An experiment in the weight of a domain name for SEO.

[SoftWareRevue]

I thought it might be interesting to document the weight of a domain name when it comes to SEO.

So, I purchased page-cannot-be-displayed.com today and tossed it up on the web at, surprise! . . . http://www.page-cannot-be-displayed.com

Of course today, at the time of this entry, there was only one direct connection.

Let's see how it goes.

I don't expect much to happen. But, this should give an indication of how important a domain name is when it comes to indexing by search engines.

<<< Edited to unlink domain >>>

Last edited by SoftWareRevue on 11/23/2005 8:28 pm

[Scott]

I think domain name has quite an important factor to play in SEO. However, I don't know whether a domain on its own will be enough to make much changes to the search results.

A google search on "page cannot be displayed" gives a site with quite a low page rank. Once your site gets indexed it'll be interesting to see what happens.

Nice Idea :)

[writespeak]

Keywords in domains are believed to be just a small factor in SEO, while incoming links and anchor text play a much larger role.

From this article (http://www.highrankings.com/issue016.htm):

Perform some searches in any spidering search engine and sure, you may very well find keyword-rich domains in the top spots. But upon closer inspection, you'll see that the same keywords are also in the Title tags of those sites. Title tags *are* very important to high rankings; I daresay that they're as important as body copy. Many see the keyword-rich domain name, and assume that's what is causing the high ranking. Yet it's much more likely that the high ranking is a direct result of the Title tag and/or the body text, along with the other usual SEO suspects.

OTOH, a subdomain of mine shows up in searches for the name of that subdomain when that word is not used anywhere on the page or in the title tag. AFAIK there are no incoming links that contain that word in the anchor text.

Lois

[SoftWareRevue]

And, as that article states, it is speculation.

I, of course, titled the page the name of the domain (without the dashes).

So, should I leave the title blank for the experiment?

[writespeak]

So, should I leave the title blank for the experiment?

If you want it to be a valid experiment, you'll have to leave it blank or enter something that isn't in the domain, e.g. "Experiment." You'll also have to avoid using the domain words in anchor text and on the page itself (oops?), and you'll have to hope that no one links to the site using the words in the anchor text. If you give the page a title, anyone who might link to it will at least have an option other than the domain to use in the anchor text.

Suggestion: you could use that page to list sites that you want to link to, such as this one. Then it'll have content but not the domain words. :)

Lois

[Lurker]

Keying in page-cannot-be-displayed brings up a heap of page cannot be displayed entries, so it might take a bit for you to feature in the top 20 :)

[SoftWareRevue]

That was one of the reasons for choosing that domain name. :blush:

[writespeak]

I'm not sure I understand this experiment. You want to see how well a domain does based just on the keywords in its domain, right? But even if those keywords are highly important in ranking results, if the site has no (or only a couple of) incoming links, sites with several incoming links may rank higher because of those links. How can you weigh keyword domains without incoming links against non-keyword domains with incoming links? I think your site would have to have the same number of incoming links, from pages with the same PR as the other pages in the search results, for the results to mean anything.

Also, you have the keyword text as the first words on the page. Let's say that your page reaches #1 for those words. The fact that those words appear on the page will probably be more important than the keywords in the domain. But if you remove those words, and pages with those words on the page rank higher, what does that tell you? I don't know how you can assess the value of keywords alone unless you're comparing the page against pages with identical conditions except for those domain keywords.

At the site I maintain that gets the most traffic, I made changes one at a time and watched the traffic grow for those keywords. The changes involved putting keywords in the <title>, <h1>, and <h2> tags. That's the order in which keywords appear to have the most importance. But experimenting with domains is more difficult.

The only way I see to make your experiment work is to get the site in the search engines and then change one variable at a time to see what difference it makes. But you can't change the domain itself, so...I don't know. :spin:

What do other people think about this?

You inspired me to try my own keyword experiment. I've just changed my sig to have expressions that I use on 2 of my pages in the anchor text. I occasionally get hits for those expressions, and now that they're in anchor text for that site (although not in the domain), I'll see how much difference (if any) that makes.

Good luck with your experiment. :)

Lois

[writespeak]

I just had an idea. Dennis, you could leave the page as it is. When (if?) it finds itself in a nice position in search results, you could then remove the keyword text from the page itself and see how much of a difference that makes. If the page drops a lot or disappears, you'll know that the text gave it its position on not the domain. OTOH, if it makes only a little difference, that would suggest that the domain has some value in SEO.

Whatever you do, I suggest tracking the SERP in several search engines because they behave differently.

Lois

[Scott]

I think Lois has a point, there are a lot of other factors which can't really be tracked at this point. If you leave it as is, I'm sure that eventually it will get indexed by the search engines, however it could be a considerable time before it gets there. If you link to the page to try and get it into the search results a little bit quicker, and then do as Lois suggested and remove factors (eventually leaving a blank website) this would probably be an effective assesement of how much value a domain name has.

The problem with that method is that linking to the page will in itself be a factor, so you'll need to be able to remove the links at some point (Therefore hoping noone else links to the site).

Perhaps there is a better method which we haven't thought of. :confused:

[writespeak]

...and then do as Lois suggested and remove factors
(eventually leaving a blank website) this would probably be an effective assesement of how much value a domain name has.

If the site is completely blank, there'll be no reason for it to be indexed at all. I suggest giving it content without the keywords (e.g. links to all of Dennis' other sites or maybe one of his favourite recipes), and then seeing where it's positioned in SERP for the keywords in question.

I'm not sure if "keywords" is applicable here because does anyone search for those words?

The problem with that method is that linking to the page will in itself be a factor, so you'll need to be able to remove the links at some point (Therefore hoping noone else links to the site).

Everything will be a factor. I think the only thing you can do is compare it against itself before and after changing other factors, e.g. having the keywords on the page and then not having them. If you make just that one change, incoming links won't be a factor in the change.

Lois

[SoftWareRevue]

Well, we could do different things on different, but very similar, domains.

I have page-cannot-be-displayed.com and the-page-cannot-be-displayed.com

[writespeak]

Well, we could do different things on different, but very similar, domains.

I should've known you'd come up with another alternative. :) If you treat the sites exactly the same and then change only one variable for one of them, your experiment should work.

I wonder if it makes any difference that one domain has protected whois details and the other one doesn't?

And you know that having "The page cannot be displayed" in your <title> tag wrecks your experiment completely because search engines rank <title> words highly, but you're probably getting to that. :goggle:

Lois

[SoftWareRevue]

:think:

Maybe the approach is wrong. :blush:

Perhaps the way to test the strength of a domain name in regards to SEO, is to have the same content on something like the-page-cannot-be-displayed.com and abcdefghijklmn.com.

What you think?

[writespeak]

:think:

Maybe the approach is wrong. :blush:

Perhaps the way to test the strength of a domain name in regards to SEO, is to have the same content on something like the-page-cannot-be-displayed.com and abcdefghijklmn.com.

What you think?

Now that I've thought about it more, I think you're right that we were wrong. The variable needs to be the domain itself.

Do you have an extra domain for the purpose? If not, I might be able to help with that.

Lois

[SoftWareRevue]

I have too many domains. :(

One of the newer ones I own (within a week of the page cannot domains) is forumaholic.com

I would think that should serve the purpose, since one name has nothing to do with the other.

Now to give it some more thinking? :dog:

[Scott]

New ideas are sounding like a better way to go about this experiment :)

Exact same content on the 2 domains would work, but you'll need to control the offpage factors as well. This is probably most important in linking to the sites. Anchor text will need to be exactly the same for both sites, something like "My Site" ought to do it. You could, of course, not link to the site...but I think you'd be waiting too long for anything to take effect.

This'll be interesting to watch once the kinks are ironed out.

[Aussie Bob]

bah. SEO gives me a headache. It's like the technical version of voodoo. :eyep:

:bag2:

[SoftWareRevue]

. . .You could, of course, not link to the site...but I think you'd be waiting too long for anything to take effect. . ..Yeh. I think, even just a couple links would help kick start things a bit.

Maybe forumaholic.com would be the wrong domain because there are results for forumahoic returned. It was just purchased in that same time frame. But, now that I think about it, it may skew the results more than an older domain that there are no search results for.

I do have a domain that only returns one result back; the domain name.

I think it would work to put the same content on page-cannot-be-displayed.com and blogorvlog.com. :ghug:

[SoftWareRevue]

Now that I think about it, we can have a test within a test.

With the same content and, hopefully amount of links, tied to page-cannot-be-displayed.com, the-page-cannot-be-displayed.com, and blogorvlog.com, it will be interesting to see if there's a difference between page-cannot-be-displayed.com and the-page-cannot-be-displayed.com.

[Scott]

Sounds like it should work pretty well

:thewave:

[John D]

I'm very interested to see how this turns out..

I have never really been into SEO although I know it's very important.

I guess this will help to see if the domain name helps much.

I just bought a webmaster domain yesterday for a large sum hoping it will help with this.

[SoftWareRevue]

I'm very interested to see how this turns out. . .Yes. It would be nice to have a definitive answer to the question.

To put some more thought into this . . . . Does anyone have a suggestion on what we should use for test results?

[Scott]

Google Pagerank?

I'm not sure what else would provide a consistent value to ensure this is fair.

[SoftWareRevue]

Google Pagerank? . . .I think that probably updates the middle of January.

The thing about pagerank is, I can see them all being a PR 2. That certainly wouldn't prove nuthin.

I guess we can wait and, if it's a tie, we'll just figure out another way to weigh them.

[writespeak]

...we'll just figure out another way to weigh them.

SERP (search engine results position) for the targeted words?

Lois

[SoftWareRevue]

Update. Stats per month.

December

page-cannot-be-displayed:

Unique visitors: 36
Number of visits: 44
Pages: 48
Hits: 150
Keyphrases: page-cannot-be-displayed.com, this page cannot be is displayed.com


the-page-cannot-be-displayed:

Unique visitors: 24
Number of visits: 26
Pages: 26
Hits: 79
Keyphrases: the-page-cannot-be-displayed.com


blogorvlog:

Unique visitors: 32
Number of visits: 37
Pages: 40
Hits: 97
Keyphrases: this page cannot be displayed.com, the page cannot be displayed, the page cannot be displayed secure site

[Scott]

Now to come up with a way to measure them. Pagerank is 0 on all of them :iunno:

[SoftWareRevue]

I think we just have to give it more time.

[Scott]

Over a month since the last post, thought I'd take a look at how the sites were coming along. Interesting, all of them have a google PR of 2. As far as SERPs are concerned, the phrase "page cannot be displayed.com" seems to have the most effect with a number 1 result in google.

With only these very limited results, I guess it's clear that domain name has an effect or SERPs but not PR. The prefix "the" doesn't seem to help much here.

More conclusive results might appear over a longer period of time...

[SoftWareRevue]

Thanks for the update, Scott. I was going to post something when I noticed one had a PR 2 a few days ago. But, thought we better wait until google gets done dancing. :dancin:

Here are some traffic and search stats to date:

page-cannot-be-displayed.com:

January Uniques: 46
February Uniques: 16

Search Keyphrases:

January: the page cannot be displayed.com
this page cannot be displayed.com
page cannot be displayed page currently available
page cannot be displayed.com
page cannot be displayed lan
the page cannot be displayed pct
page cannot be displayed 128
page cannot be displayed technical ssl

February: www.page cannot be displayed.com

the-page-cannot-be-displayed.com:

January Uniques: 37
February Uniques: 11

Search Keyphrases:

January: this page cannot be displayed.com
this page cannot be displayed.com
the page cannot be displayed.com
page cannot be displayed.com
cannot be displayed.com

February: this page cannot be displayed.com
blogorvlog.com:

January Uniques: 39
February Uniques: 27

Search Keyphrases:

January: this page cannot be displayed.com
page cannot be displayed
this page cannot be displayed
networking the page cannot be displayed some sites
page cannot be displayed secure site

February: this page cannot be displayed.com
page cannot be displayed
the page cannot be displayed secure connection
www. this page cannot be displayed.com
this page cannot be displayed
this page can not be displayed.com

[SoftWareRevue]

Updating with latest stats.

All three domains are PR 2. (Without the www prefix)

page-cannot-be-dsplayed.com is the only one with a PR (2) when including the www.


Here are the latest traffic and search stats:

page-cannot-be-displayed.com:

June Uniques: 46
July Uniques: 30

Search Keyphrases:

the page cannot be displayed.com
www.this page cannot be displayed.com


the-page-cannot-be-displayed.com:

June Uniques: 39
July Uniques: 27

Search Keyphrases:

the page cannot be displayed.com


blogorvlog.com:

June Uniques: 46
July Uniques: 29

Search Keyphrases:

NA

[Jan]

Time to whack some adsense ads on there :banana:

[Scott]

I'm struggling to come up with any conclusions on those, doesn't seem like domain name is having a massive impact on the weight of PR anyway. What is impressive is how they've got a PR of 2 when according to google no sites actually link to them.

It would be interesting to try out Jan's suggestion and put adsense on one of them, I don't think we're going to see any more activity on domain names alone. :iunno:

[AjiNIMC]

hell, if your domain passes phone and bill board tests then it is good and well. Forget SEs

[SoftWareRevue]

Kinda interesting.

http://www.page-cannot-be-displayed.com PR 3
http://page-cannot-be-displayed.com PR 0

http://www.the-page-cannot-be-displayed.com PR 3
http://the-page-cannot-be-displayed.com PR 3

http://www.blogorvlog.com PR 0
http://blogorvlog.com PR 3


Not sure what that means yet. But there it is. :crazy:

[SoftWareRevue]

Time to whack some adsense ads on there :banana:That's actually not a bad idea. It would be interesting to know if any of them got traffic that actually clicked on an ad. :idea:

[SoftWareRevue]

Here are the latest traffic and search stats:

page-cannot-be-displayed.com:

August Uniques: 46
September Uniques: 55

Search Keyphrases:

page not displayed help
page not displayed
the page cannot be displayed adjust browser settings
www.the page cannot be displayed.com
the page cannot be displayed detect network settings
this page cannot be displayed.com
the page cannot
the page cannot be displayed.com
the page cannot be displayed
web page cannot be displayed
secure page cannot be displayed
page can not be displayed ssl
page cannot
microsoft the page cannot be displayed



the-page-cannot-be-displayed.com:

August Uniques: 74
September Uniques: 241

Search Keyphrases:

the page cannot be displayed
page cannot be displayed
this page cannot be displayed
the page cannot be displayed secure sites
cannot automatically detect settings
the page cannot be displayed error
page cannot be displayed domain the-page-cannot-be-displayed.com
page cannot be displayed secure
this page cannot be displayed error
page cannot be displayed lan setting
autosearch
the page cannot be displayed secure site


blogorvlog.com:

August Uniques: 46
September Uniques: 51

Search Keyphrases:

the tools menu and then click internet options..com
select automatically detect settings and then click ok.
the refresh button or try again later. if you typed
the page cannot be displayed.com

[SoftWareRevue]

STATS UPDATE:


page-cannot-be-displayed.com:

October Uniques: 87
November Uniques: 55
December Uniques: 58

Search Keyphrases:

www.the page cannot be displayed.com
page cannot be displayed for all secure site connection
this page cannot be displayed.com
the page cannot be displayed
page cannot be displayed
the page cannot be displayed.com
web page cannot be displayed
this page cannot be displayed what is that about?
this page cannot displayed
the page cannot be displayed for secure sites




the-page-cannot-be-displayed.com:

October Uniques: 139
November Uniques: 35
December Uniques: 41

Search Keyphrases:

the-page-cannot-be-displayed
page cannot be displayed
this page cannot be displayed
\the page cannot be displayed
the page cannot be displayed error
search
on the connections tab click lan settings
the page cannot be displayed secure site
cnn
page cannot be displayed 128




blogorvlog.com:

October Uniques: 61
November Uniques: 83
December: 82

Search Keyphrases:

ssl automatically detect page cannot be displayed
www. this page cannot be displayed.com
this page cannot be displayed.com
this site cannot be displayed
this page can not be displayed.com
some sites require 128-bit connection security
page cannot be displayed secured site
secure site cannot be displayed
page cannot be displayed - automatically detect settings
browser back button the page cannot be displayed
help page cannot be displayed when on a secure site
page cannot be displayed cannot find server or dns error
page can not be displayed.com
this page can not be displayed.com

[SoftWareRevue]

Wonder what happened in October. :think:

[Webmaster7]

Nice experiment :) Another thing we could do is with non-existing words.

For example:

register: zxcvbnmh.com
create a subdomain: zxcvbnmh.yourdomain.com
create a blog on blogger: zxcvbnmh.blogspot.com
create a subdirectory: yourdomain.com/zxcvbnmh
create a page: yourdomain.com/zxcvbnmh.html

now put the same content in all these 5 sites, and link to all of these from a single place. Hum... I'm going to do this :)

[SoftWareRevue]

... I'm going to do this zxcvbnmh.com is available! :guitar:


Let us know when you start the experiment. Sounds fun. :clap:

[SoftWareRevue]

.. I'm going to do this :)Where'd ya do it at?




Runs to get stats for this thread. :out:

[SoftWareRevue]

Runs to get stats for this thread. How's that goin' for ya? :discuss:

[vtabletop]

This is an engrossing thread! I just went over all the five pages and checked the PRs for the domains in mention.I still can't see any conclusive results based on the stats.:confused: I am truly interested in the results of your experiment,if any and the other experiment also sounds interesting.Hope google is not intentionally spoiling the party by tampering the results ,lol.

[Scott]

I still can't see any conclusive results based on the stats.:confused:
I'm having the same problem. I think perhaps picking a subject which already has many many results hasn't helped us judge this easily. Using a few search engine analysers it appears that none of the domains are in the top 100 of any common search engine. I don't think we're going to get much more out of the experiment as it is currently set.

A few possible ideas for expanding this:

Add more links. At the moment (to my knowledge) only SWRs blog is linking to these sites. Adding more should (I believe) improve their ranking. It is important though to add all 3 links together to ensure the number of links remains the same.
Change the content. Instead of the current content perhaps something someone might find useful - reasons for getting that kind of error for example. Again, needs to be identical on all of them.
Adsense. Jan mentioned this above. It might have no effect what so ever. However, for whatever reason, it may push them up slightly so that a comparison could be made.By doing something like this, we could hopefully improve the rankings to a level where they could be measured. Out of them all adsense is perhaps the most interesting as it fuels the debate of whether using adsense impacts rankings. If adsense was put on all of them and within a month (or more?) they all shot up to under 100 that would be a very interesting result though I'm sceptical if such a thing would happen.

Thoughts?

[SoftWareRevue]

Thoughts?:flirt: Maybe.

This experiment was to evaluate the weight of a domain name when it comes to SEO.

So far, I see it as holding absolutely no weight. (The domain name).

We have three domains, all carrying the same content -- All having the same backlinks -- And all displaying the same ads.

Traffic varies negligibly for all of them and all hold the same PR.

I'm not sure that optimizing them keywords or phrases would make any difference on how they perform. But, it might be interesting to try. After all, we obviously know a little something about SERP (http://global-warming-2007.com).

That contest (http://www.seoworldchampionship.com/) doesn't support that a domain name has bearing on placement as well. I suppose you could argue that the top three had variations of the key phrase, but there were much bigger factors that played in their placement. And, they were not in those positions for most of the contest.

If adsense was put on all of them and within a month (or more?) they all shot up to under 100 that would be a very interesting result though I'm skeptical if such a thing would happen.They've had adsense on them ever since we slapped adsense on our Globalwarming Awareness2007 project (http://global-warming-2007.com). :P

I don't think higher placement would show us anything other than we could move a domain up.

But, hell, I'm game for that!

Maybe we should see where they place now, and simply try and get them on that first page of results. If we can move them all, then this experiment (the weight of domain name) will be verified with the conclusion that a domain name doesn't matter. However, if we move only the two related, we'd have to reconsider.

Whatcha think? :nurse:

[vtabletop]

Linking back to the domains with the anchor text will definitely increase their PR ,but then the purpose of the experiment,the importance of domain name will remain a mystery.Perhaps adding adsense doesn't really influence Pr ,but here I am guessing.Speaking of experiments,I came across a neat experiment a few weeks back.The person had purchased a domain named pagerank10.He set a goal of achieving a PR of 10 within 730 days from the date of purchase of the domain that being 20th April 2006.Now,if the experiment was purely based on the domain name ,then this could have been very useful ,but I suspect that many sites are linking back to his site because he intends to gift off the domain to the top referrer on the site and that brings in the other variable in the experiment-backlinking,lol.Nevertheless,it's an interesting experiment.I am not sure if I am allowed to post a live link.The domain is pagerank10.co.uk and has already climbed up to pr6.That leaves him another 305 days to cover the remaining distance ,lol.

[Scott]

They've had adsense on them ever since we slapped adsense on our Globalwarming Awareness2007 project (http://global-warming-2007.com). :P

Oops, that's what happens when I'm blocking google ads :bonk:

[SoftWareRevue]

...the importance of domain name will remain a mystery...I don't think it is a mystery.

Given the parameters of this experiment, it is safe to conclude that the domain name has no bearing on SEO positioning.

The question becomes, have we done enough to call this experiment conclusive? :think:

[SoftWareRevue]

... I'm blocking google ads Silly Scotts. :joker2:

I mean Scotts. Not Scotts. :scatter:

[Scott]

Silly Scotts. :joker2:

I mean Scotts. Not Scotts. :scatter:

Offending Scot(t)s? (http://bebo.com/watch/2987482421) :cmere:

[vtabletop]

Considering that all three domain names have backlinks from the same sites, the same content and similar ads , I think the experiment is scientific enough to reflect on the importance of domains.It sure seems that domains are irrelevant for SERP , but many gurus claim that they are as important as the title.

[PinkFloyd]

Although keywords in domains do play a "small" factor in SEO, its still pieces of the puzzle that is beneficial for your domain. There are some large factors that your website needs to be successful in marketing it to search engines, but don't forget the little factors that can add up quickly.

Keywords in domains is one of the small factors, but it does help a bit. Might not be a big boost, but it will help out just a bit.


but many gurus claim that they are as important as the title.
I don't see how it would even be close to as important as a title. Besides back links, Google itself shows the some of the most weight off of titles.

As long as you have "some" back links, and "some" relevant content, using appropriate titles along with your keywords will get you to the top of Google for most niche keywords, or at least compete with general keywords.

I have built several websites in the last year simply putting my keywords into the title, meta, and domain.. while roughly 80% of the content was relevant to the keywords and got them into the top 5 on Google. And at least in the top 10 for MSN and Yahoo in maybe a month. I truly think the biggest factor was the titles. I set the titles to match the exact keywords I wanted people to find my websites under.

[vtabletop]

You are probably right.I have read that domain names are as important as the title , but not being an expert myself I can only report what I read and yes, that could be wrong.I read this in a pdf of an SEO expert.I have no way to know about his real expertise ,lol.

[sharrychrist]

Hi, this post is very informative; however I would like some specific information. If someone can help me then please send me a private message. Best Regards,

[SoftWareRevue]

I'm thinking of kicking this up a notch.

What if we try to get others to link to The Page Cannot Be Displayed (http://www.the-page-cannot-be-displayed.com/) by inviting them to "get involved" in the TPCBD Project (http://www.the-page-cannot-be-displayed.com/)?

Whatever that project is. :think:

I made some changes to that one site. Because, I'm pretty sure we've decided that this study has been made. Now it's on to . . . . other stuff.

[beamup]

Hello

I agree Keywords in Domain name has small effect as I generally try to choose from the research. Meta tags has no effect now I guess, Keywords are also playing very important roles. Meta Description is also important and most of all one way link is the key factor.

i like the article though!
Thanks
Steve

[jerry555]

You'll also have to avoid using the domain words in anchor text and on the page itself and you'll have to hope that no one links to the site using the words in the anchor text.