What is best practice for "Sorting" URLs to prevent indexing and for best link juice ?

[lcourse]

We are now introducing 5 links in all our category pages for different sorting options of category listings.
The site has about 100.000 pages and with this change the number of URLs may go up to over 350.000 pages.
Until now google is indexing well our site but I would like to prevent the "sorting URLS" leading to less complete crawling of our core pages, especially since we are planning further huge expansion of pages soon.

Apart from blocking the paramter in the search console (which did not really work well for me in the past to prevent indexing) what do you suggest to minimize indexing of these URLs also taking into consideration link juice optimization?

On a technical level the sorting is implemented in a way that the whole page is reloaded, for which may be better options as well.

[katemorris]

With canonicals, I would not worry about the incoming pages. If the new content is useful and relevant, plus linked to internally, they should do fine in terms of indexation. Use the canonical for now, and once you launch the new pages, well a month after launch, if there are key pages not getting indexed, then you can reassess. The canonical is the right thing to do in this case.

As for link equity, you are right, that is a simplistic view of it. It is actually much more intricate than that, but that's a good basic understanding. However, the canonical is not going to hurt your internal link equity. Those links to the different sorting are navigational in nature and the structure will be repeated throughout the site. Google's algo is good at determining internal, editorial links versus those that are navigational in nature. The navigational links don't impact the strength nearly as much as an editorial link.

My personal belief is that you are worrying about something that isn't going to make an impact on your organic traffic. Ensure the correct canonicals are in place and launch the new content. If that new content has the same issue with sorting, use canonicals there as well and let Google figure it out. "They" have gotten pretty good at identifying what to keep and what not.

If you don't want the sorting pages in there at all, you'll need to do one of the following:

• Noindex, disallow in robots.txt - Rhea Drysdale showed me a few years back that you can do a disallow and noindex in robots. If you do both, Google gets the command to not only noindex the URLs, but also cannot crawl the content.
• Noindex, nofollow using meta robots - This would stop all link equity flow from these pages. If you want to attempt to stop flow to these pages, you'll need to nofollow any links to them. The pages can still be crawled however.
• Noindex, follow - Same as above but internal link equity would still flow. Again, if you want to attempt to cut off link equity to these sorting pages, any links to them would need to be nofollowed.
• Disallow in robots - This would stop them from crawling the content, but the URLs could technically still be indexed.

Personally, I believe trying to manage link equity using nofollow is a waste of time. You more than likely have other things that could be making larger impacts. The choice is yours however and I always recommend testing anything to see if it makes an impact.

[lcourse]

Kate. The domain has  100.000 pages and will scale to over 1 million unique pages during the next couple of months. I do not want the Sorting URLs have any negative effect on the new indexing of the new 900.000 unique pages in the next months.

Regarding link equity. My simplified understanding of link equity is that if a page has 10 links then each link carries 10% of the total link juice of the page. If now 5 of the 10 links do link to a canonical version of the same page (=sorting URLs), I may be losing out on 50% of the potential link juice the page carries. This is my concern. Therefore my doubt is if I should rather try to hide these sorting URLs from google (same as was also recommended by Rand for facetted navigation pages that one does not consider important for being indexed).

[katemorris]

Is your issue with crawling or indexing? Those are two separate issues. Why don't you want Google having the canonicals in the index? If you can give me some more insight, I can try to recommend the best option.

And I'm not following your last question. Can you try to ask it another way?

[lcourse]

Hi Kate, thanks lot. Yes canonical is something we should definetly do and we have implemented.

Still I had the experience in the past that google also indexed lots of canonicalized URLs with near identical content. Any additional step I could do to minimize indexing of these URLs further?

Wouldn't then the basically "self referencing" URLS of sorting links (going to canonicalized versions of the same page) be lost for link equity?

[katemorris]

This one would need a canonical. For one category page with 5 different sort options, you'd need one canonical URL (one without any sorting or the default sorting) and point all others to that URL using a canonical tag.

https://support.google.com/webmasters/answer/139066?hl=en

Would that work for your setup? If I understand your situation correctly, this should work. It consolidates link equity and allows Google to choose what needs to be indexed and served.