IT.COM

information Domain Valuation: The Length and Frequency

Spaceship Spaceship
A cursory glance at the domain market will suggest that shorter domains sell for more money than longer names. So, I examined the connection between domain name length - meaning the number of characters in the name - and domain name value to see if there is in fact a relationship between the two, and if so, why? Here are the results.


Does the Number of Characters in a Domain Name Impact Its Value?

I’ve done a fair amount of research as to why one domain will sell for $1,000 and another $10,000, and I have found that all other things being equal, the length of a domain name does influence the cost and the ROI. My “sweet spot” when I shop for domain names is between five and ten characters in length. Shorter names are too expensive for me, and after ten characters, the ROI falls too much in my experience.

Here’s a chart of “all time” domain name sale price averages for domain names between three to fifteen characters at DNPric.es and NameBio. I took the DNPric.es data from their stats page, and the NameBio stats from their database using the dashboard on NameBio.com.

See table below.

length-vs-price.png


Length vs. Avg. Sale Price
correlation.png

As you can see, the average domain name value falls as the number of characters increases. I did not use the data for one and two letter domain names as they are outliers; although, the correlation between length and price holds true even when accounting for one and two letter domains. Many people already know that shorter domain names are more valuable, but much fewer understand why. So, let’s find out.


Why Does Domain Name Length Influence Price?

There are a host of reasons why domain X sells for $1,000 and domain Y for $5,000, and not all of them are tied directly to the domain. Factors like bidding wars impact domain name sales for reasons that may or may not reflect the inherent value of a domain. If domainer A thinks a domain is worth $50,000 and is willing to pay for it, that sets the price in the market, and domainer B will have to pay more than domainer A to acquire it. In bidding wars, the market ultimately decides the price of the domain. So, not all domain sales data can be reduced to a formula.

However, Thies Lindenthal, the creator of the domain stock index (IDNX) posits an interesting theory on domain values. In his paper,
Monocentric Cyberspace, he compared domain names to real estate and emphasized that the “cost of commuting,” meaning the cost that a customer incurs travelling from his/her home to a store, is relatively similar to the effort that an internet user puts into remembering and entering a URL into a browser. According to this logic:

An easier to remember domain name would be more valuable than one that people struggle to recall, because it is easier for them to “commute” to it.

We can apply this to domain prices, and see what happens. Zipf’s law shows that the most frequently used word is twice as common as the second place word, three times as common as the third most used word, etc. It seems to me that the more frequently a word is used in real life, the easier it would be for people to remember. Thus, according to Lindenthal's logic, a more common word would be an easier "commute" for internet users, which should make it more valuable as a domain name.

Of course, there are some exceptions, such as two or three character domains that are not words sometimes having more value than slightly longer domains containing common words, but again, those domains are outliers. In those outlier cases, their value is often based on scarcity, and they're still memorable with only a single character or two to remember, aside from the TLD.

Google’s Ngram Viewer allows you to find out the usage frequency of any word in the totality of every book printed from 1800 to 2008. So, we can use the Ngram Viewer to compare words in domain names based on their frequency in our culture’s texts, and compare their frequency to their prices to see if there is a correlation between how often a word is used and its value on the domain market.

I used pat.com and paty.com as an example.

ngram.png

As you can see, the word “pat” appears exponentially more than “paty” in our culture. This data would suggest that “pat” is a much more expensive domain than “paty.” In fact, using Namebio sales data, we have proof that it was considerably more expensive. Pat.com sold for $63,000 while paty.com sold for $3,500. The domain name sales data and word usage rates suggest to me that the frequency of a word being used impacts its value as a domain name, because it shortens the mental “commute” that people need to take to remember it. Because shorter words are more frequently used and often more memorable, shorter domains are more likely to be valuable than longer ones.

I hope this article has helped you understand why domain name length matters.
 

Attachments

  • Lindenthal-2016-The_Journal_of_Real_Estate_Finance_and_Economics-VoR.pdf
    550.3 KB · Views: 193
58
•••
The views expressed on this page by users and staff are their own, not those of NamePros.
Excellent research and writing. Size matters and it's never been more true than in today's mobile device economy. Nice work Jay!
 
7
•••
3
•••
Enjoyed your writing style, well written and your research was easy to understand.
 
2
•••
Thanks guys.

More thanks to the namepros editor.
 
3
•••
frequency of any word in the totality of every book printed from 1800 to 2008

I bet the total of books printed in 2008 was much higher then in 1800
so that spoils the stats a little
 
1
•••
Very nice article indeed. That Ngram Viewer was a new tool to me, love it already, much thanks.
 
4
•••
I bet the total of books printed in 2008 was much higher then in 1800
so that spoils the stats a little
Very few new words have been uncovered in English for decades
 
1
•••
I've said the following a few times but it seems to get brushed aside.

Another huge benefit of having a short domain name is it will give you greater reach on signs and billboards.
 
3
•••
I've said the following a few times but it seems to get brushed aside.

Another huge benefit of having a short domain name is it will give you greater reach on signs and billboards.


But, doesn't a short domain have a greater reach on signs and billboards because it is memorable; because of the shorter 'mental commute'? So then you are saying the same thing Jay's saying, no?
 
1
•••
But, doesn't a short domain have a greater reach on signs and billboards because it is memorable; because of the shorter 'mental commute'? So then you are saying the same thing Jay's saying, no?
I'm only using shorter as a factor. Meaning bigger letters farther reach utilizing the same space.
 
2
•••
Hi Jay
Quite good study with logic conclusions. It will educate the domainer for their benefits.
Regards
Brandsks
 
2
•••
Excellent piece of writing!
 
1
•••
1
•••
I've said the following a few times but it seems to get brushed aside.

Another huge benefit of having a short domain name is it will give you greater reach on signs and billboards.


it looks much better on ( smaller cars [smart] )
 
1
•••
1
•••
Good write up, but i still wonder why shorter domains names with no meaning at all still trump longer keyword rich domains
 
2
•••
2
•••
Good write up, but i still wonder why shorter domains names with no meaning at all still trump longer keyword rich domains
He never made that direct comparison. Thats like comparing Apples and Oranges.
 
2
•••
it looks much better on ( smaller cars [smart] )
Not sure what you mean, was just trying to explain that more people will be able to read the url from farther away so you get more exposure for the same size billboard.
 
1
•••
Honestly, you would sell paty.com for $3,500 ?
 
2
•••
Thanks for a well written article.
However I don't understand why you chose "pat" "paty" as a typical example of differential value based on domain length. The ngram data is comparing a word that's a noun, a verb and a personal name (Pat) with one that's none of the above (paty) - of course the former is going to show dramatically higher usage. Apples and oranges.
Also the "collectibility" factor and "rarity" of LLL names comes into play. Lots of factors at work there, would have been interesting to compare several 5 and 6 character names with otherwise similar characteristics AND similar ngram results. That would minimize other factors and give a better indication of price bias relevant to character length.
 
4
•••
Thanks for a well written article.
However I don't understand why you chose "pat" "paty" as a typical example of differential value based on domain length. The ngram data is comparing a word that's a noun, a verb and a personal name (Pat) with one that's none of the above (paty) - of course the former is going to show dramatically higher usage. Apples and oranges.
Also the "collectibility" factor and "rarity" of LLL names comes into play. Lots of factors at work there, would have been interesting to compare several 5 and 6 character names with otherwise similar characteristics AND similar ngram results. That would minimize other factors and give a better indication of price bias relevant to character length.

I am not entirely sure what you mean, but ..

  1. In regards to the first paragraph, pat and paty where used for the sake of "apples to apples" comparison. "of course the former is going to show dramatically higher usage" - has been cited as a strong reason why domain x would cost more than domain y.

The most importance factor here is frequency and rank. Analyzing the results or the application in "contextual frames" would be a futile effort, because language is highly contextual. And the user (person communicating) and the context (situation) determine what certain words or sentence mean. When someone says "alcohol", how do you deduce what is been suggested. Did they mean alcohol/rubbing alcohol or alcohol/vodka, this can only be accurately inferred in context or with more information in a written sentence. And, you would certainly have a higher failure rate should you attempt to this with one word or two word domains.

2. The second paragraph, is another research projects which would require a different approach. What feature would you be observing ?
 
1
•••
1
•••
1
•••
Back