When Google, Yahoo! and Microsoft joined forces and endorsed the rel=canonical tag back in 2009 SEOs across the land jumped for joy. Rel canonical does fulfill a very useful purpose buying a site owner time to correct their content duplication issues. Unfortunately, many site owners and developers are beginning to use the rel canonical tag as a band-aid for just about every URL problem. This is a dangerous practice that could produce a devastating result.
First, let’s discuss what the rel canonical tag is and why it is needed. Search engines want a unique piece of content to reside on one URL on your website. That means the search engines do not want to find your wildly popular article about Kony2012 republished on your site at multiple URLs:
Having the exact or very similar content on multiple URLs will trigger a duplicate content penalty. The above example of URLs highlight the SPAMMY way of duplicating content. Often times duplicate content issues are caused by a CMS that is running amok. Below is an example of a CMS creating a new version of your URL each day by appending a paramater:
Notice the nasty parameter code being added everyday? This is not an intentional act of the site owner trying to create multiple URLs for the Kony2012 article. Nevertheless, a duplicate content penalty will soon loom over this site. The best course of action is to slap a rel canonical tag in the <head> section of this page essentially informing the engines, “Yes, we do have duplicate content for this page. However, please treat the specified URL in our rel canonical tag as the sole location for this content.”
For our example the rel canonical tag would appear like this in the <head> section of the source code:
<link rel=”canonical” href=”www.bestsocialmediaever.com/kony2012/”/>
Every duplicate page would have this rel canonical tag in the source code so the engines would know the intent of the site owner is to assign that content to one URL. After you have added that tag your job is not complete. Now you must FIX YOUR CMS URL WRITTING PROBLEM. Think of rel canonical as a real band-aid.
You cut your arm and blood comes pouring out. You clean the wound and slap a band-aid on the gash. In a few days your body repairs the cut and you no longer need the band-aid. That is rel canonical. Site owner discovers CMS URL writing problem and uses the rel canonical tag. Then the site owner resolves the URL writing problem, gets rid of all the dupe URLs so only the original URL is present and then the site owner can remove the rel canonical tag.
Unfortunately, I am seeing the rel canonical tag “resolving” many more issues on a permanent level. The frequent and most common use is to resolve the problem above, but the site owner never fixes the CMS parameter URL issue.
A deceptive and diabolical use of the tag can be found with sites that use content provider partners. Site A publishes celebrity gossip and becomes insanely popular. Site A cannot keep up with the demand for creating content. They reach out to Site B and strike a content syndication deal where Site B will allow Site A to reuse content from Site B. The same articles on Site B will also reside on Site A at the same time. I have seen contracts where Site B requires Site A add a rel canonical tag on the Site A content pointing to the original URL on Site B. This ruins any and all search value of having Site B’s content on Site A. The content on Site A will not surface on the search engine results pages. Ensure your content syndication agreements do not require the use of a rel canonical tag.
That last example is a sneaky way to manipulate the rel canonical, but the worst use is substituting 301 redirects for rel canonical. When you change your URL structure or move your entire site to a new domain you will ALWAYS NEED TO UTILIZE 301 REDIRECTS.
I recently saw a high revenue website change their URL structure and then use rel canonical to point to the new URLs. So the legacy URLs are still live and the new URLs are live. The site owner did not use 301 redirects, but slapped a rel canonical on the legacy URLs pointing to the content on the new URLs. Do not do this. Rel canonical was never intended to supplant the redirect process. What would happen if the rel canonical tags were put on the new site pointing back to the legacy site? The new site would simply not exist in the eyes of the engines.
Do not mess around with rel canonical because if used improperly it can become a shriveled up, and foul smelling band-aid. You do not let your band-aids for cuts become that way, so do not let that happen to your website.