IT.COM

analysis What Does Domain Name Sales Data Really Say?

Spaceship Spaceship
Afternic recently provided data insights from the 2022 sales year. While precise numerical values are not given, graphical results are presented for total dollar volume, average price, sell-through rate and average revenue per domain. This allows one to compare the performance of different extensions.

Total dollar volume was released for the the 20 TLDs with the highest dollar volumes during the year, while the other measures were provided for the top 10 from that list.

Afternic data is not normally included in NameBio. When I do an analysis based on NameBio data, I often wonder how it would have changed if Afternic data had been included. In this article I look at corresponding measures using NameBio to the data recently released by Afternic.

Most Domain Sales Are COM

Afternic showed sales volume, for 2022, in the following graph where the size of the square corresponds to volume. The huge blue square is .com, while the other top TLDs are the rectangles shown on the right. Most domain sales are in .com.

IMG-Afternic-SquaresVolume.jpeg


The labels are difficult to read on the original, but after .com, the next three, not that different in volume from each other, are .org, .net and .co in the Afternic data.

I took the top 10 extensions from the Afternic data, and looked up the 2022 dollar volumes using NameBio, with results shown below. Note that .com would be far off the graph scale, with $143.8 million in total sales volume, compared to about $8.0 million for second place .org.
IMG-NB-Volume.jpg

While the two datasets agree that .com dominates and .org comes next, after that there are differences. In the NameBio data .xyz comes next, followed by .net, .io and then .co. Also, .us and .ca are much smaller in NameBio compared to the Afternic data. Some other extensions, like .de, do far better in the NameBio data.

The Afternic Top 20

Based on 2022 Afternic sales data, here are the top 20 extensions in order.
  1. .com
  2. .org
  3. .net
  4. .co
  5. .xyz
  6. .io
  7. .us
  8. .co.uk
  9. .ca
  10. .info
  11. .me
  12. .cc
  13. .de
  14. .tv
  15. .shop
  16. .app
  17. .pro
  18. .eu
  19. .fr
  20. .club
Should One Just Invest In .COM?

In both datasets .com dominates. In the NameBio data for 2022, 81.4% of the entire volume is due to .com.

While one can certainly invest only in .com, that does not mean there are not opportunities in other extensions.

An important point is that the vast majority of names listed for sale are also in .com. In the next section we look at the ratio of sales to listings.

Sell-Through Rate

Afternic shared sell-through rate (STR) data for their top 10 extensions. While we don’t have numerical values, the lengths of the columns provide a relative measure.

IMG-Afternic-STR.jpeg

In the Afternic data, .io has the best STR, followed by .org, .co and then .com.

I looked at the 2022 STR using NameBio sales data and Dofo data for the number of sales listings for each TLD. While Dofo no longer exists, I had computed and saved the data from an earlier analysis. The results are shown below.
IMG-NB-STR.jpg

The NameBio data is consistent with the Afternic data in that .io has the best STR, with .org next. There is also agreement that .com is better than .net. In contrast to the Afternic data, .co is lower in STR in the NameBio-Dofo data. I was not able to get the number of .co.uk listings from Dofo, so that value is not given.

Average Price Matters, But Challenging To Interpret

While the sell-through rate is important, average price is also a key metric. Afternic presented average prices for the top 10 extensions.

IMG-Afternic-AvgPrice.jpeg

in the Afternic data, .io has the highest average prices, followed by .xyz and then .com. Average prices for .co, .org and .ca are very similar to each other.

What about the NameBio data? Average prices vary widely according to the selection criteria. If I simply use all sales $100 and up, one obtains the data on the top left.
IMG-AveragePrices.jpg

But that dataset is dominated by wholesale acquisitions, especially auction sales. If I exclude all auction sites, and a few other venues known mainly for wholesale acquisitions, the average prices change to the values shown in the top right graph. Note that the scales are different on the three graphs.

To stress how dominated raw NameBio data for legacy extensions is with auction and other wholesale transactions, the top right graph is based on 7091 .com sales compared to about 124,600 in the top left overall distribution. The top right graph is based on private sales in NameBio, as well as Afternic (not many), BuyDomains, DomainMarket, Sedo and Uniregistry.

In the retail-only NameBio data, .xyz has the top average price, with .io and .com next, almost equal to each other. This is somewhat different from the Afternic data, that had .io first, followed by .xyz then .com.

Another way to concentrate on retail sales is to set a price minimum. If I look only at NameBio sales $1000 and up, I obtain the bottom graph. This still shows .xyz with the highest average price, with many other extension not that different from each other. A higher efficiency of reporting of high-value .xyz sales may be part of what we are seeing.

The main message from this data, though, is that average price values are very sensitive to the details of the filters used. The next section looks at a more robust measure, the average revenue per domain.

Revenue Per Domain

In January in the NamePros Blog, I introduced a simple measure that I called Dollar Volume Per Listing (DVPL). This is simply the average dollar return per domain in a portfolio, or a subset of a portfolio, such as a particular extension.

The Afternic Revenue Per Domain (RPD) is similarly calculated, with 2022 results for the top ten extensions shown below.

IMG-Afternic-DVPL.jpeg


The .io extension clearly dominates the Afternic data in terms of the Revenue Per Domain metric, with a value just over 3x that of .com or .org. According to Afternic Revenue Per Domain, .xyz trails .com, .org and .co, and is slightly lower than .net.

Let’s compare this to the values I obtained for NameBio sales data and Dofo listings data. Note that these were calculated using an average of 2021 and 2022 sales data, so are not exactly comparable to the Afternic data.
IMG-RevenuePerDomain.jpg

Again .io dominates, with a significantly higher DVPL than the other nine extensions. The NameBio+Dofo data indicates that .org is significantly higher than .com in Revenue Per Domain, and .xyz slightly leads .net and .co.

It should be kept in mind that the annual holding cost for .io is just over 3x times that for .com, while .co is about 2x.

Michael Cyger’s Insights

Recently industry expert @Michael Cyger interpreted the Afternic data. After noting the obvious domination of .com in terms of total sales volume, he provided this perspective on .io:
Domains in the .io TLD lead the way on RPD, STR and ASP, but domains in the .io TLD have low volume. This suggests that there were some very large .io outlier sales in 2022.
Note that RPD is revenue per domain, STR is sell-through rate, and ASP is average sales price.

Next, he looked at .xyz.
Although .xyz has a high ASP, its RPD is low. So that implies that you shouldn't buy 10,000 random .xyz domains because your RPD will be low. But if you have a lower quantity with higher quality, both your RPD and ASP should be high.
He then commented that Swetha (@DNGear) has both high quality .xyz names and many of them, and that explains her strong sales record during this period.

He returned to the quality issue in his concluding comments.
Quality is key in non- .com TLDs. Quality sells. And if you have quality in large quantity, you'll sell more.

As well as the obvious need for memorable words that make good brands, he commented on the importance of short words, both in terms of length and number of syllables.
If you're buying alternative extensions, try to focus on single words that are one syllable like Wrap, Bolt, Move and Keys. Then expand to single words that are two syllables like Profile, Pixels, Pluto and Chroma. Then expand to single words that are three syllables like Capital, Conduit, Momentum and Artemis.Two words can be good too, just not as high of an ASP. And try to find words that companies would want for their brand name.

Here is the link to his full commentary on social media. The commentary was repeated in his free weekly newsletter - you can subscribe and see the latest newsletter here.

My Thoughts

Here are some things that struck me while researching this topic.
  • With tens of millions of domain name listings, Afternic has a lot of sales data, and even with the restriction of relative instead of numerical data, there is much to be learned from the data shared.
  • It was encouraging that many of the trends were similar in this comparison of NameBio data with Afternic. This helps validate that analyses based on NameBio data probably adequately represent the overall market most of the time.
  • But there were differences too. It seems that Afternic do better with .us and .ca, while Sedo, responsible for much of the NameBio retail data, sells various European country codes more effectively.
  • The differences in .co surprised me a bit, with the extension doing significantly better in the Afternic data.
  • The analysis with different selection factors convinced me that average price, by itself, will vary widely with outlier sales and different retail dividing lines. We should be wary of reading too much into average price data.
  • The total dollar volume, though, is much less sensitive to the precise cutoff for retail sales.
  • If one looks at the top 20 Afternic extensions, it is interesting that .shop is one new gTLD that made the list along with .xyz, .club, and .app. I would not have predicted .shop on top 20 list, and it does not seem to carry over to the NameBio data. So if selling .shop domains, it seems that Afternic listing may be important.
  • Both datasets substantiate the inroads that .xyz has made, but it is important to keep in mind that the overall .xyz STR is not that high, an indicator that many investors have names that do not sell in that extension.
  • The .org extension continues to shine, with solid metrics in both datasets.
  • As @Michael Cyger commented, especially in extensions other than .com, quality, usually represented by short single words that would make a great brand, is key.
  • Both sets of data are missing many sales. Most brandable marketplace sales are missing, and probably the majority of private sales.
Please share your thoughts in the comments!

The Afternic data usually gets reported here on NamePros. Follow @James Iles. Also, Afternic tweet new data releases – their official account is Afternic if you are on that platform.



I would first like to thank Afternic for providing the data. Also, thanks to @James Iles who answered several questions I had regarding interpretation. My sincere thanks to NameBio for the sales data used in this article, and to Dofo for listings data accessed earlier. Thanks to @Michael Cyger for the insights he provided on interpreting the data, and for permission to quote for this article.
 
Last edited:
89
•••
The views expressed on this page by users and staff are their own, not those of NamePros.
Thanks Bob - an interesting read :)
 
7
•••
Thank you Bob . Data is very important.. Imagine when Afternic publish and report domains sales regularly , weekly or monthly, especially for all sellers when its not possible for public.
 
Last edited:
5
•••
What Does Domain Name Sales Data Really Say?

Hi

to me, it's more about what it doesn't say.
as you can "read or derive" any conclusions into it or from it, that you want.

but the data doesn't say anything about the actual transaction and/or negotiation activity occurred or even mention the parties involved.

it doesn't mention specific domain names, nor does it show how long each name was held, before it sold.

it doesn't tell whether it was purchased by a domainer or by an end-user who plans to develop the name.

there is some info missing or is unknown, so the data really isn't saying that much.

it's like somebody saying they sold blank.blank for $3k, but they don't supply any other details.

still, what i did hear it say, was that "no domain extensions longer than 5 letters was in the top 20." :)

imo....
 
Last edited:
13
•••
4
•••
Thanks for the insightful post. Good to see the .ca in the top 10.
 
5
•••
4
•••
Thank you Mr.Bob for all your theses and topics of high resonance and level. :xf.smile:
 
Last edited:
3
•••
Great job as usual Bob.
This analysis in my opinion undermines the certainty of the .com kingdom.
If there were no premium pricing policy I am sure the data would have been even better in favor of the new extensions.
We will see how the scenario changes in a few years time
 
3
•••
What Does Domain Name Sales Data Really Say?

Hi

to me, it's more about what it doesn't say.
as you can "read or derive" any conclusions into it or from it, that you want.

but the data doesn't say anything about the actual transaction and/or negotiation activity occurred or even mention the parties involved.

it doesn't mention specific domain names, nor does it show how long each name was held, before it sold.

it doesn't tell whether it was purchased by a domainer or by an end-user who plans to develop the name.

there is some info missing or is unknown, so the data really isn't saying that much.

it's like somebody saying they sold blank.blank for $3k, but they don't supply any other details.

still, what i did hear it say, was that "no domain extensions longer than 5 letters was in the top 20." :)

imo....
Leanne's Crunch.id shows a breakdown of what seems to be wholesale vs end user sales (the former, linked to Gd and the latter, with Buy Domains).
 
2
•••
to most it says their folios are crap and need work
 
2
•••
what seems to be wholesale vs end user sales
Hi

unless you know who the buyers were and what they did/do with the domain, then you still don't know.

because endusers can buy at wholesale and domainers will buy at what appears to be enduser prices.

imo...
 
4
•••
Leanne's Crunch.id shows a breakdown of what seems to be wholesale vs end user sales (the former, linked to Gd and the latter, with Buy Domains).

Yes, @LeanneMac in her excellent work "assumes all sales from Park.io, Sav.com, Register.TO, GoDaddy, NameJet, DropCatch, Dynadot, SnapNames, Whois.ai, Catched.com, Docky.ly and NamesCon are wholesale and all others are retail." That retail/wholesale division is almost exactly the same as for my top right graph in this article looking at average prices. Obviously any division system, based on sales venue or price, will not be perfect. There are some retail buyers who compete at $$$$ and far above in GoDaddy and other auctions, and sometimes domainers buy names at Sedo. But for the most part excluding these sites will mean that most of the sales are retail.

All of the Afternic results reported here does not include any GoDaddy auction sales.

Thank you for pointing out the work of Leanne.

-Bob
 
Last edited:
7
•••
Hi

unless you know who the buyers were and what they did/do with the domain, then you still don't know.

because endusers can buy at wholesale and domainers will buy at what appears to be enduser prices.

imo...
Realistically, the vast majority of domainers don't have deep pockets, and so it can be safely assumed, will buy at wholesale prices.
 
3
•••
Thank you Bob, your work very important for our domain industry but few people still thinks "Domaining should be restricted to over the 50's"
I'll say to those just watch this thread this made by the over 50's
 
3
•••
I wonder if the higher holding cost of .io's has led to pickier investors, and thus higher quality names for sale and higher STR/sale prices.

.com also has a lot of "bread and butter", cheap sales - like geo domains, whereas I would assume .io/.xyz are being purchased almost exclusively by tech companies.

I also wonder if the investor/wholesale market for .io is as robust as it is for .com. If not, that could cut out a huge swath of the wholesale pricing tier and make the average sale price for .io weighted more towards end user pricing.

Great post, thanks @Bob Hawkes.
 
Last edited:
5
•••
Excellent read, thanks for sharing this, it's a refresher.
 
Last edited:
3
•••
Daily sales volume of the #1 marketplace is under $500,000

It explains me why selling a domain is difficult.

Thanks for the post Bob.
 
Last edited:
8
•••
I liked your Analysis From the data Bob. Thanks !
 
Last edited:
3
•••
What Does Domain Name Sales Data Really Say?

Hi

to me, it's more about what it doesn't say.
as you can "read or derive" any conclusions into it or from it, that you want.

but the data doesn't say anything about the actual transaction and/or negotiation activity occurred or even mention the parties involved.

it doesn't mention specific domain names, nor does it show how long each name was held, before it sold.

it doesn't tell whether it was purchased by a domainer or by an end-user who plans to develop the name.

there is some info missing or is unknown, so the data really isn't saying that much.

it's like somebody saying they sold blank.blank for $3k, but they don't supply any other details.

still, what i did hear it say, was that "no domain extensions longer than 5 letters was in the top 20." :)

imo....

Why must you always be the eternal skeptic and always minimizing someone else's work, for benefit of others on NP. @Bob Hawkes did great work here. Data is data, it's never 100% precise - but nothing else is.

Otherwise there are unknowns even in whether any of us will live till tomorrow.

I don't expect, or need a response.
 
Last edited:
2
•••
Why must you always be the eternal skeptic and always minimizing someone else's work, for benefit of others on NP. @Bob Hawkes did great work here. Data is data, it's never 100% precise - but nothing else is.

Otherwise there are unknowns even in whether any of us will live till tomorrow.

I don't expect, or need a respons

I think @biggie 's guided-note is an adrenaline rush for @Bob Hawkes. Without researchers and observers we can barely see beyond dot com.

Thx Bob

Regards
 
6
•••
Why must you always be the eternal skeptic and always minimizing someone else's work, for benefit of others on NP. @Bob Hawkes did great work here. Data is data, it's never 100% precise - but nothing else is.

Otherwise there are unknowns even in whether any of us will live till tomorrow.

I don't expect, or need a response.
Hi

the work that was done, is what it is.
though it may be entertaining or educational to you-
to me, it is incomplete, as far the info i would want to know.

the numbers don't mean much, if/when you can't attach them to names and persons.

being analytical by nature, i will question anyone or anything that i think, is questionable.
i don't follow the crowd or always nod in agreement, because i am an individual and can think for myself.

sure, i raise points, that are of interest to me and have even given Thanks and Like point to some of his posts, but i have never given @BobHawkes a "disagree" point.

but by reading your reply, at least now, i know why you "disagreed" with my post.

imo....
 
1
•••
I wonder if the higher holding cost of .io's has led to pickier investors, and thus higher quality names for sale and higher STR/sale prices.
I think that is exactly it. Because it costs so much to hold each year, people only register .io names that they think are very high quality. That means the STR is substantially higher. I know I do that personally.

The STR or the average net price, or both, need to be better to make domains with higher holding costs worthwhile investments. Even if they do not think of it in numerical measures, I think most domainers instinctively factor these into their investment decisions.

I also agree with your point that the wholesale market for anything is not as robust as the major legacy extensions, although definitely more .io wholesale than used to be the case, and Park.io have developed a strong market, with the presence of both retail and wholesale bidders.

Thanks for your observations.

-Bob
 
Last edited:
2
•••
Daily sales volume of the #1 marketplace is under $500,000

It explains me why selling a domain is difficult.

Thanks for the post Bob.
That's what happens when the world economy is in a slump.
 
1
•••
Back