Do Utm Tagged Links For Google Analytics Pass Pagerank?, indexable link tagging with Google Analytics: link waste or not Options

[A.N.Onym]

Hi there.

Link tagging is when you add various ?variable=value parameters to the link you use in an email newsletter, a banner ad or AdWords (which are autotagged now). The variables are campaign names, source, medium and others, which lets you track ROI of your non-indexable campaigns with Google Analytics. (Learn more about link tagging for Google Analytics here.)

Naturally, if you place such a link on your site as an internal link, it'll be considered as a different URL and will have its own link value.

The question I have is, since Google knows it's a Google link and knows the original URL (by stripping the tags), does it pass link value to the original URL? In theory, Google is able to do it, but practically, they might've not implemented this in the algorithm yet.

Do you have any knowledge/experience with this?

Thank you.

This post has been edited by A.N.Onym: Nov 13 2009, 05:45 AM

[JohnMu]

Hi Yura

There are a few things you could do here, if you were worried about that:

- Move to "#" for these parameters, which will effectively hide them from search engines. There are a few articles on this, eg http://esev.com/blog/tutorial/hiding-googl...aign-variables/

- Use the rel=canonical link element: http://googlewebmastercentral.blogspot.com...-canonical.html

- For Google, use the URL parameter handling tool to tell us to ignore these parameters: http://googlewebmastercentral.blogspot.com...helps-with.html

In general, we'll try to figure these things out on our own (as will the other search engines), but personally I like to be in control so I'd probably try something like that.

Cheers
John

[jonbey]

Ooh, I think this answers a question I had before. I use Tynt.com which adds your URL to the bottom of text when it is copied, and it places some gumpf after a hash, e.g. .html#ixzz0WkR41RPu which then allows the copied text to be highlighted. So I guess that the # is ignored and pagerank is passed to the proper page, not a weird duplicate hashy page?

Re. Tynt, it seems most of the people that copy my site remove the URL anyway. But shocking to see how many people are coping.

[JohnMu]

Hi Jon
Yes, search engines generally ignore the "#" and everything afterwards, mostly because that is not a part of a normal HTTP request for the page.

The one exception is our proposal regarding making AJAX content crawlable by using a "#!" to signal when a URL has crawlable content (which is available via a slightly different URL):
http://googlewebmastercentral.blogspot.com...-crawlable.html -- however, that's not something you'd run into on a day to day basis, so see it as a bit of trivia and not something to worry about :-).

I don't know tynt.com so I can't say much about that.

Cheers
John

[jonbey]

All tynt.com does is give you a bit of javascript to add to your site which then, whenever someone copies a chunk of text adds the URL to the bottom. DailyMail.co.uk use it -

eg. copy the first paragraph of a story, and you get this:

Brazil nuts have been hailed as a tasty way of building up the immune system and even protecting against cancer.

Read more: http://www.dailymail.co.uk/health/article-...l#ixzz0Wkah8ISy

The "Read more: bit is automagically added. Then on your control panel in Tynt you can see how many times your work has been copied, and clicks from active URLs. Same for images. Since having Tynt on my site it has recorded 1,562,241 word copies. That is a lot of theiving!

[A.N.Onym]

John, thanks for the tips, especially about the # tagging.

What concerns me about canonical tags is that we don't have those pages (the pages are actually files. Yes, I know. It's how it works in a *paid* CMS here), so that's not a very scalable solution.

Can you please confirm that, if you do understand that the destination URL is actually a shorter URL, you do pass the link juice to the original URL? I mean, I really don't have to fiddle with all that rewrite rules to keep the URL short and pretty?

That being said, would it still be a good idea to use a 301 redirect at all (with GA loading before the redirect or after, as it is in the article)?

Thanks

This post has been edited by A.N.Onym: Nov 16 2009, 06:40 AM

[JohnMu]

Hi Yura
I'm not someone you'd want to ask for a definitive answer for Google Analytics, but as far as I know if you use redirects you effectively lose the tracking information in Google Analytics (it runs as JavaScript on the page, which won't trigger if you do a 301 redirect).

In practice, yes, we do try to pick the better URL and show that one to users. Many webmasters prefer not to leave that up to the search engines though and try to make sure that it happens in the way they want :-).

So personally, if I had the choice (and time to tweak :-)), I'd work to make sure that only my preferred URLs are indexed. If I didn't have the choice (or time), I wouldn't worry about it and just assume that it'll work out. There are usually bigger issues to worry about :-).

Cheers
John

[A.N.Onym]

John,

I do try to control the links that exist on our site, but when it comes to campaign tracking and content that people may make available on other sites (such as whitepapers that we encourage to share), that's beyond our control. In fact, we'd rather encourage sharing our freely available information online, than otherwise.

The article you shared did describe the way to rewrite URLs without losing Google Analytics data. I'm guessing doing a (delayed) PHP redirect after a JS script is served might work for Analytics (but possibly not for a properly execute 301 redirect, but it's another matter), too.

Like you said, I'm now trying to provide a roadmap to make sure things that are partly in our control don't yield duplicate content. But yes, if I can't do it, then yeah, I'll essentially wave goodbye to those anchored internal-page links from our republished whitepapers. Or, perhaps, not use URL tagging there (oh, I've just thought how to stop making the links count for link weight w/o nofollow - I might as well just go with that).

This post has been edited by A.N.Onym: Nov 16 2009, 11:20 PM

[Sebastian]

Thanks for pointing me to this thread, Yura.

I totally disagree on nearly everything said here. Of course it's helpful to pass tips on tweaking this URI clutter on arrival, but that's the wrong approach. There's only one solution: Google must develop a method that doesn't screw URIs in the first place. That's possible, I've posted a rough spec on my blog.

Oh, and I agree that UTM clutter won't produce many dupe issues. Google's canonicalization is top notch and the other engines didn't bother indexing UTM query strings so far.

The problem is that manipulating URIs breaks some functionality in general, so that's a no-no.

[A.N.Onym]

Hey, of course, Google has to get things right without us tinkering with ? and # parameters.

What functionality does manipulating with URLs break?

Thanks.

[Sebastian]

That's a long list. I've posted a few examples, but searching for it you'll find more.

Cool URIs don't change: http://www.w3.org/Provider/Style/URI?iseewhatyoudidthere is worth a reread every once in a while. Part of the message is "don't change cool URIs". So to speak, a resource is not supposed to implement routines that correct misuses of its URI. Google's very, very evil URI manipulations require just that. Also, Google reps keep on telling us that we've to fix their mess on our side. It would be cool if they'd admit "we've f***ed up, expect a revamp soon".

[A.N.Onym]

Hmm, could you folks please click a few times on this Google Analytics tagged link? I need to test, whether it works in GA

http://bcs-it.com/books/connected-enterpri...paign=connected

Thanks

This post has been edited by A.N.Onym: Jan 22 2010, 07:25 AM