Cross-Domain Canonical and duplicate content

[crvw]

A cross-domain canonical will help with Google.  (make sure the pages truely are duplicate or very close), however, I haven't found any confirmation yet that Bing supports Cross Domain Canonical.

If the other sites don't need to rank at all, you could also consider no-roboting the job pages on the other sites, so that your only Site A's job listings get indexed.

[RyanKent]

According to Dr. Pete Bing does support cross-domain canonical.

If you disagreed I would first recommend using rel=author to establish "Site A" was the source of the article.

[seoug_2005]

You mean rel=author on site A ? How does it help ? Where should rel=author points to ?

[MaartenvandenBos]

Thanks Ryan!

So link is like:

On the site a i will use the canonical to point everything to site A.

[MaartenvandenBos]

Thanks!

It's for a site in the Netherlands and google is about 98% of the market. Bing is comming up so a thing to check.

No-roboting is a way to do it i didn't think about! thanks for that. I will check with the client.

[RyanKent]

That is correct. If you choose to read the information directly from Google it can be found here:

[RyanKent]

I would advise NOT using the robots.txt file if at all possible. In general, the robots.txt file is a means of absolute last resort. The main reason I use the robots.txt file is because I am working with a CMS or shopping cart that does not have the SEO flexibility to noindex pages. Otherwise, the best robots.txt file is a blank one.

When you block a page in robots.txt, you are not only preventing content from being indexed, but you are blocking the natural flow of page rank throughout your site. The link juice which flows to the blocked page dies on the page as crawlers cannot access it.

[MaartenvandenBos]

Thanks Ryan!

So it will be a Canonical tag

[RyanKent]

My apologies Atul. I am not sure what I was thinking when I wrote that. Please disregard.

[seoug_2005]

I  was bit confused when i read that. You put my mind to rest !

[crvw]

Ryan, at the end o the thread you linked to, it seems like both Dr. Pete and yourself, agreed that there wasn't much evidence of bing support.  Have you learned something that changed your mind?

I know a rep from Bing told Dr. Pete there was "some" support, but what does that mean?  i.e. Exactly Identical sites pass a little juice/authority, or similar sites pass **a lot **juice/authority?

Take a product that has different brands in different parts of the country.  Hellmanns's and Best Foods for example.  They have two sites which are the same except for logos.  Here is a recipe from each site.

http://www.bestfoods.com/recipe_detail.aspx?RecipeID=12497&version=1

http://www.bestfoods.com/recipe_detail.aspx?RecipeID=12497&version=1

The sites are nearly identical except for logo's/product names.

For the (very) long tail keyword "Mayonnaise Bobby Flay Waldorf salad wrap" Best Foods ranks #5 and Hellmann's ranks #11.

I doubt they have a SEO looking very close at the sites, because in addition to their duplicate content problem, neither pages has a meta description.

If the Hellmanns page had a

[http://www.bestfoods.com/recipe_detail.aspx?RecipeID=12497&version=1](http://www.bestfoods.com/recipe_detail.aspx?RecipeID=12497&version=1)"/>

I'd expect to see the Best Foods page move up and Hellmanns move down in Google.  But would Bing appears to not like the duplicate pages as much, currently the Best Food version ranks #12 and the Hellmann doesn't rank at all.  My own (imperfect tests) lead me to believe that adding the rel="canonical" would help in google but not bing.

Obviously, the site owner would probably like one of those two pages to rank very high for the unbranded keyword, but they would want both pages to rank well if I added a branded term. My experience with cross-domain canonical in Google lead me to believe that even the non-canonical version would rank for branded keywords in Google, but what would Bing do?

I'd be very cautious about relying on the cross-domain canonical in Bing until I see some PUBIC announcement that it's supported.
```

[RyanKent]

One aspect of the SEOmoz Q&A structure I dislike is the ordering of responses. Rather then maintaining a timeline order, the responses are re-ordered based on other factors such as "thumbs-up" and staff endorsements. I understand the concept that replies which are liked more are probably more helpful and should be seen first, but it causes confusion such as in this case.

Dr. Pete's response on the Bing cross-canonical topic appears first, but it was offered second-to-last chronologically speaking. We originally agreed there was not evidence indicating Bing supported the cross-canonical tag, then he located such evidence and therefore we agree Bing does support the tag.

The statement Dr. Pete shared was that "Bing does support cross-domain canonical". There was no limiting factor. I mention this because you said they offered "some" support and I am not sure why you used that qualifier.

[crvw]

Thanks Ryan, I didn't noticed that about the reply sequencing, and you're right, I read them in the wrong order.  It makes much more sense now.

By "some" support, I meant that even Google via Matt Cutts says that they don't take cross domain canonical as "a directive" but rather a "hint" (and even that assumes Google agrees with you, that your pages are duplicates).

So the magic question is how how much authority do Bing and Google give the rel="canonical" and is it similar between the two engines?

[RyanKent]

Every document I have seen all agrees that canonical tags are followed when the tag is used appropriately.

The tag could be misused either intentionally or unintentionally in which case it would not be honored. The tag is meant to connect pages which offer identical information, very similar information, or the same information presented in a different format such as a modified sort order, or a print version. I have never seen nor even heard of an instance where a properly used canonical tag was not respected by Google or Bing.