Recovered from the Wayback Machine
Victor is trying to see how we can influence Google page rankings, using weblog links. He’s calling the process Googlestak. It will be interesting to see if this effort works. Target link is Stand Out Training — Victor’s company.
Because of my interest in the semantic web, I’ve always been interested in our friend Google. Through research I’ve found that you can, with effort, influence Google, but the circumstances have to be just right. There are entire mailing lists and web sites devoted to how Google works, and how the page ranking algorithm works.
The number of links to a particular page are only part of the Google search algorithm. The importance of those links can also influence the scoring. If you have one page that is linked by 10 pages that are themselves ranked fairly low, it won’t rank higher than a page that has one or two links from very high ranking pages.
Unfortunately, weblogs are — for the most part — not the highest ranking pages. Links are sporadic outside of the blogrolls, and these tend to be incestuous: page A links to page B which links back to page A. Put a lot of links within your blogroll linking back to you, and you’re still going to be low ranking because of that “importance” ranking catch 22. You would need to break this cycle by getting linked by an A-lister such as Scripting News or Doc Searls or Rageboy or JOHO — people that are linked by a disportionate share of the weblogging community.
As an example Jonathon will usually get more blogroll hits from my weblog rather than from the Scripting News blogroll because I’ve made his link part of the “Australian Delegation” — I made it stand out. However, within Google, Jonathon’s weblog will get more buzz because of that Scripting News link than anything I could do.
In addition, because our pages roll off into archives, the ranking algos start to break down because they don’t necessarily figure in the temporal nature of weblogging.
If you have a static page that generates mild interest and covers a topic that isn’t time sensitive, over time that page will rise slowly to the top of Google because the interest is cummulative. Weblog pages can rise to the top (especially when helped by strange words such as googlewhacking), but they’ll fall away quickly once the initial intense interest fades, and as the weblog pages roll off into archives (which are not directly linked to from blogrolls, remember).
I’m linking to a specific post of Victor’s. Others will join me. Eventually, this post will fall off into the archives and less pages will link to it. And the page where I created the link will also fall off into the archives, and less pages will link to it. Instead of a cummulative effect, with weblogs you get a decremental effect — over time the rank will decrease rather than increase. Regardless of the time sensitive nature of the topic or not.
The decremental effect of weblog postings is a good reason to use a Story mechanism for more important and wordy postings rather than the usual weblog posting — the story will remain accessible rather than get lost in the archives. However, this doesn’t help you when the pages that link to you eventually fade away. Still, the hope is that new, fresh links will come along. It helps to mention or reference the story occasionally in newer postings — call it a form of weblog pinging.
If you’re trying to raise to the top of the rankings associated on popular search terms, good luck. You can’t influence the rankings with something like Linux; it’s been around too long, and the major sites associated with this term are too firmly entrenched. You would have to invent a Microsoft buster within the Linux world to generate enough buzz to even get to the first page of the result set.
As for other subjects, I imagine you might have a shot if you’re diligent, work hard, get tricky.
I found a page with some interesting and clever tips for influencing Google. From this, I found that one thing you can do is find pages that link to you and make sure that Google is aware of them. Interesting idea, isn’t it? Because the more buzz these pages get, the higher ranking your linked to page gets.
Another thing you can do is package word sets to capture specific searches.
I’ve started an online C# book (that I really need to finish). Now, searching for C# is not going to return my book pages in the first result set — I haven’t finished the book and only have so many sites linking to the pages. However, if you search on “C# book”, I’m the second from the top in the listings (yasd.com is my domain as well as burningbird.net and several others). I’m using the term “C# book”, and I have sufficient links — combined, I’m close to the top.
Another way to influence Google is to do something to drive people to your site. If you write, write articles for online publications and insist they use a link to your weblog or site within the author bio. This is probably the number one reason why I get so much Google traffic — I’ve been writing online articles for years.
Can’t write for an online publication? Then start joining newsgroups, MetaFilter, whatever and post, post, post! Post to anything you can attach a signature to that has your weblog or site address. Believe it or not, this can influence Google — I’ve seen it happen with content in my own web sites. I still see it and surprises the heck out of me (as well as made me a bit more cautious about why I say at said newsgroups et al).
Fascinating tool Google. If we harness this type of connectivity and attach it to “meaning” rather than ranking, then you have the semantic web — one short step away. Exciting stuff.