How use Rel="canonical" for our Website

[ZZNINTERNETMEDIAGROUP]

How is the best way to use Rel="canonical" for our website www.ofertasdeemail.com.br, for we can say goodbye for duplicated pages?

I appreciate for every help. I also hope to contribute to the SEOmoz community.

Sincerely,
Amador Goncalves

[MikeRoberts]

First, determine if those duplicate content pages need to exist or if your users would be better served by another page. If that page doesn't need to exist then you may want to consider a 301 redirect to the better page. If a page is an exact replica of another page then you need to ask yourself "Why do we have it?" If its only a duplicate because of thin content then you might want to consider adding more, relevant content to the individual pages to better differentiate them.

If the duplicate page needs to stay for whatever reason then you can consider adding a canonical tag pointing to the primary page. Some cases in which canonicals have worked best on the sites I work on have been relating to parameters. E.G. example.com/product and example.com/product?model=4 are basically the same page but they each serve a purpose. In this case, example/com/product?model=4 is a subset of the one without a parameter and was given a canonical tag pointing to the primary page.

Canonical tags are a signal, not a directive though... which means that the search engines may choose to listen to it or ignore it as they see fit.

I apologize if any of that seems confusing. Here's a link to the SeoMoz guide on canonicals: http://www.seomoz.org/learn-seo/canonicalization and a blog post on the subject: http://www.seomoz.org/blog/canonical-url-tag-the-most-important-advancement-in-seo-practices-since-sitemaps

[ZZNINTERNETMEDIAGROUP]

Thanks Mike,

All pages, and the offers pages need to existe.

When I enter in the store's offers page, show a list with last offers of store (so users will can use the offers selector on left side of website, so after click, the offers will appears on the right side of website).

For example, the Submarino Store:
http://www.ofertasdeemail.com.br/desconto/submarino

Please do a test, and help us.

Thanks

[MikeRoberts]

Ah... Googlebot can't see those changes on the page from when you click the different offers so each page looks almost exactly the same and like thin content. In which case I'd suggest something along the lines of adding more written content to pages like _http://www.ofertasdeemail.com.br/desconto/submarino/_ so they look different than page like _http://www.ofertasdeemail.com.br/desconto/submarino/so-o-cartao-submarino-indica-as-melhores-ofertas-para-voce-10733.html_ and then adding a canonical tag to show that the offers are a subset of _http://www.ofertasdeemail.com.br/desconto/submarino/_

They only problem with this solution is that the individual offers won't really rank for anything in the SERPs and will likely be replaced by the primary page in the index (assuming Google follows your canonical signal). So that may not be a perfectly solution for your needs but it could alleviate problems associated with duplicate & thin content.

[ZZNINTERNETMEDIAGROUP]

All offers of our website open through ajax.

How is the best use of canonical in this case?

[Dr-Pete]

Yeah, I'm with Mike - these are prone to cause you some real trouble. Given how many there probably are and how often they change/rotate, I'd strongly suggest using rel=canonical or some not indexing the alternate offers somehow.

They may be necessary for users, but these pages aren't all necessary to have in your index. By trying to rank for every single one, you risk harming your more important rankings. Honestly, as Mike said, Google can't even really tell these are different, except for the URLs, so even the long-tail ranking benefits are nearly zero, I suspect.