In the second part of this blog series, we take a look at the difference between First user, Session and Event level dimensions and how each of these can be employed to gain insights on your prospecting, retargeting and conversion strategies in turn.
There are a multitude of options available to us when it comes to acquisition reporting in GA4, with many of the standard dimensions you’d expect now split out by scope and platform. This can make it tricky for analysts and marketers to work out where, when and how to use these dimensions.
In the first blog in this series, we saw how in GA4, each traffic dimension like Source, Medium and Campaign, among others, has been broken down as follows:
Crucially, every traffic dimension is now set at one of three scopes: First user, Session or event level. So, what is scope, and how does it affect our reporting?
In a nutshell, scope dictates how long a particular dimension’s value will last, and what dimensions and metrics it will play nicely with in the reports. The scope you use for a given dimension will depend on where you are in the interface and what you are looking to find out. Working from the widest scope to the narrowest:
First user scope dimensions in GA4
Any dimension beginning with ‘First user’ has its value set the moment a new user lands on your site or app and is scoped to the user, meaning it persists for the user’s entire lifetime. It does not update. It is in effect how a user discovered your online business presence, and as such is best suited to reporting on how prospective users are being attracted to your site, or how users are finding, installing and first opening your app. It should not be used to analyse retargeting campaigns or returning users, since it will not accurately reflect how users return to your site after that first visit.
First user scoped traffic dimensions are available in the standard reports ‘User acquisition’ and ‘User acquisition cohorts’, in custom Explores listed under Traffic source dimensions, and in the GA4 BigQuery schema under traffic_source, where they can be used to understand the value of each of your channels for discovery. Be aware, due to the nature of this field common upper funnel channels such as Organic Search are likely to rank higher here than in breakdowns of the same dimension at other, narrower scopes.
Pairs with:
Metrics like New users, User key event rate, Average 120d value. Dimensions like User Id, First session date.
Session scope dimensions
For more conventional traffic analysis, you’ll want to use the dimensions starting with ‘Session’. The values for these dimensions are set each time a new session begins in GA4 (which, to recap, is when a user comes to the site either for the first time or after a greater-than-30-minute absence). As such, they show how users are coming to your site on a visit-by-visit basis. If you did any traffic analysis in Universal Analytics, it will have been using dimensions set at this scope.
Session scoped traffic dimensions are available in the standard report ‘Traffic acquisition’, or in custom Explores listed under Traffic source dimensions. They are also available in the GA4 BigQuery schema under the session_traffic_source_last_click column
Pairs with:
Metrics like Users, Sessions, Session key event rate, Engagement rate. Dimensions like Landing page, Device category.
Event scope dimensions
You may also have seen a version of each dimension without any kind of prefix. Although these look as though they may be the natural successors to Source, Medium and Campaign as they appeared in Universal Analytics, they are not exactly the same. In GA4, a traffic dimension without a prefix is set at the event level, meaning that its value can update on an event-by-event basis, potentially multiple times within a single session. Specifically, it will be collected and updated with each new key event recorded.
Currently you will not be able to find event-level traffic dimensions in the standard reports. However, they are available for use in the Attribution reports in the Advertising area of the interface, as well as in custom Explores. They are present in the GA4 BigQuery schema under collected_traffic_source.
Pairs with:
Metrics like Key events, Total revenue. Dimensions like Page path and screen class, Event name.
Putting It All Together
By not settling on a single approach, GA4 allows for more flexibility in the way we report on acquisition and attribution. But without guidance on what to use, and when, it is also easier than ever to accidentally wind up reporting on the wrong thing. Here are some examples of how these three scopes differ, and the results you would expect to see using each of them.
Example 1 - The Rebound
A display ad brings a user back to your site after an extended absence. They try to log in but have forgotten their password. If the forgotten password email is manually tagged, then any key events the user completes after successfully logging back in, such as making a purchase, can be seen in one of two ways.
At the event level, the purchase will be tied to the forgotten password email, e.g.
- Source: mailchimp_account
- Medium: email
- Campaign: Password_Reset
However, at the Session scope, everything the user does, including the purchase, will be tied to that initial display ad click, e.g.
- Session source: thedailybugle.com
- Session medium: cpm
- Session campaign: clearance_menswear_q4_2024
The First user scope, set the first time the user ever came to the site, is liable to be different again. In fact, since the user successfully logged in, giving you the ability to match them to past sessions with the same user_id, the First user Source, Medium and Campaign could have been set as far back as when you started collecting data in GA4! e.g.
- First user source: mailchimp_newsletter
- First user medium: email
- First user campaign: Best_Looks_from_Oscars_2022
Example 2 - The Savvy Shopper
A user searches on Google for a pair of shoes, and sees multiple Shopping listings to products on your site. They click to open each one in a new tab, and add the ones they like to their wishlist, which is another key event on your site.
In this case, although the Session and event-level traffic dimensions start off the same, they will gradually diverge as the user opens and interacts with more tabs. The important thing to note here is that in any reports using Session-scoped dimensions, each time they fire a key event, whether that’s adding an item to their wishlist, purchasing or something else entirely, it will be recorded against the campaign details for whichever pair of shoes they clicked on to start this session, which might not have been the pair they ultimately liked or bought. The same point goes for First user, which could be different again if the user has previously visited the site.
At the event scope, you will know the specific shopping ad they clicked on last, and, depending on your chosen attribution model, give this all or some of the credit for the conversion.
Scope in the Realtime Reports
One pain point with GA4 is the relatively stripped back Realtime report, and specifically the options that are made available (or rather, not made available) for you to break down live data by traffic channel. You are not currently able to see where your live users are coming from - observant users will have noticed that the only options available in the Realtime report for traffic analysis are all at First user scope!
This means that what this scorecard is actually telling us is what proportion of users who first ever came to the site by a given channel, whether now or several weeks ago, are on the site right now. As such, care should be taken with this report, such as switching the metric in this widget from Users to New users, which is a little more intuitive in this case.
Realtime information on session-level traffic dimensions is not currently available in the GA4 UI, though with the ability to stream your data to BigQuery you may be able to set up a custom solution.
Final Thoughts
In this blog we saw how scoping impacts both how and where we can report on our traffic channels in GA4. Whatever your focus, there’s a set of dimensions for you. Those interested in how prospects become new users can turn to the First user dimensions for insights on how users discover your brand, where they download your app, and which channels generated the highest LTV through metrics like Average 120d value, while those more interested in more traditional traffic analysis can use Session scoped dimensions to see how users came to the site on a visit-by-visit basis. Going further, those interested in which channels ultimately led to onsite conversions can use the Event-level dimensions to see how much credit different attribution models give to each of your channels each time a key event occurs.
For more information on everything covered here, Google’s Analytics support site provides a fairly extensive breakdown of traffic scopes here.
Still stuck on your best options for traffic reporting? Join our Introduction to GA4 course for hands-on exercises using the acquisition reports, or our GA Advanced Analysis course where we look at advanced reporting and attribution. We run regular, open classrooms as well as private sessions customised to your own data and use cases.