Exploring GA4 One Year On

Despite Google sunsetting Universal Analytics, their 3rd version of their analytics platform over a year ago now, its replacement remains decidedly unpopular.

The main complaint basically distills down to the fact that it is not user friendly. At all.
Sure, there’s now more data, but that seems to come at the cost of user-friendliness. We wrote a post last year about reporting on this new platform, focusing particularly on metrics and how the names have altered in the new platform. But it’s not only the names that have been affected, but how they are presented by the software.

It’s no longer as easy to alter these graphs or play with the data in the same way that it once was. This lack of intuition means that unless you can afford to spend a great deal of time getting to grips with the platform, GA4 simply does not have the data accessibility of its predecessor. Unless you’re looking for something that can be found in one of the standard built-in reports, you’re going to struggle.

So how do I see specific data in GA4?

This lies in the function of explorations. UA had a custom report function that worked in a similar way, but with the loss of the Universal Analytics intuition, users have now become more dependent than ever on the custom report (now named exploration) function. They are also the only place you can construct and permanently save filters.

Metric scope

Before we begin, an important note about how data is collated in GA4. Several times when creating a report, you’ll find either that a new variable is either greyed out and you are unable to add to the report, or (more commonly in looker studio), you’ll get an error stating that the data is ‘incompatible’ without a clear explanation as to what that means. However, we’ve found that the majority of the time, this error is caused by a misalignment in metric scope.

Whereas Universal analytics provided every piece of data in relation to the session it was associated with, GA4 focuses solely on the individual event. To help categorise the data, each ‘event’ on GA4 is assigned one of three scopes: user, session, or event. As a general rule, any piece of data, is only compatible with other events of the same scope.
This can be frustrating, which is why it’s important to understand the role that scope plays in these dimensions. A key example of how scope works can be seen in the three types of source/medium dimensions:

Session source/medium: session-scoped, breaks down how each individual session came to the site.

First user source/medium: user-scoped, breaks down how each user first came to the site.
Source/medium: event or attribution-based, breaks down where the credit for key events is assigned.

Let’s say a user arrives on your website for the first time after a google search for a product. This first visit triggers both a new session and a new user and both get assigned the source/medium google/organic. While visiting your website they sign up to your company newsletter.

A few days later, this same user returns to your site, this time using a link from a marketing email. Enough time has elapsed between this and the previous visit that this is counted as a new session. In this case the Session source/medium is email/newsletter, but the First User s/m remains google/organic.

If then this user exits the site again, only to return 10 minutes later via a retargeting ad to purchase a product, this would trigger Source/medium, which splits the credit for this conversion as dictated across all channels. If the attribution model is last click, then all credit for that purchase would go to the retargeting ad. However, as the previous session hasn’t elapsed, the Session source/medium would still show as email traffic.

One good thing about explorations is that it makes it very clear when you’re trying to create a report with incompatible data types by greying them out and making you unable to interact with them at all. However, outside of this it’s difficult to tell which variable specifically matches which scope, so you’ll need to keep that in mind when selecting data for your exploration, and potentially pull two or three different variables where possible so you can play around with the data in case of sudden incompatibility.

Variables

Open an exploration by going to the explore tab and opening a blank exploration (there is also a gallery of templates available if you’re looking for something specific). Usually, we go for free form reports whose default visualisation is a traditional table (or pivot table depending on the number and location of dimensions added). If you’re looking for something a bit different then there are several other data presentation options in the settings column, but for the purposes of figuring out how GA4 data works, we’d recommend sticking to a table.

Once you’ve opened your blank report you want to go about collecting your variables. Similar to a recipe, when you want to create an exploration, you want to have all the variables (ingredients) needed to hand to facilitate the creation process. Of course, you can go in and add things later, but it’s easier if you can see everything you need in one place, especially if you’re playing with the table and trying to find out which variables are suitable for providing the necessary information (again this is a good time to consider the scope of what you’re looking at and how that can affect the variables you use).

Variables are split into two categories: metrics (blue) and dimensions (green). Simply put, metrics are numbers (e.g. sessions) and dimensions are usually text (e.g. channel group), essentially, what numbers you want vs how you want to split it.

Clicking the ‘+’ icon next to Dimensions and Metrics sections on the variable’s column will open the selection menu. This has all available metrics neatly categorised by function as well as a search bar. And tabs for pre-defined (built-in) and custom metrics.

Adding dimensions and metrics from here is as simple as ticking your required metrics from the list and then clicking the blue ‘import’ button at the top right of your screen when you’re done.

When you’re finished, the dimensions and metrics column should look something like the above, with the dimensions and metrics available for you to use in the report. You can also go in and add/remove metrics as necessary at any point during the process.
The above variables will create a table that will allow me to view the number of purchases by transaction ID to verify that each event is only firing once, and then to further verify by checking the overall count of the ‘purchase’ event against the number of assigned Purchases (these should be equal).

Segments

Another thing to consider at this variable-gathering stage is any filtering that needs doing. Seeing data relating to specific events or user types, especially custom values is a process that either involves using a filter in the report settings column, or a segment from the variables.

Filters can be a quick and easy way of narrowing down data matching one particular condition. This section can be found at the bottom of the ‘Settings’ column and uses the contains/exactly matches-style logic to specify/eliminate your chosen condition. For example, if you only want to see the event count for the Purchase event, take a look at the total list of Event names, find the relevant event (purchase), and then use the filter function so only those events that meet this condition are included in the report.

However, if you want to compare filtered sections, use two-or-more points of comparison, or apply a specific set of conditions to create an audience, event group, or session type, then using segments is the way to go.

The first thing to think of is whether you want it to have a user, session, or event scope, which will affect compatibility going forward as discussed earlier. Once you have decided this, the setup uses simple and/or logic to narrow down its data similar to the filters. In the below example, this segment is designed to show sessions that occurred over approximately a month and came from google. This is then contrasted with a similar segment for the month after, and applying both of these to the table allows a direct comparison of data between these two time periods.

Settings

Segments, dimensions, and metrics are all then used to construct the table either by a simple drag-and-drop, or by clicking the plus sign and selecting the appropriate variable from there. The drag and drop options are colour-coded, making it easy to tell where metrics and dimensions belong in order to populate the report. A report will populate to the right of the settings column once at least one metric has been added to the ‘values’ section.

Troubleshooting current problems

Why am I only being shown 10 rows of data?

Explorations as a rule do not have the capacity for endless scrolling. If you have successfully created an exploration but you don’t seem to be returning the breadth of data you expected, then try upping the number of rows being shown from the default 10 using the dropdown in the settings tab of the report.

The report will let you display up to 500 rows of data. Any data set larger than that will need to be exported using the icon at the top of the table on the right before you can see the full breadth of data.

This is something we always reliably forget to do until the table is generated and our data is falling short. Fortunately, a lot of exploration types should generate significantly more than 10 values, so this problem can be immediately obvious, especially if the amount given as a total far exceeds the sum of the variables you’re being presented with. However, be wary with smaller data sets where it might not be so apparent that your data is missing one or two key rows.

I can’t apply the report to today’s data/yesterday’s data is anomalous

Officially, GA4 takes up to 72 hours to process data so even though you have the option in the date range of selecting up to the day before you are currently working on, data may not have had time to fully process.

This is a pain but there’s not a lot we can do except to wait for those three days to elapse. The built-in realtime report provides some limited insight but anything more in-depth needs time to configure.

I can only use data from the last 2 weeks

GA4 as a default save 2 months’ worth of data. If you want access to this on longer-term then you can change the retention window to 14 months. First, go to Admin (at the bottom of the left-hand menu) and then find ‘Data retention’ in the ‘data collection and modification’ settings under property settings to do this.

However, this doesn’t apply retention retroactively and new data will only start collecting from the date of change. Longer-term data is still available on standard, built-in reports on GA4 and via other data presentation software such as Looker Studio.

In addition to GA4 support and analytics, Remarkable offers a full range of digital marketing services. To view these, please click https://remarkable.net/digital-marketing/