One of the sessions that stood out from the SMX Advanced 2009 conference in Seattle last week was "Duplicate Content Solutions & the Canonical Tag". This was an informative session with some excellent examples of its benefits revealed by optimisers. More importantly a number of factual confirmations were made by the search engines in its use & benefits.

The fact that Google's Matt Cutts was following this session on Twitter and felt compelled to leave his social media session and crash the last 20 minutes of this session to take over the microphone, provides some real insight into exactly how important Google consider the use of this tag.

A little bit of recent history, in April this year at SMX Sydney, the search engines announced the accepted use of the Canonical Relationship tag. The idea being that if a website were somehow producing duplicate content pages, it could use this tag to point or link the duplicate page or pages back to the source page so that the engines would recognise & index only the source page, not the duplicate page.

This seemed all very simple and looking at the stats, it seems it has been well accepted by webmasters.

?The no follow tag was suggested as standard practice by Google some three plus years ago, with only 1% of websites currently using this tag.

?The REL Canonical tag was suggested as standard practice only 3 months ago with Bing confirming 1.6% of websites already using this tag.

That's a staggering uptake in just 3 months.

Some good reasons for using the tag are:

?Sites with load balancing issues creating duplicate content across ww types or sites with HTTP & HTTPs versions of pages.

?When the wrong page ranks in the SERP, tag the wrong one and direct the search engines to the desired page.

?If two results are appearing in the SERP, tag one to the other to create one good listing.

?CMS issues causing keyword friendly & dynamic URLs to the same page.

?Just like a 301 redirect, tag an old page URL through to a new replacement page URL

Now here's some important facts confirmed by Google during the session:

?The tag works within a domain across sub domains and sub folders, but DOES NOT work across root domains.

?The tag does pass PageRank just like a 301 redirect.

?The tag does pass history, just like a 301 redirect.

?The tag does pass anchor text value of hyperlinks.

?REL tag to page a URL that 301 redirects blocks the flow, this was stated by Matt Cutts.

The Canonical REL tag has been described as 'the poor man's 301'. This is something worth considering for its general use by website owners.

If you have a website on a server that you don't have a level of access that allows you to set 301 redirects when moving or changing pages, you can instead leave a page on the old page URL and insert the Canonical REL tag to pass the PageRank & history from the old page to the new page.

Something worth noting is that Google called the tag a "Strong Hint" to the engines. Now I believe they do this more so because they are unsure that the tag works 100% of the time, rather then the fact they don't want to use the tag and don't trust it.

This theory is supported by Matt Cutts stating that they are happy to be provided examples of when the tag has not worked correctly by website owners or optimisers leaving examples on Google's webmaster forums. Google will then actively follow through and look to correct these issues and adjust the robots to stop further issues.

Several examples of the tag not working correctly were provided to Matt Cutts during the session and he seemed committed to understanding the issues to why it hadn't worked and fixing them.

So I believe the engines are firmly committed to using the Canonical REL tag and see it as a realistic way of cleaning up duplicate content on the web. Duplicate content is a major issue for the engines as they have to continue to build data centres to handle the content of the web.

Duplicate content is also an issue for website owners as it does lead to website penalties and lost rankings.

What seemed to come across in the session is that perhaps website owners should not over use the tag and perhaps be a little more restrained in its use?

Certainly I think Google hinted at this as perhaps there are still unknown ways of manipulating its use. Top optimisers seemed to agree that overuse of the tag or a lack of planning in its use could lead to negatively effecting internal link structure of a website.

Website owners should also consider doing everything else possible to clean up there websites and stop duplicate content issues and only use the tag when other best practices cannot be implemented, thus using the tag smartly and not just as a work around for lazy development.

The session also lead onto the use of the no follow tag in conjunction with the REL tag as well as how the no follow tag is treated by Google these days. Some important changes have happened on the no follow which I will cover in my next post.