Are search engines using link-data over time?

Every once in a while, you encounter pages in search results that haven’t changed for quite a while. When reading Aaron’s post about major ranking changes in Yahoo! I started doing one of my favorite queries: the one for CSS3 (screenshot). I noticed two strange things: first of all, result #1 and #3 are from the same site, and #2 and #4 are as well, and they’re not indented, which every other search engine would do. I thought that was weird and so did Aaron, he even called it “a step back”.

The second thing I noticed was the result #5 in there, the geocities page. That page was created back in 2003, and hasn’t been updated since. For a subject like CSS3, that basically means the page is completely out of date. Now it’s a known fact that search engines tend to like “old” things, whether it be pages or domains, and favor them in their rankings. But you’d think, that if these pages don’t get linked to anymore, they’ll slowly drop in the rankings. Consider this history:

The geocities page is created in 2003 and acquires 27 links in the six months after it is created. It rules the SERPs for the keyword it targets and is left there to age. In 2006, is created. It acquires an enormous amount of links, and is updated every few weeks. Yet it takes until the beginning of 2007 for that page to “beat” page 1. Since 2003, the keyword it targets has become way more popular, and people have been writing all kinds of great articles about it. And still, that page is there at #5 in Yahoo!, mostly because the page and it’s links are old and the geocities domain is strong.

What’s going wrong? Well, the geocities page might have old backlinks, and some authority due to it’s age, but it hasn’t been acquiring any links any more for the last 2 years. On the other side, a lot of other pages about the same keyword have been gaining lots and lots of links. Search engines should use link-data over time to determine whether a page should be ranking or is not “up to date” anymore. However, from the fact that the page ranks #13 in Google, and #19 in the new Live search, I determine that the search engines aren’t using this data as much as they could…

Now it could be that the geocities domain is just too strong, but I think this is something the search engines should fix…

Tags: runs on the Genesis Framework

Genesis theme frameworkThe Genesis Framework empowers you to quickly and easily build incredible websites with WordPress. Whether you're a novice or advanced developer, Genesis provides you with the secure and search-engine-optimized foundation that takes WordPress to places you never thought it could go.

Read our Genesis review or get Genesis now!

5 Responses

  1. AshishBy Ashish on 1 October, 2007

    I think yahoo is making changes in algorithms. Because when i searched a top money keyword, i found out that it ranks many pages of the same site for that keyword in its serps.
    many sites are ranking for more than 2 positions for a single keyword.

  2. David HopkinsBy David Hopkins on 4 October, 2007

    This is certainly strange. I would have said it was just a Yahoo shortcoming until you mentioned it ranked well in Google and Live too.

    Despite all the people who say that search engines record your entire link histroy just because it has been mentioned in a few patents, I haven’t seen too much to suggest this.

    The only thing I have noticed is Yahoo seems to like to wait a while before moving your site up the ranks. What I mean is that you wil have a site that has some good incoming links and it will be sitting below page 10 for a search, but after several months it will start to rise. Also Google have some way of detecting sites that get lots of very low quality links in a very short period of time. The result is they will slow or stop visting your site.

  3. Sint SmedingBy Sint Smeding on 8 October, 2007

    This reminds me of an article several months ago that described the Google algorithms, especially the way lots of small things determine the rankings. Factors like what Google calls Signals, Topicallity and Classifiers.

    Shouldn’t these technologies ‘know’ that CSS3 is a subject that is changing all the time and therefore providing recent information is more important than for, say, a search query about World War II (which should be really fixed information)?

    It looks like current-day SE-technology, or at least all the talk around it on blogs and news websites, concentrates mainly on to interpret the user’s search query and how to get as much information about what the user is looking for. My opinion is that there is no universal ratio of all factors that works perfect for every query and topic. Therefore I agree with Joost that for CSS3 it’s wrong to give the age of pages and links such a great influence.

    And while I’m typing this, I’m thinking: Maybe the search engines are putting in this kind of old results in an attempt to give a complete view of the subject. If someone is looking for information about the ongoing development of CSS3, this result might be very useful!