homepage Welcome to WebmasterWorld Guest from 54.205.254.108
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member
Home / Forums Index / Google / Google News Archive
Forum Library, Charter, Moderator: open

Google News Archive Forum

    
Measuring External Site Backlinks
Google off-site links only
Sally Stitts




msg:107757
 9:58 pm on Jun 21, 2003 (gmt 0)

Sorry for the neophite question, but I have searched the usual suspects (this site and Google), to no avail.

Is there a Google command that will display only those incoming links that are coming from "off-site", instead of the mixture of off-site and on-site links shown with the
link:www.url.com command?
I noticed that -site doesn't work as on ATW-FAST.

Does the term "backlinks" refer to ALL links coming in, including links from your own pages, or just "external" incoming links?

Thank you very much in advance.

 

Alphawolf




msg:107758
 6:51 am on Jun 22, 2003 (gmt 0)

Go to ATW and type in www.yourdomain.com and hit search.

Regards,

AW

vincevincevince




msg:107759
 9:34 am on Jun 22, 2003 (gmt 0)

unfortunately it is not possible to combine search terms with the link: operator in google.

otherwise it would be great to write "link:www.mydomain.com -site:www.mydomain.com".

my solution for google is just to go through them manually ignoring the urls from my site.

Net_Wizard




msg:107760
 1:37 pm on Jun 22, 2003 (gmt 0)

I thought everybody know about this already but here it goes again

inurl:domain.com site:domain.com - all your internal URL(www,index,etc..)

inurl:domain.com -site:domain.com - all external links pointing to your domain

inurl:domain.com - all URL(external & internal)

Dolemite




msg:107761
 1:44 pm on Jun 22, 2003 (gmt 0)

inurl:domain.com -site:domain.com - all external links pointing to your domain

This doesn't nearly show all external links, for whatever reason. Doing link:http://www.domain.com/ and ignoring internal links shows more. There's no usable command combination for that, though.

Net_Wizard




msg:107762
 2:09 pm on Jun 22, 2003 (gmt 0)

What you mean it doesn't show 'all' external links?

Are you using it like this...
inurl:www.domain.com -site:domain.com
if so then it will only look for links having 'www.domain.com' in the URL and ignore others such as [domain.com...]

The correct way of doing it is without the www. like so...
inurl:domain.com -site:domain.com
This will show all external links to domain.com including external links for URLs with www. and all your internal pages.

Cheers

annej




msg:107763
 2:14 pm on Jun 22, 2003 (gmt 0)

All the web always shows more backlinks when I use
inurl:domain.com -site:domain.com
while Google shows about a third less.

If you are simply trying to find who is linking to a domain the above works.

If you want to know what Google is counting for an individual page it's pretty much impossible because Google counts more external backlinks than they show when you use the backlink function on the tool bar.

vincevincevince




msg:107764
 2:15 pm on Jun 22, 2003 (gmt 0)

inurl:domain.com site:domain.com - all your internal URL(www,index,etc..)

inurl:domain.com -site:domain.com - all external links pointing to your domain

inurl:domain.com - all URL(external & internal)

this DOES NOT show links, it only shows the url of your page written out in a readable way. if your link text is your domain name (<a href="domain.com">domain.com</a>) then it will pick it up, but it won't get (<a href="domain.com">Widgets</a>) and it WILL get "I went to www.domain.com which was useful..." which isn't actually a link at all.

Net_Wizard




msg:107765
 2:17 pm on Jun 22, 2003 (gmt 0)

And just for clarification, using...

link:http://www.domain.com/

would not show all your external links for your site. The above format will only show links to www.domain.com(internal and external) but not links to domain.com nor any links to your other internal pages.

The only practical purpose of link: is when you want to find out the specific links to a specific page

i.e.,

link:mydomain.com/directory1/page12345.html
link:www.domain.com
link:domain.com/index.html

Dolemite




msg:107766
 2:18 pm on Jun 22, 2003 (gmt 0)

Net_Wizard,

I'm not arguing with your logic. Although logically, I'm not sure what the inurl and allinurl commands should have to do with backlinks, but anyway...try it out, it just doesn't show all the external links that are visible in a link:http://www.domain.com/ search. And no, I'm not using www.domain.com anywhere.

Net_Wizard




msg:107767
 2:26 pm on Jun 22, 2003 (gmt 0)

Vince,

I don't know if this would violate the TOS here but to better demonstarte what I'm talking about check this out...

inurl:yahoo.com site:yahoo.com - this shows 'ALL' yahoo URLs at Google including the www. and all those subdomains(chat.,my.,groups.,etc.,etc..) no external links at all

whereas

inurl:yahoo.com -site:yahoo.com - will show 'ALL' external links to yahoo.com excluding yahoo.com internal pages

Try it

Cheers

vincevincevince




msg:107768
 2:36 pm on Jun 22, 2003 (gmt 0)

netwizard

you missed my point

inurl will match a text string within the page content
link will match the location of an <a href="LOCATION"> tag

when you use inurl:domain.com with a site: filter you get all the pages which have written the words domain.com in their sites content.
Example:
1 - I went to domain.com and found it useful.
2 - Domain.com is an interesting place.
3 - <a href="domain.com">Domain.com</a>
4 - <a href="foo.bar">Domain.com</a>

All the above would be matched by inurl:domain.com, and you can see that only one of those is actually a link to domain.com.

Assuming those pages have sufficient PR, only no. 3, the real link, would be picked up using link:domain.com

You are correct about adding -site:domain.com to remove from the results those which are at domain.com. Just you're missing the point that the list you are filtering by site is not a list of backlinks.

Dolemite




msg:107769
 2:39 pm on Jun 22, 2003 (gmt 0)

OK, after looking at more sites than my own, I stand corrected that inurl:domain.com -site:domain.com generally returns more results than link:http://www.domain.com/, but it returns results that contain no links to any page at domain.com, so it can't be considered valid as a backlink check.

Net_Wizard




msg:107770
 2:53 pm on Jun 22, 2003 (gmt 0)

inurl will match a text string within the page content

I understand what you are saying but how do you account results for yahoo such as

2002 FIFA World Cup (tm)
Official Site of The 2002 FIFA World Cup TM ¦ May 31
- June 30, 2002 FIFA World Cup - Korea and Japan. ...
fifaworldcup.yahoo.com/en/ - 40k - Cached - Similar pages

Get Local - Regional:US States
All of United States Change Location. Major Cities. Atlanta, GA
Austin, TX Baltimore, MD Boston, MA Chicago, IL Cleveland ...
local.yahoo.com/ - 19k - Cached - Similar pages

Historical Prices
Historical Prices, Start: ---. End: ---. ...
chart.yahoo.com/d - 4k - Cached - Similar pages

You don't see yahoo.com in those results except in the URL

I have a dynamic site which uses? and ID, my URls look like this mydomain.com/index.php?id=2342342 when I query Google like this...

inurl:?id site:mydomain.com

Google return all my URLs that have ?id which is not anywhere in the page text but only in the URL

I used this method find out what are my dynamic urls that have been indexed by Google

Net_Wizard




msg:107771
 2:59 pm on Jun 22, 2003 (gmt 0)

Dolemite,

Exactly :)

just one of the benefit of that format, with that format you can catch sites that supposedly have link to your site(possibly cloaking) and redirecting users to somewhere else. and also those that have hijack your title and descriptions.

doc_z




msg:107772
 3:37 pm on Jun 22, 2003 (gmt 0)

Google's definition of 'inurl' is:

If you include [inurl:] in your query, Google will restrict the results to documents containing that word in the url. For instance, [inurl:google search] will return documents that mention the word "google" in their url, and mention the word "search" anywhere in the document (url or no). Note there can be no space between the "inurl:" and the following word.

However, the 'inurl' command seems to work only in the correct way as long as it is used without any other word(s) or command(s). For example, inurl:google will return documents with 'google' in the URL, while 'inurl:google search' seem to be nearly the same as 'google search'.

Therefore, 'inurl:domain.com -site:domain.com' is similar to 'domain.com -site:domain.com', i.e. will return documents that contain 'domain.com' (in the text, url, ...) from any source except for the site 'domain.com' (as already explained by vincevincevince).

link:http://www.domain.com/
would not show all your external links for your site. The above format will only show links to www.domain.com(internal and external) but not links to domain.com nor any links to your other internal pages.

This is definitively wrong. If these URLs return the same page then you'll get in most of the case the same backlinks. For example, I get the same backlinks for:
link:mydomain.com
link:www.mydomain.com
link:www.mydomain.de
link:www.mydomain.de/index.htm

(which is always the same page). Of course, if these URLs return different pages, you'll get different results for the 'link' command.
Unfortunately this doesn't work in any case (only if Google identifies the different URLs as the same page).

Net_Wizard




msg:107773
 4:09 pm on Jun 22, 2003 (gmt 0)

However, the 'inurl' command seems to work only in the correct way as long as it is used without any other word(s) or command(s).

Exactly, it's a one word query, that's is why when you query 'domain.com' it treated as 1 word(could be a bug, otherwise it would look for 'com' in the content/anchor of the page)

In programmers terminology 'domain.com' is the base expression of all your URL regardless how you construct your links(i.e. www.domain.com, domain.com/index.html) and I would venture that Google would consider this as 1 word instead of 'domain' and 'com'

while 'inurl:google search' seem to be nearly the same as 'google search'

Two words or more 'in URL' queries you should be using 'allinurl' that is if you are looking for all the words inside the URL.

This is definitively wrong. If these URLs return the same page then you'll get in most of the case the same backlinks. For example, I get the same backlinks for:
link:mydomain.com
link:www.mydomain.com
link:www.mydomain.de
link:www.mydomain.de/index.htm

(which is always the same page). Of course, if these URLs return different pages, you'll get different results for the 'link' command.
Unfortunately this doesn't work in any case (only if Google identifies the different URLs as the same page).

I think you better check Google again ;)
For instance if...

5 links are pointing to domain.com
and 150 links are pointing to www.domain.com
and 1 link is pointing to domain.com/index.html

if you query

link:domain.com -> you will get 5 links
link:www.domain.com -> you will get 150 links
link:domain.com/index.html -> you will get 1 link

From Google itself
The query [link:] will list webpages that have links to the specified webpage.

specified webpage and not specified site or domain

caine




msg:107774
 4:21 pm on Jun 22, 2003 (gmt 0)

yep, its another way off playing the PR game.

Unfortunately the backlinks that G shows, are limited by the incoming links PR, whereas the ATW url investigator shows them all, so a bit of cross correlation, and knowing the last crawl / updates of both can give you a reasonable view of what links are being seen by G, best doing when ATW have just updated, though pinning the crawl date down other than logs, is quite hard. Could turn out to be a lot of work. But due to the intrinsic nature of the hyperlinking web, i pretty much work on the fact what the url investigator displays, G is also seeing and algo-ing, with obvious de-linked association for banned sites.

aravindgp




msg:107775
 4:29 pm on Jun 22, 2003 (gmt 0)

I personaly till date find msn.com to be the best judge of backlinks.

linkdomain:www.widget.com -domain:www.widget.com

Atw gives you all affiliates urls as well.

Google gives your site pages as well.

But msn.com is best in this category, now do you judge if the backlinks is counting in google, is simply check if PR is not zero there's 95% chance of that url counting as backlink.

The one disadvantage is google also gives hypothetical PR for sites not indexed based on assumptions.

Hope this helps
Aravind

doc_z




msg:107776
 5:04 pm on Jun 22, 2003 (gmt 0)

Two words or more 'in URL' queries you should be using 'allinurl' that is if you are looking for all the words inside the URL.

What I want to say is 'inurl:google search' doesn't work as described by Google. It should return 'documents that mention the word "google" in their url, and mention the word "search" anywhere in the document'. However, that's not the case. It seems that the results are nearly the same as for a search of 'google search' (i.e. documents that contain 'Google' and 'search'), while 'inurl:google' works correctly (i.e. yields results that mention 'Google' in their URL). I was never referring to a two word search in the URL (which is done by allinurl). Also, I was never referring as 'domain.com' as two words (that's not my point). All what I want to say is: a search for 'inurl:whatever -site:domain.com' is something similar to a search of 'whatever -site:domain.com' and has nothing to do with the search of backlinks.

I think you better check Google again ;)

Sorry, but that are the results that I get for my own domain. The results for the 4 mentioned cases are always the same. Thus I have no possibility to distinguish between pages linking to www.mydomain.com, www.mydomain.de, mydomain.de or a relativ internal link to my index page. (Google identifies these URLs. And the results for the 'link' command are always the same - this hasn't changed for months, even if the backlinks changed.)

Obvioulsly, the 'link' command refers to pages not sites (I never said this). However, if these pages are the same - as in the example - then you'll get often the same backlinks even if the URLs are different (I already mentioned this clearly). And (as I already said) it doesn't work in any case.

Net_Wizard




msg:107777
 5:20 pm on Jun 22, 2003 (gmt 0)

Obviously you have a very good internal redirect method that Google is actually looking at all your URL structure as one.

Where for instance if 1000 links are pointing to yourdomain.com and you do a query like...

link:yourdomain.com -> 1000 links result
link:www.yourdomain.com -> 1000 links result
link:yourdomain.com/index.html -> 1000 links result

Kudos to you for achieving that ;)

Bottomline: Use whatever method you think works for you all I'm saying is that IMO inurl: for me is more accurate, of course I'm just learning the complexity of this issue ;)

Sally Stitts




msg:107778
 7:12 pm on Jun 22, 2003 (gmt 0)

Thank you everyone!

Wow! Asking what I thought was a simple question seems to have ignited quite a bit of discussion. Evidently, the simple answer to the question seems to be "no".

I too have used the manual "eyeball" method, but it becomes somewhat tedious when doing one's top 20 competition (most with large incoming link numbers).

Ahhh, but the discourse! Great stuff. So many combos.
Like so many other subjects, further study is indicated.
It will take me a while to work through all the permutations, and tabulate all the ones I like, with notes.
I believe I will create a "test sheet", and run down the list.
Thanks again for all the suggestions!

jeyval




msg:107779
 10:05 pm on Jun 22, 2003 (gmt 0)

Net_wizard I desagree with you too

Either one of these queries give my site the very exact number of links

Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Google / Google News Archive
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved