If you run online marketing activity and use Google Analytics on your site, campaign tagging is a crucial aspect of understanding the impact of activity you put out into the market – as opposed to organically occurring links and search traffic.
There are some great online resources covering campaign tagging from Google, Justin Cutroni and Kissmetrics – but, while these cover the mechanics of tagging links, I want to look at some thoughts on how to approach tagging so you get easily useful data, not just segmented data at the end of your campaign.
I spend a lot of my time unpicking badly, inconsistently or thoughtlessly tagged campaigns after the fact. Normally under pressure to deliver reports brimming with insight and ROI-laden justification. The process of trying to clearly illustrate the impact of a campaign where the activity is completely fragmented is very like pulling teeth.
So just what is this tagging thing?
But before we jump straight, I just want to give a quick overview on what campaign tagging is.
Any links back to your site can have parameters manually appended to the URL that allows you to set the information Google Analytics captures on where traffic comes from and what type of link it is.
What are the benefits of doing this?
GA, by default, uses the following ‘buckets’ to segment traffic sources:
- Traffic that enters the site having enter a URL directly into the address field on the browser.
- Link from another site.
- Split into organic or paid, this bucket records all links coming in from recognised search engines and will include paid search advertising.
- Set values defined through manual or automated tagging.
Tagging the traffic that is responding to your marketing activity allows you to quickly separate anything you are driving, because it will stop showing up in referrals and instead be put into the campaigns bucket. This allows you to very quickly isolate the traffic you are probably paying for and understand how these visitors behave.
The more detail you include in your tags, the greater your ability to segment and analyse this traffic throughout, and at the end of, a campaign.
How tagging works
Initially, use the URL builder that Google provide for free. This gives you a simple form to fill in and generates your final link at the click of a button. Use the full tagged URL for any online ads, links in emails etc and you will see your marketing activity separated from your organic traffic.
The screengrab above shows the URL Builder tool and how the highlighted fields are used to generate the full URL with tracking parameters included.
It’s really that simple.
Analytics is a great tool, it gives a ton of data but that data doesn’t always equate to intelligence. Thinking about your campaign tagging up front makes it more likely that it will help you make intelligence-based decisions about your campaigns later.
On a slight tangent here, if someone else reports on your analytics, talk to them when you are developing campaign tags. What makes sense to you, may not be helpful when they’re reporting. I’m a firm believer in anyone tagging doing the reporting (at some point) as well, so you can see how your tags affect reporting.
5 (and a bit) Tips for intelligent campaign tagging
1. Be consistent
Google Analytics doesn’t know that when you tag one link as
utm_source=facebook and another one as
utm_source=Facebook that you mean the same thing, so it creates two different results because of the difference in capitalisation. So setting guidelines for consistency is important. Easiest way is to set clear rules.
- Every thing lowercase.
- All spaces replaced by hyphens.
- No TLDs included (e.g. .co.uk) so I can’t get campaign sources mixed up with naturally occurring referral URLs.
- Name all media by what they are (e.g. leaderboard not 728x90px banner).
- Use consistent media types (e.g. Don’t blend email, emailer, enewsletter etc. Pick one and stick with it).
2. Use the most useful level of granularity in your tags
utm_medium=social-media, or variations of this, cropping up a lot. The problem here is that social media could describe a lot of things.
Is it a link in a Facebook status update, a link in a tweet or a link on my YouTube profile. These are a few of the many social media sites the link could be on and, when I’m reporting on my campaigns, I may not be viewing Source and Medium information at the same time, so having everything bundled under social media might not be useful.
I make a distinction in tags between status updates, tweets, profile links so I can see at a glance what is working best without having to change my report view.
3. Tag every link you are shortening using a URL shortener
Chances are, if you’re URL shortening, you are looking to add that link to a space-limited medium, such as a tweet. Because a lot of social media content is accessed via mobile apps you need to make sure that everything is tagged because, while a web browser would include referrer data with that link (the snippet that shows which website the visit came from), apps do not send referrer data.
So tagging the link becomes vital to allow you to see collate all your tweet/status/post-respondent traffic in one place. This way you can keep the focus on your content first, rather than the software people are using to access it.
4. Apply a testing mind-set when tagging
Campaign tagging allows you to capture great data, but you need to make sure that the tags you set at the start of a campaign allow you to make performance decisions during and after the campaign.
The best way to ensure this is to think about what you would test if you could.
Want to see how a leaderboard that runs on the homepage of a site compares to a leaderboard that is run of site? Use tagging to capture that info, possibly as
This way you can compare performance over something concrete, especially where there is a cost difference for premium advertising space.
Look at it as if it were a split test. Use the tagging to define difference in elements based on the following (where appropriate):
- Where the ads/links appear
- What the ads/links are
- What the ads/links say
- What the ads/links are trying to get people to do
If you are looking to test two headlines, use
utm_content=your-second-headline so you can see immediately which one worked better. If you’re testing calls to action, put the calls to action in.
If you’re testing design elements, use clear descriptions of the tested difference. For example green button and red button, female shot and male shot etc.
If your tagging allows you to split successful campaign elements from unsuccessful ones, you’re on the right track.
4 (and a bit) Don’t use tags like Ad1 and Ad2
Because then you need to keep a separate record of what Ad1/Ad2 etc correspond to, and what you were trying to understand as part of the test. It’s redundant, open to errors and risks not being visible to everyone who needs to see it.
5. Connect tags to goals
If you are testing calls to action that encourage respondents to sign up for a newsletter, make sure the ad_content tag and the goal title relate to each other. This way you can immediately see if the activity is having the desired effect.
If you have a content tag of
objective is what you want to activate and
call-to-action reflects the call(s) to action, you can build a quick report which shows you:
- How many people actually signed up for the newsletter from activity aimed at driving that (objective)
- How many people signed up from each call to action (allowing you to refine messaging)
- How many people completed a different goal than the objective for this activity (are you giving people the right options or selling the right benefits?)
Wouldn’t that kind of report let you make decisions faster, justify your decisions to others more easily and generally screw as much value out of your activity as you can?
Give it a try and let me know what you think.