Thursday, January 23, 2014

Google Canonical Quirk?

How well does Google handle the canonical tag?

We just found a quirk that makes us wonder. Here's what piqued our interest - we noticed some landing pages in our Google Analytics that were not consistent with our url structure. Check the image below:

As you can see, two of our top landing pages appear twice under slightly different urls. One version includes a 4-digit number (the post id) at the end of the url and the other does not.

Several days ago, before we published these articles, we changed to our current url structure where we added the post id to the end of each post url (in the hope of getting into Google News). So it could make sense that some source of traffic might get the old urls. But we published these particular articles after the change, so they were never "published" to search engines under the old structure.

Our Facebook, Twitter and Google+ posts were all linked using the new url. And we do include the tag in our posts, pointing to the url with the 4-digit number in it, as you can see below:

At first we weren't sure what source was sending users to the link without the post id. So we searched for the article title and sure enough, we saw that Google was directing users to the old url:

We don't understand why, but the first search result has my authorship, and the link does not include the 4-digit code. The second and third results do include it, and the 2nd one is from before we changed our url structure.

Is this important? It doesn't seem to be a huge deal. But it's an interesting quirk in Google's search results

No comments: