Welcome to the Q&A Forum

Hi everyone,

Here's a duplicate content challenge I'm facing: Let's assume that we sell brown, blue, white and black 'Nike Shoes model 2017'. Because of technical reasons, we really need four urls to properly show these variations on our website. We find substantial search volume on 'Nike Shoes model 2017', but none on any of the color variants.

Would it be theoretically possible to show page A, B, C and D on the website and:

• Give each page a canonical to page X, which is the 'default' page that we want to rank in Google (a product page that has a color selector) but is not directly linked from the site
• Mention page X in the sitemap.xml. (And not A, B, C or D).

So the 'clean' urls get indexed and the color variations do not?

In other words: Is it possible to rank a page that is only discovered via sitemap and canonicals?

Hi everyone,

Here's a duplicate content challenge I'm facing: Let's assume that we sell brown, blue, white and black 'Nike Shoes model 2017'. Because of technical reasons, we really need four urls to properly show these variations on our website. We find substantial search volume on 'Nike Shoes model 2017', but none on any of the color variants.

Would it be theoretically possible to show page A, B, C and D on the website and:

• Give each page a canonical to page X, which is the 'default' page that we want to rank in Google (a product page that has a color selector) but is not directly linked from the site
• Mention page X in the sitemap.xml. (And not A, B, C or D).

So the 'clean' urls get indexed and the color variations do not?

In other words: Is it possible to rank a page that is only discovered via sitemap and canonicals?

Thank you Andy, this does answer, and confirm, my situation/expectation. But, what you are suggestion is quite a high-maintenance solution, because if my client adds another couple of plates, they'd all need new canonicals to the 'main product page' that we want to rank.

Hi James,

With articles, I meant 'products for rent', they are not written articles.

The point in having them in the first place is because this is how the website works. Every product that clients can rent has its own page, so you could rent a brown plate from brand X or a blue plate from brand Y and both would have their own url and page.

I would really like to combine every version/variation of the product in a single page/url, but that has a large technical impact, and will probably be too expensive.

Hey there,

A client rents all kinds of party articles, like plates, bowles, etc. Currently, al his article pages have canonicals to their parent category pages, supposedly to have any pagevalue flow to these category pages, (which are much more relevant for SEO).

Is there anyone who agrees with this method? I think a noindex,follow would be a better measure to prevent Google from accessing all these 'low value' article pages. Besides, a canonical should indicate that page A and B are (almost) identical, which they most certainly are not in this case.

What are your thoughts?

Thanks for the response Davinia. We don't use the mentions methods to block crawlers.

Thanks for your response Sean. I do know that the use of canonicals is correct here.

My question though, is if it would be better to reduce the amount of actual pages (introduce a uniform URL structure, so to speak) because this would reduce the amount of pages the Google crawler needs to crawl drastically (over 65% on some of my clients webshops). As far as I know, they do crawl every canonicalized url?

Hi there,

I've been thinking a lot about this lately. I work on a lot of webshops that are made by the same company. I don't like to say this, but not all of their shops perform great SEO-wise.

They use a filtering system which occasionally creates hundreds to thousands of category pages. Basically what happens is this: A client that sells fashion has a site (www.client.com). They have 'main categories' like 'Men' 'Women', 'Kids', 'Sale'.

So when you click on 'men' in the main navigation, you get www.client.com/men/. Then you can filter on brand, subcategory or color. So you get: www.client.com/men/brand. Basically, the url follows the order in which you filter. So you can also get to 'brand' via 'category': www.client.com/shoes/brand

Obviously, this page has the same content as www.client.com/brand/shoes or even /shoes/brand/black and /men/shoes/brand/black if all the brands' shoes happen to be black and mens' shoes.

Currently this is fixed by a dynamic canonical system that canonicalizes the brand/category combinations. So there can be 8000 url's on the site, which canonicalize to about 4000 url's.

I have a gut feeling that this is still not a good situation for SEO, and I also believe that it would be a lot better to have the filtering system default to a defined order, like /gender/category/brand/color so you don't even need to use these excessive amounts of canonicalization. Because, you can canonicalize the whole bunch, but you'd still offer thousands of useless pages for Google to waste its crawl budget on.

Not to mention the time saved when crawling and analysing using Screaming Frog or other audit tools.

Any opinions on this matter?

I'm not sure why you would want to do this, you use a canonical when page A and B are more or less the same and you want Google to index only page A.

When page A and page B are different, a canonical would make no sense and Google will probably ignore it.

You can read more here: http://searchengineland.com/faq-google-mobile-first-index-262751 

I don't think different banners, navigation or footers have real impact as you'd expect this to be different across screen sizes. I'd be concerned if my texts or internal linking are considerably different between both versions, or when there's content on the desktop version of the site that's not on the mobile version.

If 'stuff is hidden with CSS' means you're hiding texts, you should really look for a different solution as hiding text is against Google's guidelines.

Placing a rel=canonical on your own blogposts to HuffPost is essentially telling Google that you're not unique enough to be indexed and that you want all ranking juice to flow to HP. Which isn't the case I guess. I would leave it alone. Nevertheless it will be very hard to outrank HP in this case.

This is what Yoast tells us about cross domain canonicals: _"You might have the same piece of content on several domains. For instance, SearchEngineJournal regularly republishes articles from Yoast.com (with explicit permission). Look at every one of those articles, and you’ll see a rel=canonical link point right back at our original article. This means all the links pointing at their version of the article count towards the ranking of our canonical version. They get to use our content to please their audience; we get a clear benefit from it too. Everybody wins." _(https://yoast.com/rel-canonical/#cross-domain-canonical)

So IF you can get HP to put a canonical to your own blog, this would be highly favorable for your own ranking. You could use the 'everybody wins' argument to try to get this working.

As long as your organic traffic is doing fine I shouldn't be too concerned. That being said:

• Is your robots.txt or search console disallowing crawler access to parameters like '?count=' or '?color='?
• Is your robots.txt disallowing crawler access to urls that have a 'noindex' but were indexed before they got noindex?
• You can also take a couple of parameters from your site and test if any url's have been indexed, by using the 'inurl:parameter site:www.site.com' query.
• Are some of the canonicalized urls indexed anyway? This may indicate that page content is different enough for Google to index both versions.
• If there's a ton of articles that go in and out of stock and use dynamic ID's, Google may keep these in their index. Do out of stock articles return a 404 or are they kept alive?

You can read more here: http://searchengineland.com/faq-google-mobile-first-index-262751 

I don't think different banners, navigation or footers have real impact as you'd expect this to be different across screen sizes. I'd be concerned if my texts or internal linking are considerably different between both versions, or when there's content on the desktop version of the site that's not on the mobile version.

If 'stuff is hidden with CSS' means you're hiding texts, you should really look for a different solution as hiding text is against Google's guidelines.

Alt tags are more important for SEO as far as I know. I think titles do help, but it can get complicated really fast if you want to have unique filenames for several languages. I'd go with the unique alt tag solution.

A strong site structure, some decent backlinks and a strong, optimized site are going to have much more of an impact on rankings anyway.

Then again, if you have just a couple of images and are really concerned about everything being perfect to the last detail, you could use different image filenames.

Hi everyone,

Here's a duplicate content challenge I'm facing: Let's assume that we sell brown, blue, white and black 'Nike Shoes model 2017'. Because of technical reasons, we really need four urls to properly show these variations on our website. We find substantial search volume on 'Nike Shoes model 2017', but none on any of the color variants.

Would it be theoretically possible to show page A, B, C and D on the website and:

• Give each page a canonical to page X, which is the 'default' page that we want to rank in Google (a product page that has a color selector) but is not directly linked from the site
• Mention page X in the sitemap.xml. (And not A, B, C or D).

So the 'clean' urls get indexed and the color variations do not?

In other words: Is it possible to rank a page that is only discovered via sitemap and canonicals?

I'm not sure why you would want to do this, you use a canonical when page A and B are more or less the same and you want Google to index only page A.

When page A and page B are different, a canonical would make no sense and Google will probably ignore it.