A Dreaded Canonical Tag Question

[Colan Nielsen]

I apologize ahead of time for dropping a question about the dreaded canonical tag. But, since we have such a bright community I thought I would give it a shot

I'm wondering if removing a canonical tag from a URL takes away some "juice" or domain authority from the URL that it was pointing at?

Very curious to see if this is a straightforward answer or if it depends on a bunch of other factors. I'm guessing the latter.

 

[Linda Buquet]

Sorry Colan but canonical conundrums are not my cup of tea.

I'll leave this to some of the more techy types in the house.

 

[Simon James]

Why would you want to remove the tag?

 

[Colan Nielsen]

Hey Simon, for the most part it's just a hypothetical question. What are your thoughts on it?

 

[Robert Ramirez]

I apologize ahead of time for dropping a question about the dreaded canonical tag. But, since we have such a bright community I thought I would give it a shot

I'm wondering if removing a canonical tag from a URL takes away some "juice" or domain authority from the URL that it was pointing at?

Very curious to see if this is a straightforward answer or if it depends on a bunch of other factors. I'm guessing the latter.

So if I'm understanding the question, you're essentially asking if canonical tags pass some sort of page rank to the page they are pointed to? That if you had a bunch of pages with canonical tags pointed to another page and then you removed those paths, would the page that lost the canonical tags be affected?

The short answer is no. In my experience, canonical tags do not pass any sort of "juice". The page will rank the same as it did previously.

 

[Eric Marshall]

I've always thought that rel=canonical doesn't pass any sort of juice/authority/rank, but I found this post from Dr. Pet at MOZ, where he claims:

"We suspect it passes authority/PageRank for links to the non-canonical URL, with some small amount of loss (similar to a 301)."

Rel=Confused? Answers to Your Rel=Canonical Questions - Moz

 

[Robert Ramirez]

I've always thought that rel=canonical doesn't pass any sort of juice/authority/rank, but I found this post from Dr. Pet at MOZ, where he claims:

"We suspect it passes authority/PageRank for links to the non-canonical URL, with some small amount of loss (similar to a 301)."

Rel=Confused? Answers to Your Rel=Canonical Questions - Moz

So I respect Dr. Pete, he certainly knows his stuff, but this is contrary to my experience. I've had the opportunity to test this and the difference between the amount of equity that passes when you 301 redirect a page vs. when you use a canonical is extremely different.

Lots of variables here obviously, but the fact that a 301 is a directive and a canonical is a recommendation kind of highlights the difference in my mind. If you're trying to pass link juice to another page, I would always recommend using a 301 redirect.

 

[anim8tr]

...it's just a hypothetical question. What are your thoughts on it?

It's highly unlikely they would penalize for non-use of a canonical.

Google has said that they use the canonical value as a reference and they may not use it at all if other indicators point back to the original URL. We've seen the canonical ignored by Google many times.

 

[Colan Nielsen]

Thanks for reviving a two year old thread! Great insight all.

 

[Eric Marshall]

Funny...I just now realized this was and old question...really old

 

[anim8tr]

Thanks for reviving a two year old thread! Great insight all.

Colin, I work day-in and day-out with people that are not as up-to-date when it comes to canonicals. This is especially true with larger corporations that have smaller divisions with a global presence. There is still a huge misunderstanding of canonicals, even two years after this thread was started. This might not be a question you see today, but the reality is that it's still a relevant question globally.

 

[Colan Nielsen]

Couldn't agree more

 

[Eric Rohrback]

So what would be the effect of using both a 301 and a canonical tag? Here's the case - we see URLs that are showing a shortened technical string (eg; www.website.com/?xyz=123) but we want to redirect that to the user friendly URL (eg; www.website.com/this-is-the-correct-page).

My question is whether a 301 would be sufficient or whether you would want to use both? I think you would want to use both just to be sure, since you wouldn't want search engines to index both and risk duplicate content.

 

[anim8tr]

Here's the case - we see URLs that are showing a shortened technical string (eg; www.website.com/?xyz=123) but we want to redirect that to the user friendly URL (eg; www.website.com/this-is-the-correct-page).

My question is whether a 301 would be sufficient or whether you would want to use both?

Eric, the bot would never see the canonical on the "xyz=123" page with a server-side redirect because it would have already been redirected to the "correct-page" URL. So having a canonical on the "xyz" page wouldn't do anything.

Canonicals were intended for pages represented by multiple URL query parameters, or in situations where the parent page was the original page before it was filtered... but primarily for duplicate content situations. The two (canonicals, redirects) are used for different purposes.

 

[Eric Rohrback]

That's what I thought, but we're having a internal discussion about this issue and can't find the best way to get it resolved. I assumed if the server side redirect was put in place then the tag would be useless, but the weird thing is we're seeing the xyz=123 page indexed over the friendly URL. Very odd things happening.