How to avoid duplicate content with product filters?

[GrumpyGoddess]

Hi. We are completely rebuilding our website from scratch, and I am taking the opportunity to optimize the SEO. I have spent the past two days researching and reading, trying to figure out what the best indexing/noindexing strategy would be for our ecommerce site, particularly when it comes to duplicate content.

I would be very appreciative for any recommendations. Here's a specific example about brand names. We use a filter system in our shop that is based on attributes. The filter sidebar is displayed on four different category pages, plus the general shop (all products). I have also created a page called "brands" with a filter just for brands. In addition to products that can be filtered by brand, this page will also have some content on it, as well as links to each brand name. So, for instance, to find products of a specific brand, a customer can use the filters and will get a variety of search parameter urls. And they can use the filters on each of the category pages. So for instance:
https://example.com/categoryA/?pa_brand=A
https://example.com/categoryB/?pa_brand=A
https://example.com/categoryC/?pa_brand=A
And many more variations...
And you can also get to the products of a certain brand by using the filters on the page that I created specifically for brands, and that would be a better page to be indexed between Google because it has useful content:
https://example.com/brands/?pa_brand=A

Would the solution to avoid duplicate content here be to write a canonical tag on the page https://example.com/brands?

[edited by: phranque at 10:28 pm (utc) on Apr 20, 2020]
[edit reason] exemplified domains [/edit]

[phranque]

welcome to WebmasterWorld [webmasterworld.com], GrumpyGoddess!

[GrumpyGoddess]

Thanks, phranque. :)

[tangor]

Glad you could join us, @GrumpyGoddess ...

Some amount of "duplication" can't be avoided, especially on ecommerce sites with large groups of the same thing with minor variations.

Canonical and noindex can help with some of those issues, but not all. Best way to find out is to experiment. Since this is an existing site you have nothing to lose in trying different aspects.

JUST GIVE IT SIX MONTHS AT LEAST between each change to get any kind of "real data" to know if something actually works in the long term. Else you might end up chasing false positives (there was a change) that doesn't stick down the road.

[GrumpyGoddess]

Thanks, tangor. I swear, the more I research, the more confused I get. And the more stupid I feel. For instance, all the articles I read say to noindex/nofollow internal searches. That makes sense. But they never tell us noobs how to do it. I mean, there are endless search variations, so it would be impossible to think of them all specifically. So am I supposed to noindex and nofollow something like https://example.com/?s* ? I know that's probably really dumb, but at this point I don't care what I sound like...lol

[edited by: not2easy at 3:08 am (utc) on Apr 21, 2020]
[edit reason] Please use example.com for readability [/edit]

[not2easy]

It is not uncommon to have a line in your robots.txt file to disallow search results URLs with " /?s" parameters, something like:

Disallow: /?s= 

If you can set up your results pages to noindex, that makes sense, though I would not bother with using nofollow.

You can use canonicals to show Google which version of a page you would prefer to index as well as using only that one version for your sitemaps. Not a guarantee they will agree, but it helps avoid the confusion when there is more than one way to see the same content.

This topic comes up fairly often. This one is from a few weeks ago: [webmasterworld.com...]
and other threads, in case it helps you see the duplicate content issues and solutions inherent to ecommerce sites: [webmasterworld.com...]
[webmasterworld.com...]

[GrumpyGoddess]

Thank you both, tangor and not2easy!

[RedBar]

Whose template are you using for your new site or is it an in-house custom one?

[GrumpyGoddess]

Hi, Redbar. My husband and I are building it ourselves. We use the Divi builder on Wordpress/Woocommerce, and I have Rank Math installed for SEO guidance.

[not2easy]

Be sure to update (if you haven't) that Rank Math plugin, it was reported : [wpvulndb.com...] and the 1.0.41 version update about a week ago fixed it.

[GrumpyGoddess]

Yes, not2easy. I've got1.0.43 installed now, thanks.

[GrumpyGoddess]

It's me again. I just discovered a setting on the plugin I use for my product filters which allows the url/permalink not to reflect parameter queries. It was on the default setting, which is to "use product filter redirects," so if a person searched for Attribute A under Filter Type 1, the permalink would be something like https://example.com/category/?pa_filter-type-1=attribute-a

But there is also a setting to "disable url changes," so no matter what filter is clicked on, the url stays at https://example.com/category/

Would this be a good solution to avoid the duplicate content issue when using filters? I would probably have put canonical tags on all my categories in order to avoid getting to a product through a zillion different parameters/permalinks, but I think this is achieving the same result and even better, is it not?

[edited by: not2easy at 12:57 am (utc) on Apr 23, 2020]
[edit reason] Please use example.com for readability [/edit]