- Impact
- 1,097
Happy Easter everyone
As I am doing some kind of recap of my different ventures right now, I thought I'd share the story of my journey through brandables domain names. I think some of the data and resources can help those starting in this field!
I grew interest in brandables about the end of 2015 ago because I thought it involved a bit of creativity in finding the names AND you could still handreg good brandables. Pretty quick I opened accounts with Brandroot and Namerific and tried to get a feeling of what was a "good" brandable, especially one that could get listed (and hopefully sold!) on these platforms.
1) Building lists from marketplaces and past sales
First I built lists of domains found on marketplaces. My goal was to understand the different types of brandables and their price tags.
In order to do this, I wrote some simple php code to browse and parse the marketplaces' websites.
This approach was a bit too comprehensive. I realized a better goal was to actually find the kind of names that SELL!! So I used past sales tools like Namebio and DNpric.es and scoured brandable sales lists like the ones compiled and published by users here (e.g. @Doron Vermaat : 1134 domains sold at Brandbucket)
2) Analysing the lists
Once I had these huge lists of names (about 70k listed names and 3k sold names), I tried to categorize the names.
2.1) Sold at Brandbucket
To give you an example of the kind of analysis I did, here is the results of the "1134 domains sold at Brandbucket" list:
a) 84% of the names are 13 Letters or shorter
b) Among the "very short names", you can have:
Ad
Pay
Car
Sys
Ink
Ego
Web
Joy
Tap
Bank
Hire
Free
...
d) Words can be madeup by replacing letters (i => y...)
2.2) Past CVCVC.com sales from Namebio VS. Brandbucket listing
This is a second example of the kind of analysis I performed.
Analysing 2278 CVCVCV on Brandbucket, I determined the most frequent first letter was V (13% of the total), followed by Z (8%) and M (7%). And that names started by V started mostly with VO (35%).
Analysing 1265 CVCVCV sales from Namebio I gathered the following type of informations:
- more half of the names (54%) started with one of 7 letters (M: 11%, S:10%, C, T, L...)
- the most frequent syllables were MA, RE, SA, CA etc...
With only the first Brandbucket analysis in mind I would have started immediately looking for available voCVCVs, but the analysis of actual sales showed me that these accounted only for 0,6% of total CVCVCV sales, VS 4,8% for maCVCVs. 8 times more!!
3) Choosing a strategy: what kind of names to buy and where to list?
With data analysed, I had now to choose a strategy to buy names.
I decided I would apply the following rules:
4) Generating the names
4.1) Patterns
I wrote simple pieces of PHP code to generate 5L and 6L patterns and checked availability on Namebright bulk checker. Unfortunately this led to HUGE lists of available domain names. I tried to go through these but after a few minutes I felt every name was okayish and always ended up with hundreds of possible registrations... I decided to narrow the possibilities, for example, with 6L.com I'd look for repeating patterns (makoma.com) or patterns with same vowel like vabaza.com. With 5L.com I'd look for CVCVCs or VCVCVs or paterns with double letters at the begining or end : ooLLL, LLLoo...
I also set up email alerts for dropping domains through the excellent website ExpiredDomains.net. To catch these I used Desktop Catcher, a great piece of software I found through Namepros.
4.2) Combos
What I call combos are names built by sticking two words together : Bitcoin is a combo of Bit + Coin...
Remember in 2.1. I built lists of frequent words ? I wrote a small php page where I could enter a word (list of words) and add to it any prefix / suffix from a predefined list.
For example, here is a short part of my current prefix list :
my
go
bio
geo
car
max
zen
bit
I'd then enter the word I want to generate names for (e.g. "Trade") and get a list like the following in the bulk availability tool.
MyTrade.com
GoTrade.com
BioTrade.com
GeoTrade.com
CarTrade.com
MaxTrade.com
ZenTrade.com
BitTrade.com
I would choose my main word ("Trade" in the exemple above) either from my list of words appearing frequently in combos or from what is trending now.
After a few months I trimmed the list so it generates about 1200 names for the main word. At the start I had more about 5k names. I realized over time that it's better to have a shorter but more efficient list of possible candidates to register.
Today, on a strong main word like "Pay", I have only 3 possible handregs (and they are not good names I'm afraid). I consider that indicates my list of suffix / prefix is good enough that it generated about 1200 names worth registering, since they are already registered!! With less strong words, of course, I'm left with more names I could register...
As a side note, I've also used some website name generators, but I've always prefered my homemade solution because I know how I built it and why I chose these suffix / prefix! You'll find these sites in the references below.
5) Buying and listing names
With all these rules and tools in hand, I started buying and listing names. I used coupons whenever I could and focused on one brandable marketplace, Brandroot, with the following reasoning. They disclosed somewhere (here on NP I guess) that they sold about 6% of their inventory of about 12,000-15,000 names. I thought I could list about 100-150 domains and thus about 1% of their inventory. Ithought I would have some visibility on the platform and thus have a chance to sell between 3% and 6% of my inventory there. $10/listing + $4 to $8 per handreg depending on the coupon. The experiment would cost me roughly $2k. If I sold 3 to 6 names and get about $1-1,5k after commission, I would be OK.
Back then, Brandroot had a long approval process, and a lot of my submited names were refused. That was OK as I listed them on Afternic as I wished to compare both platforms.
6) Results
These are the results, some of the names have been listed for 2 years now (started dec 2015 but seriously in feb 2016).
Brandroot : 144 names listed / 5 names sold
Afternic / Public Whois : 200 names listed / 2 sold
7) Conclusion
The experiment is clearly in favor of the brandable marketplace over regular marketplaces. One could argue that the names I listed elsewhere were the leftover from the brandable marketplace, but I had many CVCVCV names that seem really equivalent to me in terms of quality on both Afternic and Brandroot and I sold more of them on the latter.
I must admit I've lost interest in trying to list names on Brandroot though. They have now over 20k names and I'd need to add about 70 to reach back my 1% exposure. I hoped to sell between 3 and 6% of my names there, but the end result is 3,4% in two years => less than 2% per year. In short, the waiting game gets a bit boring... I don't know. Maybe I should liquidate these and buy myself some bitcoins
I tried to launch my own marketplace to see if I could have more success than with Afternic. Called it BrandCombo.com. So far I only got a few price inquiries after 1 year of it being online. I'm pretty sure I don't have the critical mass for it : I listed xx / low xxx names. My guess is that below 1,000 names it's hard to get a marketplace effect where you can generate traffic and get visitors interested in more than one of your names.
That was quite a long one! Thanks for reading this far. Hope that gives food for thought to some of you
8) References
Past sales brandables
DNGeek: Brandable domain name sales of past week
Past sales websites
Namebio: link to past CaCaCa.com sales
DNpric.es : link to 6L.com sales above $1000 in 2018
Bulk check for available names
Namebright
Expired names email alerts
ExpiredDomains.net
Dropcatcher software
DesktopCatcher
Brandable platform I listed on
Brandroot
BrandCombo (my own)
Generate Names
Lean Domain search: generate names for the word Robot
Wordoid
Portmanteau generator: Travel + Velocity = Travelocity
Invent-a-word
WerdMerge
As I am doing some kind of recap of my different ventures right now, I thought I'd share the story of my journey through brandables domain names. I think some of the data and resources can help those starting in this field!
I grew interest in brandables about the end of 2015 ago because I thought it involved a bit of creativity in finding the names AND you could still handreg good brandables. Pretty quick I opened accounts with Brandroot and Namerific and tried to get a feeling of what was a "good" brandable, especially one that could get listed (and hopefully sold!) on these platforms.
1) Building lists from marketplaces and past sales
First I built lists of domains found on marketplaces. My goal was to understand the different types of brandables and their price tags.
In order to do this, I wrote some simple php code to browse and parse the marketplaces' websites.
This approach was a bit too comprehensive. I realized a better goal was to actually find the kind of names that SELL!! So I used past sales tools like Namebio and DNpric.es and scoured brandable sales lists like the ones compiled and published by users here (e.g. @Doron Vermaat : 1134 domains sold at Brandbucket)
2) Analysing the lists
Once I had these huge lists of names (about 70k listed names and 3k sold names), I tried to categorize the names.
2.1) Sold at Brandbucket
To give you an example of the kind of analysis I did, here is the results of the "1134 domains sold at Brandbucket" list:
a) 84% of the names are 13 Letters or shorter
b) Among the "very short names", you can have:
- 4L: I decided to not even consider these as they were way out of my buying power and a niche of their own
- 5L: 13% of total sales. 95% madeup / 5% short words.
- 6L: 11% of total sales. 55% madeup / 45% combo or sentences
- 7L: 45% madeup / 55% combo or sentences
- etc...
Ad
Pay
Car
Sys
Ink
Ego
Web
Joy
Tap
Bank
Hire
Free
...
d) Words can be madeup by replacing letters (i => y...)
2.2) Past CVCVC.com sales from Namebio VS. Brandbucket listing
This is a second example of the kind of analysis I performed.
Analysing 2278 CVCVCV on Brandbucket, I determined the most frequent first letter was V (13% of the total), followed by Z (8%) and M (7%). And that names started by V started mostly with VO (35%).
Analysing 1265 CVCVCV sales from Namebio I gathered the following type of informations:
- more half of the names (54%) started with one of 7 letters (M: 11%, S:10%, C, T, L...)
- the most frequent syllables were MA, RE, SA, CA etc...
With only the first Brandbucket analysis in mind I would have started immediately looking for available voCVCVs, but the analysis of actual sales showed me that these accounted only for 0,6% of total CVCVCV sales, VS 4,8% for maCVCVs. 8 times more!!
3) Choosing a strategy: what kind of names to buy and where to list?
With data analysed, I had now to choose a strategy to buy names.
I decided I would apply the following rules:
- Stick with short names <= 10 Letters
- Put emphasis on 5L CVCVC / VCVCV and 6L CVCVCVs because they provide more "easy to spell" names
- Generate combo names with one "strong" word that appears often in the sold / listed lists I had built
4) Generating the names
4.1) Patterns
I wrote simple pieces of PHP code to generate 5L and 6L patterns and checked availability on Namebright bulk checker. Unfortunately this led to HUGE lists of available domain names. I tried to go through these but after a few minutes I felt every name was okayish and always ended up with hundreds of possible registrations... I decided to narrow the possibilities, for example, with 6L.com I'd look for repeating patterns (makoma.com) or patterns with same vowel like vabaza.com. With 5L.com I'd look for CVCVCs or VCVCVs or paterns with double letters at the begining or end : ooLLL, LLLoo...
I also set up email alerts for dropping domains through the excellent website ExpiredDomains.net. To catch these I used Desktop Catcher, a great piece of software I found through Namepros.
4.2) Combos
What I call combos are names built by sticking two words together : Bitcoin is a combo of Bit + Coin...
Remember in 2.1. I built lists of frequent words ? I wrote a small php page where I could enter a word (list of words) and add to it any prefix / suffix from a predefined list.
For example, here is a short part of my current prefix list :
my
go
bio
geo
car
max
zen
bit
I'd then enter the word I want to generate names for (e.g. "Trade") and get a list like the following in the bulk availability tool.
MyTrade.com
GoTrade.com
BioTrade.com
GeoTrade.com
CarTrade.com
MaxTrade.com
ZenTrade.com
BitTrade.com
I would choose my main word ("Trade" in the exemple above) either from my list of words appearing frequently in combos or from what is trending now.
After a few months I trimmed the list so it generates about 1200 names for the main word. At the start I had more about 5k names. I realized over time that it's better to have a shorter but more efficient list of possible candidates to register.
Today, on a strong main word like "Pay", I have only 3 possible handregs (and they are not good names I'm afraid). I consider that indicates my list of suffix / prefix is good enough that it generated about 1200 names worth registering, since they are already registered!! With less strong words, of course, I'm left with more names I could register...
As a side note, I've also used some website name generators, but I've always prefered my homemade solution because I know how I built it and why I chose these suffix / prefix! You'll find these sites in the references below.
5) Buying and listing names
With all these rules and tools in hand, I started buying and listing names. I used coupons whenever I could and focused on one brandable marketplace, Brandroot, with the following reasoning. They disclosed somewhere (here on NP I guess) that they sold about 6% of their inventory of about 12,000-15,000 names. I thought I could list about 100-150 domains and thus about 1% of their inventory. Ithought I would have some visibility on the platform and thus have a chance to sell between 3% and 6% of my inventory there. $10/listing + $4 to $8 per handreg depending on the coupon. The experiment would cost me roughly $2k. If I sold 3 to 6 names and get about $1-1,5k after commission, I would be OK.
Back then, Brandroot had a long approval process, and a lot of my submited names were refused. That was OK as I listed them on Afternic as I wished to compare both platforms.
6) Results
These are the results, some of the names have been listed for 2 years now (started dec 2015 but seriously in feb 2016).
Brandroot : 144 names listed / 5 names sold
- Kitono: $2,990, got $1,993 after commission. Held 7 months. Handreged because easy to spell CVCVCV with repeating "o"'s.
- Vamaza: $1,290, got $901 after commission. Held 10 months. Handreged CVCVCV easy to remember with A's only.
- ExoCode: $1,450, got $815 after commission. Held 1.5 years. Handreged because was an available combo for the word "Code" generated by my php generator.
- Traverce: $1,625, got $896 after commission. Held 11 months. Handreged because was on a list of brandables previously listed on one of the platforms and I felt it had potential.
- TradeViser: $2,995, got $1,855 after commission. Held 2.5 years. Handreged because was an available combo for the word "Trade" generated by my php generator.
- Sold 5 names for $6,460
- Payed in registrations + renewals + listing about $3,515
Afternic / Public Whois : 200 names listed / 2 sold
- Mokono: $500 direct offer, talked up to $900. Held 3 months. Got from Snapnames for $85 because I thought it was a good CVCVCV easy to remember with O's only.
- Proofus: $1,475 Afternic BIN, got $1179 after commission. Held 8 months. Handreged because was on a list of brandables previously listed on one of the platforms and I felt it had potential.
- Sold 2 names for $1,679
- Payed in registrations + renewals + listing about $3,449
7) Conclusion
The experiment is clearly in favor of the brandable marketplace over regular marketplaces. One could argue that the names I listed elsewhere were the leftover from the brandable marketplace, but I had many CVCVCV names that seem really equivalent to me in terms of quality on both Afternic and Brandroot and I sold more of them on the latter.
I must admit I've lost interest in trying to list names on Brandroot though. They have now over 20k names and I'd need to add about 70 to reach back my 1% exposure. I hoped to sell between 3 and 6% of my names there, but the end result is 3,4% in two years => less than 2% per year. In short, the waiting game gets a bit boring... I don't know. Maybe I should liquidate these and buy myself some bitcoins
I tried to launch my own marketplace to see if I could have more success than with Afternic. Called it BrandCombo.com. So far I only got a few price inquiries after 1 year of it being online. I'm pretty sure I don't have the critical mass for it : I listed xx / low xxx names. My guess is that below 1,000 names it's hard to get a marketplace effect where you can generate traffic and get visitors interested in more than one of your names.
That was quite a long one! Thanks for reading this far. Hope that gives food for thought to some of you
8) References
Past sales brandables
DNGeek: Brandable domain name sales of past week
Past sales websites
Namebio: link to past CaCaCa.com sales
DNpric.es : link to 6L.com sales above $1000 in 2018
Bulk check for available names
Namebright
Expired names email alerts
ExpiredDomains.net
Dropcatcher software
DesktopCatcher
Brandable platform I listed on
Brandroot
BrandCombo (my own)
Generate Names
Lean Domain search: generate names for the word Robot
Wordoid
Portmanteau generator: Travel + Velocity = Travelocity
Invent-a-word
WerdMerge