I’ve been keeping an eye on a very interesting discussion about the difference between the Google pagerank of pages on a “www” domain name and the same page without the “www” prefix (e.g. “www.intuitive.com/index.html” versus “intuitive.com/index.html”). Google calculates pagerank based on quite a few different criteria – as detailed in my article How does Google figure out what pages are more relevant? Pagerank. – and if you care about that, and you should, you should be automatically mapping all non-www URLs to their www equivalent.
And here’s how to do just that…
In your httpd.conf file, look for an entry like this:
<VirtualHost bestbookbuys.com> ServerAlias www.bestbookbuys.com
That’s the problem in a nutshell: you are defining the “real” name of the Web site to be the domain name without the www prefix. To fix it, change the above entry to something more like:
<VirtualHost www.bestbookbuys.com>
without any ServerAlias entry at all.
Failing that, if you have an entry in your httpd.conf that’s more like this:
<VirtualHost www.auctionincome.com> ServerAlias auctionincome.com
just remove the ServerAlias line.
Either way, you’ll also want to, later in the httpd.conf file, add this:
<VirtualHost auctionincome.com> Redirect permanent / http://www.auctionincome.com/ </VirtualHost>
This is explained in some detail in
the Apache documentation too, if you want more detail.
One cool additional feature: a Redirect of this nature copies everything after the first slash, so a URL like
auctionincome.com/abc/def.html will redirect to
www.auctionincome.com/abc/def.html, which is exactly what you want.
aw
das
Betty White
New interesting theme
Phil, I’ll add it’s worth doing for more reasons than just Google. I’ve noticed for some time now that the Daypop Top 40 robot makes a distinction between www and no-www.
http://www.vardaman.org/archive/2003/11/26
See the rankings for adbusters.org in the graphic? Their ranking is fragmented at #16 and #19, where combined properly they should rank a good deal higher.
JD Hodges covers this as well:
http://www.jdhodges.com/log/1373
If you read through the comments you’ll notice I got a big difference comparing the reason.com site.
What I should say, as a followup, is that it doesn’t really matter whether you use the “www.” or not, what matters is that you’re consistent. If you prefer the non-www version of your domain, just make sure that all inbound links (like comments on Weblogs!) also use that form too.
The main idea is that Google indexes *pages* not *sites*, so as far as it’s concerned, http://www.phord.com/page3.html is completely different from phord.com/page3.html. By splitting your links between the two, you don’t gain the combined value of all the incoming links on the one page and you end up with a lower search engine ranking that you want.
Make sense?
Why would I want to use http://www.site.com instead of just site.com? You implied in your opening that this was somehow Googlebot-friendly, but I can’t find anything about that in the post you referenced. Can you clarify or point to some more complete info?
mod_rewrite also does a good job, and avoids the extra vhosts. Here’s what I use to map http://www.site -> site (this is for ccsacertification.com, substitute as appropriate)
RewriteEngine on
RewriteCond %{HTTP_HOST} !^ccsacertification\.com [NC]
RewriteCond %{HTTP_HOST} !^$
RewriteRule ^/(.*) http://ccsacertification.com/$1 [L,R]
Sean