• http://www.SEOcritique.com SEOcritique.com

    If only this were required reading before anyone could register to post on an SEO forum…

  • http://www.searchenginemarketinglead.com Tilak

    Very nice write up.

    I have a question, does there any impact of ranking if Google more regularly revisit the pages/site?

  • http://www.mattcutts.com/blog/ Matt Cutts

    Danny, nice in-depth write-up. Just some small things that I would clarify:
    – “You had to manually choose to do it [turn on PageRank], and it was mostly search engine optimization people that did so.” That’s not accurate, in my experience. Certainly many SEOs turn on PageRank, but a *lot* of regular people choose to opt-in to PageRank as well. I believe the quantity of people that opt-in to see PageRank is much larger than the set of active SEOs.
    – “PageRank is only a score that represents the importance of a page.” I’d just clarify that we consider it Google’s opinion of the importance of a page.
    – “poured” -> “pored”. Okay, that’s probably more detail than you wanted. I’ll stop. :)

    Nice reference for folks. I agree that getting everyone to read this before joining an SEO forum would be a Good Thing. :)

  • http://seonirupam.blogspot.com/ MrRoy

    Hi Danny

    Excellent post. Gr8 stuff

    I have a very simple question for you.

    Say for an example, If my Credit Card website gets a link from New York Times, say they gave me a link from a page which is not relevant to my website, do you think that link will have much more importance as far as anchor text is concerned in comparison to 20 links that i got from relevant websites ?


  • http://www.bodyabcs.com Purposeinc – Dr. David Klein

    As far as you know is this the prioritization of the value of links? 1. most valuable 10. least valuable

    1. link from site about fruit, and page about grapes, to a page on your site about grapes when your site is about fruit.

    2. link from site about fruit, and page about boxes, to your page about boxes, on your site about cardboard.

    3. link from site about fruit and page about boxes, to your page about televisions on your site about electronics.

    Any other data you have picked up about this would be very helpful. I apologize if you have yacked about this before, I am still coming up to speed in the field :)

  • http://www.feedthebot.com feedthebot

    this post is very feedthebot-ish i like it. Thank you for writing that up.

    Of course, if we are citing Google, it would be a shame not to mention this often overlooked Google librarian article that helps explain the basics of ranking. I bet even Danny hasn’t read this…

    How does Google collect and rank results?

  • http://www.feedthebot.com feedthebot

    Actually I should have said that better. There is a rather old article on Google’s website citing Matt Cutts named
    How does Google collect and rank results? it is unique because it is Google stating clearly (a long time ago) that Pagerank isn’t the biggest factor in ranking.
    to quote it…

    “As a rule, Google tries to find pages that are both reputable and relevant. If two pages appear to have roughly the same amount of information matching a given query, we’ll usually try to pick the page that more trusted websites have chosen to link to. Still, we’ll often elevate a page with fewer links or lower PageRank if other signals suggest that the page is more relevant. For example, a web page dedicated entirely to the civil war is often more useful than an article that mentions the civil war in passing, even if the article is part of a reputable site such as Time.com.”

    I found this, as I find all sorts of other juicy bits in Google documentation using the Ask Google tool on feedthebot.com.

  • http://www.searchmarketingtrends.com chris boggs

    wow Danny well done…dugg this and will certainly bookmark.

  • chris boggs

    lol Matt the spelling police. Is “poured” the UK version?

  • http://sethf.com/ Seth Finkelstein

    Any Kreminology regarding what the update to the Technology page means, if it means anything at all?
    Perhaps quasi-official confirmation that PageRank has dropped significantly in terms of weighting of factors?

  • http://searchengineland.com Danny Sullivan

    Tilak, regular revisits alone shouldn’t impact your ranking that much. Many do feel that these days, Google does consider freshness as one of those many factors for ranking. If your pages are regularly updated, that might add a bit to the rankings. But I wouldn’t start changing things just to be pretending your fresh. Focus on your content. Good content, you’ll gain PageRank, that will bring you regular revisits but more important help your site stand out as an authority.

    Matt, thanks for all the comments. Fixed the typo (I am American, after all) and incorporated some of what you said into the story above, now.

    MrRoy, the first thing with any link is generally the anchor text. Are the words you’re hoping to be found for in it? If so, that one link from an authority site like the NY Times may indeed count more than 20 links that don’t use the words from elsewhere. Google and other search engines more and more do seem to be trying to assess the overall “relevance” of a particular page, rather than just look at a link’s anchor text. But still, anchor text remains important.

    David, you’re already over analyzing far too much for someone who is coming up to speed. Stop! Seriously, I mean well with that. :) Don’t worry which links are the best links, in the way you want. Think like this.

    Where’s your audience going? If they’re interested in grapes, search for “grapes” on Google. The top pages listed are where they’re going, because many of them will search for grapes and show up on these pages. So those are the pages you want links from.

    First, you want links from those pages because if your audience clicks from Google to them, they may then see you and click over to your site. Second, Google just told you these are important pages on grapes. So that importance, both PageRank and contextualy, could get passed along to you.

    Feedthebot — ha! Not read the article. Yep, and I blogged about it to the world when it first came out. It remains excellent. Thanks for the reminder. I’ve incorporated it into the story.

  • Jeremy Chatfield

    Superb article. Thanks.

    You’ve pointed to the idea that Google is adding more and more techniques to evade untrustworthy data sources. This is directly analagous to the experience of many programmers… Code gets more and more complex until the burden of understanding the exceptions to the main flow makes the whole system collapse.

    Doesn’t the baroque ornamentation of filtering based on speed of link addition, the assumption that word matching is equivalent to conceptual matching, etc, etc… suggest that this system is likely to fail in the face of increasingly sophisticated automated link building methods?

    The model is based on information, but has been used to drive the economy… Economic value of pages doesn’t appear to factor. Like it or not, Search Engines dominate new traffic to many sites. Occasional bleating from Google about why businesses have to put more information on the web in order to rank well, is putting the cart before the horse. The value to the searcher is the economic value of the business for their needs, not the quantity of web pages and links the business has developed.

    The whole basis is flawed, for a system that influences so much spending. Isn’t it? Doesn’t seeing CTR of 50, 60% or even more on paid search, suggest that, at least for certain classes of search, the organic results are economically less relevant than the paid search results?

    Pages with more links and more information on them aren’t intrinsically economically more important. They may be more important to a researcher, though… it’s the intent that counts.

    Hmm. More to think about… Thanks.

  • http://searchengineland.com Danny Sullivan

    Seth, I think the changes just reflect Google catching up with the times. When they originally launched, PageRank was an easy way to get buzz on being different. It was never the only way they ranked pages, but it sounded a heck of a lot cooler than saying “we do link analysis.” So that page never really properly explained the Google technology, as far as I was concerned.

    In fact, during the big Florida update crisis of November 2003, I did a mini-rant about this:

    Q. Does this mean Google no longer uses the PageRank algorithm?

    A. Google never used the PageRank algorithm to rank web pages. PageRank is simply a component of that overall algorithm, a system Google uses to measure how important a page is based on links to it. It has always — ALWAYS — been the case that the context of links to the page was also considered, as well as the content on the page itself.

    Unfortunately, some writing about Google have called its system of ranking PageRank, and Google itself sometimes makes this mistake, as seen in its webmaster’s information page:

    “The method by which we find pages and rank them as search results is determined by the PageRank technology developed by our founders, Larry Page and Sergey Brin.

    In reality, the page describing Google’s technology more accurately puts PageRank at the “heart” of the overall system, rather than giving the system that overall name.

    By the way, PageRank has never been the factor that beats all others. It has been and continues to be the case that a page with low PageRank might get ranked higher than another page. Search for books, and if you have the PageRank meter switched on in the Google Toolbar, you’ll see how the third-ranked Online Books Page with a PageRank of 8 comes above O’Reilly, even though O’Reilly has a PageRank of 9. That’s just one quick example, but I’ve seen others exactly like this in the past, and you can see plenty first-hand by checking yourself.”

    As I said, the tech page did talk about PageRank at the “heart” of things, but I suppose you’re seeing Google trying to ease people think they’re all about PageRank.

    Actually, that is kind of a bigger move if you want to get all into it. I mean, the Google Web History story I referenced in the article talks at depth about Google moving to use personal data to influence search results. What Google does NOT want happening is for people to think they only are about links (AKA, PageRank) and so aren’t keeping up with the times as links get bought, sold, traded influenced and more important, may not reflect the big users ratings wave going on.

    I doubt many people themselves would think this, ordinary searchers — but Google competitors would try to play it up. For example, in the enterprise space, I continue to roll my eyes when some enterprise search company tries to make a clueless reporter believe that Google can’t do enterprise search since it can’t count links there. Google, of course, has excellent basic search engine textual analysis capabilities. But reporters gobble that “weakness” up.

  • http://pardonmyfrench.typepad.com Eric Frenchman

    This is great thanks Danny. Quick question. I have both Alexa and Google toolbars loaded in firefox and when I visit popular blogs sometimes I notice a very high (popular) Alexa ranking and a zero page rank. How do you explain that other than Alexa’s ratings are dubious at best?


  • http://sethf.com/anticensorware/ Seth Finkelstein

    Thanks, Danny (by the way, I should have said great article too), I was thinking along technical lines, but your explanation of the public-relations aspect makes sense.

    I’ve actually been seeing aspects of the move from other angles. PayPerP*st has made keyword-buying a commodity, and the user-rating aspect is something I suspect the (so far, vaporware) Wikipedia-like search engine wants to exploit.

  • http://www.daviddalka.com/createvalue/ DavidDalka


  • http://www.feedthebot.com feedthebot

    Danny, I never said you didn’t blog about it. :) oops.

  • http://www.seo-theory.com/ Michael Martinez

    You had me nodding in agreement right up to the point where you linked to Phil Craven’s travesty of an explanation for PageRank. That is one of the absolute worst treatments of the subject ever published and is so misleading I am surprised to see you recommend it.

  • http://searchengineland.com Danny Sullivan

    Eric, Alexa’s rankings are based on people who visit you and have the Alexa toolbar. Google’s rankings come from counting up link votes. Two compeltely different methods, so you just can’t compare them. Perhaps, for example, lots of people come to a page but for whatever reason, there are few links pointing at it. Alexa would say it was great; Google, not so much.

    Feedthebot, just playing with you. For the record, I have both read and blogged about it :)

    Michael, Phil’s papers is one of the earliest out there trying to play with the math and is notable from that account. Didn’t say it was correct, only that some might want to check it out. The bigger point was that I was more strongly saying that none of these types of analysis are that useful in my view, working off a paper from 1998.

  • http://www.seo-theory.com/ Michael Martinez

    Danny: “The bigger point was that I was more strongly saying that none of these types of analysis are that useful in my view, working off a paper from 1998.”

    Fair enough.

  • http://www.mattcutts.com/blog/ Matt Cutts

    Danny, many thanks for adding those extra details to the article. This is a great resource, and judging from the comments already, it looks like this will be a popular one. :)

  • MrRoy


    Thanks for the reply.
    Keep up your good work :)


  • http://www.hobo-web.co.uk/ Sandpetra

    How long did that take you ? :)

  • Jill

    Slightly off topic, but I’m curious as to why you no-followed the link to Google, but no other links?

  • http://searchengineland.com Danny Sullivan

    That nofollow was on a link to Google Librarian Central, and I think it had nofollow because I copied it out of a comment, where all the links get nofollows. I didn’t notice it was there, in short. I removed that now.

  • http://www.seo4fun.com/blog/ Halfdeck

    “High PageRank does NOT guarantee a high search ranking for any particular term. If it did, then PR10 sites like Adobe would always show up for any search you do. They don’t.”

    I ran a relevance VS importance test here:


    My conclusion: when relevance (as determined by authoritative/trusted IBL anchor text) is low, PageRank/domain authority comes into play. When relevance is high, it leaves other signals in the dust.

  • http://seonirupam.blogspot.com/ MrRoy

    Hi Danny

    Your article came at a time when the toolbar PR update has just happened. Great coincidence. :)


  • http://www.inLoughborough.com Loughborough

    A nice summary of how page rank works. Presumably though it is a moving target as people try to optimise their site to get ahead of the competition.

  • http://learnonly.com/ Dr Umesh R Bilagi

    thank you 

  • seo2012

    top seo tips