodwill of a trademark belonging to someone else.

Source: http://en.wikipedia.org/wiki/Cybersquatting

After seeing how good domain names could help me better market my software as a startup on a tight budget, I became intrigued and am a part-time domain investor (domainer).

Uh, I don't know what it's like where you live, but where I am "real estate investors" are some of the shadier individuals in society...

I've never squatted on a domain.

Is the idea here that each year every domain would be subject to auction to the highest bidder? If so, that hardly seems fair that after having spent thousands of hours promoting a domain and invested in it that a spammer can come along and outbid me for my own domain.

If the idea is that this should only happen when you first buy, it's similarly absurd. When you buy a domain there is only one person that wants it - you. Otherwise the domain is "not available".

Also, no one made the domain. Why am I paying some third party for a domain anyway? It's not like their labor manufactured the domain.

The problem really only exists for the .com TLD. What really needs to happen is to deprecate .com entirely. ICANN should stop registering them, and all existing .com's should be migrated, or re-registered in their country of origin (.co.us, .co.ca etc). Each country can then decide on the restrictions of their .co domains. I would prefer .co domains to be of officially registered companies and trademarks. If you're not a business, then you don't need a .co.* plain and simple.

That's stupid simply because the Internet removes barriers and many companies operate on a global scale.

It is trivial for a multinational corporation to register company.co.kr and company.co.us. If they don't really want to run separate websites one can redirect to the other.

Why would the physical location of a company matter if that company provides online services?

While trivial, it's not good marketing.

The only instance I see country TLDs as being actually useful is for localized versions of said sites ... Google already does it, but it's not like it is useful for an email address to end with "gmail.ru", quite the contrary.

A better suggestion would be to use custom TLDs based on some category your website fits into ... like ".fm" for radio, or ".info" for hosting your personal resume, or ".doc" for archives of documents, and so on.

Sure, people are very accustomed to .com domains, but who's to blame for that?

Also a brand is a brand ... say you have the luck of registering "google.nu" ... most likely Google can take it from you by force ;)

ooh I want to bid on some stuff...http://ebay.com...404? WTF

your plan would pretty much eliminate 99% of websites.

There of course would be a period of migration.

Ebay would tell ICANN redirct ebay.com to ebay.co.us indefinitely. And maybe at some point in the future .com is entirely eliminated.

What happened with domains over the last 20 years (landrush) happened with property hundreds of years ago. Governments basically gave it away to ex-convicts (in Sydney, Australia for example) to own and do with as they please. Over time, the land was bought and sold and now modern day Sydney sits on that land. This will happen with domains over time. And, unlike land we can create new tlds like .jobs .travel etc.

Domains are archaic and waiting for a better solution, give me my semantic web already.

I think $10/mo is fine for domains. Actually I think it should be $20 a month.

The small time owners will have no problem paying that. Who doesn't have $20 a month for their blog? Eat one less taco bell meal and you are good to go.

And it won't do much for the small time consumers who get 5-6 domains. It's bad...but not by much. The only people such a price would hurt are the guys with 2,000 domains who'll go from paying $20,000 a year while making $180K of selling domains, to paying $480,000 a year, while still making the same $180K a year(prices will increase, but not by much...can't very well ask $500,000 for strawberries.com)

This will free up a ton of short term domains from the guys who get exact match domains hoping someone will buy one(whiteweddingdresses.com, ipadsoftware.com, askhackernews.com, goodbookstoread.com). At $10/yr it doesn't matter if someone buys...at $240 a year you can only afford to keep the really good domains.

And this can actually happen. The registars don't control domains...ICANN does. All they need to do is bump up the ICANN fee from 30 cents to 18 bucks for the .com and bam instant market correction.

And with all the money ICANN will get, they can create a free domain extension...i.e. something like .web. And the rest can be spent on improving internet infrastructure world wide...laying fiber, so anywhere in the world you can get 1gb/s.

ICANN is a non-profit...it should be doing more than just solving domain disputes.

>The small time owners will have no problem paying that. Who doesn't have $20 a month for their blog? Eat one less taco bell meal and you are good to go.

If all you have is actual people you know reading your blog (and that's most people) there's no reason hosting should cost more than $5/month. For those people, 2/3rds to 4/5ths of the cost would be something that basically costs nothing, and is only priced because there needs to be a mechanism for assigning ownership.

I mean, where would all this money go exactly?

those people can be on the free .web extension...or the .nets/.orgs. The new pricing would only apply to .coms.

as far as where money goes, infrastructure improvement.

It also could go to subsidies the free (or cheap) extensions.

For example: .com could cost 15$ instead of 7$, but .net could go from 7$ to 3$ or even less.

Yes, you would end up with Internet ghettos, and "high-street" areas, but given it's only the extension, and a .com address is not really surrounded by other .com addresses, then it's not such a bad thing.

> Who doesn't have $20 a month for their blog?

Vast majority of people in the world. Everyone who is below middle class income in the USA (which is a lot more than you probably realize).

The Internet is not just for rich people and it should certainly not just be for making money.

fine...first 3 domains at current prices...everything else $240/yr...or they can just get a .net...or get a domain on their country's TLD

Their country's TLD?

I assume you are American, but when did exactly .com became the USA's TLD? .us is your country's TLD

1985...when it was administered by U.S. DoD

You don't seem to know what you're talking about, I mean that in the nicest way.

ICANN doesn't solve domain disputes at all. That is handled through UDRP providers (WIPO, NAF, etc).

ICANN doesn't manage any of the registries themselves, they give the contracts in some meaningless manner (different problem, shitty organization ICANN is).

You eat some of the biggest taco bell meals I have ever heard of. As others stated, you have a very rich mindset. Even 7$ is expensive for some people, I've met startups in third world countries that have trouble paying $20/month for hosting (that can be 20% of a monthly salary or more!)

UDRP = uniform dispute resolution policy...ICANN was the one that created it.

ICANN manages the fees so they can make them whatever they want...no?

If $7 is expensive for someone...they can just get a non .com name. Your blog that gets 5 visitors a day, will work just as fine on a .net. Same for businesses. If you can't afford $240/yr then you are probably not serious enough about your venture. So your third world startups can just stick on their country's domain name...or get a .net or do the whole start.ups type name.

Yes I overdid it with the TB...but that's because I usually get it for 2 people.

This is getting way off topic, but...

> Yes I overdid it with the TB...but that's because I usually get it for 2 people.

Where do you live and who do you eat with? I can feed like 8 people for $20 at Taco Bell here in Columbus, OH

as far as food: 4 supreme beef chalupas, 2 chicken quesadillas, 2 drinks = ~20

That's a silly argument. .com isn't yours or mine. It's everyones. Why don't you go register .US? Bringing ccTLDs only clouds this even more (and ICANN doesn't have control of them).

ICANN doesn't do anything themselves, it's all done by registrties, udrp providers, etc. ICANN is also supposed to be a non-profit, raising the fees for an arbitrary purpose serves who?

Reply on news.ycombinator.com