Duane Forrester from Bing and Matt Cutts from Google, who are the SEOs view into search engines, have given us conflicting canonical advice. Or have they?
Friday morning Duane Forrester from Bing published a blog article entitled Managing Redirects – 301s, 302, and Canonicals. Within this article, Duane explains how Bing sees and may treat different redirects. The part about 301s and 302s is informative and interesting, but the most interesting part of the post comes when he talks about canonicals and says:
Something else you need to keep in mind when using the rel=canonical is that it was never intended to appear across large numbers of pages. We’re already seeing a lot of implementations where the command is being used incorrectly. To be clear, using the rel=canonical doesn’t really hurt you. But, it doesn’t help us trust the signal when you use it incorrectly across thousands of pages, yet correctly across a few others on your website.
A lot of websites have rel=canonicals in place as placeholders within their page code. Its best to leave them blank rather than point them at themselves. Pointing a rel=canonical at the page it is installed in essentially tells us “this page is a copy of itself. Please pass any value from itself to itself.” No need for that.
We do understand that doing work at scale requires some compromises, as it’s not easy to implement anything on a large site page by page. In such cases, leave the rel=canonical blank until needed.
Conflicting Advice Re: Rel=Canonical?
This advice seems to differ from this Webmaster video from Matt Cutts where he says:
” ”We built in support to make sure that that doesn’t cause any sort of problem. So I can’t speak for other search engines, but it’s definitely a very common case. Imagine if you had to check every single URL and then do a self check to see whether you were on that URL. If you were then you couldn’t have a rel=canonical tag. That would be a lot of work to generate all those tags. So for our part, we said you know what? Go ahead and you can put a rel canonical on every single page on your site if you want to. And then if it points back to itself, that’s no problem at all. We handle that just fine.”
Hat tip to Anthony Nelson for finding the video and the above transcript.
Watch the video if you desire:
Does the advice actually conflict?
At first the Bing post seems to contradict the Webmaster video, but does it? I actually do not think that the two are offering different advice, but rather that Google is telling us that they have a more sophisticated system for it. Google seems to be making our lives easier as webmasters and SEOs since we do not have to worry if Google treats self-referencing canonicals correctly. Are they necessary/mandatory? No, as Forrester correctly says. Are they a good idea to have? Yes, as Cutts says.
When should I definitely use the canonical tag?
Here are the cases where I think that you should use the canonical tag:
- If you use tracking parameters, or may at any point in the future; or
- If you enable people to permalink to comments in your blog posts; or
- If you might ever link to your site using the non-www version; or
- If others may ever link to your site using your non-preferred www version.
The above scenarios are pretty well all-encompassing, so I say to use the canonical tag on every page! Duane does not go so far as to say that using it will hurt your Bing results, just that “it’s not necessary.” Google says they are smart enough to know when it is self referencing. So why worry if someone is going to link to your site using the non-www when you use the www? Use the tag.
When should I definitely NOT use the canonical tag?
A few times exist when you should not use the canonical tag, and instead use a different tactic:
- On paginated results (use rel=prev or rel=next instead
- When the page is no longer necessary. Use a 301 redirect to a relevant page instead.
What’s the takeaway?
Bing’s advice in the article is unnecessarily confusing for webmasters and especially for novice SEOs and webmasters. SEO best practices regarding using canonical tags, even self-referential ones, on every page should still be followed, I think. In this case, it’s better safe than sorry. If the search engines tell us that they are smart enough to figure it out, let’s go with that. It’ll save us time in the long run.