For questions on support for a new web implementation within the search engines, it’s always a good idea to start with the search engines themselves. Here are a few contributions from the search engine guys:

Google’s own Webmaster Central blog has a great discussion on reasons to use the tag, how to implement it and how Google handles it. Several questions from users have also been addressed in the comments section. Scroll down – Google does answer eventually:
link

Here’s MSN’s blog, with some information on how MSN’s Live will be implementing support for the canonical tag:
link

And here’s where Yahoo Search weighs in with their comments and information on how their search engine will treat the tag:
link

If you prefer to listen to an expert talk about the subject, Web Pro News provides this video interview with Matt Cutts of Google, Inc:
link

SEO experts are another good source to turn to when a new optimization tool becomes available for use by developers. SEOmoz provides this blog entry that explains the tag and its usage with some nice graphic visualizations of canonical relationships. It also quotes the search engine pages and raises some interesting questions:
link

For those developers or webmasters who utilize Content Management Systems like WordPress, Joomla and Drupal, contributing developers have created plugins for support of the new tag. Here are a few direct download links for the more popular CMS:

WordPress: link
Magento: link
Drupal: link

And one for Joomla that includes a tutorial: link

Last, but not least, it’s always good to hear both sides of the story. Here are few comments from “the Oracle”. We’ll let you be the judge on whether they’re pertinent:
link

The first step in providing a canonical tag for the search engines is to decide where you may have duplicate content within your website. This isn’t always as simple as it may sound. Let’s recap a few ways that duplicate content might be a problem and what the canonical URL for those pages should be:

Gateway pages: Many sites have multiple entry pages. There are a number of valid reasons for configuring a site this way, one of which deals with tracking scripts. It’s easier to know which search engine or linking page a visitor comes from if you list a separate welcome page for each search engine or link. An example might look like “http://mysite.com/yahoo_index.php”. In this case, the canonical page will probably be “index.php”, “index.html” or “default.htm”, etc. – the entry page someone would see if they typed your site address directly into the browser.

Printer-friendly versions: There’s often a need to supply visitors with separate versions of pages, formatted to print nicely. In this case, you’re definitely duplicating the content that the search engine spiders will index. You’ll want the search engines to list the pages you’ve designed for web viewing, so your printer-friendly versions should contain references to the web versions in the canonical tags.

Pages with dynamic content: Dynamic content can often create duplication for the search engines that isn’t always obvious to the developer. An example might be a dynamically created menu that builds links to a product with different color choices or sizes. These links may all lead to the same product page, with only a query string as the difference, such as: “http://mysite.com/ballcaps/1234?color=black”. This means the search engine spider can follow each of the menu links to the same product page. Pages that might be called with a query string should contain canonical tags pointing directly to themselves.

There are certainly more cases than those listed above, but the examples will hopefully get you started off in the right direction. Remember that the less duplicate content the spiders find, the better your chances for high page ranking.

So, now that we know where, let’s get to the “how”. A canonical URL is specified with the tag. It belongs in the section of your pages, and the structure looks like this:

The href value should be the page that best represents the duplicate pages, in other words, the page you want the search engines to index. This value can be an absolute reference, such as “http://www.mysite.com/index.htm” or a relative reference, like “index.htm”.

One other point to note is that the URL in the tag must be located within the same domain as the current page. Pages within subdirectories (mysite.com/products) or subdomains (products.mysite.com) will work, but if you need to reference a page outside the domain, you’ll need to learn to use the 301 redirect directive.

Implementing this new hint for the search engines is just that simple. Consider that it may help improve your site’s search engine positioning and we think you’ll agree it’s probably worth the time and effort.

First and foremost, let’s look at what the canonical tag actually is. The first thing to realize is that it isn’t actually a tag. It’s a new value for an attribute of a tag that’s been in use for quite some time: . You’re probably familiar with the use of this tag for linking external stylesheets, to use the CSS styles for multiple pages. In fact, the purpose of the link tag is to define relationships between pages in just that way. The exact type of relationship is specified by the “rel” attribute of the tag. Up to now, the only really practical use for the tag was to link stylesheets to documents.

So, now we have a shiny new value for that attribute – canonical. What good is it? Well, it lets us tell the search engines that a specific URL may be a duplicate of another page. That’s important because it’s not advantageous for a search engine to index pages with duplicate content. The fact is, having several pages indexed with identical or very similar content will probably harm your search engine ranking. It’s also extremely annoying for users to click on multiple search results only to end up on pages with the same content. By stating the canonical relationship to a page in the link tag, we tell the search engines not to index this page, because it’s really the same as the other. We’re also informing the search engine that the other page is the one to use for calculating rank.

There are instances where it may be to your advantage to use duplicate pages. For instance, some sites track where their visitors come from by creating and linking multiple index pages, often referred to as “gateway” pages, such as: “http://myplace.com/google_index.html” or “http://myplace.com/index.php?se=google”. In this case, the content of the pages will probably be identical to the index.html or index.php page, which would be the page we would point to in the link tag. In other instances, the duplicate pages may simply be unavoidable, as in a dynamically generated menu with options in a query string, for example: “http://myplace.com/tshirt?size=L”. There may also be a need for pages with the same content in different formats, as in “printer-friendly” version of a page. A session ID may add a query string that can cause multiple instances of pages. In all these cases, adding the canonical relationship will let the search engine know which to index.

The canonical relationship should prove to be a very useful tool for developers for optimizing their sites. Google, Yahoo and Live have not only embraced it, but also published their support and thorough explanations of how their engines will handle page indexing with implementations of it. All in all, it should bring leaner, cleaner search results and more accurate ranking to the Web.