Anyone who has worked with SEO has run into the term ‘canonical tag.’ There has been a great deal of confusion over the proper use of canonical tags: when to use them, how to use them, and why would anyone need them anyway? In this blog article, I will discuss how to properly use canonical tags to cut down on duplicate content, and how the improper use of canonical tags can be detrimental to your site’s search engine rankings, as I found out working on Accelebrate’s website.
Why Use Canonical Tags?
Having duplicate pages in your site can negatively affect your search engine rankings, especially in Google. With Google’s release of Panda and Penguin (algorithmic updates for ranking websites) came some major changes for how sites are analyzed and ranked. One of those changes was that Google would have a much lower tolerance for duplicate content within a site. Having identical, or almost identical, pages is seen as spam and will result in those pages being on Google’s naughty list (and possibly harm your site’s standings as well).
What are Canonical Tags?
Canonical tags tell the search engines which page is the primary page, and which pages are duplicates. By putting a canonical tag in a webpage, you are telling search engines ‘don’t index this page; instead, only index the parent page.’
For example, many online shopping sites will use canonical tags to cut down on duplicate pages for a product. Let’s say you are selling a scarf in many different colors and you have a dynamically generated web page for each different color. The individual scarf pages are going to be almost identical; same materials, same dimensions, same description, but a different color. You wouldn’t want Google to index each of the individual color pages, just the main page with the scarf’s description and links to the colors that could be chosen.
Main page for the scarf – http://www.madeupsite.com/accessories?category=scarf
By putting a canonical tag in each of individual color pages, search engines will consolidate those individual URLs to point to a single, preferred URL. In this case, the target would be the main scarf landing page.
The tag below would go within the <head></head> of each of the color pages:
This would cause the search engines to ignore the color pages, and instead serve up the main scarf page in a search.
The Accelebrate website uses canonical tags in a similar way. Our site has different pages for different versions of the same technology. For example, we have three pages for RoboHelp because we have three versions that we offer (RoboHelp 9, 10, and 11). The pages are quite similar because the outlines for the classes and the objectives are almost the same. Our solution was to use canonical tags, and it has worked well.
Canonical versus NOINDEX
So why not just use a noindex tag on pages you want Google to ignore? A noindex tag simply tells the search engines ‘ignore me and I stand alone.’ A canonical tag tells the search engines, ‘you can ignore me, but look, there’s my dad!’ This is important so that in Google’s eyes the duplicate pages in a group will all collapse into a primary page, which gets served up in the search results. Canonical tags are recognized by Yahoo!, Bing, Google, and other search engines.
Improper Use of the Canonical Tag
While canonical tags are very useful if implemented correctly, as in the examples I have shown you, misuse of the canonical tag can have dire consequences on your site’s SEO. We did not use canonical tags correctly and there were consequences. I will divulge the ugly details of what happened in our little experiment, but remember, we are professionals: do not try this at home.
Accelebrate’s site contains many pages for certain types of training in certain markets that we serve, and naturally we would like to show up well for those combinations. For example, we have pages for Atlanta Java Training, Boston HTML5 training, Washington DC XML training, Chicago JavaScript Training, etc. The only problem is that the content in these pages is a bit redundant and we felt we were being penalized by Google for having duplicate content on our site.
We decided to put canonical tags in these city pages and have them point to the technology’s landing page. The theory was that by reducing the duplicate content in our site we would have an overall boost in our search rankings. Using a noindex tag would guarantee that the pages would be ignored, but if we used the canonical tag, then anyone searching for training in specific markets could at least be pointed to the relevant landing pages, and our overall rankings would go up due to less duplicate content.
This did not happen.
Our site did not gain any SEO ground at all overall, and the city pages for which we had canonical tags took a nose dive in search results. We put the canonical tags on in May, and just took them off in September. The chart below shows the 2014 Organic Search Traffic to the City pages of the site. As you can see, there was the large drop in May of 2014; however, there is an increase in September of 2014.
2014 Organic Search Traffic to City Pages
With the canonical tags in place, if someone were to search on a particular city+technology, at least a related landing page should have shown up on search. Unfortunately, our landing pages were also being negatively affected. Perhaps part of the issue was that the technology landing pages were not similar enough to the duplicate city pages. Google only permits minor differences between the canonical page (e.g., the parent page) and the duplicate pages. Ours were quite different, so the city pages were never good candidates for canonical tags. Since taking off the canonical tags, we have seen the rankings for our landing pages go up.
Also, our city pages weren’t being indexed, but that should have been expected. Using a canonical tag is supposed to be a strong hint to Google to ignore certain pages, so it is theoretically possible that some of those duplicate pages may have been acknowledged anyway. However, this was not our experience at all. I have to admit we wanted to have our cake and eat it, too. We wanted to rank higher in Google overall by getting rid of duplicate pages, but we still wanted those duplicate pages to somehow still show up in a search. Poof!
To solve the problem of having city pages with similar content, we are now in the process of creating useful, unique content on these pages, rather than trying to hide them via the canonical tag. This is where the Google juice is, and I’m afraid there are no shortcuts.
Canonical tags should only be used on pages that are truly duplicate pages that collapse into a parent page that is very similar to the duplicate pages. The parent page will then be served up on search results.
Tips for Using Canonical Tags
Use an absolute path rather than a relative path.
Canonical tags should go in the <head></head> of the document.
Be sure that the link you are providing is correct and exists. Otherwise, your duplicate pages will not lead the search engines to any other page and will be lost.
Make sure that the target page of your canonical tag does not have a noindex tag in it.
Do not put more than one canonical tag in a page. Otherwise, Google will ignore it.
Author: Anne Fernandez, one of Accelebrate’s SEO team members
Anne is the web content specialist and instructor manager at Accelebrate. She manages digital marketing initiatives and search engine optimization, makes regular updates to the website, and oversees all instructor travel.
