Rel=Canonical vs. No Index
-
Ok, this is a long winded one. We're going to spell out what we've seen, then give a few questions to answer below, so please bear with us!
We have websites with products listed on them and are looking for guidance on whether to use rel=canonical or some version of No Index for our filtered product listing pages. We work with a couple different website providers and have seen both strategies used.
Right now, one of our web providers uses No Index, No Follow tags and Moz alerted us to the high frequency of these tags. We want to make sure our internal linking structure is sound and we are worried that blocking these filtered pages is keeping our product pages from being as relevant as they could be. We've seen recommendations to use No Index, Follow tags instead, but our other web provider uses a different method altogether.
Another vendor uses a rel=canonical strategy which we've also seen when researching Nike and Amazon's sites. Because these are industry leading sites, we're wondering if we should get rid of the No Index tags completely and switch to the canonical strategy for our internal links. On that same provider's sites, we've found rel=canonical tags used after the first page of our product listings, and we've seen recommendations to use rel=prev and rel=next instead.
With all that being said, we have three questions:
1)Which strategy (rel=canonical vs. No Index) do you recommend as being optimal for website crawlers and boosting our site relevance?
2)If we should be using some version of No Index, should we use Follow or No Follow?
2)Depending on the product, we have multiple pages of products for each category. Should we use rel=prev & rel=next instead of rel=canonical among the pages after page one?
Thanks in advance!
-
Oleg, I like your thought process on this.
I am dealing with this exact issue and have 2 brilliant minds arguing over what is best approach. In reviewing the above, I agree with the approach. Canonical links to the first page of "Honda-civic-coupe" makes perfect sense.
Total we use prev-next, but self-refer rel=canonical the URL's on subsequent pages, but are not no-indexing page 2+. The negative impact is that Google will from time to time, add as site-links to the #1 search result a pagination page (e.g., 6 ) and some pagination pages are indexed. Landing page traffic to these is near zero. Our decision is determining whether to non-index or rel-canonical to the first page.
The pages in my case are new home communities where we might be listing all the different communities that are luxury communities in the specific city. While they are all this same category, as a group can be described similarly, and will have near duplicate metas, each community (list element) is unique. So, page #1 can be viewed as quite differentiated.
Here are the arguments:
-
Rel=canonical to the first page. As much as we think each shingle (i.e., page of 15 communities) is unique. The 15 Descriptions, amenities, location, what it is near, things you can do there are unique, As a group it can be considered just a list of communities. By pointing back to page #1 we are saying this is a collecting list of 3 pages of luxury communities in a given city. This will concentrate authority to the page that is most relevant.
-
No-index the subsequent pages. When Google said near duplicate, they really were considering limiting that scope to pages where the items are exactly the same or nearly the same. If the individual page content due to the differentiated product can be seen as unique content simply due to the in-page list elements, they are not really duplicate and rel=canonical is inappropriate. To use rel=canonical would at some point be viewed as manipulative and over-reaching use of rel=canonical. While this may cause this page to rank better, it may be considered not okay at some point.
Option #1 would seem to have a better immediate rank impact, but is there some real risk that it would be considered manipulative since the pages would not look to Google as near enough duplicates?
Glad to hear what you or others have to say.
-
-
Hey Oleg,
Thanks for the input - we'll look into making those updates!
-
Yes, you would canonical to that searchnew.aspx page.
In this scenario, I would set up mod_rewrite to create "Category" page for each specific model so you can rank for more pages.
e.g /model/Honda-Civic-Coupe/ would be a static page and you can canonical all of the other filters to their respective pages.
-
Hey Everyone,
Thanks for the answers and advice - here's an example of a filtered inventory listings page on one of our sites that isn't currently using a rel=canonical on it. Would you just have the canonical point back to the main "searchnew" page? If you have any other insights to improvements to this page's structure, please feel free to send suggestions.
http://www.leithhonda.com/searchnew.aspx?model=Civic+Coupe
Thanks all!
-
I would say using rel canonical would be the best. I am guessing your filter system is using a anchor or a hashbang? We only do ecommerce work and we typically just have the canonical of the filter page pointed to the category that is being filtered. The reason being is that you don't want to reduce the chances of the category ranking in the serps.
But honestly like Oleg said, the site would need to be seen to give a 100% best possible answer. We have used several different strategies with our clients. Some involve actually rewriting the filter urls as landing pages and trying to rank them as well.
-
Hey Oleg,
Thanks for the response. We're actually looking for info on our product listings pages, or search results pages within the site. Would this advice still apply to those pages?
-
Hard to give answer without seeing the site... ideally, you don't use canonicals or noindex and instead have 1 page per product.
-
Canonical is better overall i'd say - as long as the two pages you are merging are (almost) identical
-
keep the follow, doesn't hurt and only boosts pages it links to
-
Again, tough to understand but sounds like you should use canonical (pagination basically "merges" the paginated pages into 1 long one so to speak, so if you have the same content over and over again, best to canonical)
-
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
PDF best practices: to get them indexed or not? Do they pass SEO value to the site?
All PDFs have landing pages, and the pages are already indexed. If we allow the PDFs to get indexed, then they'd be downloadable directly from google's results page and we would not get GA events. The PDFs info would somewhat overlap with the landing pages info. Also, if we ever need to move content, we'd now have to redirects the links to the PDFs. What are best practices in this area? To index or not? What do you / your clients do and why? Would a PDF indexed by google and downloaded directly via a link in the SER page pass SEO juice to the domain? What if it's on a subdomain, like when hosted by Pardot? (www1.example.com)
Reporting & Analytics | | hlwebdev1 -
Switch to www from non www preference negatively hit # pages indexed
I have a client whose site did not use the www preference but rather the non www form of the url. We were having trouble seeing some high quality inlinks and I wondered if the redirect to the non www site from the links was making it hard for us to track. After some reading, it seemed we should be using the www version for better SEO anyway so I made a change on Monday but had a major hit to the number of pages being indexed by Thursday. Freaking me out mildly. What are people's thoughts? I think I should roll back the www change asap - or am I jumping the gun?
Reporting & Analytics | | BrigitteMN0 -
Index.php and /
Hello, We have a php system and in the MOZ error report our index.php shows up as a duplicate for / (home page). I instituted a rel canonical on the index.php because the / gets better rank than the other. This said, the error report through MOZ still shows them as duplicates. Should I be using a 301 instead? Please help! Also, I would love a good technical SEO book (for bridging the gap between SEO and programmer) if someone can recommend one? Thanks in advance!
Reporting & Analytics | | lfrazer0 -
Webmaster Tools Indexed pages vs. Sitemap?
Looking at Google Webmaster Tools and I'm noticing a few things, most sites I look at the number of indexed pages in the sitemaps report is usually less than 100% (i.e. something like 122 indexed out of 134 submitted or something) and the number of indexed pages in the indexed status report is usually higher. So for example, one site says over 1000 pages indexed in the indexed status report but the sitemap says something like 122 indexed. My question: Is the sitemap report always a subset of the URLs submitted in the sitemap? Will the number of pages indexed there always be lower than or equal to the URLs referenced in the sitemap? Also, if there is a big disparity between the sitemap submitted URLs and the indexed URLs (like 10x) is that concerning to anyone else?
Reporting & Analytics | | IrvCo_Interactive1 -
Show item status in GA - LIve vs sold vs unsold
Hi guys, I am working with SEO for an auction site with B2C and C2C. (similar to eBay) All item\products on our page have an unique url. The auction can last from hours to weeks.
Reporting & Analytics | | helgeolaussen
When the item is live, sold or (finished but)unsold it still has the same url.
So when I take a look at SEO traffic to items in Google Analytics, I can't tell if the item was live, sold or unsold at the time the user landed on the page. Which makes it diffucult to analyse the traffic. Is there anyway I can make GA show the status of the item for the time user landed on it? Best regards, Ceran0 -
Moz Rank & Trust | Page vs Sub vs Root
Hey guys, Just need some help deciphering my OSE link metrics for my site theskimonster.com . Page MozRank: 5.51 (highest among my competitors) Page MozTrust: 5.74 (#2 among my competitors) Subdomain MozRank: 4.19 (#4 among my competitors) Subdomain MozTrust: 4.63 (#2 among my competitors) Root Domain MozRank: 3.89 (#5 or last place among competitors) Root Domain MozRank: 4.1 (#5 or last place among competitors) What does this mean? What am I doing right, what do I need to do?
Reporting & Analytics | | Theskimonster1 -
Is Google Webmaster Tools Index Status Completely Correct?
So I was thrilled when Webmaster tools released the little graph telling you how many pages are indexed. However, I noticed with one of my sites, that it was actually going down over the past year. I've only been doing the SEO on the site for a few months, so it wasn't anything I did. The chart is attached to this post. Now here's the funny part. I haven't really noticed anything out of the ordinary for keyword ranking dropping off. I also tested my most recent page that I put up. 3 days after I posted it, I could find it in a Google search. Shouldn't this mean that's it's indexed? I can also find any other page I've posted in the last few months. Another oddity is that I submitted a sitemap a while ago when the site was only 22 pages. The sitemap index count says 20 of those pages are indexed. The chart only says that there are 3 indexed pages right now. However, I can clearly find dozens of pages in Google searches. Is there something I'm missing? Is my chart for this website broken? Should I report this to Google? Has anyone had a similar issue? decreaseIndex.png
Reporting & Analytics | | mytouchoftech0