UTM tagging is a popular but nevertheless mysterious method of tagging URLs to edit or augment visitor data in Google Analytics. Most Analytics experts know that Google purchased the Urchin log/web analytics system (along with its “Urchin Tracking Module”) in 2005 and turned it into Google Analytics, and that has something to do with why these links work with Analytics today, but there’s generally very little discussion or understanding (even in the Analytics industry) regarding why or how UTM tagged links influence data in Analytics.
This is unfortunate because UTM tags have a significant influence on the two most essential variables for every pageview : source & medium.
Unfortunately, UTM tags are frequently understood as a means to conveniently create easily customizable segments in Analytics. They’re often used as a tool that allows people to easily add a layer of superficial customization on top of Analytics. This is simply false. As I’ll explain, the only justification for using UTM tags is when you need to change information in Analytics in order to solve blatantly inaccurate data.
The Problem with UTM Tags
The main problem (and benefit) with UTM tags is that they overwrite the most important data provided by the Analytics script.
Two of the things they influence are:
- the utm_source variable (e.g., reddit.com, google.com), and
- the utm_medium variable (e.g., referral, organic, cpc, social, email, etc.).
These are the two primary data points used in web metrics (going back well before Google Analytics). If you fill out UTM tags on a link, the Analytics script won’t find and pass this data on its own; it will simply pass the values you’ve defined in the URL.
Here are a few examples of how using UTM tags as a data segmentation tool can result in inaccurate data:
Bad UTM Use-Case #1:
Let’s say your company is working with 5 different industry bloggers to promote a new product or service. In order to easily segment this data in Analytics (to see if the bloggers are generating traffic and leads), you’ve shared the following link with these websites and instructed them to use it when linking to your website:
Because UTM tags override data from the Analytics script, doing this means that you won’t be able to separate traffic from those 5 bloggers in Analytics. The utm_source is filled in, and that’s the variable that’s used by the Analytics script to store the domain name of each referral source. The Analytics script will simply pass the defined utm_source (“bloggers”) to Analytics. All visits from the 5 bloggers will be lumped together with absolutely no differentiation.
Bad UTM Use-Case #2:
Let’s assume the marketing apocalypse has arrived and your company is experiencing a horrible PR crisis. In response you work with your PR company to quickly craft a blog post. You distribute the link via social media, and because you want to easily view the resulting website traffic in Analytics you use similar but different tags for each social network, like so:
After you post these links, the one remaining MySpace user in the world (Tom, obviously) takes the first link and shares it on Reddit. The post blows up on Reddit and ends up generating more traffic than MySpace, Twitter and Facebook combined. Unfortunately, you’ll never see Reddit as a referral source. Tom shared a link with utm_source defined as myspace.com; the UTM tags override the Analytics script, and consequently your largest source of social traffic is MySpace.
When to Use UTM Tags
Because of how UTM tags work, they’re only useful when you need to intentionally override Analytics data. There are two situations where this happens frequently: advertising and email.
Display advertising is a great example. If you run display ads with untagged URLs through any ad network ( including Google’s DoubleClick), this traffic will show up in Analytics with a medium of ‘referral,’ and the source will be the website where your display ad was clicked (or a domain associated with the ad network). This is why UTM tagging is so popular in advertising – in order to get (non-AdWords) ads to show up correctly in Analytics, it’s necessary to overwrite the utm_source and utm_medium variables.
The situation is similar with email marketing. If your company publishes an email newsletter or runs marketing automation drip campaigns, you want to track how many visitors these email communications are generating. The problem here is that visitors who click on untagged links in email show up in a variety of ways in Analytics. If they use a browser to access their email they’ll typically show up as referral traffic. If they use an application such as Outlook to access their email they’ll show up as direct traffic. Consequently, using tagged links in email communications is the only way to reliably segment email traffic in Analytics.
That said, there are a few things to keep in mind even when using UTM tags for advertising and email links:
- If a visitor shares the link, anyone who visits using that link will show up as ad or email traffic.
- Your UTM tags should follow standard Analytics values as much as possible. For example, there’s no reason to use a utm_medium value other than “email” in email newsletters.
UTM Tagging Best Practices
- Don’t use UTM tags unless you explicitly need to overwrite source and medium data in Analytics to make it more accurate, as outlined in the section above. This happens most commonly (and almost exclusively) with advertising and email traffic.
- If you need to use UTM tags but you don’t have the time to understand how they work, the most important thing you can do is use utm_source and utm_medium values that already show up in your Analytics. Examples of source are ‘google.com,’ and ‘smartsearchmarketing.com.’ Examples of UTM_medium are ‘referral,’ ‘cpc,’ ‘organic,’ and ’email.’
- Keep UTM tags in mind when planning your advertising landing pages. If there’s a chance that someone will share the link, provide them with easy, compelling methods of doing so that don’t carry over the UTM tags (such as social sharing widgets).
- Use alternative tagging conventions for social posts, referral links and any content that you intend to be shared. Make up your own tags (for example: smartsearchmarketing.com/?ssm_source=FB&ssm_medium=social). Query variables show up in Analytics, allowing you to create custom segments or reports for these types of tags. Doing this allows Analytics to do its job and correctly determine the source and medium for each pageview.
Read more here.