Ecommerce Canonical Question

[DSCarl]

Hi all, first question (eek)

Could I pick the brains of fellow users around an issue we are having with canonical urls on a magento website. At the moment we do not have these enabled as it seems to break our indexing. Cut a long story short, we have thousands of products but haven't rewritten many of the descriptions from the manufacturers yet and so have noindexed all the product pages (freeing them as we go). The goal, for now, is to pull in traffic via the filtering options we have on the site

The goal, for now, is to pull in traffic via the filtering options we have on the site. For example, if you go to Dresses, there then are several filtering options which would allow you to choose a colour, shape and material - if you wished to filter that precisely.  These filtering options are all crawlable and so we would then have a page that google could index for, for example,  Green Lace Maxi Dress. All good there, few people search for specific products and a lot search for types of products so we are covered.

To get back to the issue at hand. If we enable the canonical option on our magento plugin it will stop us from being able to target these terms. Whereas the filtering option would create domain.com/dress/green/maxi/lace with the page title of Green Lace Maxi Dress, if we enable the canonical part of the seo plugin the canonical link which would be added to the page would be  - instantly removing our ability to rank for longer tail dress related searches (we are not going to compete with the big players on the premium terms, yet!).

There are alternative plugins we can buy for magento to add the correct  tag, however, if every page's canonical just points back it itself like this, is there really much point spending nearly $1000 on the 4 licences we would need to cover our range of sites.  Is it really necessary, in this case, that we have a canonical for the product filtering?

Sorry for the long post, hope it made sense.

Thanks for any assistance.

[DSCarl]

hmm not sure how 'The goal, for now, is to pull in traffic via the filtering options we have on the site' ended up twice. Oops

[becole]

DSCarl, taking care of the duplicate content that the site appears to be generating is a big deal. So it definitely needs to be fixed--and that's good that you've identified it.

Ideally, you really do need to be able to canonical the sizes of the dress, for example, to the product page, which is "Green Lace Maxi Dress", assuming that you will have a unique page (along with a unique product description) written for the Green Lace Maxi Dress, which would be different than, say, a Red Lace Maxi Dress.

There are generally two ways to deal with duplicate content like this. One way is to deal with it using canonical tags. But before we had the canonical tag, we certainly did have duplicate content--and we dealt with it using the robots.txt file. You can deal with this issue with a canonical tag or robots.txt.

With the robots.txt file, you would need to identify which pages (for example by looking at your URL parameters) and stop the search engines from indexing URLs with certain parameters in them. This is pretty easy to do if you understand your site structure, your parameters in your URLs (or how you have those set up in folders in the URLs), and can add those to the robots.txt file. Using the robots.txt file sounds like it would be a cheaper option for you (rather than spending $1,000 on plugins or add-ons to your CMS).

Alternatively, the canonical tag is the way to go if you can get it to work properly. Oftentimes if it's not working properly you can contact the developer of the plugin or add-on and see if they'll help you install it or get the settings right so that it works properly on your site.

Either way, it's definitely an issue that you need to deal with, as it will have a dramatic effect on your site's rankings. The canonical tag option is probably preferred if you can get it to work properly, as all of the "link juice" and other "credit" will be passed onto the page you're canonicaling "to".

[DSCarl]

Bernadette, many thanks for the reply.

This is a very tricky issue, and I'm still very confused. If I go to dresses and then 'red' i get a set of products, if i go dresses and then 'blue' i get a different set of products so I cannot see any potential for duplication there. All of these products may appear on other pages in small numbers - for example one of the red dresses would appear in the lace dress filter as it is a red lace dress, however, the other 15 dresses on that page would not match the red dress page.

There are about 250,000 products in the store, so while products will appear under several attributes, the chance of the same 16 products appearing on the same page for multiple attributes is tiny.  I guess this is why I'm thinking there wouldn't be a need for a canonical tag as no page is really duplicating any more than a tiny % of another.  There is only one way to access the list of products on any one page.

If i borrow Amazon as an example - would rather not share my url at the moment as it isn't 100% finished.  On amazon we have a page for Black Prom Dresses at https://www.amazon.co.uk/s/ref=sr_nr_p_n_size_two_browse-_0?fst=as%3Aoff&rh=n%3A83450031%2Cn%3A!425571031%2Cn%3A!425573031%2Cn%3A9780907031%2Cp_6%3AA3P5ROKL5A1OLE%2Cp_n_size_two_browse-vebin%3A1946173031&bbn=9780907031&ie=UTF8&qid=1476896938&rnid=1836991031&lo=clothing  (whoa that's a big url!) with the canonical tag of https://www.amazon.co.uk/Prom-Dresses-Edit-Black-Clothing/s?ie=UTF8&bbn=9780907031&page=1&rh=n%3A83450031%2Cn%3A!425571031%2Cn%3A!425573031%2Cn%3A9780907031%2Cp_6%3AA3P5ROKL5A1OLE%2Cp_n_size_two_browse-vebin%3A1946173031" />

If we go to red dresses we have a url https://www.amazon.co.uk/s/ref=sr_nr_p_n_size_two_browse-_0?fst=as%3Aoff&rh=n%3A83450031%2Cn%3A!425571031%2Cn%3A!425573031%2Cn%3A9780907031%2Cp_6%3AA3P5ROKL5A1OLE%2Cp_n_size_two_browse-vebin%3A1946185031&bbn=9780907031&ie=UTF8&qid=1476897017&lo=clothing

https://www.amazon.co.uk/Prom-Dresses-Edit-Red-Clothing/s?ie=UTF8&bbn=9780907031&page=1&rh=n%3A83450031%2Cn%3A!425571031%2Cn%3A!425573031%2Cn%3A9780907031%2Cp_6%3AA3P5ROKL5A1OLE%2Cp_n_size_two_browse-vebin%3A1946185031" />

Both of these two sets of dresses were reached by going Dresses and then clicking on the colour filter. Both have unique urls. This is pretty much the same situation as my site - the products on each of those pages is different so is not a duplicate of any other page on the site, so would they really need that canonical tag?

I'm not sure if I have managed to confuse myself. My understanding of the canonical tag is that it is to be used if you have the same, or very similar content on more than one page. You point all versions to just one page to avoid duplicate content penalties and to make one page have all the links which might point to alternative versions. Therefore, if the page isn't a duplicate of another, is the tag really necessary

Thanks, Carl