So Who’s Canonical Now?

by Aaron Bradley on April 21, 2009

in SEO

It's now been over two months since Google et al. announced support for <link rel="canonical">, so I thought I'd do a spot check to see who's using the new <link> tag attribute and who isn't.  As you'll see, so far the response has been, well, underwhelming.

SEO Industry

http://www.seo.com/
Not observed.

http://www.seochat.com/
Not observed.

http://www.seobook.com/
Not observed.

http://www.seomoz.org/

Not observed.

http://www.davidnaylor.co.uk/
Not observed.

http://sphinn.com/
Not observed.

http://searchenginewatch.com/
Not observed.

http://www.mattcutts.com/blog/

http://www.mattcutts.com/blog/canonical-link-tag/

Not observed.

Search Engines

http://www.yahoo.com/

http://ysearchblog.com/

Not observed.

http://www.live.com/

http://blogs.msdn.com/livesearch/

http://www.google.com/

http://googleblog.blogspot.com/

http://googlewebmastercentral.blogspot.com/2009/02/specify-your-canonical.html

Not observed.  But…

http://www.google.com/support/webmasters/bin/answer.py?hl=en&answer=35291

<link rel="canonical"
href="/support/webmasters/bin/answer.py?hl=en&answer=35291″ />

Online Shopping

http://www.shopping.com/
Not observed.

http://www.shopzilla.com/

http://www.shopzilla.com/clothing-accessories/10000000/browse

Not observed.

http://shopping.yahoo.com/
Not observed.

http://www.overstock.com/
Not observed.

http://www.bizrate.com/
Not observed.

http://www.amazon.com/

http://www.amazon.com/books-used-books-textbooks/b?ie=UTF8&node=283155

Not observed.

Blogs and Social Media

http://en.wikipedia.org/wiki/Main_Page

http://en.wikipedia.org/wiki/Blog

Not observed.

http://technorati.com/
<link rel="canonical" href="http://technorati.com/" />

http://www.myspace.com/
Not observed.

http://www.facebook.com/
Not observed.

http://wordpress.org/
Not observed.

http://dailykos.com/

http://www.dailykos.com/

Not observed.  No 301-driven canonical domain name, either (see my note on nytimes, below).

Poker Sites

I added this category because online gaming sites are supposed to care so much about SEO.

http://poker.bodoglife.com/
Not observed.

http://www.pokerstars.com/
Not observed.

http://www.partypoker.com/
Not observed.

http://www.partypoker.com/
Not observed.

Media and Gateway Sites

http://www.latimes.com/
Not observed.

http://www.bbc.co.uk/
Not observed.

http://www.theglobeandmail.com/
Not observed.

http://www.suite101.com/

http://colleges.suite101.com/

Not observed.

http://www.about.com/

http://drugs.about.com/

Not observed.

http://www.nytimes.com/

http://nytimes.com/

http://www.nytimes.com/pages/politics/index.html

http://nytimes.com/pages/politics/index.html

http://www.nytimes.com/pages/politics/

http://nytimes.com/pages/politics/

Not observed.  Just as an aside I pasted all of the above URLs because, aside from not supporting the <link rel="canonical> tag, nytimes.com isn't even doing supposedly industry-standard 301 redirects to posit canonical folder URLs and – gasp – a canonical domain name.  Marshall, baby, you're my hero:  what gives? :)   At least it's there for about.com.

If you were counting TLDs, you'd see that 36 sites aren't employing the tag, and two are (one and a half, really, since only one of the google.com URLs I checked was carrying it).

Make the nay count 37 if you add SEO Skeptic.  By the way, I'm actually a proponent of the tag, but haven't got around to employing it here (after I upgrade, after I upgrade).  Of course, with WordPress, I'd need a plugin, so did a quick search:  sure enough, I found one on yoast.com.  And that brings the "yea" count three (or two and a half):
<link rel="canonical"
href="http://yoast.com/wordpress/canonical/"/>

{ 3 comments… read them below or add one }

1 Ani López April 25, 2009 at 12:06 pm

canonical means you have problems in a site, better mend them than add patch don’t you think?
[See Ani's excellent article on web structure, duplicated content and the canonical tag.]

Reply

2 Ani López April 25, 2009 at 12:06 pm

canonical means you have problems in a site, better mend them than add patch don’t you think?
[See Ani's excellent article on web structure, duplicated content and the canonical tag.]

Reply

3 admin April 26, 2009 at 1:44 pm

Good point, Ani. And certainly every effort should be made to avoid the need for the use of the canonical tag in the first place (and, by extension, in the creation of problematic duplicate content in the first place, as you outline in your post).

However, there are instances where issues of canonicalization arise specifically from the URL structure, rather than the interlinked page structure of the site. The chief example of this is URLs that render with different parameters that are caused by exposing those parameters in the linking source.

And yes, it is better in these cases to repair rather than patch the problem, most straightforwardly through the use of 301 redirects. Even at that, though, there are two issues involved. In the real world, unfortunately (at least from my perspective) SEO types don’t necessarily control web development, and using the canonical tag may be the only way of giving the search engines some sort of canonical guidance when the development solution is behind, or even absent. Secondly, even when you have properly-functioning 301 redirects, not all search engines necessarily handle these correctly (in different ways, both Yahoo and Live Search are notorious for either misinterpreting 301 redirects, or failing to update their indexes so variant URLs are collapsed in a single form via redirects).

The situation is analogous in part to the introduction of the XML sitemap protocol, when many webmasters and SEOs were against employing sitemaps on the principle that if there were spiderability issues with a site, they’d rather know about them and fix them, rather than use sitemaps as a workaround. However, it has since turned out that there are probably advantages to using sitemaps, even if your site is well-structured for spiderability (e.g., apparently faster addition of new URLs to indexes, and at least the theoretical ability to use spidering resources more efficiently through the update frequency and priority settings). So, in my opinion, if there’s little overhead required to do so something that may help optimization, such as sitemaps or <link rel=”canonical”>, I think it’s better to add those functions rather than to ignore them.

Reply

Leave a Comment

Previous post:

Next post: