URLs with parameters + canonicals + meta robots

[JessicaZylberberg]

Hi Moz community!

I'm posting a new question here as I couldn't find specific answer to the case I'm facing.

Along with canonical tags, we are implementing meta robots on our pages (e-commerce website with thousands of pages). Most of the cases have been covered but I still have one unanswered case:

our products are linked from list pages (mostly categories) but they almost always include a tracking parameter (ie /my-product.html?ref=xxx)

products urls are secured with a canonical tag (referring only to the clean url /my-product.html) but what would be the best solution regarding the meta robots?

For now we opted for a meta robot 'noindex, follow' for non canonical urls (so the ones unfortunately linked from our category/list pages), but I'm afraid that it could hurt our SEO (apparently no juice is given from URLs with a noindex robots), and even maybe prevent bots from crawling our website properly ...

Would it be best to have no meta robots at all on these product urls with parameters? (we obviously can't have 'index, follow' when the canonical ref points to another url!).

Thanks for your help!

[GlobeRunner]

Jessica, whenever you think of adding a meta robots noindex, follow tag, I prefer to try to determine if you need the page at all on the website. If you're using a canonical tag, then that's fine--but we prefer to remove pages entirely from the site if you're going to use the noindex, follow tag. A page with that tag on it generally doesn't provide any SEO value to the site, it only allows engines to continue to crawl the site.

even maybe prevent bots from crawling our website properly
When you mention that, the follow tag will actually allow the site to be crawled.

If the page on your site is useful for users, then keep it (and use a canonical tag if necessary to prevent duplicate content issues). Otherwise, consider removing the page if you don't want it indexed.

[JessicaZylberberg]

Hi Eric,

Thanks for your answer, but as said in my original post, I can't get rid of these URLs because of tracking (these tracking parameters are used all across the website in order to know from where products are the most clicked etc). One of the only spot where the product URLs are 'parameter free' is in the sitemaps xml.

Most of the time, a link from a list page to a product URL will look like /style/cuff-gold/804-item.html?ref=by-shop%3afashion-and-lifestyle%3a, while the 'true' URL is /style/cuff-gold/804-item.html. In order to prevent duplicate content from these tracking codes (I have seen some products being indexed twice or more because of this), the 1st URL has a meta robots 'noindex,follow' and has for canonical the 2nd one (which has a robots 'index, follow').

I just wanted to make sure this could be the best solution in our case (as we unfortunately can't get rid of these tracking codes) in order to have only clean product URLs indexed, and only once!.