Do you canonicalise UTM URLs?

[Eric_S]

Hi Rhys,

I had the same question and got help on Moz! Here is some practical suggestions for you on this:

Add the following lines to your robots.txt file and you should be good to go. This will prevent Moz (aka RogerBot) from crawling and reporting on URLs that are tagged with UTMs:

User-agent: rogerbot
Disallow: /*?utm

Even better, add self-referring canonical tags on your homepage, preventing versions of the same page from being indexed. For example:

http://www.yoursite.com/some-page?utm_source=facebook&utm_medium=social should have a canonical tag that looks like this:

Thanks,
Eric

[SwanseaMedicine]

Hi Eric,

Thanks for your help.

I don't quite follow; how can you add a canonical from http://www.yoursite.com/some-page?utm_source=facebook&utm_medium=social

to

http://www.yoursite.com/some-page

If the former doesn't actually exist as a page, but only as a referrer, like a bit.ly?

All of our pages are automatically self-canonicalised, but I've read that it's best practice to canonicalise all URLs back to the main.

Cheers,

Rhys

[Alick300]

Hi Rhys,

No need to do any thing if all of your pages is self canoncalised. If you want you can use

User-agent: *
Disallow: /*?utm

in robots.txt as well.

Thanks

[LoganRay]

Good answer Eric!

[LoganRay]

Rhys,

The self referring canonical on http://www.yoursite.com/some-page will also be applied when parameters are appended to that URL. If your pages are already self-canonicalized, you don't have anything to worry about.

Also, what do you mean by this statement? "All of our pages are automatically self-canonicalised, but I've read that it's best practice to canonicalise all URLs back to the main." Are you saying canonicalize all pages to the homepage??

[SwanseaMedicine]

Hi Logan,

Sorry, not the homepage. I meant canonicalise all UTM URLs, like

http://www.yoursite.com/some-page?utm_source=facebook&utm_medium=social

to

http://www.yoursite.com/some-page

[davebuts]

You definitely shouldn't block all user agents from crawling UTM URLs.

This will mean that links from external sites to you UTM URLs are not crawled - so you'll be losing out on any potential benefit from those links.

If you want to block UTMs to clean up your Moz crawl report, go with what Eric mentioned above:

User-agent: rogerbot
Disallow: /*?utm

Cheers,

David

(note: edited comment to help avoid confusion about using UTM parameters internally)

[ThompsonPaul]

RE your first point - UTM parameters should NEVER be added to links pointing to other pages within the same site. There aren't a lot of absolutes in Analytics, but that's one of them. Doing so totally borks you analytics of any visitor who clicks on of those internally-marked links.

[Eric_S]

Thanks Logan! Of course, it was yourself that taught me.

Rhys, as some have said, if you use internal links my suggestion may not be for the best. Good luck with deciding how to move forward and don't forget to update us on what you did too!

[Eric_S]

I think this is an excellent point. If there is benefits from using the UTM on internal links, what should we do - is it a matter of either blocking all UTMs with the script or allowing search engines to crawl all UTMs?

[ThompsonPaul]

You CANNOT use UTM parameters internally on a site, Eric. Doing so will ruin your Analytics data regardless of whether you block crawling or whatever.

To track behaviour on internal links, you apply GA's Event Tracking instead. You can also gather a great deal of useful link-clicking behaviour quite easily by utilising Enhanced Link Attribution in GA as well.

Paul

[davebuts]

Hi Paul,

Thanks for the input. I'm definitely not saying to use UTM tracking on internal links, but I do see it happen quite a bit.

This has gone down a different path than what I think was intended so I'll update my first comment to help avoid any confusion.

Cheers,

David

[SwanseaMedicine]

Hey all,

Glad this has generated a lively discussion! For clarity, my UTM URLs are exclusively external (primarily from social or AdWords).

Based on the advice above, because the canonical page is self-canonicalising, I don't think I'll bother with canonicalising the UTM URLs.

Also, for what it's worth, I'm not a huge fan of event tracking, always found it rather clunky. I tend to use Next Page Path in analytics, or even heatmapping.

Cheers,

Rhys

[Eric_S]

Thanks Paul. Makes a lot of sense and clears things up.