Canonical Expert question!

[artdivision]

Hello,

I am looking for some help here with an estate agent property web site. I recently finished the MoZ crawling report and noticed that MoZ sees some pages as duplicate, mainly from pages which list properties as page 1,2,3 etc. Here is an example:

http://www.xxxxxxxxx.com/property-for-rent/london/houses?page=2
http://www.xxxxxxxxx.com/property-for-rent/london/houses?page=3 etc etc

Now I know that the best practise says I should set a canonical url to this page:
http://www.xxxxxxxxx.com/property-for-rent/london/houses?page=all

but here is where my problem is.

http://www.xxxxxxxxx.com/property-for-rent/london/houses?page=1 contains good written content (around 750 words) before the listed properties are displayed while the "page=all" page do not have that content, only the properties listed.

Also http://www.xxxxxxxxx.com/property-for-rent/london/houses?page=1 is similar with the originally designed landing page http://www.xxxxxxxxx.com/property-for-rent/london/houses

I would like yoru advise as to what is the best way to can url this and sort the problem. My original thoughts were to can=url to this page http://www.xxxxxxxxx.com/property-for-rent/london/houses instead of the "page=all" version but your opinion will be highly appreciated.

[GPainter]

Have you considered the paginated tag ?  you could also have a page with a view all option and canonical to that and thus get all the content listed. Why wouldn't the view all page have the same content as each page ?

[artdivision]

i have but i was hoping to know if this is solved by adding rel=canonical to the original content landing page? http://www.xxxxxxxxx.com/property-for-rent/london/houses

all page have same content but the text content for some reason appears only on http://www.xxxxxxxxx.com/property-for-rent/london/houses page and on http://www.xxxxxxxxx.com/property-for-rent/london/houses?page=1 page

[GPainter]

It would work but the content after that e.g http://www.xxxxxxxxx.com/property-for-rent/london/houses?page=2 would but lost as they would not be indexed. so if there is content on those pages you feel is valuable might want to look int alternatives however is the strongest content is on http://www.xxxxxxxxx.com/property-for-rent/london/houses you will be fine to set that as the tag location.

[Dr-Pete]

Do "/houses" and "/houses?page=1" have exactly the same content? I'd definitely want to see rel=canonical on the "page=1" version - those are just duplicates. Google has expressly said that they don't want you to canonical pages 2, 3, etc. back to page 1. That doesn't mean it never works, just that it's a bit dicey.

As Chris said, rel=prev/next is another option. Theoretically, it would allow all of the results pages to rank, but let Google know they're a series and not count them against you as thin content. In practice, even my enterprise SEO colleagues have mixed feelings. There's just very limited evidence regarding how effective it is. It is low-risk.

The other option is to go a bit more old-school and META NOINDEX anything with "page=", and just let the original version get indexed and rank. This can help prevent any dilution and would also solve your "page=1" issue. The biggest risk here is if that cut off PR flow across your site or if you had links to the paginated results. In most cases, that's unlikely (people don't link to or tweet page 17 of your search results), but it's a case-by-case thing.

Unfortunately, the "best" solution can be very situational, and even Google isn't very clear about it.