The company I work for has 5 domain names for 5 different markets within the country. Those domains have their own branded sites and share content between them. We place canonicals on each piece of content so we don't get pinged for duplicate content. Everything is above board. All SEO is white hat, nothing fishy going on.

If domainA.com/articleA has a canonical of domainB.com/articleA and is placed on marketA's Facebook page Facebook will use the domainB.com address!! This is highly undesired because each market have their own managed Facebook accounts and domainB getting the link juice for domainA's work is just not going down well.

I can see that I can have an og:url tag which doesn't help us as there will still be only a single URL in that method.

Next to removing the canonical, I don't see how we can do this and I'm desperately hoping that there is some method for doing what we need.

  • So Facebook is looking at the canonical tag and using that as the link when posted, and not the actual URL?
    – Max
    Oct 7 '14 at 11:35
  • Is it possible to show an example? Also you mention the og:url wont help, but I'm not sure I follow, how do you mean only 1 URL if using the og:url? If you define domainA.com/articleA as the og:url does it not just use the domainA.com/articleA as the link?
    – Max
    Oct 7 '14 at 11:44
  • Hi Max, first answer; Yes. Facebook uses canonical and not actual URL. Second answer; Yes I can define that as og:url but what happens when domainB.com/articleA is posted to FB? The canonical is used and it posts domainA.com/articleA.
    – Christian
    Oct 7 '14 at 22:49

One solution would be to not show the rel canonical meta tag when Facebook is crawling the page. Facebook's bot uses a user agent string that contains "facebookexternalhit". If "facebookexternalhit" is in the user agent, then don't show the meta tag.

Note that showing different content to a robot is called "cloaking". I haven't seen any evidence that Facebook doesn't allow this type of cloaking for their user agent. If you were to cloak for Googlebot like this, then Google could very likely remove your site from the Google search results.

  • I had considered cloaking but every time it is referred to its as a black hat practice. We're thinking about dynamically setting og:url to whatever the current domain is which allows us to keep canonical.
    – Christian
    Oct 9 '14 at 1:33
  • Cloaking for Googlebot is black hat to Google. Other cloaking is not so black hat, especially when you are not trying to deceive users. Oct 9 '14 at 2:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.