By Matthew Magain

Google To Abandon PageRank?

By Matthew Magain

The Googlified blog reports that Google Labs are experimenting with a new style of search result customizations: digg-style voting buttons that allow users to customize their search experience. If the experiment is a success, it’s conceivable that the results of such voting could be folded back into the general search results, complementing and potentially even replacing the PageRank algorithm.

Google search results with digg-style voting

From the experiment’s description page:

This experiment lets you influence your search experience by adding, moving, and removing search results. When you search for the same keywords again, you’ll continue to see those changes.

Google’s I like this voting buttonLike it?
This button will move the result to the top of the page and add this orange marker (fig. 1a) next to it so you can easily recognize it. The result(s) you promote will appear at the top whenever you search for the same keyword(s) in the future.

Google’s Bury This buttonDon’t like it?
This button will remove the result, and it will remain hidden when you search for the same keyword(s) in the future.

While initially an enhancement that is confined to that user’s results, you can bet that Google will be monitoring very closely the data they track as a result of this experiment, and comparing it to how well their existing algorithm performs.

PageRank can only go so far in delivering customized search results, and this is one interesting way of improving the individual experience for Google’s users.

  • Who’s going to spend the time to vote up or down search results? Also, I believe this is to cutomize your own search results – it won’t replace PageRank.

  • Just as an FYI, so people aren’t mis-led by the title. Nowhere in any Google’s PR docs or blogs does it many an abandonment of Pagerank. It’s simply an experiment to gather more useful information to produce better, more relevant results.

  • Folding this data into the algorithm is purely me hypothesizing on the outcome of the experiment, hence the question mark in the title.

  • skyjuice

    Google 2.0 T_T

  • I meant “mention” not “many” – my goof.

  • jont17
  • If people come to know that their vote has value, they will surely attempt to manipulate the results.

    However, this method is useful to eliminate the most unwanted pages.

    The more the negative votes for a page from different IPs and geographical locations, a small dip in PR can be a possibility, with strict scrutiny from an automated system.

    Positive voting can be difficult to consider.

  • studiobe

    this really has nothing to do with pagerank, so why use the misleading headline?

  • Dr Livingston

    it won’t replace page rank though i could see how it’d compliment it in many ways. there is a lot of complete and utter rubbish out there, and we need to get rid of it, collectively as users of the internet, and if we can vote it off in a public way, then great.

  • Eyespi20

    Hmmm….. well, what’s rubbish to one is treasure to another (or so the adage goes).

    IF Google does away with page rank, I don’t see this “voting machine” taking it’s place. One of the gripes (if you will) about page rank is that there are so many ways to “game it” which led to this most recent revision of PR algorithims and the subsequent demotion of a lot of sites’ rank (much to the vocal dismay of their owners!)

    I don’t really see how voting for or against a site can really be counted into the grand scheme either as once again, it’s too easy to manipulate votes. You see it happening all the time on these “vote for my cute dog” websites.

    Just my .02


  • Who’s going to spend the time to vote up or down search results?

    No one is going to spend a significant amount of time organizing their search results.

    But I tell you I loved the idea as soon as I heard it. Now whenever I get a search result with one of the many sites I don’t like I can block them from ever appearing again.

    It’s looking like search results will be delivered in two stages; first you’ll get the results based on page rank, then they’ll be reordered to your preference.

    I applaud the idea, I hope it works out.


    and SEOers will get their armies to put false votes and instead of selling text links, they will sell 100 positive votes for $10 …

  • binjured

    I certainly hope that headline was suggested by some PR person to get more hits because the question itself is completely asinine. Yes, Google is going to abandon PageRank, a proprietary algorithm they’ve put countless man-hours into that not only helps provide their high-quality search results but which is also a basis for how many marketers, advertisers, affiliates and so forth gauge the quality of a site… for user rankings that can easily be manipulated and ignored all together by users.

    Are you on drugs? Drunk? Seriously, give me a plausible excuse for post. Please.

  • @skyjuice: you are right, it’s google 2.0 :D

  • It could lead to lots of Spam. For example some culprit could use proxy and stuff to vote his own web page. Which will bury the most relevant results below you know.

  • SEO Canada

    How would they determine the top 10 to begin with if they drop PR? The idea of how likely a random surfer is to visit a given page is good, it’s just measuring that likelihood that needs refinement. I think that’s where these thumbs up/down will come in to use.

  • SEO Canada

    Like what Doug said: It’s looking like search results will be delivered in two stages; first you’ll get the results based on page rank, then they’ll be reordered to your preference.
    Also, the title is a lame attempt to linkbait.

  • I read this is a personalised search, getting rid of the subject pages that you recognise as not being relevant to you personally. Eventually when you do a search at a later time, a more refined search probably – say three keywords instead of two, the “better” results from the earlier search will be the first ones to be checked for the extra keword, or influence which pages with all three keywords are more likely tobe returned. It’s not a case of some people voting for a site as a better one and then the rest of us being influenced by these people’s choices.

    I don’t think it is a knew idea either, I’m sure I readsome time ago something about personalising your searches over time, to give you more of what you want.

    For example I’m an ex-chemist. If I searched on alcohol from uni, it’s chemicals which are alcohols that I’d be after. where as you might be searching for beer. As a chemist, my search on nmr (look it up) would be very different from a medic searching on nmr, and I and the medic could filter out over time the pages that were not our normal intended targets. So future similar searches would be tailored to our individual needs.

  • I think PR can not be replaced , but algo should be sharpened with user’s votes.

  • Mr Misleading

    Change the misleading title of this POINTLESS blog. Why are you even letting this guy blog, SitePoint?

  • Who’s going to spend the time to vote up or down search results?

    err Simon, I think that Digg gets a lot of traffic & there are quite a lot of users who vote up for stories on it. so if that can work, why not this? I agree that not all will vote on search results but then not all the visitors to digg vote as well. but even a few hundred thousand visitors who vote will make a difference imho! :)

  • APatcher

    I have a number of websites. It seems to me that PageRank is virtually dead at this time. As long as the page has some PR, it can rank just as well in the results as other pages that have much more PR. For example, my PR3 and PR4 sites unexpectedly became my most popular sites on Google. It used to be the PR5 and PR6 sites that always did well. More than ever before, I believe Google is now weighing how many people stay on pages versus how many people “bounce” back to Google after looking at the page for only a few seconds. This makes sense to me because the “bounce” factor would be a difficult thing to fake on a grand scale (unlike PR, which is relatively easy to fake/spam up the rank). Google Analytics and AWstats both offer a measure of “bounce” (but in different ways/terms). Pages with high bounce factors rank low because people are looking and the page for a couple seconds and then going right back to the search results by clicking the back button on their browser. Pages with low bounce factors rank high because if someone stays on the page or clicks a link on the page to go to another area on the web, they probably have actually FOUND what they were looking for in the search engine. Bounce factor is also a easy thing to measure (unlike PR).

  • Shrinidhi Hande

    How will the user know before visiting the link that it is useful or not?

    After he visits the page he wont care to go back and rate…

    Shrinidhi Hande

  • Alby10

    Surely, this would be open to abuse, it must purely be for the individual within their own browser to modify and manipulate the results.It sounds like a “digg” type of system, also owned by google.

Get the latest in Front-end, once a week, for free.