First of all, ask yourself if you really have no duplicate content problems (or, even better, use a tool such as Screaming Frog SEO Spider to find out for sure). You may be surprised – an awful lot of CMS and e-commerce packages produce duplicate content problems without meaning to (Magento creates three crawlable URLs per product, for example). Almost any kind of information architecture that uses categories, tags or other taxonomies can be a minefield.

But let’s say you really, definitely have no problems – do you still need the tag? The answer is yes. The reason is that although your own platform may not be creating duplicate content, other people or platforms might be causing them on your site, by linking to you with non-canonical URLs (typically by adding tracking parameters).

That may sounds like an unlikely scenario, but consider this:

  • You send an email out to your subscribers. That email, following standard best practice, adds tracking information to the links, so you can analyse its performance.
  • That email is automatically syndicated to your site, Facebook, or somewhere else where Google can crawl it: you have instantly created duplicate content problems.
  • Even if you are not syndicating your emails, one of your subscribers may copy and paste one of the links somewhere public (you’d surely hope that they will be doing so!) Again, that’s a duplicate content problem.

You could further argue that Google can detect and discard tracking parameters automatically, and yes, it can (you can even give it a helping hand in Google Webmaster Tools). However, tracking parameters are just one example of something that can cause a problem, and either way, it is usually best practice to solve these problems yourself rather than trusting Google to do the job. Considering that canonical tags are usually easy to implement across your site, there’s no real reason not to.

Canonical tags are also used for other specific purposes, such as dealing with paginated lists, but that’s another question…