Yesterday Google announced that it is now supporting the rel=”canonical” link element (sometimes referred to as the canonical URL tag) across different domains. This means that in addition to using the tag to help sort of duplicate content issues on a single domain, it can also be used in dealing with duplicate content on more than one domain. (If you’re not familiar with rel=”canonical” see Publishers: Solve Tracking Code, Duplicate Content Issues with the Canonical URL Tag).
In the Q&A in Google’s announcement, one question in particular caught my eye:
Q: I’m offering my content / product descriptions for syndication. Do my publishers need to use rel=”canonical”?
A: We leave this up to you and your publishers. If the content is similar enough, it might make sense to use rel=”canonical”, if both parties agree.
At Define we work with a lot of newspaper and magazine sites, many of which have syndication deals with multiple partners. “How can we prevent our syndication partners from outranking us for our own content?” is one of the most common consulting questions we get.
As I covered in my post on syndication best practices, to some extent publishers have to except that they can’t have their cake and eat it too; if you allow your content to be published on other sites there is always a chance that they will outrank you for that content. Since most syndication partners will not agree to block their duplicate versions from search engines, the current best practices are built around reducing the risk of being outranked as opposed to eliminating it.
Now cross-domain support of rel=”canonical” is the best available option to publishers for dealing with duplicate content caused by syndication and establishing the content on their sites as the original source.
It will be interesting to see if publishers are able to get a rel=”canonical” requirement added to future syndication contracts. Most likely the partners will push back, but if you can get them to agree to it you absolutely should. Though it is important to note that Yahoo and Bing are not yet supporting rel=”canonical” across domains, so this solution only applies to Google.
As you can see from the Q&A above Google is going out of its way to avoid offering a specific recommendation on this matter. I guess in the current climate they want to avoid telling news sites how to run their businesses. 🙂 But they certainly recognize the tag’s value in helping to sort out duplicate content issues caused by syndication.
UNRELATED SIDE NOTE
Right after I published this post I noticed a mistake with the URL, so I immediately re-published with a different URL. Unfortunately that seems to have interfered with the ability for the TweetMeme button to register retweets of this post. I tried testing and playing around with it but I couldn’t sort it out. Anyone ever have this issue? It’s obviously not that big a deal but that zero doesn’t look so hot.
David Mihm says
Adam,
I seem to remember Matt Cutts expressing some concern when this was announced (at SMX Advanced?) that SEOs would see this as a way to canonicalize non-identical pages across multiple domains…and get de facto link juice flowing into a page they control. Presumably this latest announcement means there is no linkjuice passed by this tag?
Adam Sherk says
Good point David, there’s certainly a lot of potential for abuse and too much misuse will render it useless, so it’ll be interesting to see how Google handles the tag moving forward. The cross-domain announcement doesn’t make any mention of link-related properties. No mention would imply that the tag works the same cross-domain as it does on the same domain, but I haven’t seen anything definitive on that yet. (The original rel=canonical announcement back in Feb simply said “Additional URL properties, like PageRank and related signals, are transferred as well.”) Hopefully people will start sharing their results with cross-domain testing soon.
john andrews says
We need to notice that Google mentions indexing very carefully whenever discussing rel canonical. It is an indexing management tool… label a page non-canonical and you tag it for de-indexing. Keep that in mind when considering scenarios.
I can see how Google might like to keep that low profile among the general population (so it can safely drop 2ndary/tertiary content versions from the index), yet go higher profile in the seo/webmaster community (so as to influence the use of rel canonical).
We do live in interesting times.
Adam Sherk says
A related thought to that John is that cross-domain rel=canonical does help in creating the most positive user experience both from a search engine and an on-site perspective.
For users discovering content via search engines filtering out duplicates makes sense, as showing multiple versions of the exact same article in the top results is not a useful experience. But since lots of users still discover news via direct navigation to particular sites, syndication with rel=canonical allows the content to gain a wider audience without impacting the search experience.
The catch is that most syndication partners won’t be willing to implement the tag, in the hopes that they can still get some search referrals for their versions of the content.