Pagination on a product page with reviews spread out on multiple pages

[Don34]

Our current product pages markup only have the canonical URL on the first page (each page loads more user reviews). Since we don't want to increase load times, we don't currently have a canonical view all product page. Do we need to mark up each subsequent page with its own canonical URL?

My understanding was that canonical and rel next prev tags are independent of each other. So that if we mark up the middle pages with a paginated URL, e.g:

Product page #1http://www.example.co.uk/Product.aspx?p=2692"/>http://www.example.co.uk/Product.aspx?p=2692&pageid=2" />**Product page #2 **http://www.example.co.uk/Product.aspx?p=2692&pageid=2"/>http://www.example.co.uk/Product.aspx?p=2692" />http://www.example.co.uk/Product.aspx?p=2692&pageid=3" />Would mean that each canonical page would suggest to google another piece of unique content, which this obviously isn't. Is the PREV NEXT able to "override" the canonical and explain to Googlebot that its part of a series? Wouldn't the canonical then be redundant?Thanks

[DirkC]

Hi Don,

Normally it's not really necessary to put a canonical on each the pages with different pageid - the most important reason to put a canonical url on a page is to let Google know that if the same content is published under different url's which url needs to be indexed. In your example - a canonical url on http://www.example.co.uk/Product.aspx?p=2692&pageid=3 would be needed if the same content would also be available under http://www.example.co.uk/Product.aspx?p=2692&pageid=3&sessionid=123456 (or any other additional parameter). Strictly speaking the canonical is not to indicate Google that the content is unique but to indicate on which preferred url you want the content to be indexed. That is also the reason why you can implement both together.

If you check the example from Google  - they use the rel next/previous to indicate that the different pages belong together - the canonical is basically used to indicate that the session id needs to be ignored when indexing the page.

If you are sure that there is only one version of http://www.example.co.uk/Product.aspx?p=2692&pageid=3 the canonical url is not really necessary. Note that e-commerce platforms have a tendency to add additional parameters to url's depending on the view - example when sorting on price etc - if this is the case the canonical would be needed.

Hope this helps,

Dirk

[Don34]

Hi Dirk,

Thanks for the answer but I'm still a tiny bit confused.

Currently we only have the Product.aspx?p=2692 first page including a canonical link, the rest of the variations don't have canonicals, ex:

Product page #1 http://www.example.co.uk/Product.aspx?p=2692"/> http://www.example.co.uk/Product.aspx?p=2692&pageid=2" />

**Product page #2 **http://www.example.co.uk/Product.aspx?p=2692" /> http://www.example.co.uk/Product.aspx?p=2692&pageid=3" />

Should we go ahead and add a canonical link for these pages with each of the unique pageid which specifies the page number? Because that was always my understanding of the Google support page for pagination. But then like I said we are basically telling Google that these pages are all unique, when if fact just the additional 10 reviews on them is new content.

[DirkC]

Hi,

If you want all pages with the same product id p=2692 and different &pageid=x to be considered as one big page your implementation is correct. Canonicals can be used in parallel with rel next/previous. The example you give is quite similar to the example given by Google: https://support.google.com/webmasters/answer/1663744?hl=en (example at the bottom of the page). The canonical you use on the pages aren't really necessary, but they don't hurt either so you can leave them there.

The one thing you have to avoid when mixing canonicals & rel next/previous is when you implement it like this:

**Product page #2 ** http://www.example.co.uk/Product.aspx?p=2692"/>
http://www.example.co.uk/Product.aspx?p=2692" /> http://www.example.co.uk/Product.aspx?p=2692&pageid=3" />  => in that case you would be sending mixed signals to Google - on one hand indicating that all the pages with different pageid's should be considered as one big page & on the other hand saying that Google should only index the first page as the pages with different pageid's are duplicates.

Hope this clarifies,

Dirk