How to choose the best canonical URL
-
In a duplicate content situation, and assuming that both rel=canonical and a 301 redirect pass link equity (I know there is still some speculation on this), how should you choose the "best" version of the URL to establish as the redirect target or authoritative URL?
For example, we have a series of duplicate pages on our site. Typically we choose the "cleanest" or shortest non-trailing-slash version of the URL as the canonical, but what if those pages are already established and have varying page authority/backlink profiles? The URLs are:
example.com/stores/locate/index?parameters=tags - PA = 54, Inbound Links = 259
example.com/stores/locate/index - PA = 60, Inbound Links = 302
example.com/stores/ - This is the version that currently ranks. PA = 42, Inbound Links = 3
example.com/stores - PA = 40, Inbound Links = 8
This might not really even matter, but in the interests of conserving as much SEO value as possible, which would you choose as either the 301 redirect target and/or the canonical version? My gut is to go with the URL that's already ranking (example.com/stores/) but curious if PA, backlinks, and trailing slashes should be considered also.
We of course would not 301 the URL with the tracking parameters.
Thanks for your help!
-
I like to keep the canonical neat so it looks better in the serps so as to encourage future users to click on it (people like clean, readable URLs), so I try not to use unnecessarily complex URLs. I have had very good luck in canonicalizing messy URLs with decent authority to new, completely non-ranking but clean, URLs and not only keeping but growing page authority.
And I have a fairly large site which is crawled regularly so I do value consistency--whichever page is set as the canonical will eventually become the ranking page, so current rank is not the biggest issue to me. In the long run, consistency will make your life (and the life of anyone who follows you in your present position) easier. (That being said, it wouldn't hurt anything if you prefer to use what is currently ranking.)
-
Thanks Linda! With regards to the trailing slash, typically we do set the non-trailing slash version as the canonical version across our site. So in this case, would you recommend that we stick with the non-trailing slash version for consistency's sake, even if it seems like it has lower SEO value?
Or would you go with the trailing slash version since it's the one currently ranking, and seems to have the more value; or even the longer URL with the highest PA/backlinks?
Thanks again!
-
In general, you will get the most links to a page via whatever URL is easiest for the linker to grab, so often it is not the prettiest, with parameters and such. But you really don't want that showing up in the serps, so don't use that as a canonical.
As far as the trailing slash/no slash issue, most people will use the no slash version if they are choosing how to add a link, even when the slash version would be more correct (as in your example) so you would also expect to see more links there.
But you mention that the slash version is the ranking version in your example. Is this mostly true throughout the site? (Maybe it is a Wordpress site that ends everything with slashes?) Then I'd stick with that. (I myself use the rule you first mentioned, the cleanest, non-slash version, but my site doesn't use trailing forward slashes.)
Do keep it consistent though. It would be a pain to always be checking pages to see what the canonical should be. The amount of SEO value you might lose is minimal.
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
-
Do we need rel="prev" and rel="next" if we have a rel="canonical" for the first page of a series
Despite having a canonical on page 1 of a series of paginated pages for different topics, Google is indexing several, sometimes many pages in each topic. This is showing up as duplicate page title issues in Moz and Screaming Frog. Ideally Google would only index the first page in the series. Do we need to use rel="prev" etc rather than a canonical on page 1? How can we make sure Google crawls but doesn't index the rest of the series?
Moz Pro | | hjsand1 -
Tools which scan urls for social data
Hi can anyone recommend any tools out there, which can allow me to scan a list of pages (urls) and give me back social data for each page (e.g. number of facebook likes, shares, twitter data, google plus, etc) Cheers, Chris
Moz Pro | | monster990 -
How to remove URLS from from crawl diagnostics blocked by robots.txt
I suddenly have a huge jump in the number of errors in crawl diagnostics and it all seems to be down to a load of URLs that should be blocked by robots.txt. These have never appeared before, how do I remove them or stop them appearing again?
Moz Pro | | SimonBond0 -
Batch lookup domain authority on list of URL's?
I found this site the describes how to use excel to batch lookup url's using seomoz api. The only problem is the seomoz api times out and returns 1 if I try dragging the formula down the cells which leaves me copying, waiting 5 seconds and copying again. This is basically as slow as manually looking up each url. Does anyone know a workaround?
Moz Pro | | SirSud1 -
Keyword Difficulty Tool - How to use it the best way?
Hi, I am freshman, both, here on SEOmoz and in SEO generally and have a question concerning the assessment of KW difficulty. I did browse through the Q&A-Section (great content!!) but could not find a relevant answer to my problem. So I am currently building my initial keyword list, our website is only 2 months old, so we are still in the very early stage. Fashion is a very competitive area, so the KW Diff. Tool indicates high difficulty for a lot of words and phrases. however, I identified some with percentages <50 or even lower than that. Then I compared the results of the SEOmoz tool to Google Adwards and the Google results for competitiveness differed significantly. For example: for the KW Personal Shopping I got KWD from Seomoz 33% (in Germany) and from Google 0,5 for broad and 0,64 for exact search. I am quite confused how to make the right choices for my KWs now. Which metrics should I consider most? What else apart from the competition factor is behind the metric KW diff.? Does it matter in any way that I search from Germany for Germany in German? Do you have any further recommendations for the process of identifying the best Kws? Thanks a lot in advance, best from Berlin Tani
Moz Pro | | TaniBogi0 -
What's been the best SERP rank tracker for you?
I do a lot of my SERP ranking tracking via Aaron Wall's Rank Checker, but I'm realizing it's taking up too much time. I was looking at Raven's tools and a few others. What's been your favorite thus far? If you do reply, can you talk about: price capabilities time savings b/c of it graphical UI other thoughts I like SEOMoz's UI, but I can only manage 5 domains and I don't have a huge budget yet. Also, if there's a rank checker out there that incorporates google places results, that would be great too! Maybe too good, eh?
Moz Pro | | SeattleOrganicSEO0 -
Crawl Diagnostics finding pages that dont exist. Will Rel Canon Help?
I have recently set up a campaign for www.completeoffice.co.uk. Im the in-house developer there. When the crawl diagnostics completed, i went to check the results, and to my surprise, it had well over 100 missing or empty title tags. I then clicked it to see what pages, and nearly all the pages it say have missing or empty title tags, DO NOT EXIST. This has really confused me and need help figuring out how to solve this. Can anyone help? Attached image is a screen shot of some of the links it showed me on crawl diagnostics, nearly all of these do not exist. Will the relation Canonical tag in the head section of the actual pages help? For example, The actual page that exist is: www.completeoffice.co.uk/Products.php Whereas, when crawled it actually showed www.completeoffice.co.uk/Products/Products.php Will have the rel can tag in the header of the real products.php solve this?
Moz Pro | | CompleteOffice0 -
Can someone explain why I have been seeing an increase in the number of Linking Page URLs in OSE that link directly to downloads?
Ever since the last couple Linkscape updates when doing competitive back link analysis I have noticed a large increase in the number of URLs of Linking Pages in OSE that result in an immediate file download. The majority of the time these downloads are not common files ie PDF, DOC files. For example, these were all in a competitors back link profile: http://download.unesp.br/linux/debian/pool/main/i/isc-dhcp/isc-dhcp-relay-dbg_4.1.1-P1-17_ia64.deb http://snow.fmi.fi/data/20090210_eurasia_sd_025grid.mat http://www.rose-hulman.edu/class/me/HTML/ES204_0708_S/working model examples/Le25 mad hatter.wm?a=p&id=145880&g=5&p=sia&date=iso&o=ajgrep These are just a few I came across for a single competitor. Is this sketchy black hat SEO, some sort of error, actual links, or something else? Any information on this subject would be helpful. Thank you.
Moz Pro | | Gyi0