SERPs Analaysis
Observations and thoughts
Notably there are no hyphenated domains in the top ten except for position 10 although pages 2,3,4 and on are littered with them.
It seems clear from this that seperation of keywords is no longer an issue and Google is able to easily extrapolate keywords from a text string. This has been the case for some time but this seems to be even less of an issue now.
There is a lot of speculation around the length of domain name registration and its effect on SERPs. I am a strong believer in the fact that it doesn't make much if any difference. That being said some of the top ten results are registered for long periods. It should also be noted that there is a 12 month only registration that is in the top ten so I don't think it makes much difference if any.
Having the keyword in the domain name does not seem to make much diference but having the domain name in the URL and thus in the file name does seem to make some difference. This is obvious as you would name a page aptly for the topic it was about.
Having high keyword density in the title is still a big factor. It also appears that having high keyword density in Google's description of your site is a fair indicator.
To what extent I am not sure. Google uses a combination of (or either or) the meta description and the body text for its description. It uses strings of text it sees as relevant. Quite how it decides upon the exact description is still unclear but it does seem to have a reasonable bearing on things.
Backlinks still factor highly in the algorithm. To determine exactly how much a more in depth look at the back links is needed.
Here is proof to dispell a common myth only two out of the top ten (20%) use Google's XML sitemap technology and none of the top 3 seem to.
Some of these sites were specificly registered and built for the competition where as some were just sub directories and new pages on existing domains. This will obviously effect the relevance of the links to each site. There is also an odd ball which is a sub domain of an existing domain which falls into both categories. Positions 1 & 2 were topial specific sites.
I am considering wether Google's link command shows a random selection of links and if this selection gets bigger with age. The reason for this is that there is not much correlation between either Google and Yahoo or Google and MSN values, where there is reasonable correlation between Yahoo and MSN's back links.
I have noticed some sites in other SERPs that look like a directory but are
actualy just a file without an extension.
An example of this is
www.watching-paint-dry.com/v7ndotcom-elursrebmem
which does not work as
www.watching-paint-dry.com/v7ndotcom-elursrebmem/
there for it is a file called v7ndotcom-elursrebmem not a directory. I
can't see how this would have much bennefit but I am noticing more and more of these comming up.
<- Back SERPs analysis | Next -> Conclusions