Rel canonical on other page instead of duplicate page. How Google responds?

[vtmoz]

Hi all,

We have 3 pages for same topics. We decided to use rel canonical and remove old pages from search to avoid duplicate content. Out of these 3 pages....1 and 2 type of pages have more similar content where 3 type don't have. Generally we must use rel canonical between 1 and 2. But I am wondering what happens if I canonical between 1 and 3 while 2 has more similar content? Will Google respects it or penalise as we left the most similar page and used other page for canonical.

Thanks

[Igor.Go]

I wouldn't do that.
By adding a canonical tag to the page you kind of tell google to ignore it. So, in this case, you have pages 1 and 2 with similar content and 3 with different content. You add canonical to page 3 referring to page 1. Google will now ignore page 3 (the one with different content) and will still index pages 1 and 2 which are duplicates.
You will not solve the duplicate problem, and you will also harm the unique page.

[vtmoz]

HI,

Thanks for the immediate response. I agree with your analysis and conclusion. What if the duplicate page we are leaving is redirected to the page we are pointing?

I meant "1 will be pointed to 3 instead of 2" and "2 will be redirected 3"

How this works?

Thanks

[Linda-Vassily]

This sounds like you will be pointing/canonicalizing the two similar pages to the third one that is different from them? I am not quite sure why you would want to do that.

If you don't want the 1/2 content available but the pages have some authority (good links), 301 redirect those pages to 3 (if the topic is close enough and you don't have a more similar page) or if they are not strong pages, just remove them and let them 404.

If you do want the 1/2 content available on your site, but don't want it competing with page 3 in search, you could redirect 2 to 1 and rewrite 1 to make it stronger for whatever it is that makes it different from 3, so both 1 and 3 could potentially rank (for different things). Or you could redirect 2 to 1 and noindex 1.

Canonicals are intended for pages with very similar content, however people sometimes do use them as a type of redirect for not-so-similar pages. The problem with this is that a canonical is just a suggestion to Google and, as you mention, Google may ignore the canonical, especially in a situation like this.

[EGOL]

We know of a case from early 2017  in which Google stopped honoring rel=canonical for a large number of pairs of pages that were not verbatim duplicates.  Shortly after that all of those pages were indexed and displayed in the SERPs.

[vtmoz]

Thanks for the answers and suggestions. I have more questions raised in my mind and I put them in the below different thread very clearly. Please reply there.

https://moz.com/community/q/what-happens-when-we-canonical-and-point-to-a-page-which-has-been-redirected-to-another-page-google-response