WWW vs Non-WWW: Why Does Google Even Care?

A Brief Intro to the Domain Name System

A common bane of those who care about their rankings is duplicate content. The basic idea is that a search engine only has limited space to store pages, and thus doesn’t want to store duplicates if it can help it. Plus, having the same content appearing in the the result pages makes for poor user experience - a big no no in the Search Engine world (you listening, Microsoft?).

Now, I know that I’m preaching to the choir here, but for those who don’t know what I’m talking about, the most common source of duplicate content is from your own site - namely the “www vs non-www” problem. I won’t go into detail about what the problem is, as it’s been done to death before. What I will go into, though, is just why the search engines look at things this way.

The root of the ‘problem’ is that a search engine doesn’t automatically associate the subdomain with the parent one. This is because every domain is a subdomain which can point to it’s own server, meaning that the search engines kinda have to treat each subdomain as it’s own domain. Confused? Don’t worry, it’s pretty straight-forward. It all revolves around DNS (Domain Name System), which is the telephone directory of the internet.

Basically, the internet is one giant computer network. Computers need to know the IP address of the website servers they want to talk to, just like how the computers in your office or home network need to know the IPs of the other computers if they want to communicate. The catch is, though, that while remembering a few IP addresses is no big hassle, having to remember the IP addresses of all your favourite sites can become a bitch, especially if they move servers frequently or have multiple servers slaved together.

Thus, DNS is a way to allow people to find websites quickly and easily. DNS translates domain names to IP addresses, so that your computer can talk to the others (and thus isn’t the weird freak the other computers avoided in High School) and you get to remember a nice name. As I stated before, every domain is a subdomain. This is because domain names are divided up into ‘zones of authority’, meaning that only parent domains contain the DNS info for their immediate subdomains. Using www.dsj.za.net as an example, this means that the DNS info for www.dsj.za.net is kept by the dsj.za.net servers, the info for dsj.za.net kept by the za.net servers, etc, etc.

As you can see, this basically means that every ‘.com’ site is merely a subdomain of .com, every ‘.net’ site is a subdomain of .net, etc, etc. As we all know, these sites get seen as seperate ‘entities’ by the search engines - so what’s to stop them from treating the subdomains of these sites the same way? Nothing - and that’s exactly what they should do, as they are keeping to DNS standards. Remember, the ‘www’ subdomain gets no special treatment in DNS. The fact that it it used to display web content is merely one of convention, which is why Google’s Webmaster Central gives you the option of stating which version is ‘official’. Hell, you can make the ‘www’ version point to whatever you want, even completely unrelated things such as your mail servers (though that may be a bit confusing for people, lol).

So, from an webmaster (especially SEO) perspective, it is important to know this when requesting that a client does redirects from one to the other. If the main domain of a site does not point to a web server (eg, no relevant DNS entry for it), there is no point in asking the client (or host) to put up a redirect to the www subdomain (or vice versa)! You’d first need to have that domain to point to a webserver (many hosts do this by default), then you can tell it to do the redirect.

By now many will probably be wondering how to check the DNS records of a domain. I’ll go into that in a future post, however for now it’s easiest just to type in the URL and see if it returns an error. Hell, that’s all most webmasters (and SEOers) care about - does a domain show a webpage and will it be able to rank?