SitePoint Sponsor

User Tag List

Page 1 of 3 123 LastLast
Results 1 to 25 of 63
  1. #1
    I <3 Internet Tekime's Avatar
    Join Date
    Dec 2003
    Location
    Maine
    Posts
    189
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Lightbulb PageRank Means Nothing (See for yourself...)

    I recently ran up a report on Google search referrals vs. PageRank on 37 different sites... these are up to the minute stats. Very interesting results. Note the complete lack of a relationship between Google search referrals and PR.

    Somebody please convince me that PageRank actually has anything to do with performance now.

    Last edited by Dan Grossman; Apr 17, 2009 at 18:44. Reason: URL snipped, you included the full chart here, linking to your site added nothing to the discussion
    Scriptalicious SEO Scripts
    Save 20% with coupon code SPROCKS


  2. #2
    SitePoint Wizard bronze trophy hooperman's Avatar
    Join Date
    Jan 2006
    Location
    Manchester, UK
    Posts
    4,301
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Nobody ever said that there was a relationship between PR and traffic. Nobody that knew what they were talking about anyway.

  3. #3
    I <3 Internet Tekime's Avatar
    Join Date
    Dec 2003
    Location
    Maine
    Posts
    189
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by hooperman View Post
    Nobody ever said that there was a relationship between PR and traffic. Nobody that knew what they were talking about anyway.
    Just to clarify - it's not PR and traffic - it's PR and Google search performance that I'm addressing here.

    I'm inclined to say I fully agree with you - but I think the misconception extends beyond just the newbie's. If it were so clear that PageRank doesn't mean good SERPs, I don't there there would be so much discussion of it.

    Edit: I should note that Sitepoint is one of the more "educated" communities; as opposed to certain other forums where the misconception is quite rampant.
    Scriptalicious SEO Scripts
    Save 20% with coupon code SPROCKS


  4. #4
    Follow Me On Twitter: @djg gold trophysilver trophybronze trophy Dan Grossman's Avatar
    Join Date
    Aug 2000
    Location
    Philadephia, PA
    Posts
    20,580
    Mentioned
    1 Post(s)
    Tagged
    0 Thread(s)
    While I am in complete agreement that there is no direct correlation between PR and SERP positions (although PR still informs many parts of the algorithm, so it's still important), this experiment doesn't show anything useful.

    The problem is that you examined the relationship between PR and referrals instead of SERP rankings. All your high PR sites could have #1 results for their target keywords, and none of your PR0 sites appear on page 1, and still get the result you showed. Simply make sure the high PR sites are targeting keywords fewer people search for than the low PR sites. Search volume would completely obscure the relationship between PR and SERP rankings if there was one.

    Traffic does not indicate anything about the effect, or non-effect, of PR on SERPs.

  5. #5
    SitePoint Enthusiast
    Join Date
    Apr 2009
    Location
    Ljubljana, Slovenia
    Posts
    36
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    I went to boston seo meetup and heard interesting story: somebody got lots of traffic just by selling some kids stuff....
    the idea behind seo is to find keywords that will produce hits. monitoring live on site is a good idea to quickly update it. one of the most searchable words is "web".... no comment on that.
    Pagerank will help you getting results when you are trying to sell something that everybody else does!
    Gregor Grajzar, web developer
    http://xweblabs.com
    http://grajzar.info

  6. #6
    Programming Since 1978 silver trophybronze trophy felgall's Avatar
    Join Date
    Sep 2005
    Location
    Sydney, NSW, Australia
    Posts
    16,788
    Mentioned
    25 Post(s)
    Tagged
    1 Thread(s)
    PR is ONE of the MILLION or so factors in SEO and has an appropriate level of affect on the result.
    Stephen J Chapman

    javascriptexample.net, Book Reviews, follow me on Twitter
    HTML Help, CSS Help, JavaScript Help, PHP/mySQL Help, blog
    <input name="html5" type="text" required pattern="^$">

  7. #7
    Follow Me On Twitter: @djg gold trophysilver trophybronze trophy Dan Grossman's Avatar
    Join Date
    Aug 2000
    Location
    Philadephia, PA
    Posts
    20,580
    Mentioned
    1 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by felgall View Post
    PR is ONE of the MILLION or so factors in SEO and has an appropriate level of affect on the result.
    I think you might be exaggerating the number of factors, and the level of affect, just a bit...

    A good resource on Google search algorithms, if you're interested, is the US Patent & Trademark Office. It may run behind the latest and greatest due to the delay between filing a patent and it being publicly viewable, but it's always better to learn from the source than all the bogus theories people come up with.

    Links are involved in virtually every major patent they hold on search algorithms. Let's take one of the better known examples, the so-called "Hilltop patent" from 2004.

    It's what initially introduced context and what some call "authority" (though I prefer not to use that word as people make too many assumptions based on it) to the reranking step. Rather than rerank the initial result set (simple scoring of page text against the search phrase, by position, density, etc.) based on PageRank alone, only involve those links which are deemed topically relevant to the search. Once that subset is identified, only those links participate in adding weight to pages in the initial result set, adjusting their scores for a (near) final ranking to display as the search results. But how do you weigh how much contribution each link in the identified subset offers to the candidate pages? PageRank would be an obvious choice.

    Hilltop also introduced methods for identifying affiliated sites; detecting networks of sites owned by the same organization, cross-linking, etc. It's what sparked the whole "get a separate class C IP address for every site!" craze 4-5 years ago.

    Now I'm sure much has changed since then, and there have been many newer patents, but the theme of using links to let the web define itself has not changed since Google's inception. Nobody's come up with anything better. It's still at the heart of not only Google but Yahoo! and Live.com as well.
    Last edited by Dan Grossman; Apr 18, 2009 at 01:54.

  8. #8
    I &lt;3 Internet Tekime's Avatar
    Join Date
    Dec 2003
    Location
    Maine
    Posts
    189
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by Dan Grossman View Post
    While I am in complete agreement that there is no direct correlation between PR and SERP positions (although PR still informs many parts of the algorithm, so it's still important), this experiment doesn't show anything useful.
    Dan, thanks for such a valuable post. I understand that having no involvement with these sites, or watching their performance every day (for as many as seven years in some cases), it's probably hard to extrapolate any value from the data here.

    If it were one or two sites I reported on, I'd agree that it was totally useless. I felt the size of the test group was at least large enough to give people an idea of what's happening on my sites, and is in my opinion is at least indicitive of a disparacy here between actual Google performance and the PageRank my sites are being assigned.

    I do hope PageRank still has some purpose in Google's algorithms aside from being a gimmick and publicity tool, but it's just not evident to me anymore. For all intents and purposes, PageRank means nothing to me now. If PageRank had any influence on my Google performance, then why does it go from 5, to 0, to 3, to 2, to 4, to 0 with each update with no observable impact on SERP rankings or Google search referrals? (Not even exaggerating here about the fluctuations on some of my sites).

    Furthermore, I have a few sites that are unequivacobly "flagship" sites. As in, first page positions for the most competitive terms in my market, a few hundred thousand backlinks (including many relevant PR6, PR7 links), DMOZ listings, etc. Yet, in the last year PageRank has plummeted to as low as PR3 (from PR 5/6) with sustained growth in traffic and improved SERP rankings within this time period. One site in particular has 2,000 google backlinks and PR 3. Another has 60 Google backlinks and PR 4. Just some more food for thought.

    The trends I have observed for a long time now tell me that if PageRank does indeed influence how a site performs, it's not influencing my sites. If it only affects how much "juice" it passes, then many of my sites have been made an exception. If the case is simply that my PageRank has been penalized for some reason, it doesn't seem to affect anything except that little number in the toolbar. Who knows, maybe the "penalties" are just to the public PageRank, and they're using the "real" PageRank number internally. A "scare tactic" (for lack of a better term..).

    The problem is that you examined the relationship between PR and referrals instead of SERP rankings.
    Without having verifiable data about search popularity and competition for each phrase being analyzed, this seems like an even less useful way to do it. If I could establish a reliable baseline and a method for determining this data, I'd love to expand on the tests. I obviously don't plan to share private performance statistics that involve commercial sites, which is an added challenge.

    All your high PR sites could have #1 results for their target keywords, and none of your PR0 sites appear on page 1, and still get the result you showed. Simply make sure the high PR sites are targeting keywords fewer people search for than the low PR sites. Search volume would completely obscure the relationship between PR and SERP rankings if there was one.
    They could, I suppose... but they don't. In fact it's quite the opposite in most cases.

    Perhaps I'm misreading your post, but it seems like you may be insinuating that the data was engineered in some way. It was not. The tests were run because I am deeply involved in developing a wide range of statistics analysis tools, and I've been seeing completely unexplainable PageRank behavior on my sites for too long.

    Traffic does not indicate anything about the effect, or non-effect, of PR on SERPs.
    I agree with this on a small scale. When you take a wide enough sample of data, I believe that obvious trends (or lack thereof) are at least indicative, and worth a second look.

    This is, inarguably, a first look for anybody but me.

    A good resource on Google search algorithms, if you're interested, is the US Patent & Trademark Office. It may run behind the latest and greatest due to the delay between filing a patent and it being publicly viewable, but it's always better to learn from the source than all the bogus theories people come up with.
    Just food for thought - companies also patent thousands of ideas that are never implemented, and many ideas are implemented which are never patented. Usually the most important ideas are the ones that don't get patented - the trade secrets.

    Not to say we should defer to bogus theories from the blogosphere, of course.

    Now I'm sure much has changed since then, and there have been many newer patents, but the theme of using links to let the web define itself has not changed since Google's inception. Nobody's come up with anything better. It's still at the heart of not only Google but Yahoo! and Live.com as well.
    This is what I feel is being challenged. If PageRank is no longer derived purely from an algorithmic analysis of links, it is no longer fair to all sites, which makes it unreliable. I believe that this alone should discourage people from using PageRank as a basis for performance, or for how a backlink will influence their performance in Google SERPs.
    Scriptalicious SEO Scripts
    Save 20% with coupon code SPROCKS


  9. #9
    Follow Me On Twitter: @djg gold trophysilver trophybronze trophy Dan Grossman's Avatar
    Join Date
    Aug 2000
    Location
    Philadephia, PA
    Posts
    20,580
    Mentioned
    1 Post(s)
    Tagged
    0 Thread(s)
    What you see and call PageRank is not PageRank, and any effect a change in PR would have on your SERPs would occur weeks to months before you saw it on the toolbar. The evidence that links are still the driving factor in determining SERPs is overwhelming to those that actually observe results.

    Put up a new site, throw some links to it from totally unrelated sites, watch it go nowhere.

    Put up a new site, do something which attracts a couple dozen topically relevant blogs and sites to link to the page, watch it jump to the top of SERPs for phrases whose anchor text was used often in those links and is topically relevant to the pages those links are on.

    The effect would happen in just days, as Google's index and associated metadata like PR are continually updated, while the toolbar would show PR0 for 3-6 months. PR is not something you can observe or test by looking at the useless, dated abstraction Google hands you. It's only observed by monitoring linking activity and changes in SERP positions.

    The PageRank reported by the Google toolbar is a grand distraction to keep most people from doing any work that would actually effect the SERPs. It's out there, despite its uselessness in SEO work, to deceive people into believing links don't matter. Google's successfully keeping you from manipulating their SERPs with this simple trick.

  10. #10
    SitePoint Member
    Join Date
    Apr 2009
    Posts
    14
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Saying "Google PR means nothing" is a bit to the extreme.
    However, I agree that Google PR is not that important, far from the most important.

  11. #11
    SitePoint Zealot busylinks1's Avatar
    Join Date
    Nov 2008
    Posts
    160
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Simply saying, Pagerank is just a factor to determin actual SERP along with other nearly 200 factors...
    If you other factors are at weaker side then pagrank do likely nothing for imroving you SERP....

  12. #12
    SitePoint Enthusiast
    Join Date
    Apr 2007
    Location
    inside the modem
    Posts
    99
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    are all these websites in the SAME EXACT NICHE ? if not then it doesnt take a rocket scientist to figure out that traffic will vary regardless of PR

  13. #13
    SitePoint Guru
    Join Date
    Feb 2008
    Posts
    655
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    It's also much too generalised to say "PageRank means nothing." If my sites didn't have the toolbar PageRank they have, I wouldn't be able to sell text links for as much as I do.

  14. #14
    SitePoint Member
    Join Date
    Apr 2009
    Posts
    9
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by corbyboy View Post
    It's also much too generalised to say "PageRank means nothing." If my sites didn't have the toolbar PageRank they have, I wouldn't be able to sell text links for as much as I do.
    The idea is that pake rank doesn't matter for serp

  15. #15
    SitePoint Addict
    Join Date
    Jan 2008
    Posts
    203
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    i have pr5 site with 120k visitors a day

    and a pr0 site with no pages in google index with 100k visitors a day


    if only webmasters concentrated on making useful sites for people not please googlebots ...

  16. #16
    SitePoint Guru
    Join Date
    Feb 2008
    Posts
    655
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by omidiu View Post
    The idea is that pake rank doesn't matter for serp
    Yes, I know. I was just pointing out that PR does have some uses, it's just not particularly relevant for rankings.

  17. #17
    SitePoint Enthusiast simmo_13's Avatar
    Join Date
    Apr 2009
    Posts
    74
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Might I add that I think your analysis of the numbers is rudimentary at best. A sample size of 37 hardly removes the possibility of natural variation.

  18. #18
    SitePoint Member jg123's Avatar
    Join Date
    Mar 2005
    Posts
    0
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    It seems that PR has little importance and I agree with this because my highest traffic/ranking sites have PR0

    PR is ONE of the MILLION or so factors in SEO and has an appropriate level of affect on the result.
    Then why does Gooogle have a spot for this value on their toolbar and Google mentions many times that PR is important?

  19. #19
    SitePoint Zealot russianzio's Avatar
    Join Date
    Nov 2007
    Posts
    109
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Who knows, maybe the "penalties" are just to the public PageRank, and they're using the "real" PageRank number internally.
    This is exactly the answer. I was sure about it for at least a year now and many SEO experts would agree with your assumption as well! Google has always had a secret file on PageRank even if they publicly stated how it worked. The green PR toolbar has been a useless toy for everyone for ages. Search results do not show PR and i think 99% of Google users do not have the feature in their toolbar anyways...

  20. #20
    SitePoint Wizard angilina's Avatar
    Join Date
    Jan 2008
    Posts
    1,011
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    PR is one of the 200 factors that Google consider while ranking a website. So I would say that PR is not worthless but at the same time, it is not everything. It is just one factors out of 200.

  21. #21
    SitePoint Addict kiduka's Avatar
    Join Date
    Mar 2008
    Location
    UK
    Posts
    265
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    I have seen websites with pr1 that rank above pr5 websites, google only looks for relevancy and backlinks

  22. #22
    SitePoint Member altopino's Avatar
    Join Date
    Apr 2009
    Location
    Montreal / New York
    Posts
    22
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    PR is not THE factor, it's part of the solution! Having a high PR has lots of advantages for a website! But having low PR does not me that the site with fail

  23. #23
    SitePoint Member
    Join Date
    Apr 2009
    Posts
    3
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Personally, I hate the PR thing. It's become so much of an obsession with webmasters lately that a lot are refusing to let -PR4 or 5 sites link to theirs!

  24. #24
    Non-Member
    Join Date
    Oct 2003
    Location
    Australia
    Posts
    1,478
    Mentioned
    2 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by Tekime View Post
    Somebody please convince me that PageRank actually has anything to do with performance now.
    Where did you get the information from that it does?

  25. #25
    Follow Me On Twitter: @djg gold trophysilver trophybronze trophy Dan Grossman's Avatar
    Join Date
    Aug 2000
    Location
    Philadephia, PA
    Posts
    20,580
    Mentioned
    1 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by cbp View Post
    Where did you get the information from that it does?
    It's easy to prove. Put up two identical pages, on some free host or something not associated with your other sites, with made-up words. Place several links to one of the pages from your existing sites, and only one link to the other page from one of the same pages the other links are on. Don't use any specific anchor text in the links.

    The one with more links will rank higher for the made-up words. Other factors are all equal, so PageRank still affects SERPs.

    Read my earlier posts for why I think it's not just there but relevant.

    And don't confuse PageRank with the toolbar icon.


Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •