A canonical tag is a piece of code you can add to a page to instruct search engines that a more important version of that page exists. If the canonical page is a near-duplicate, search engines will consolidate the equity from the canonicalised page and attribute it to the canonical version. The canonicalised page will no longer appear in search results, and the canonical version may perform better in terms of rankings, this is because it has inherited the equity (i.e. it benefits from the links) of the canonicalised page. Also, if there are fewer very similar pages on the same domain targeting the same (or similar keywords), search engines will be clearer about which one should rank for the target keyword(s).
The canonical version of a page is the primary or ‘most important’ version.
The canonicalised page is the one that has had its equity attributed to the canonical version.
If canonicalised URLs genuinely are exact duplicates of the canonical page, then a better solution might be to 301 the canonicalised URLs to the canonical version and then change all the internal links to point at the canonical version.
Many URLs can be canonicalised to a single URL, and URLs can be canonicalised to themselves; and that’s what we’re going to explore in this article. There are several benefits to canonicalising URLs to themselves and in my opinion, this should be the default for most websites; having pages canonicalise to themselves by default is a much better option than no canonical tags (John Mueller also recommends self-referential canonical tags https://www.youtube.com/watch?v=XOGOhWyNSf8).
Here are a few of the reasons;
Websites that use parameterised URLs leave themselves open to many SEO issues unless they’re diligent about the way that they handle such URLs. One of the issues with parameterised URLs is that many very similar URLs can end up getting indexed, in many cases they will be competing for the same keyword. All these URLs may also pick up external links, of course this is a good thing, but link equity distributed across many similar URLs may not be as good as all that link equity pointing at one URL.
If self-canonicalising URLs is a standard across the domain from the outset, then fragmented link equity across several URLs will be less of an issue, since any parametrised versions of URLs will simply canonical back to the non-parameterised version
Here’s an example of Pretty Little Thing following this practice; the parameterised URL that displays all orange clothing is canonicalised to the ‘all clothing’ page by default, meaning that any links the parameterised version picks up will benefit the canonical URL
Of course, there may be some situations where a custom canonical tag could be tested, for example, If Pretty Little Thing thought they could rank well for ‘black dresses’, they may want to change the canonical tag of this URL
To point to itself rather than the main dresses page, and test if this yields any incremental traffic/conversion uplifts.
2. URL CHANGES
Check out this URL on the Australian version of Wendy Wu Tours
For some reason, it has been canonicalised to the UK version of the page, i.e.
Unfortunately, the canonical URL is a 404 page. Whether this was a mistake or something intentional that hasn’t been picked up on since the UK version was culled, it would certainly be beneficial for Wendy Wu to regularly audit their canonical tags and reset this page (and any others with similar issues) to self-canonicalise.
3. HTTP/HTTPS CONFLICTS
Anecdotally, I’m still seeing a fair few sites in the SERPS that aren’t using HTTPS to deliver their content (even from some big brands – see the O2 example below). These sites are potentially missing out on a small ranking boost by not using the secure protocol, not to mention that their users could be put off by the ‘Not Secure’ notification shown in Chrome for HTTP pages.
What’s worse than using HTTP to deliver content, however, is using both protocols to deliver the same content; this presents the same old potential typical duplicate content issues such as fragmented link equity and search bots not being able to determine the primary or canonical version of the content.
In many cases, an HTTPS migration may have been carried out, i.e. redirecting all HTTP pages to their HTTPS equivalents, but Google may still be able to access the HTTP versions if they are referenced in the canonical tags of the HTTPS page.
This gives Google mixed signals about which version of the page is the correct one, and another potential issue that could be circumvented with self-referential canonical tags.
WHAT SHOULD YOU DO ABOUT IT?
Firstly, this potential issue is more likely to occur on larger, older sites where URLs have been changed or migrated many times, or on sites that use parameters. Having said that, even for new or smaller sites, implementing self-referential canonical tags from the outset won’t do any harm and could futureproof against some of the issues above.
Of course, there may be situations where canonical tags need to be modified, but it’s highly likely that the vast majority of your pages should self-canonical, so implementing this from the start and making changes where necessary could mitigate the impacts of any of the issues above – going to the effort of doing this is better than having no canonical tags.
To implement this, speak to your dev team, CMS manager, or if you are both of those things, you’ll most likely be able to find a WordPress plugin that does this for you. I’m currently looking for one, so if you know of a good one, please tell me!