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This topic has been archived.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

#1 Jill

Jill

    Recovering SEO

  • Admin
  • 33,244 posts

Posted 28 July 2003 - 12:08 AM

Interesting article by Chris Beasley, where he tested out some SEO theories.

Truth, Lies, and Search Engines

Jill

#2 markymark

markymark

    HR 4

  • Banned
  • PipPipPipPip
  • 143 posts

Posted 28 July 2003 - 04:56 AM

Hmm...not sure about that article, particularly the stuff on outbound links. It's not especially clear what his conclusion is.

Some questions spring to mind:
i) Were all the pages identical in terms of content, use of tags, etc. or did they just have the same 'keyword density' ?

If not, the test is skewed from the start.

ii) When he says they all had the same 'keyword density' - is he including in this equation the keyword rich anchor text for the one page that ranked higher.

If not, all he has proved is that the page with the more mentions of the targeted keyphrase ranks higher. Nothing to do with outbound links, anchor text or anything else.

Personally, I would like to have seen the pages so I could assess for myself the veracity of what he's trying to prove.

#3 qwerty

qwerty

    HR 10

  • Moderator
  • 8,695 posts

Posted 28 July 2003 - 08:35 AM

I can appreciate his analogy to fortune tellers. It's true that a large part of the SEO's job seems to involve convincing the potential customers that what we believe about the way SEs work is correct. Only a few of us have a reputation that's developed to the point where a customer is going to just tell us right at the start of our relationship with them, "do what you think is best".

However, I continue to have a problem with people who insist that SEO is all about figuring out mathematical formulae. IMO, it's not. Yes, it's important for us to have a good understanding of what's important on a page -- where are the important places to put the right content. But SEO, in my mind, is much more about a more general idea of making a page better. That's much more nebulous than plugging your calculations of the weighting of particular tags into an algo, but it still makes more sense. Taken to an extreme, my method creates a well-written, informative page, and the other creates an incomprehensible machine-written page. Obviously, it's a bad idea to go to extremes either way, but I continue to believe that content is what matters most.

The author refers to his data as "final proof," and I think that's his biggest mistake. Whether or not one agrees with his conclusions, they're certainly not final. Ranking algos change all the time.

#4 Scottie

Scottie

    Psycho Mom

  • Admin
  • 6,294 posts

Posted 28 July 2003 - 01:40 PM

I'll have to toss in here that I believe in "optimization". Period.

User optimization, search engine optimization, design optimization, page load optimization... A web professional who can manage all of the above is the future of the indusrty; creating an excellent site that ranks well, converts well, and has excellent ROI.

You can "optimize" a site for search engines... but if no one hangs around or buys anything, you are paying for traffic that is useless, IMO.

#5 peter_d

peter_d

    HR 5

  • Active Members
  • PipPipPipPipPip
  • 287 posts

Posted 28 July 2003 - 07:01 PM

There's those that believe marketing is a science. I feel it has more to do with art.

Science

You can test theories, but without knowing that the search engine algo remains static during the test the results are still debateable. And if they are one thing today, they might be another thing tomorrow. General patterns are of value, however.

Art

Example: Outbound links do help ranking, but possibly not in the way seos think of them. One way they help is that they initiate awareness. Think referrer logs. Think karma. People may then link to you because they see value in what you do after having become aware of you. Google looks at those links. The circle completes.

Link forward :thumbup:

As Scottie says, ranking and traffic aren't good metrics in the sem game. User action and conversion are important metrics. That is some science and a lot of art :)

Edited by peter_d, 28 July 2003 - 07:04 PM.


#6 ChrisB

ChrisB

    HR 2

  • Active Members
  • PipPip
  • 25 posts

Posted 29 July 2003 - 09:01 AM

ii) When he says they all had the same 'keyword density' - is he including in this equation the keyword rich anchor text for the one page that ranked higher.


Yes, if you copy and pasted each rendered page into Notepad they'd look identical. One page has anchor tags around an instance of keyword rich text, one page has anchor tags around an instance of generic text. All the pages have the exact same text though.

If not, all he has proved is that the page with the more mentions of the targeted keyphrase ranks higher. Nothing to do with outbound links, anchor text or anything else.

Personally, I would like to have seen the pages so I could assess for myself the veracity of what he's trying to prove.


Actually there were three trials. One with keyword rich anchor text, one with generic anchor text, and one with no anchor text. The one with generic anchor text did not outrank the one with no anchor text. Ergo outgoing links do diddly for you. I wonder how closely you read my article since I not only mention this but right at the top I link to the site where you can see the experiment in action.

The author refers to his data as "final proof," and I think that's his biggest mistake. Whether or not one agrees with his conclusions, they're certainly not final. Ranking algos change all the time.


Yes, algorithms can change. But for now those results are final. Arguing would just be an act of desperation. Claiming that "that was yesterday, today could be different, so I'll still advise people that outgoing links/meta tags help" is also an act of desperation. As long as no one links to my experiment site inappropriately then the experiments will stay valid as search engines continue to update and so with each passing update it will be possible to test for an algorithm change.

Of course the whole reason I made the site public is because I know how people in the SEO industry hate being told how things work when they all know that they know what the truth really is. The site explains how to do the experiments yourself, I link to the site right in the article (and for convenience sake its http://www.searchenginelabs.com )I want you to go there and draw your own conclusions. You'll find its pretty cut and dry, but you'll find that out for yourself, which should be easier to stomach than having it told to you.

Edited by ChrisB, 29 July 2003 - 09:03 AM.


#7 qwerty

qwerty

    HR 10

  • Moderator
  • 8,695 posts

Posted 29 July 2003 - 09:56 AM

Yes, algorithms can change. But for now those results are final. Arguing would just be an act of desperation.

I'm not trying to get into an argument on semantics, but you're basically saying that this is permanent for the time being.

My main disagreement with your article is simply that I feel that no matter how much we may know about the algorithm, it's more important to me that a page be relevant to the user. No matter how easy we make it for the user to find the page, we need to give them something to appreciate ON the page.

So I'm not questioning your experiment. AFAIK, your methods are valid and so are your data. It's just not what I consider to be the most important part of my job.

#8 ChrisB

ChrisB

    HR 2

  • Active Members
  • PipPip
  • 25 posts

Posted 29 July 2003 - 10:03 AM

saying that this is permanent for the time being.


Oxymoron. Actually I'm saying that this is factual for the time being.

My main disagreement with your article is simply that I feel that no matter how much we may know about the algorithm, it's more important to me that a page be relevant to the user.


Sure, I argee, but the article is a SEO article. I don't consider making a quality site to fall under the auspices of SEO. I advocate making a quality site, and if someone wants a successful website I tell them that above all else then need a good design, their own domain, and quality content. Sure the practice of building quality websites does have some traits in common with SEO, but that doesn't make them the same discipline.

The article isn't there to give advice even, its not there to tell you how to do SEO, its there to publicize an experiment I did on two issues within the SEO field.

So you say you don't like it because you feel a page has to be relevant to the user, I ask and what does that have to do with the price of tea in China?

#9 qwerty

qwerty

    HR 10

  • Moderator
  • 8,695 posts

Posted 29 July 2003 - 10:06 AM

I don't consider making a quality site to fall under the auspices of SEO.

That is where we disagree. I think that a huge part of SEO is about content. The proper use of tags and keyword density is another. And I know the "permanent for the time being" thing is an oxymoron. That was my point.

#10 Jill

Jill

    Recovering SEO

  • Admin
  • 33,244 posts

Posted 29 July 2003 - 10:13 AM

Welcome, ChrisB! :aloha:

Thanks for stopping by to clarify.

I think the article was great for what it was able to show. Obviously, it doesn't tell someone all the ins and outs of creating a good site, or anything like that. It wasn't supposed to!

And I totally agree with the conclusion that outbound links are not part of the ranking algorithm. As others have stated here very clearly, they do have a side benefit, which is why they are always good to use and should never be hidden from people or search engines. :D

Jill

#11 markymark

markymark

    HR 4

  • Banned
  • PipPipPipPip
  • 143 posts

Posted 29 July 2003 - 10:23 AM

I wonder how closely you read my article since I not only mention this but right at the top I link to the site where you can see the experiment in action.

I read it closely enough to have noticed that link and also to have noticed that while you said that was where you were running the tests, you made no mention that you could still see the experiments in action. Maybe I should have clicked through anyway, though :aloha:

Anyway, now that I've seen the stuff, I find it hard to disagree with your conclusions. I would like to, of course, just to be argumentative.

Your conclusion that keyword rich anchor text on a page gives greater weight to a search term is an interesting one, though. Given that - in many cases - anchor text says nothing at all about a page itself but rather the page it points to, I would have thought it equally logical to give less weight to anchor text when assessing on-page factors. A good example would be text navigation on a site. Often, the text navigation will be the same on all pages and the words in the anchor may give little clue to the content of the page they reside on.

#12 ChrisB

ChrisB

    HR 2

  • Active Members
  • PipPip
  • 25 posts

Posted 29 July 2003 - 10:25 AM

Thanks Jill. You know... I think I like you now. :aloha:

That is where we disagree. I think that a huge part of SEO is about content. The proper use of tags and keyword density is another.


Sure, I do too, but its not the same thing. You have make a wonderful site with a great design and pages and pages of quality content and you can perform horribly in the search engines. Likewise you can take a poor site off Geocities and if you know what you're doing you can make it perform very well in the search engines.

Having a quality site and knowing SEO are both part of building a successful website, but they are not the same thing. In fact the biggest way that a quality website will play a role in your search engine rank is an indirect one. That is the better your site the more third party incoming links you will accumulate.

But like I said, the article wasn't meant to cover any of that. SEO is far more than meta tags and outgoing links.

I've got hundreds of pages of articles on my site that cover SEO and building a successful website in a more general sense. In fact my building a successful website section started off as a book and one of the section headings in it was "Quality Brings Quantity" or the higher quality your site is the more traffic you will get.

So we agree, just not semantically. I also don't consider PPC advertising to be SEO, but search engine marketing (and with Adsense and Overture's distribution to MSN content sites and whatnot its not even strictly search engine marketing anymore.)

Again though, the topic isn't really related to my article. The article is not about SEO in general.

#13 ChrisB

ChrisB

    HR 2

  • Active Members
  • PipPip
  • 25 posts

Posted 29 July 2003 - 10:32 AM

Your conclusion that keyword rich anchor text on a page gives greater weight to a search term is an interesting one, though. Given that - in many cases - anchor text says nothing at all about a page itself but rather the page it points to, I would have thought it equally logical to give less weight to anchor text when assessing on-page factors. A good example would be text navigation on a site. Often, the text navigation will be the same on all pages and the words in the anchor may give little clue to the content of the page they reside on.


Oh I agree. I don't think its very intuitive of Google to do that. I wasn't giving my opinion of how things should be though, I was simply reporting on how things are.

Though I am glad that my personal theories (that meta tags aren't used and that outgoing links don't help) were vindicated by the experiment. I wrote an article some time ago about common Google myths and while many people whom I consider knowledgable, like Jill, thought it was right on lots of people thought it was full of inaccuracies (especially around those two topics). I got into some really long arguments with people who believes outgoing links or meta tags helped.

#14 markymark

markymark

    HR 4

  • Banned
  • PipPipPipPip
  • 143 posts

Posted 29 July 2003 - 10:43 AM

Yeah, I understand that you were only reporting how things are. I was looking to move this thread into a discussion of why you or anyone else reading might think they do this and whether a better algo would actually give lesser weight to anchor text when assessing on-page factors.

#15 Alan Perkins

Alan Perkins

    Token male admin

  • Admin
  • 1,648 posts

Posted 29 July 2003 - 11:09 AM

I was looking to move this thread into a discussion of why you or anyone else reading might think they do this and whether a better algo would actually give lesser weight to anchor text when assessing on-page factors.

A different algo, or even the same algo in different circumstances, might...

Most link-pop algos contain a mix of "authority" and "hub" factors and the determination of whether a hub is more relevant than an authority, or vice versa, is certainly open to across-the-board tweaks over time but is also, IMO, open to being made query-dependent.




We are now a read-only forum.
 
No new posts or registrations allowed.