 After just a month, our Search Engine Optimization test is already producing some really interesting results.
Our Wordpress test site already has a sizeable lead with Drupal in second place and Joomla quite a way behind. On the most important ranking (top ten rankings in Google), the Wordpress site is twice as successful as Drupal and three times more so than Joomla.
The overall score is calculated using WebCEO, and is based upon a website's top-5, top-10, top-20, top-30 rankings. January Update|
| Top Ten Rankings in Google.com | 6 | 3 | 2 | | Top Ten Rankings in Yahoo.com | 0 | 0 | 0 | | Top Ten Rankings in MSN.com | 0 | 0 | 0 | | Top Ten Rankings in Google.in | 6 | 6 | 2 | | Top Ten Rankings in Yahoo.in | 0 | 2 | 7 | | Visitors from Search Engines | 56 | 136 | 101 | | Overall WebCEO Score | 180 | 87 | 59 |
So, is this a fluke so far?
Yes - Jabalpur has fewer results in Google than either of the other two cities which means it may be easier for it to rank more highly. This may help explain Jabalpur's low number of visits despite it's high rankings.
Yes - As these sites only have 8 pages each, the success or failure of just one page can make a big difference. Yes - Its too early to produce real results. MSN and Yahoo have not begun to rank the sites yet.
No - We ran Wordpress and Drupal out-of-the-box but added an extra search engine-friendly URL component to Joomla. The results for Joomla might have been even lower had we truly used it out-of-the-box or with the default search engine-friendly URLs.
No - We ran two more sites, one in Wordpress and another in Joomla. The results were very similar. No - Jabalpur actually had a much larger lead early on. After two weeks the results were even further apart than they were after four weeks. Does Google perhaps realise the difference between software designed for a blog and software designed for a CMS, and then punish the blog if it is not regularly update? We shall see in the coming months... - As promised we will shortly be adding a Drupal 5.0 site and also a Joomla 1.5 site when it is released.
- We hope these test results will prove to be interesting and are open to suggestions for improving the study.
While these initial results seem to sugges that Joomla and Drupal are not as well optimized for SEO as they might be, I certainly don't believe they mean that a Joomla or Drupal website is doomed to rank lowly in the search engines. As has been often said, content is what will bring you good rankings. A good Drupal or Joomla site will provide a much better long-term platform for large amounts of content than will Wordpress. However, those considering moving to Drupal or Joomla need to bear in mind that you may need to focus on closely on your new sites's SEO.
|
Comments
I know that someone could probably drive a truck through the methodology here (so we're working to improve that now). However, I hope it will cause some people to stop and think about how to SEO their Joomla sites.
We get a lot of people coming to us who thought about it a little later than they might have.
Better to be surprised on the upside rather than the downside of things.
The easiest way would be to just update the site to drupal 5.0, which is a ridicolously simple process for such a simple site, and then use Drupal's new default theme, Garland.
Alternatively you might want to browse the available themes at http://drupal.org/project/Themes, and pick one that seems to be fitting the site's purpose.
Furthermore, the Drupal site is the only one that has no images at all, and it is the only one that doesn't have a "Hotels" or "Accomodation" page, which is probably something people often search for in search engines.
True - it could look better but images and a sexy theme wouldn't help too much in the search engine rankings. Or are you saying that the Friendselectric theme is not very well optimized for SEO?
Quote:
A Drupal 5.0 site will be coming this month
The problem is there is no "test input data" that mirrors real world Google activity. And, there is no way to make perfectly similar sites. If you start modifying the base installs too much, then you are swaying the test.
Frankly, what I would like to see is *more* of this type of testing. As more independent parties attempt to isolate variables and create comparative test beds, people will get better at doing it.
Tough job - I am quite certain there is a more than a little bit of time invested to make this happen - freely shared results; way cool!
PS - congratulations to the Drupal community for v 5.0 & birthday 6.0!
These results are as expected from the testing I've done previously.
Maybe one city is more popular than the other, gets mentioned more often on the web and therefore has higher relevance for search engines and gets ranked higher...
In 2005 a famous IT magazine here in Germany startet a SEO contest. They created a new fantasy animal (the Hommingberger Gepardenforelle ) and encouraged people to do their best to create a as high as possible ranked website for it. Maybe you should do something similar - create websites using different CMS but about the same topic, with the same content and find out which works best with standard settings. Later switch to optimized settings (document them!) and compare again after a while.
It should be clear though, that you cannot rule out off-site factors and therefore results will remain doubtful. But wouldn't SEO be boring if it was different?
Thanks for the report.
Ben
RSS feed for comments to this post