Why Are There "Canonical Disasters" - Is Google Messing Up?

[tedster]

In the past week or so I've read several accounts around the web of so-called "canonical disasters." We also had several recent threads here describing problems that seemed to be caused by the canonical link - including one where there was an accidental ".com.com" so the canonical tag read example.com.com. And com.com resolves any wild card subdomains, so that messed with this members indexing and ranking.

I don't get it. This tag was introduced a year and a half ago and it should be relatively straightforward, if Google follows through on what they originally said:

Is it okay if the canonical is not an exact duplicate of the content?
We allow slight differences, e.g., in the sort order of a table of products. We also recognize that we may crawl the canonical and the duplicate pages at different points in time, so we may occasionally see different versions of your content. All of that is okay with us.

Is rel="canonical" a hint or a directive?
It's a hint that we honor strongly.

[googlewebmastercentral.blogspot.com...]

So what kind of "slight difference" allows a deep internal page to be canonicalized to the home page - with totally different content? Or to a page on com.com? Google is supposed to IGNORE the tag - treating it as a "strong hint" but only if the content has only slight differences, right?

I've used the canonical link tag with no apparent problems, and in some cases it put an easy band-aid on a nasty infrastructure knot. But now I'm reading some SEO blogs that warn against serving the canonical link on the "original" URL. How could that be a problem? For smaller websites without many infrastructure assets to work with, surely it's an easy way to say "don't let any backlinks mess with this URL by adding query strings, playing around with case, double slashes, etc, etc.)

Are these articles "crying wolf" when canonical link problems rare and most everyone is having smooth sailing? Or are lots of people really getting into trouble with their canonical links?

[tedster]

Just to develop my earlier point further about placing a canonical link on the "original" page, I've done it routinely with no problem. The only way that doesn't make sense to me is if you think of a canonical link as a 301 redirect.

[gn_wendy]

My two cents:

I am not a big fan of the canonical tag. I have not, until recently, advised a client to use it.

The case where it makes perfect sense:

The client wanted a forum on their page. They have had this for a long time, but it was outside of the indexable content. Finally they have consented to letting it be indexed. They are using an open source platform and are not willing to fix certain issues. Each thread has a canonical for each reply in that thread.

Implementing the canonical tag was the only thing that made any sense at all, because they have no resources to fix the real issues.

The index was a real mess. Since we placed the tags, everything is peachy and the linkjuice is flowing as it should.

[AnkitMaheshwari]

I am using the canonical tag on the original pages for past 8-10 months with no problem at all as all these pages are used/served with camp ids for different marketing techniques (email, banner/display, ppc).

[suratmedia]

Canonical issues are not less, Wrong relative linking can cause severe penalty of 9-12 months.

i.e On http://example.com/category-1/page/2 If you link category-2 page without checking base url, example.com/category-1/page/2/category-2* (with 200 status) will create duplicate cycle and once those problem is created, it's really hard to solve it through 301 or robots.txt

[g1smd]

If you add a canonical tag, make sure it references the correct URL!

What is the correct URL?

It is the minimal length path and minimal number of parameters to reach that content...

www.example.com/ not www.example.com/index.html

www.example.com/?page=46 not www.example.com/index.php?page=46

www.example.com/?page=46 not www.example.com/?action=show&page=46&page=next

etc.

It is not good enough to pick one URL, any URL, as the canonical URL; you must pick the correct canonical URL.

[edited by: tedster at 1:19 am (utc) on Aug 20, 2010]
[edit reason] fix typo [/edit]

[emmab21]

Is it worth putting a canonical tag on every page in case there are alternative paths to a page that you had not considered?

I've been in my current role about 4-5 months now and had a lot to get to grips with in that time as there is a lot of content on our website. I'm discovering multiple URL's all the time so would it be worth using this tag as standard practise or can this have a negative effect?

Also what are the pros/cons of using this over a 301 redirect?

[Top_Hat]

I screwed up my canonical tags and have suffered in google

2 months on and no recovery.

Does anyone have any ideas as to how long we will need to wait?

Should I use 301 redirects to correct the incorrect canonicals

[tedster]

@Top_Hat - would you share some detail on how your canonical links were wrong? Otherwise any advice you get may just make things worse for you - especially when you start throwing 301 redirects into the mix.

[tedster]

Is it worth putting a canonical tag on every page in case there are alternative paths to a page that you had not considered?

That's exactly why I use the canonical link tag. Here's a key historical thread that catalogs maybe 40 potential canonical errors [webmasterworld.com]. Many websites, especially those with a complex server infrastructure, cannot address them all on the server. A server based fix is still the ideal, but just LOOK at some of those potential canonical problems our members turned up!

[freejung]

I've had no problems with using the canonical link on the original URL. However, I did get into trouble with my very first use of it - I accidentally canonicalized a URL that was 301 redirected to another URL or something like that, I don't remember the details, but it completely removed that URL from the index for several days just as if I had installed an incorrect 301. So I think Google may well be honoring the "hint" a little too strongly even in cases where it makes no sense whatsoever.

[freejung]

Oh, I remember what it was now. I put a canonical link at URL A that pointed to URL B, but URL B was 301 redirected to URL A. This created an infinite loop: A -> B, B -> A. It's obvious why that's a problem, and I have no idea what would happen if you created a structure like that with just 301 redirects, but I'm sure it would be ugly. However, what seemed odd to me is that Google didn't just ignore the canonical link, which was obviously wrong, and just canonicalize URL A. Instead Google removed the page from the index until I fixed it so that the canonical link pointed to URL A. Since then, there have been no problems that I've seen.

[tedster]

what seemed odd to me is that Google didn't just ignore the canonical link, which was obviously wrong

That's exactly what I thought they were promising back at launch time. And it is apparently the cause for these "canonical disasters".

[incrediBILL]

I've used canonical tags without any problem whatsoever but you have to be careful.

Most errors I've seen so far weren't Google's problems but either misuse of the canonical tag itself with a typo in the URL or some redirect problems.

Reminds me of all the people that complain about bad AdSense targeting yet their SEO on the page is a complete mess.

[Sgt_Kickaxe]

Canonical is still a second best solution, 301 duplicate content to what you want indexed and sleep well at night.

The main problem I see with canonical is that it requires a webmaster to tell a search crawler.... hold it right there, search engines do not trust webmaster input, at all, ever.

[incrediBILL]

301 redirects are also webmaster input...

[Nuttakorn]

I implemented 301 and also update XML Site Map. For the load balance server , I use canonical tag.

[maximillianos]

We use the canonical tag for our mobile pages. To me, that is the perfect use for it. Print pages, mobile pages, etc.

Regarding the quote from the Google engineer who said "slight" differences are okay. That concerned me for a long time until Google finally updated their help page on the use of canonical to include mobile pages. They previously only mentioned print pages in the help section.

[CainIV]

Use it on every website, on most pages including the homepage, with no problems - and we are talking about a considerable number of websites.

I agree with g1smd - if you are going to use it, get it right, check twice, then get someone else to recheck.

Most problems I have heard of tend to be a conflict in internal redirection versus the tag, the use of the base tag and canonical, or simply brainless errors on the part of the person implementing the tag.

[Sgt_Kickaxe]

301 redirects are also webmaster input...

Indeed, but when a webmaster places it on a page that page is removed from search. Google will oblige removing pages but take our word on the page receiving the 301 benefit? nah.

In terms of effect, when compared with canonical issues, 301 kills the dupe page entirely. Rel="canonical" is still hit and miss.

[bouncybunny]

I've just started using canonical tags for some pages on one of my sites. I've had a few pages where external (or occasionally internal) links were incorrect and had added extra parameters to incoming links. E.G. www.example.com/pagename.html had been linked to as www.example.com/pagename.html?=garbage which still displayed the same content.

Too early to tell if there are any issues.

I'd setup a 301 redirect, but I've found my mod_rewrite skills to be lacking and the danger of conflicting with some other redirects was too great.

[live4life]

Why worry, the canonical will just go supplemental and then go out the index........ if your worried serve a 404

[Hannahness]

"But now I'm reading some SEO blogs that warn against serving the canonical link on the "original" URL. How could that be a problem?"

I'm currently working on a site that has site-wide canonical link elements on all of the original pages. And it's not a tiny site - it's about 30k pages. Even dup content pages contain canonicals on their original pages. (Obviously I'm currently working to change this). From what I've seen, the vast majority of posts/articles crying out against the canonical link element aren't crying out against the canonical itself, but webmasters who don't understand how to implement it.

[incrediBILL]

webmasters who don't understand how to implement it.

Bingo.

[Top_Hat]

@tedster My mistake was using a relative canonical tag and missing the /
So google used the canonical to point at non existent pages (web master tools reported loads of 404 errors)

So a page: blue/widgets/index.html
would appear in web master tools as a 404 error blue/widgets/blue/widgets/index.html

Hope that makes sense.

Anyway, its been 2 months since fixing with no recovery, should I 301 all those missing pages or just wait, and does anyone know how long I will have to wait

[phranque]

301 redirects are also webmaster input...

webmaster input that in the absence of cloaking equally affects the human visitor and SE crawler.
the canonical tag, however is meta information that for practical purposes is intended only for search engines.

[tedster]

@tedster My mistake was using a relative canonical tag and missing the /
So google used the canonical to point at non existent pages (web master tools reported loads of 404 errors)

OK - but did it affect your traffic and rankings?

I maintain that Google should just be ignoring such errors in the canonical link tag. Sure - they can notify you in WMT, but it should not modify the way your site was previously indexed.

[Bewenched]

I use them very cautiously. It's too easy to screw up if you have a data driven site or use isapi or htacess to rewrite urls.

[incrediBILL]

webmaster input that in the absence of cloaking equally affects the human visitor and SE crawler.
the canonical tag, however is meta information that for practical purposes is intended only for search engines.

That has nothing to do with my point which was webmasters can make canonical mistakes more than one way, including but not limited to mistakes in 301 redirects, meta canonical, internal link, inbound links, etc.

You must have consistency throughout or it just compounds the problem and doesn't alleviate it.

[trinorthlighting]

We do not even use the tag. Google has always done a good job in figuring out the right page to rank anyways for our sites. I have never seen a bad link ever tank our sites and We do have quite a few.