Jump to content

  • Log in with Facebook Log in with Twitter Log In with Google      Sign In   
  • Create Account

Subscribe to HRA Now!

 



Are you a Google Analytics enthusiast?

Share and download Custom Google Analytics Reports, dashboards and advanced segments--for FREE! 

 



 

 www.CustomReportSharing.com 

From the folks who brought you High Rankings!



Photo
- - - - -

Truth, Lies, And Search Engines


This topic has been archived. This means that you cannot reply to this topic.
50 replies to this topic

#31 Haystack

Haystack

    HR 7

  • Active Members
  • PipPipPipPipPipPipPip
  • 1,980 posts

Posted 10 August 2003 - 06:27 PM

Google has determined that these two pages, with meta description, are more relevant than the pages without meta description tags. Is that proof that Google uses meta description to rank? No. Just something to think about, at most.

That is something to think about. It looks like that's the variable being tested among the four pages listed here:

http://www.google.co...dog named Homer

#32 braindead

braindead

    HR 2

  • Active Members
  • PipPip
  • 31 posts

Posted 10 August 2003 - 06:31 PM

I doubt any of these could produce anything close to conclusive evidence. Only Google knows.

#33 projectphp

projectphp

    Lost in Translation

  • Moderator
  • 2,203 posts

Posted 12 August 2003 - 12:37 AM

I doubt any of these could produce anything close to conclusive evidence. Only Google knows.

Thanx. That was helpful <rolls eyes>

But conclusive isn't the only goal, understanding is. If you can learn something, even if you are only 75% sure, that is great in my opinion.


If you don't, because of the methodology employed, which you believe there is a way to improve, then post the specifics of how you can improve it.

If you have nothing, say nothing. If some topic doesn't interest you, pointing out a truism isn't helpful at all.

[Edited out slam to another forum. Please see our Forum Guidelines. - Jill]

Only MHO, of course, and I reallyu mean no offence when I say it, its just one of my pet peeves!

Edited by Jill, 12 August 2003 - 10:13 AM.


#34 projectphp

projectphp

    Lost in Translation

  • Moderator
  • 2,203 posts

Posted 12 August 2003 - 12:40 AM

You can often take the words out of the page titles, search for them and find the page in the SERPs, click on cache, and see:

http://216.239.57.10...&hl=en&ie=UTF-8

When in fact, links to that page do not contain those terms. It's just in the page titles.

That is true also. "Google bombing"

<WADING BACK IN>Who ever said that the Title tag wasn't counted? That is definitely worth checking seperately. The example given lacks too many controls for my liking. A good theory, but it needs more testing!!!

Specifically:
1. How do you know none of the links contain any of those words?
2. What part of the metadata is indexed?

Edited by projectphp, 12 August 2003 - 01:19 AM.


#35 braindead

braindead

    HR 2

  • Active Members
  • PipPip
  • 31 posts

Posted 12 August 2003 - 03:18 AM

Thanx. That was helpful <rolls eyes>

But conclusive isn't the only goal, understanding is. If you can learn something, even if you are only 75% sure, that is great in my opinion.



"Inconclusive" means that you cannot draw any conclusions from these tests. In other words, they are worthless. I'm sorry you did not find my reply useful. I know that you like to jump to conclusions - for example when you stated in IHY that "Google no longer indexes meta description tag". That would be a good example of a useless, amatuer post. Most professional SEO's know the Google indexes the meta description tag, as Alan Perkins and MakeMeTop pointed out for you here.

More professional types would find the statement to be useful, as it eliminates a faulty theory.

[Edited. See above post.]

Only MHO, of course, and I reallyu mean no offence when I say it, its just one of my pet peeves!


And misinformation is one of my pet peeves.

You can often take the words out of the page titles, search for them and find the page in the SERPs, click on cache, and see:

http://216.239.57.10...&hl=en&ie=UTF-8

When in fact, links to that page do not contain those terms. It's just in the page titles.

That is true also. "Google bombing"

<WADING BACK IN>Who ever said that the Title tag wasn't counted? ?


Nobody. Sure, you like to reply in a hurry, but forum etiquette would suggest that you read the posts preceding that one.

In regard to Google's "These terms only appear in links pointing to this page:" I was stating that Google does not differentiate. Sometimes those words were are not included in link text of inbound links. Sometimes they are found in page titles, but Google still uses the statement, "These terms only appear in links pointing to this page".

How do you know none of the links contain any of those words?


You act as if this is something new. It's basic stuff. I could give you a link a minute to cached pages with "These terms only appear in links pointing to this page" on them, when in fact those terms do not occur in the anchor text of pages linking to the page in question. The terms appear in the page titles.

I don't mind educating you, but you really should be taking the initiative and at least try to learn some of this stuff on your own.

Edited by Jill, 12 August 2003 - 10:16 AM.


#36 Alan Perkins

Alan Perkins

    Token male admin

  • Admin
  • 1,629 posts

Posted 12 August 2003 - 03:49 AM

Wow, it amazes me that this topic engenders so much vitriol!

Please just keep it on-topic, i.e. no personal attacks.

It's clear that Google indexes the meta description tag because Google uses the meta description tag in some SERPs. It couldn't do that if it didn't index it.

What's not clear is whether Google actually uses the meta description tag in its algo. IMO this is a non-issue. I really don't care if Google does or doesn't factor the meta description tag into its ranking calculations. I don't think it does at the moment, but so what? Whether it uses them or not, their weight should be inconsequential in the algo. Far more important, IMO, is the fact that they are seen in the SERPs.

braindead, I agree that page titles are treated as links pointing to the page for the purposes of this sentence in the Google cache header:

These terms only appear in links pointing to this page


Given that the <title> is used in Google's SERP as a link it is obviously true. :drunk: I also think it's a good way to consider the title - as an intrinsic link to the page itself.

#37 braindead

braindead

    HR 2

  • Active Members
  • PipPip
  • 31 posts

Posted 12 August 2003 - 03:53 AM

FWIW, braindead, I agree that <title>s are treated as inbound links for the purposes of that sentence in the Google cache header. Given that the <title> is used in Google's SERP as links, and that a cached page must be able to appear in a SERP, this is obviously true.  I also think it's a good way to consider the title - as an intrinsic link to the page itself.


Agreed 100%.

#38 projectphp

projectphp

    Lost in Translation

  • Moderator
  • 2,203 posts

Posted 12 August 2003 - 08:55 PM

Braindead:
Good points, and all valid! Matbe that :"words only appear in links.." is meant to indicate in the title. Never really considered that. Does seem quite likely!!

As for

Most professional SEO's know the Google indexes the meta description tag

But isn't that the point? Indexes and uses as part of algorithm are different issues. If something is indexed, but none of those words are ever used to help rank a page, wouldn't that indicate that they are relevant? They still help to develop a better snippet (if they include relevant keywords), but they may have no influence on rankings.

I don't mind educating you, but you really should be taking the initiative and at least try to learn some of this stuff on your own.

Touche!! Point taken and I concede on this one: Google does do exactly what you say it does!!

#39 Matt B

Matt B

    The modem is the message.

  • Active Members
  • PipPipPipPipPipPip
  • 558 posts

Posted 12 August 2003 - 10:15 PM

IMO this is a non-issue.  I really don't care if Google does or doesn't factor the meta description tag into its ranking calculations.  I don't think it does at the moment, but so what?  Whether it uses them or not, their weight should be inconsequential in the algo.  Far more important, IMO, is the fact that they are seen in the SERPs.

Exactly. php and Alan are right on in this.

Even if descriptions aren't utilized as part of the ranking algorithm, they may be seen by users. Users tend to use the snippet to judge a site before clicking on the link. IMO, that is much more important than a few ranking positions. A good snippet could drive much more traffic to a site that is listed below others that may have with no description or (worse yet) incomplete or misleading information.

#40 ChrisB

ChrisB

    HR 2

  • Active Members
  • PipPip
  • 25 posts

Posted 14 August 2003 - 10:20 AM

I've never said that Google doesn't use the meta description tag to create your site's abstract -- infact I say quite the opposite in many places, such as on the home page of the experiment and in my article discussing the experiment.

So bringing up this fact as some kind of counter-argument to my experiments on ranking doesn't make a whole lot of sense.

Yes, two pages with meta tags are coming up first this week. Last week they weren't. Google just updated though, with 5 of each type statistical probability would lead one to believe that if they were randomly reorganized each month then sometimes the first couple would be all from the same type.

I also don't understand how someone can claim that a site specific search is using a vastly different ranking algorithm than a normal search... however to placate those people please view this SERP:

http://www.google.co...fe=off&filter=0

It isn't a site specific search at all. You'll notice that there is no apparently order with the meta v. non-meta pages. The first two have meta tags, the next two don't, and so on. If meta tags were used then the first 5 would all have them.

Next month (or next big update) the order will likely change again. Its random.

Next I suppose you'll say that the search I linked to above isn't valid either because I selected to show omitted results with similar content. I suppose that results in a fundamental algorithm change too?

However if next month the first two don't happen to be meta tag pages I assume that you (Braindead aka John theHack Scott) will finally admit that Google doesn't use meta tags to rank pages. Or will I be treated with another far fetched explanation? Perhaps something to do with the moon's alignment at the time?

Edited by ChrisB, 24 September 2003 - 02:57 PM.


#41 ChrisB

ChrisB

    HR 2

  • Active Members
  • PipPip
  • 25 posts

Posted 14 August 2003 - 10:23 AM

Even if descriptions aren't utilized as part of the ranking algorithm, they may be seen by users. Users tend to use the snippet to judge a site before clicking on the link. IMO, that is much more important than a few ranking positions. A good snippet could drive much more traffic to a site that is listed below others that may have with no description or (worse yet) incomplete or misleading information.


No one has said not to use them, or atleast I haven't.

Afterall Teoma and Inktomi still use them last I checked.

This is simply a matter of investigating how search engines really work.

#42 Mel

Mel

    HR 5

  • Active Members
  • PipPipPipPipPip
  • 353 posts

Posted 14 August 2003 - 10:29 AM

Hi Chris
I like the idea of your experiments keep them up there is always something to be learned.

I do believe that we can learn something, but would be hesitant to say that a set of rankings under one algo compared with a set of rankings under another algo proves anything.

#43 Sharon & Roy

Sharon & Roy

    HR 3

  • Active Members
  • PipPipPip
  • 96 posts

Posted 10 October 2003 - 02:27 PM

Hi Chris & Forum-Mates,

Some interesting observations can be made by analyzing the following queries ...


This query ...

http://www.google.co...henginelabs.com

... Returns just 3 pages of about 24

Searched for pages linking to searchenginelabs.com.  Results 1 - 3 of about 24

Sites We Publish - Jalic Internet Consulting
Welcome to Jalic.com Just a Little Internet Company. Sites We Publish,
In addition to providing website development and promotion ...
www.jalic.com/webpub.php - 4k - Cached - Similar pages

The Blue Dog
WARNING! READ THIS FIRST! WARNING! READ THIS FIRST! This page exists
to run experiments on how search engines rank pages. In order ...
www.searchenginelabs.com/meta_ranking_5.php - 3k - Cached - Similar pages

Truth, Lies, and Search Engines - Website Publisher.
Printable Version. Truth, Lies, and Search Engines. By: Chris Beasley,
Published: 2003-07-26, Parent Category: Search Engine Optimization. ...
www.websitepublisher.net/article/truth-lies-search-engines/ - 10k - Cached - Similar pages

In order to show you the most relevant results, we have omitted some entries very similar to the 3 already displayed.
If you like, you can repeat the search with the omitted results included.

.............................

http://www.google.co...ilter=0&num=100

Searched for pages linking to searchenginelabs.com.  Results 1 - 24 of 24. Search took 0.19 seconds.
By default, Google searches for variations of your search terms. To search only for an exact term, place a '+' sign before it.


Some Food For Thought Questions:

1.) Why did Google return the internal page ... searchenginelabs.com/meta_ranking_5.php ... instead of any of the 21 others internal pages?

2.) Is the page ... searchenginelabs.com/meta_ranking_5.php ... ALWAYS returned as the the ONLY internal page when Google is queried for ... http://www.google.co...henginelabs.com ?

3.) The page ... searchenginelabs.com/meta_ranking_5.php ... currently has a PageRank4 (as all the internal pages do) and the Home Page currently has a PageRank5. Do you think that all of the 22 internal pages have EXACTLY the same amount of ACTUAL PageRank Points? Or do you think that there may be SOME difference in PageRank Points between all of the internal pages no matter how small they may be, (e.g. 605 for one page, 604 for another page and 598 for yet another page) or do you think that they are all the same, (e.g. all 605 points)?

4.) For this query ... http://www.google.co...henginelabs.com ... we see that the results returned are as follows ...

www.jalic.com/webpub.php
::: PageRank6 with 21 Forward Links :::

www.searchenginelabs.com/meta_ranking_5.php
::: PageRank4 with 2 Forward Links :::

www.websitepublisher.net/article/truth-lies-search-engines/
::: PageRank5 with 18 Forward Links (with 17 Unique Forward Links) :::

... So the question becomes, why that ORDER? Could it be PageRank ordered (notice the PageRank of each page above)?

Your Friends,

Sharon and Roy Montero

#44 ChrisB

ChrisB

    HR 2

  • Active Members
  • PipPip
  • 25 posts

Posted 10 October 2003 - 10:56 PM

To answer your last question.

Google has seemed to, for a long time, ordered backwards link searches by order of importance. There was a period where they did not, but now again it appears that they are.

#45 ChrisB

ChrisB

    HR 2

  • Active Members
  • PipPip
  • 25 posts

Posted 10 October 2003 - 11:05 PM

By the way.

Because of some incessant complaints from the peanut gallery I switched all the pages. The previous meta tag pages have had their tags removed and put on the previous non-meta tag pages.

Previously the pink dog and the black dog were ranked #1 and #2 in this search:

http://www.google.co...dog named Homer

A couple foolish people though that 2 out of 10 was statistically significant and that this proved Google used meta tags.

So I removed the meta tags. Guess what? The order is the same.

So what can we infer from this? Meta tags don't matter. Specifically because removing meta tags from some pages did not lower their ranks and adding them to other pages did not increase their ranks. I moved the meta tags all around and ranks did not change.

As to why Pink is #1? Who knows, but I know its not because of the presence or lack of meta tags.




SPAM FREE FORUM!
 
If you are just registering to spam,
don't bother. You will be wasting your
time as your spam will never see the
light of day!