Fixing Mis-used Rel-Canonicals

[CleverPhD]

You do not have to remove the rel=canonical before setting up the 301.  Google generally treats the rel=canonical like a 301, it is just not as strong as a signal. The rel=canonical is more of a hint/suggestion vs the 301 is a command, but they are telling Google the same thing.   So if you point the 301s to the same pages as where the rel=canonicals were pointing to, you should be fine.  Don't worry about if you need to take out the rel=canonical first.

What you do need to worry about you mention above, 301ing (or rel=canonicalling for that matter) to the "correct" page.  If you 301 to a page that is not related to the original page, then Google will not pass along any link equity.  They have to have "semantic relevance" to one another

https://moz.com/blog/save-your-website-with-redirects

Now how "semantically relevant" is the old page to the new one in your case?  You have to decide.

If the best-cat-toy page did not rank well in Google and if there are no quality links to it and you are going to take it out of your site navigation and sitemaps, why 301 it to a new page?  Just let it 404 and go away.   If it does rank well and all those good things, you need to keep it around and/or improve the page and send it to a semantically relevant location.  It may be that you can use that criterion to decide what categories your users are searching for an using on your site and what categories they are not.  Spend your time making those good categories better and lose the old categories, as nobody used them anyway.

Hope this helps!

[geteducated]

Hi Matt, thanks so much for your answer!

What about if the rel-canonical is pointing to a different page than where the 301 will point? Does that rel-canonical need to be removed to avoid confusion and general "messiness"?

[geteducated]

And what about when a page that has a rel-canonical tag pointing to itself? Does that need to be removed before 301 redirecting to the updated/current URL?

[CleverPhD]

Re pointing to a different page.   Ultimately does not matter where it was pointing before.  If you can point it to the most semantically relevant page with the 301 (and the page is worth 301ing for the reasons above) then you are good.

Re pointing to itself.  If it points to itself, you probably should not be redirecting it with a 301 as it is its own canonical page.  If that was an error to point to itself and you need to redirect it, then use the rules above.

[geteducated]

To clarify -- yes, the ones re pointing to themselves were in error (there was a better choice - more relevant/current) and they should have been 301s from the start. So, just to clarify, for those re pointing to themselves, I do not need to remove the canonicals before 301 redirecting?

Ex:

http://www.geteducated.com/careers rel points to itself and is an outdated landing page for the content and soon, it will no longer exist in our system at all.

It should have been pointed here all along and this will be the only landing page for the content going forward: http://www.geteducated.com/career-center/hottest-careers.

We want to 301 redirect http://www.geteducated.com/careers to http://www.geteducated.com/career-center/hottest-careers but try to make sure the link juice / page rank is passes as much as possible.

[geteducated]

Another example:

URL #1 http://www.geteducated.com/all-careers-articles/11-detail/43-forensic-accountant-and-fraud-examiner
has a page rank of 1 and it's rel-canonical points to
URL #2 http://www.geteducated.com/all-careers-articles/11/43-forensic-accountant-and-fraud-examiner which points to itself and has a page rank of 23.

The SAME article can also be accessed here
URL #3 http://www.geteducated.com/career-center/detail/forensic-accountant-and-fraud-examiner 
which has no rel-canonical and a page rank of 1.

We want to use URL #3 going forward and in a new site URL #1 & #2 won't exist anymore. What's the best practice here? 301 both #1 & #2 into #3 without touching canonicals? Should we add a self-canonical to URL #3?

[CleverPhD]

It depends.  See my points above.   If the other "non best" pages don't have any link equity or other worth, then you may just want to 404 them.  If they do generate traffic and/or links and are all semantically related then you can 301 the other pages to the "best" page.

The canonical to self is used by some sites if they are trying to make sure if there are any URL parameters added to the end of the main slug that all the link juice is consolidated to the main slug.  You can canonical to self on all your pages, just make sure that you have a way to QA the process so that you dont mess up any canonicals that should be pointing to another page.

If you are unsure on if you should use the canonical to self technique or not, I would not use it.  The 301s are more important at this point to consolidate across pages.

[geteducated]

I guess what I'm trying to suss out is if any of the described would create a canonical chain that wouldn't pass link juice etc.

I found this in a Moz article:

(9) Can I Chain Rel=Canonicals (+301s, 302s, etc.)?

What happens if you rel=canonical to a URL with rel=canonical to another URL, or you rel=canonical to a URL that 301-redirects to another URL? It gets complicated. In some cases, it might work and it might even pass PageRank. Generally speaking, though, it’s a bad idea. At best, it’s sloppy. At worst, it might not function at all, or you might lose significant PageRank across the chain. Wherever possible, avoid chains and implement rel=canonical in a single hop.

With what this is saying, it seems that I should take out the canonicals before 301 redirecting to try to avoid a chain situation and to give the page rank the best change of getting passed along rather than lost along the way?

[CleverPhD]

When the server 301 redirects a page, the page that used to have the canonical link on it ceases to exist.  The server simply routes the user to the new page without ever showing the old one.

What they are talking about in the chain situation above applies to rel=canonicals and 301s regardless and has nothing to do with your point about removing the rel=canonical off the page before you 301.  It is completely different.

Ideally you would not want to not rel=canonical  page1 to page2 and then rel=canonical page2 to page3, it would be better to rel=canonical page1 to page3 (go direct avoid the chain).

in the same way you would not want to 301 redirect page1 to page2 and then 301 redirect page2 to page3, it would be better to 301 redirect page1 to page3 (go direct avoid the chain).

in the same way you would not want to  rel=canonical page1 to page2 and then 301 redirect page2 to page3, it would be better to 301 redirect page1 to page3 (go direct avoid the chain).

[geteducated]

Thank you very much, I appreciate all the additional clarifications!

[CleverPhD]

You bet!  If you don't mind, please mark this as a good answer then.

Cheers!