| This 77 message thread spans 3 pages: 77 (  2 3 ) > > || |
|Does PageRank Affect Your Ranking in Google?|
| 1:10 pm on Feb 17, 2005 (gmt 0)|
Macro drew some responses to his post [webmasterworld.com] when he answered "How does page rank affect your listing in SERPs (Search Engine Result Pages)?" with "It doesn't". He qualifies that comment, but let's just follow that tangent.
In my view yes, PR has a direct influence on ranking for a given search phrase but the direct influence is very small indeed, it is not worth worrying about, it is not important. (Just to avoid confusion, I'm using "important" as definted in Cambridge/American Heritage/etc.)
There are secondary factors (e.g. people like to exchange links with high PR pages/sites) and there are related factors (e.g. sites that have high PageRank tend to have plenty of links from different sites with supporting anchor text).
Also, PageRank is useful for other things (e.g. crawl depth and frequency) but the direct influence on rankings of having more PageRank is, as Macro put it, "small enough a consideration to be ignored".
| 1:31 pm on Feb 17, 2005 (gmt 0)|
|Google uses PageRankô to examine the entire link structure of the web and determine which pages are most important ... |
... Important pages receive a higher PageRank and appear at the top of the search results. Google's technology uses the collective intelligence of the web to determine a page's importance.
So do you think that Google's continuing references to Page Rank as an important factor remain there to deliberately mislead us?
Do you think that an "ethical" company like Google would tell what amounts to such a whopper on their website?
| 1:43 pm on Feb 17, 2005 (gmt 0)|
ciml - I agree wholeheartedly - however LOL
I think the thing that most people just havent got down yet is that there is a difference between PR (displayed on the toolbar and in the Google directory) and real Page Rank.
As far as the toolbar PR (I call it little page rank) it is very useful for getting spidered more frequently (but I think you would get that with the real Page Rank anyway) and the pretty green bar does give a little indication that at least the site has been indexed (although that would also be easy to see with a site: command.
Real Page Rank - the basis of the entire Google engine and what really affects the SERPs, in my opinion, is a combination of those 100 or so other little factors that play into the algorithm and is what is transferred around the web through the linking structures.
| 1:56 pm on Feb 17, 2005 (gmt 0)|
Marval you beat me to it.
TBPR: I'm not interested in.
PageRankô to examine the entire link structure of the web and determine which pages are most important .. : I am interested in and what might be at play.
Is it 100+ factors? For example titles will only tell google what a page is about and this is specified by the owner, something which google would not want to infer too much trust. So I think we could discount this factor for starters.
What else? Anchor text has always been important to google for determining what a page is about.
Has there been a change of counting both incoming and outgoing links (instead of both together as before)?
[edited by: tantalus at 2:09 pm (utc) on Feb. 17, 2005]
| 2:04 pm on Feb 17, 2005 (gmt 0)|
Thanks ciml :)
BDW, the intention may not be to intentionally mislead but, given that their last disclosure caused everyone to chase PR and had a detrimental effect on SERPs quality, would they want to be so open as to disclose their current thinking? Or update information they've already provided? Even the act of taking it down could send a message to the masses.
Marval, most webmasters here have long abandoned thinking of tPR as the single determinant of position in SERPs. Perhaps Google have an aPR - or perhaps they don't - Allegra has thrown up a lot of interesting questions being debated all over the place. Even the "100 factors" may not* be applicable anymore. A heuristic system could obviate the need for relying on any factor beyond observing user behaviour. In theory.
* We simply do not know. It is unsafe to assume that just because they used 100 factors in the past they are still doing so now.
| 2:16 pm on Feb 17, 2005 (gmt 0)|
Macro - I tend to agree that some WMs have abandoned the tPR, however, there are loads of SEO companies out there that still advertise getting you a guaranteed PR level as well as SERPs positioning and I still see hundreds of posts a day across the web talking about link exchanges in terms of tPR.
I (again this is just my opinion based on a few years of watching Google) really believe that the original Page Rank which was made up of those 100 or so factors is still king - and I see the results of playing with those factors every day. I understand that some think that there are new factors that we don't fully know the meaning of, and there are the little tweaks to adjust the importance of each, but with all of the algo parts put together into Page Rank, I have yet to see a site that didn't do well with all of those factors in play. Granted it's not an easy thing to do in a short period of time, but over a year or two, all of the tweaking for those factors does pay off in huge returns.
One disclaimer - I don't do much in the way of short term "try to grab what you can and then waste the domain" stuff so I don't have the experience to comment on how to do well from that standpoint.
| 2:29 pm on Feb 17, 2005 (gmt 0)|
> continuing references to Page Rank as an important factor remain there to deliberately mislead us?
No, I don't think so. PageRank is an important factor, it's just the direct affect on ranking that is in dispute. Part of your quote goes further thoughm and could easily make someone thing that PR moves pages to the top:
|Important pages receive a higher PageRank and appear at the top of the search results |
Well, these two things both tend to be true. The causal link is now small though.
Keep in mind that Google (maybe Backrub?) originally ordered only by PageRank. Things have changed considerably since The Anatomy of a Large-Scale Hypertextual Web Search Engine [www-db.stanford.edu] and The PageRank Citation Ranking: Bringing Order to the Web [citeseer.ist.psu.edu] were written.
Also, from negligible to important is a sliding scale so it's easy to argue over the words. Many things can mislead, depending on interpretation.
> loads of SEO companies out there that still advertise getting you a guaranteed PR level
There are also plenty of "SEO companies" that submit to several thousand search engines. :-)
> original Page Rank which was made up of those 100 or so factors
I must disagree there, Marval. Google's implementation or PR may change (and often has, e.g. counting or not counting certain types of links) and Google's use of PR within the range of factors they use changes often. But, PageRank has a clear definition.
[edited by: ciml at 2:33 pm (utc) on Feb. 17, 2005]
| 4:33 pm on Feb 17, 2005 (gmt 0)|
Google says ...
|Important pages receive a higher PageRank and appear at the top of the search results |
CIML says ...
|Well, these two things both tend to be true. The causal link is now small though. |
CIML, I would say that the top statement is pretty unambiguous. As far as I can see they are categorically stating that there is a direct and significant link between PR and your position in the SERPs.
The quote below is the definition of PR from the toolbar help page.
|Gives an indication of the PageRankô for the page you're currently viewing. PageRank is the importance Google assigns to a page based on an automatic calculation that factors in the link structure of the web and many other variables. In order to automatically update this display for each page you visit, the Toolbar sends information about the page you are viewing to the Google servers. |
Once again this tells us quite clearly that PageRank is about the importance of the site. If PR is not now importnat one can only assume that Google is deliberately misleading their users. Surely not?
| 4:41 pm on Feb 17, 2005 (gmt 0)|
What you expect GG to say?
Yeah, sorry, we forgot to tell you guys. We've changed the algo. We don't use Page Rank anymore, we use this new system called GoogleRank. The formula for Googlerank will be posted here shortly.
Also, he he, back at the 'plex we think of the PageRank concept as funnier than, he he, Pigeon Rank [google.com]. Bird Brains!
| 5:16 pm on Feb 17, 2005 (gmt 0)|
> [...] I would say that the top statement is pretty unambiguous. As far as I can see they are categorically stating that there is a direct and significant link between PR and your position in the SERPs.
Then I think it's fair to say that our interpretations are different. How about "Billy is tall and has a blue jacket." :-)
> [...] tells us quite clearly that PageRank is about the importance of the site.
PageRank is about the importance of a page, not a site. It is absolutely the case now as it always has been. That is different from saying that PageRank is an important factor in Google rankings.
| 5:40 pm on Feb 17, 2005 (gmt 0)|
See message three [webmasterworld.com] on new G toolbar
| 5:49 pm on Feb 17, 2005 (gmt 0)|
Yes Herb, this is more proof that they are still pushing page rank as being an important factor. CIML I undertsand that people can read things differently but on this occasion I think that they are quite clearly stating that PR still affects your ranking.
If they don't intend us to believe this then they have a funny way of saying so ;)
If it is no longer important then if this is not a lie it is being economic with the truth.
Incidentally I am not saying one thing or the other. I have kind of lost the place and I just don't know anymore :(
| 6:06 pm on Feb 17, 2005 (gmt 0)|
> economic with the truth
If you mean that Google purposely hold back from telling us how they rank pages, then yes that is certainly the case.
I'm sorry if I gave the impression that PageRank is no longer about identifying important pages, I wouldn't want to help anyone think that.
| 6:51 pm on Feb 17, 2005 (gmt 0)|
What ciml brings up is an important paradigm shift that some, but not most have taken in their understanding of Google. The first significant move away from PR having a high coefficient in the ranking equation was AFAIK back in early 2003. It has gradually declined since then, and about a year ago was a smaller coefficient than the number of links with exact anchor text.
The coefficients, or knobs, or whatever change constantly and can often but not always overide one another. The constant question of what are the settings will always be with us and can be calculated as long as we know what the more calculation intensive parameters are.
PR is still calculated, and for all I know is still calculated the same way. What remains is what are the other complex variables, and are their coefficients large enough to waste your precious SEO efforts on. Variables such as "Eigentrust" or perhaps some LSI driven variable are clearly possible and potentially viable. The semantic driven possibilities are endless. Ultimately Google takes these complex variables, simplifies them, turns them into an index, and applies to the results with a coefficient. My guess is that all of these "complex variables" are calculated offline, are probably added all the time, and likey have a different team of people working on each.
I don't expect Google to add little green bars for each of them, however much I wish they would.
ciml: I have a blue jacket - I think you are on to something.
| 7:20 pm on Feb 17, 2005 (gmt 0)|
One of the problems when discussing the importance of PR in SERPs is many webmasters obsession with it.
PR is something that can easily and concretely be seen in the toolbar and something that webmasters feel they have some control over. Want to get your PR up from a 2 to a 5? Start a link building campaign, maybe change your internal linking structure, then sit and watch the green sneak up a notch every month or so. Hey, look what I did! It does give one a concrete sense of accomplishment. It's documented right there in the toolbar.
Then the next question comes: Why is this PR 3 page that was beating me in the SERPs when I was PR 2, still ranking better than my page? Shouldn't my PR 5 page trump that one?
Now the other parts to SEO aren't as concrete. What's the optimal keyword density? Well, it depends. How many links should I have on a page? Well, it depends. Etc, etc.
And we're now back to chasing PR again. It's a basic, relatively simple process that distracts many webmasters from working on all the other important factors.
| 7:23 pm on Feb 17, 2005 (gmt 0)|
There are many facts confirming the hypothesis, that Toolbar PageRank is different than actual PageRank used by Google in sorting SERPS.
Once I found even a Googler's statement, that Toolbar PR is nothing but an amusement thing. The statement has been denied by Google shortly, but there are also many opinions, that Toolbar shows the PageRank, for example, delayed for 6 months ago.
As we already know, that Google doesn't show all backlinks, perhaps Toolbar PageRank also shows incomplete and inaccurate information. Both ideas make the same sense - to make SEO work harder, to make wanna-be SEO attempts more ineffective. Because Google's purpose is to give user the best results, so any manipulation is against their interests.
I found some proofs for both (false tPR and false backlinks), after I'd launched a new website and linked it from some other sites I own for years. After some time, I found in G cache that G knows about these links. Of course, G cache is updated faster than G links index which is updated faster than PR and SERPS.
But there are months since then and link: tool still doesn't show anything, and the new site still has no PR in toolbar. But a month ago it started to be crawled much deeper and started to rank in some keywords.
It looks exactly like G updated its internal PR faster, and waited much longer to update tPR in order to make every manipulation more difficult. This theory fits better to the facts than saying that they completly abandoned using PR, and is more reasonable - PageRank works good, only the manipulation is a problem for them.
| 7:48 pm on Feb 17, 2005 (gmt 0)|
ciml - I was actually saying the same thing - att least in my mind LOL - the Page Rank does have a very precise definition - and yep - you are correct that it is for a page - not a web site.
I think where we disagree a little is in the calculation of that Page Rank. Long ago as you know, it was a pretty straightforward calc. - links in, etc. I believe that now Google has melded the Page Rank into the algorithm in such a way that it can actually become virtual and dimensional where other factors are played on the outside of it's sphere to re-shape it a little. These "distortions" are where I think we see most people stating that G has adjusted a knob for a factor - I really dont think in a mathematical world that would be the way things happen, more so it would be closer to reshaping the algo and adding in specific filters that are based on probabilities - not absolutes.
I hope I dont sound like Im wearing a tin hat or anything - it's just sometimes the mathematica comes out in me to try to figure cause and response instead of using emotions and hopes/feelings.
I do agree that of those 100s of factors - some like authority linking are so closely related to Page Rank that they have basically become a part of it - wheras others like W3C compliancy would be one of those outside modifiers.
| 7:49 pm on Feb 17, 2005 (gmt 0)|
I don't know how much PR helps or doesn't help, but I have a page with a PR of 7 that can't be found right now. Been that way for weeks. So in at least this instance, PR is not helping any. Lots of PR 2 and 3 sites listed above it, a lot of them are spam, and most of them link to my PR 7 page...
Last I looked, the page was #300 something in Google for its title, which is unique. The title also matches the domain in the URL. Clean site, no tricks. Something sure doesn't seem right about this!
| 7:50 pm on Feb 17, 2005 (gmt 0)|
Wizard - I think the Directory PR display for sites that are listed there is a little more accurate although understand it is NOT based on a log scale from 1-10 like the toolbar.
| 8:01 pm on Feb 17, 2005 (gmt 0)|
I was invited to take a look at this thread by sticky, only to find out that ciml already had pinned it down in the first post. However, there are some subtle points i can add, but that's basically semantics (pun, and intended. Not LSI).
For ranking purposes the PR value (be it the toolbar PR or the real "raw PR") is not the most important factor anymore. It is definitely not an unimportant factor however, it's just that it's being used differently now, and other factors can have greater effect on rank. PR2's can outrank PR7's - i bet most of us have seen examples of this during the past year. (Added: AndyA posted while i was writing this)
Also, there's a correlation between number and "quality" of inbound links and PR, let's not forget that. High PR pages tend to have high numbers of inbounds and/or high quality inbounds, and hence they tend to rank high. >> I'm sorry if I gave the impression that PageRank is no longer about identifying important pages
Agree again. High PR pages are "important pages" - they're just no longer per default the most important ones for every conceivable subject they choose to deal with. Analogy: The less educated sales clerk sometimes do know better than the professor. >> AFAIK back in early 2003
Freshbot boost was an eye-opener to a few of us i believe. Fresh was not the right thing to judge by, as Google learned, but the general idea was right. >> What remains is what are the other complex variables, and are their coefficients large enough
>> to waste your precious SEO efforts on.
Seriously, that sentence is so much on the money, i'm almost triggered to do a marathon post here about website life, the SE universe, and everything. I'm holding back a bit though, as this is already too long. A semantic twist (*lol* i love these puns) could be "are you able to influence these variables?"
Don't forget that the very reason that PR was such a good idea before links became a market commodity was that it's a measure that relates to a page without influence from the publisher of that specific page. These are the types of variables that Google (as well as the rest of the SE's) tend to prefer - for obvious reasons.
>> LocalRank (comment per sticky)
Yes. I have thoughts about this. This morning i also wrote a brief post to another thread but never pressed "Submit". I just can't find the right thread to post in as it relates to so many threads, not just this one. Now, did i fire six shots or only five?
| 8:14 pm on Feb 17, 2005 (gmt 0)|
|A semantic twist could be "are you able to influence these variables?" |
That's huge. Massive implications for the SEO industry. If the answer is "No" then it suggests they should pack up and go home.
Thanks for dropping in with your comments, claus. It's always interesting to read what people who do know about SEO (unlike me) think :)
| 8:18 pm on Feb 17, 2005 (gmt 0)|
I'm seeing results that look more like Yahoo and MSN in this regard. Very low PR sites, PR 3 for example, ranking above established PR 6,7, and even 8 recently. This is the first time I've seen that in Google to such an extreme.
I think Google is experimenting with a move away from PR beucae PR, while it was revolutionary in its day, it's getting stale and manipulate SERPs via buying and selling among other tactics. Google just seems to keep giving it less and less weight.
I have no idea why Google doesn't just turn it off. Does anyone have any idea what good it does other than to tempt folks to undermine Google, or worse, as a Google misinformation campaign to get webmasters to waste their time and money?
| 8:38 pm on Feb 17, 2005 (gmt 0)|
Contextual PageRank is what Google's using - It's been in place for close to 2 years in my opinion.
How Google scores your page's PageRank for a specific search changes depending on how much PageRank is flowing through the links that contain the searched upon keyword phrase in the anchor text of links pointing at your page.
Other links pointing at your page without the keywords in them are either irrelevant - or less relevant - for that specific search.
If you think about this, it makes sense.
So toolbar pagerank doesn't have much value.
But if Google is doing what I am suggesting here, then I think Contextual PageRank is a very significant factor in their algo.
| 9:12 pm on Feb 17, 2005 (gmt 0)|
|How Google scores your page's PageRank for a specific search changes depending on how much PageRank is flowing through the links that contain the searched upon keyword phrase in the anchor text of links pointing at your page. |
Hmmm, that would be somewhat similar to Teoma's 'communities' or whatever the heck they call it:
--Find a set of pages that satisfy the keyword search.
--Analyze the linking patterns among those pages/sites.
--Rank the pages according to 'authority'.
--Push the SERPs to the searcher.
Now, according to Teoma this is all done 'on the fly' for each request. I assume Google would be able to do the same with PR, though with its volume it would be on a more massive scale.
Interesting thought, egomaniac.
| 9:17 pm on Feb 17, 2005 (gmt 0)|
Page Rank is important, and more is better than less, but it's only one ingredient in the ranking recipe, and like any recipe adding more of just one ingredient might not be an improvement. It depends on the total balance.
For practical purposes, I find that if I create content and cultivate links based on RELEVANCE, Page Rank and other factors take care of themselves well enough to let me be respectably competitive.
I learn things from folks who study the fine points of algo manipulation, and I certainly keep them in mind as I work, but my over-riding focus is relevance, and the best user experience my limited technical skills can deliver.
It works for me.
| 9:46 pm on Feb 17, 2005 (gmt 0)|
"Anyone with any ideas on how the extent of PR's influence (or lack of influence) on SERPs can be best tested?"
Look at the serps.
This is such a simple issue, and reflects I think a critical difference in how people understand ranking on search enegines, specifically Google.
Pagerank is an enormous influence in ranking, but moreso for highly profitable, successful businesses, not the more marginal, throwaway low quality, no-original-content domains.
Pages with perhaps 100 PR4 to PR7 links, that have a PR6 or so, plus all allapproriate on page titles and text, will kick the butt, almost always, of PR1 or PR2 or PR0 pages that have 1000 blog-type links with approriate anchor text, etc. Search any major term, and low pagerank pages are nowehere to be found.
But as anyone reading webmasterworld for some time should know, mastering a wealth of search phrases is how you build a very successful business. Most people complaining about the serps tend to complain about ONE term where they dropped ranking or some number of junky sites moved above them.
In contrast, people who use pagerank wisely have dozens, hundreds or thousands of search terms that earn them income each day. While most searches are as pithy as a person can make them (one word, maybe two), any niche will also generate many multiword queries that would NEVER be in the link text to a page, or in the order of the words in the page title. Quite simply, this is where good pagerank on multiple pages wins, always.
GoogleGuy and others have mentioned the basic concept here many times. Breadth and volume of search terms is to be sought after and this comes from a mastery of pagerank, and more basically, the concept behind pagerank.
Maybe half my visitors come from about a dozen search terms, and link text is the key player there. But the other half comes from thousands of queries each day, where link text doesn't matter at all, and pagerank contributes considerably.
Asking if pagerank effects ranking is self evident. Put the word xysterdyust on this page, and put it on the webmasterworld main page and see which ranks higher. Unless people start linking to this page with xysterdyust text, or it gets in the title of the page, pagerank will decide that the main page will be the first result displayed. In this case xysterdyust has no value, but almost all searchers coming to my sites make me a buck.
Pagerank is just one building block, but it is one found in the foundation of a business.
| 12:46 am on Feb 18, 2005 (gmt 0)|
Interesting comments steveb, thanks a lot. It seems like we're noticing the exact opposite things. This is fascinating, as i think we both look at a lot of SERPS, but it doesn't really make any sense unless we can somehow agree on what it is that separates one class of observations from another.
|any niche will also generate many multiword queries that would NEVER be in the link text to a page, or in the order of the words in the page title. Quite simply, this is where good pagerank on multiple pages wins, always. |
These are some of the cases where i no longer see inferior pages from huge domains trump relevant pages from smaller domains. (Huge/small in terms of pages as well as PR). It's certainly interesting, though - you seem to see a PR revival while i see less importance on PR. There must be some kind of differences in the searches that we've done - or at least those we remember.
Personally, i'll have to make more searches as i'm convinced that i've seen what i've seen, but i've made a few searches after reading your post that actually support it - so where, exacly, is that separating line? Is it high-volume vs. low volume searches, or is it specialist searches vs. general searches, or "competitive" areas vs. less competitive ones - language/country/industry specific, or commercial/informational - any thoughts?
| 12:57 am on Feb 18, 2005 (gmt 0)|
>> Pagerank is just one building block, but it is one found in the foundation of a business.
It may be that there is a higher weight put on other complex paramters, but the existance of some of these parameters may be predicated on the fact that you have links from other sites (perhaps a lot of links, perhaps not) and it is difficult to be embedded in a complex link map without gaining some PR.
There are still many, many keywords, as steve points out, that can be had from PR alone. As the web, competition, and G's processing power grows I suspect that this will slowly shrink.
| 1:33 am on Feb 18, 2005 (gmt 0)|
It's what I said, where no or almost no link text or matching titling exists.
What they heck do you think they rank on when you just type a word like someone's obscure name?
Of course it's not just that, the peculiar notion that pagerank doesn't matter just dies here
Gee, where are all the PR2 pages?
| 2:19 am on Feb 18, 2005 (gmt 0)|
steveb, clever search. But I'm not sure it proves the point. A search on the kw 'www' is about as close to a straight math calc as you'll get with any search term. There is almost no optimization affecting this search. All that the algo has to go on are the basics. The incremental ranking "scores" separating any two consecutive result pages must be very small, and it stands to reason that sheer presence would win out.
The question is, what happens when you add in, for a given set of SERP's for a competitive two word phrase, the following: Title, META, Hx, kw density, kw proximity, linguistic evaluation including semantically related kw presense and density, internal backlink anchor text assessment, external backlink anchor text assessment, backlink page assessment, authority-like and hub-like references, and so on.
I could not put a number to it, but as a sheer guess, I'd say a very savvy webmaster could get a PR5 page to rank above a seemingly similar PR6 page...perhaps even without much difficulty.
Hey, wait a minute. I've done that. I guess that means one doesn't even need to be that savvy. ;-)
| This 77 message thread spans 3 pages: 77 (  2 3 ) > > |