NameSilo

strategy Can Domainers Be Replaced By AI Robots? And Why To Take A Multifaceted Approach To Domain Name Evaluation

Spaceship Spaceship
It is hard to read much without coming across an article about artificial intelligence replacing people in various jobs. What about domaining? How easily could we be replaced by a bot, or perhaps an autonomous agentic system?

If choosing names solely, or mainly, according to a few metrics, such as number of registered extensions, or how many times the term appears in a compilation of existing company names, why is the domainer needed? Couldn’t a rather simple automated bot do everything the domainer does?

As AI agents and systems get smarter, and more autonomous, even sophisticated approaches could be automated.

But this is not really an article about AI replacing jobs or automation. It is an argument for a more nuanced and multifaceted approach to domaining, one that incorporates both qualitative and quantitative considerations.

This article also deals with the topic of what to do when the signals on the quality of a name are in conflict.

Let’s first see how this article came about…

How This Article Started

Sometimes the article I publish in the NamePros Blog has been in preparation for months, or occasionally even years. I work slowly! At other times, I pivot to a topic from the news, or something that seems particularly relevant. This is one of those times when I pivoted, and the content of this article is illustrated through a specific domain name consideration from this week.

Each day I check drops of single-word domain names in several different extensions using ExpiredDomains.net filters. If you don’t yet do that, and want a guide to using ExpiredDomains, check out the articles Finding Expired Domain Names and Getting More from ExpiredDomains.

I have customized my columns in ExpiredDomains, and at a glance see how many registered extensions, whether the term is currently registered in a number of specific extensions, and search volume, cost-per-click and advertiser competition.

One night this week I quickly looked over the dropped list for .cc, and decided to take none of the dropping names. Very early the next morning, I took a second look, and decided to take one of the names I had earlier passed over.

While I prefer not to publicize the specific name in an article, it is an 8-letter dictionary noun, with an ‘er’ ending. The first six letters are also a very common noun. If anyone really must know the name, just send me a DM and I will tell you privately.

Let’s look at my thought process for first rejecting the name…

Number of Registered TLDs

ExpiredDomains.net or dotDB indicate many Exact Match TLDs (top level domains, or extensions) are registered with a term. For this name, the Exact Match number is 70 from dotDB, and 72 from ExpiredDomains. That number, the first thing I look at, is high enough that I paused when I first saw the domain name in the list, but typically for .cc I acquire names with 100+ extensions registered, so my initial decision was not to pick up the name.

Here is a tip: the number of exact match registrations that show strength in a term depends on the extension you are considering acquiring. Had I been looking at the same term in .io, .xyz or .org, I would have regarded 72 as enough to research the name carefully, and probably acquire it. If the name was the .com, even a few registered TLDs and I would have definitely considered, and probably taken the name. But the market for .cc single-word names is much less competitive, and fewer businesses consider that extension as an option, so you need terms that are more in demand.

DotDB also provides the Domain Count, which indicates the total number of domain names that contain this term, including many longer names. I covered this and more about dotDB in last week’s interview with Ken Lin, the developer of dotDB, published last week in the NamePros Blog. In this case, the Domain Count was 1083, a solid number, but not high enough, for a .cc name, to cause me to immediately acquire the name.

So why did I take a second look at this name?

Listen To Your Inner Voice

The reason I took a second look was not based primarily on any numerical metric. Rather, when I looked a second time at the list of expired names, this one struck me as a great name for a business. In short, I could without hesitation see starting a business on this name.

The name is also versatile, My mind immediately came up with many different types of businesses that might naturally name with this term.

The art of domain investing means you should listen to those inner voices. But you should also seek quantitative support for your opinion. So I dove a bit deeper into research on this name…

Number of Existing Active Companies

One of the sites I use every day is OpenCorporates. That site provides the total number of company names using, or also known as, any term. OpenCorporates covers the published directories from many jurisdictions, but keep in mind not everywhere make that information public.

So I went to OpenCorporates, checked the Active Only box, and searched the term. I found just 54 active businesses for this term. I almost never acquire a .cc domain name when this number is less than about 200. This metric seemed to validate my initial thinking that I should pass on this domain name, despite seeing many ways it could be used.

Advertiser Competition

I was almost ready to give up on the name after this disappointing result, but I had noticed that there was a fair amount of advertiser competition on the term. But even before that, I want to stress to look at whether important extensions are registered.

Without the column showing the names, I show below part of the display from Expired Domains. You can see that I have customized my columns to show whether a term is registered in .com, .org, .io, .net, .co and .xyz, as well as .me and .ca. You can see in the green box that this term is registered in all of those extensions. Don’t just look at the Exact Match and Domain Count numbers – also make sure that the term has been registered in .com, .org, .io, .net, .co, .ai and .xyz, and perhaps others.

Image-ED-markup-switcher.png

A part of a display from ExpiredDomains.net. I have marked up in red showing Exact Match domain count. The green box indicates registrations in my customized extensions, see top of table. The blue markup box shows search volume, ad competition, and CPC for the term.

If you have one of the paid plans for dotDB, you can get similar data from that; see below. We should not be surprised about the tiny discrepancy in exact match number, or a small difference in search volume, since both depend on the precise timing of data access. Note that dotDB gives a range of CPC values, while ExpiredDomains a single average. Also, in terms of competition, dotDB uses the Google Ads low, medium, high terminology, plus an indicator bar. ExpiredDomains use a value from 0 through 100.

Imagre-dotDB-CPC-switcher.png

A part of a display from dotDB. It provides a range in advertiser CPC, and a graphical display of ad competition for the term.

That relatively high value for ad competition stood out. I am no expert on this topic, but it seems to me that relatively high ad competition values, often, not always, occur when there are many relatively new/small companies using a term, and paying to get traffic from ads. You can see from the blue markup box that the value for this term is more than for most of the other dropping .cc names that day.

So I decided to check one more thing…

CrunchBase

While OpenCorporates is a fantastic resource for company names, many names are acquired by recent startups, particularly tech-oriented startups.

While there are regional biases, CrunchBase provides a better measure of tech-oriented startups. You will need a subscription to get the full search results, but can do a free search at CrunchBase with this link that can be expanded to show up to 25 company names. Many dropping terms have zero or very few names in CrunchBase. If all 25, and who knows how many more, use the term, it is usually a healthy indication of worth of the term.

Now it is a bit of an art deciding which company names are truly this term, but as I count it there are at least 10, a relatively respectable CrunchBase number, based on my experience. If I am counting for a singular term, as here, I generally include both singular and plural company names, and longer names including that term, but not names they present that are really a different base term. I also include brandable names if they are clearly based on the term.

Over the past year I have begun to reflect on whether the ratio of CrunchBase to OpenCorporates has some significance. That is, if there are relatively decent number of CrunchBase listings compared to the active OpenCorporates number, that might be an indicator of a trending term. I may do some analysis related to that, and write a future article just on that topic.

I also checked LinkedIn company name search, and found 35 with this exact term, an encouraging number.

I did an overview of active and abandoned trademarks using Furm.com.

The number of quotation mark Google search results (just over 43 million, a decent number, not outstanding) is an indicator of how well used a term is.

When I checked sales in the term using NameBio, the results were very disappointing, just $$$ level wholesale transactions, and only 4 in total.

So what we have is a somewhat conflicting picture for the domain name:
  • Modest number of exact match registrations.
  • Disappointing number of active company listings on OpenCorporates.
  • A dictionary word that could legitimately be used in many sectors and niches.
  • A noun, just 8 letters, easily pronounced and without any obvious negative factors.
  • Fairly strong ratio of CrunchBase listings to OpenCorporates.
  • Significant, but not outstanding, search volume, but moderately strong ad competition in term.
  • Fairly strong number of search results.
  • Very poor sales record from NameBio for term, both exact and contains.
But the deciding factor I have not yet covered…

The Important Deciding Factor

When seriously considering a term, I always check out the status of the important extensions. Here is what I found for this term:
  • .com is developed for a clothing outlet.
  • .io is for sale at just under $30,000.
  • .co is developed for a gaming outlet.
  • .ai is a developed site for a video related application.
  • .org has a coming soon lander.
  • .xyz is for sale at just under $8000.
  • .app is developed for business management technology.
  • .net is developed (lightly) for web hosting and domain registration.
I did look at other extensions for sale in the term, finding huge disparity from some less than $200 to the .dev being make offer with a $100,000 minimum. There were also some developed country codes, as well as a few developed or redirected new extensions. For example, the .global for the term redirects to a brandable form of the term (missing one vowel).

What I found positive was that a number of the better extensions are developed, and the others are strongly priced. But even more than that, the developed sites are in very different sectors and niches, confirming my feeling that the term was versatile. This information was the deciding factor in deciding to acquire the domain name.

Don’t depend on the color coding from dotDB to be sure which are truly developed, but use the clickable links, if you have a plan, from dotDB to visit each extension of interest yourself. Otherwise simply manually check the names.

You probably are wondering, do you really go to this much work before deciding to spend $12 (I registered it for multiple years)? Yes, yes I do.

Final Thoughts

I think the more comprehensive and sophisticated your approach to domain name evaluation, the more likely you will acquire names that will eventually sell. That will include both qualitative and quantitative indicators: domain name investing is both an art and a science.

I still think that number of exact match extensions registered is one of the most useful initial filters when looking at a list of names. This article simply argues to go beyond that one number in your consideration.

I took a rather different approach to the idea of multiple indicators in an earlier NamePros article Look Beyond a Single Domain Name Metric that complements this article.

While I briefly mention a number of indicators in the current article, it is not intended to be a comprehensive look at domain name selection and pricing. I covered that in a two-part series on How to Rate and Price a Domain Name:
Any full evaluation would consider the sales record and types of name that sell in that extension, topics not covered in the current article.

The idea of taking a second look is generally valuable in many aspects of domain investment, not just in domain name acquisition.

While the current article argues to not place too much dependance on any one metric, it remains true that each objective metric adds to the overall picture in important ways


I would like to express my thanks to the various services that provide useful data to help inform our activities as domain name investors: NameBio, ExpiredDomains.net, dotDB, OpenCorporates, CrunchBase, Furm.com, and others.
 
Last edited:
50
•••
The views expressed on this page by users and staff are their own, not those of NamePros.
hmmm
I dont really wanna become aialcy.com
 
1
•••
If there will be an AI for this business, it will belong to sector leaders at first. Like Godaddy, HugeDomains etc.

But thanks to AI, we small developers can develop small tools for free just by trying some code. Like bulk whois checker, drop catching for hand reg. There are opportunities still, just keep trying every day. Good luck to everyone.
 
2
•••
How easily could we be replaced by a bot, or perhaps an autonomous agentic system?

I reckon I could easily whip up a bot to register random names and not make any sales, perfectly mimicking the average domain namer investor.
 
12
•••
3
•••
I thought Huge Domains was an AI service of GoDaddy until HD pulled their inventory.

A certain class of domainers will be replaced by Ai with a heavily funded company supplying the funds. Many of the auctions are public and I'm sure there are Ai bots studying GoDaddy and Dropcatch auctions daily. Most domains sell around $2,500 retail.

I imagine a bot trained on human auction behavior, corporate usage of term, dictionary words & combos, extensions taken, prior sales, character length and identifying dotcom, and next four most popular extensions could build strong portfolios.

But the human touch, gut instincts, and nuances won't be replaced.
And of course a large percentage of the best dotcoms are long gone.

Ai and end users entering the competition for rare dotcoms, and Courts finding ways to help domain pirates hijack valuable domains will continue.

Making a part-time or full-time living from selling domains will become harder every year. And the number of people doing so will shrink yearly.
 
Last edited:
0
•••
They (the ai machines) would have to start buying them first though no? Domainers own (subjectively) their domains, ai is just another search/answer tool.
 
3
•••
Ai will only replace the average and mediocre in any field. That is because AI datasets are based on available knowledge. True creativity and genius is unknown to AI, which is why the best will always prevail over it.
 
5
•••
Bob,

While your article features a catchy title about AI, which is indeed a fascinating topic, I find its practical relevance even more compelling—specifically, how to analyze a name like a pro to truly "make money on the buy." Kudos to you, as always!

I’m sure you’ll follow up with a big disclaimer after I ask my question…

I noticed that in almost every metric, you mentioned a kind of "magic number" to rely on. Could you elaborate on all of them and share any others you personally use?

Thank you!
 
2
•••
Thanks Bob, I’ll ask (order?) my AI agent ’Nuances’ to absorb all of this, and your other writings for that matter.:xf.wink:

For me (and Nuances), the amount of reasoning needed depends on registration prices, though. If it’s a 2 dollar .xyz, our strategy is more like ”buy now, think later”, especially if it’s a fresh drop and requires swift action. We’ll ponder more deeply when renewals come.
 
3
•••
They (the ai machines) would have to start buying them first though no?

It's gonna be tough for ai machines to hand-reg domains without...

Hands!
80S Greeting GIF
 
Last edited:
5
•••
I noticed that in almost every metric, you mentioned a kind of "magic number" to rely on. Could you elaborate on all of them and share any others you personally use?
So when I am being careful, I use everything in the Part 1/2 series
While I briefly mention a number of indicators in the current article, it is not intended to be a comprehensive look at domain name selection and pricing. I covered that in a two-part series on How to Rate and Price a Domain Name:
When I want a quicker look to begin process, typically my order of looking at things is:
  1. Does the name sound respectable for a company name, and are there obvious potential uses.
  2. Number Exact Match TLDs (from dotDB, ExpiredDomains or a Graen evaluation)
  3. Number of OpenCorporates listings (active ones only)
Usually unless those look strong, I stop the process. Like I did initially for this name.

But if they look encouraging, the next things I would check are:
3. Google term, just to see if something I don't know
4. CrunchBase see how many listings very similar term
5. NameBio see what sales, when, where, price
6. What other extensions are for sale and at what prices

But I place a lot of importance on status of the TLDs I would rank equal or above mine. So when I found all those extensions developed in different fields, that sealed the deal for me.

This is for a one word name say in a secondary extension or a new extension. The process would be quite different for a brandable name where I would focus more on popular keyword terms, and use assistance from Atom Insights on suggesting how name might be used, and see if it pulls up similar sales, etc.

I've left out of this about evaluating extensions, since for the most part these days personally while I have a ton of one-of extension names, in terms of actively looking I am only looking at a handful of TLDs and know from the analyses etc. things about that TLD.

But I mainly choose names no one wants. This is a tough business.

The aim in most things is to get better each year. I hope my articles help people with that, and writing them has helped me with that. Many names I got 4 years ago I would not get again.

-Bob
 
Last edited:
4
•••
For me (and Nuances), the amount of reasoning needed depends on registration prices, though. If it’s a 2 dollar .xyz, our strategy is more like ”buy now, think later”, especially if it’s a fresh drop and requires swift action. We’ll ponder more deeply when renewals come.
I agree with that. Something like .ai with a stiff price and a 2 year minimum registration, I would consider more carefully than first year on something that is a dollar or two.

That said, there is time invested, so I am trying to invest in fewer names that I will later drop. The operative word is trying, I am not succeeding in that!

Bob
 
2
•••
When I want a quicker look to begin process, typically my order of looking at things is:
  1. Does the name sound respectable for a company name, and are there obvious potential uses.
  2. Number Exact Match TLDs (from dotDB, ExpiredDomains or a Graen evaluation)
  3. Number of OpenCorporates listings (active ones only)
(y) Many thanks for the advice.
These 3 are indeed very important - if the name in question even does not pass these 3 tests, looks like best to avoid.

For regging in a new TLD, can we still rely on these though?

Many thanks.
 
1
•••
For regging in a new TLD, can we still rely on these though?
For a new extension you need to know what the use and sales record is for that TLD. Things like .app, .xyz have a strong sales record. Many TLDs hardly any. But within that, xyz doesn't tend to be used much for traditional retail, but is more for decentralized web applications.

Also, it is important to differentiate new extensions that are what I would call generic, things like .xyz, .one, etc. don't imply a sector or niche. TLDs like .rentals, .homes, etc. clearly for a certain narrower application. Both can be good, but they are very different.

For the second category thigh high value terms should have a good match across the dot. One way to check this is to do a Google quotation mark search and see how many results. For example, if you are considering Example.TLD do a Google search on "Example TLD" and hit the Tools button.

You should also use dotDB with the term entered as ExampleTLD and see how many registrations. If none, it says interest in combination not that great. If .com only, a bit different, if 15 extensions have the combo registered, an encouraging sign.

But that approach does not make sense when say considering xyz. Typically there will be little results for "Example xyz" in quoations, or registrations of the combo.

Bob
 
1
•••
AI will replace domain sellers, domain buyers, content creators and content consumers. And then it will implode. ;)
 
2
•••
Thank you for the detailed answer you provided above!

Allow me to clarify my intended question:

Throughout the article, by practically every metric you provided, there was a 'magic number' used as a benchmark (as I highlighted in red). Could you elaborate on this? Perhaps, at some point, you could even create a list of your favorite ranges for all these metrics.

As always, thank you for your invaluable service to our community of domainers.

For this name, the Exact Match number is 70 from dotDB, and 72 from ExpiredDomains. That number, the first thing I look at, is high enough that I paused when I first saw the domain name in the list, but typically for .cc I acquire names with 100+ extensions registered, so my initial decision was not to pick up the name.
Had I been looking at the same term in .io, .xyz or .org, I would have regarded 72 as enough to research the name carefully, and probably acquire it.
In this case, the Domain Count was 1083, a solid number, but not high enough, for a .cc name, to cause me to immediately acquire the name.
If all 25, and who knows how many more, use the term, it is usually a healthy indication of worth of the term.
I found just 54 active businesses for this term. I almost never acquire a .cc domain name when this number is less than about 200.
I also checked LinkedIn company name search, and found 35 with this exact term, an encouraging number.
The number of quotation mark Google search results (just over 43 million, a decent number, not outstanding) is an indicator of how well used a term is.
When I checked sales in the term using NameBio, the results were very disappointing, just $$$ level wholesale transactions, and only 4 in total.
 
2
•••
Hi.

sometimes, i might be thinking if you are ai or not

cuz, you put in waay too much work, data analysis, comparables and variables.
when do you have time to eat, do you ever sleep and if so, do you dream about what you're going to write about next?

gotta give it to yah though, for the energy,
like one of those energizer bunnies
or a battery that helps run the matrix.

but for sure,
the basics of what a domainer does, if he does what everybody else does, can be done by ai/gpt/botics.
any machine can go down a list of tasks to quantify for acquisition,
and as this game has evolved, the newest have become most dependent on such lists, first.

imo...
 
4
•••
Thanks for the OpenCorporates link.
 
1
•••
do you ever sleep
I do. But admittedly this week my sleep has been pretty crazy. I am mainly an early morning person, but hopefully I can get out of this awake at 3 AM thing soon. It mainly has nothing to do with domain names.

if so, do you dream about what you're going to write about next?
I do dream,, but I never have dreamed about that and actually never have dreamed about anything domain related as far as I can remember. I do have a big disorganized jumble of dozens of ideas that I do plan to write about sometime. This was not Plan 1 or Plan 2 for this week.

ike one of those energizer bunnies
My favourite animal rabbits. Even imaginary ones. Has anyone seen Harvey around NamePros?

Thanks for your comments, and your many, many years of concise, distinctive commentary. imo. Seriously. Thank you.

Bob
 
6
•••
Throughout the article, by practically every metric you provided, there was a 'magic number' used as a benchmark (as I highlighted in[COLOR=hsl(1, 59%, 45%)] red[/COLOR]).
Ahhhh sorry to misunderstand. I would like to be able to give you a quantitative answer, but it is more just what seemed right as over the years. It is sort of if you flip the script, and decide with no metrics which are the good names, and then check the numbers, they sort of fall, for this TLD, in the range I suggest.

I only know a few extensions well enough to have a good feel for the numbers.

Bob
 
3
•••
2
•••
and decide with no metrics which are the good names
Hi

like when it was, before all the metric site builders, built their sites.
:)
before the bot, the bio, the dot do be do, and now ... how many tlds it takes to qualify as a qualifier.
could Ai do it without the metrics?

imo...
 
5
•••
like when it was, before all the metric site builders, built their sites.
:)
before the bot, the bio, the dot do be do, and now ... how many tlds it takes to qualify as a qualifier.
could Ai do it without the metrics?

AI can replace domainers who live by the metrics, but won't replace those who register names just because they're cool. AI doesn't know cool. It can try to learn, but will it succeed?

9hxo3x.jpg
 
3
•••
they are already replaced by ai, I see bots register crappy domains, and put them for sale at atom wholesale marketplace monts ago 😂😂
 
Last edited:
2
•••
  • The sidebar remains visible by scrolling at a speed relative to the page’s height.
Back