One of the most obvious: Sampling. We won’t find “(others)” in this system because we have all the data. But in return, GA4 protects itself against very expensive queries by limiting the data it extracts to approximately 10 million individual events. If you ask for more, it samples, which is like extracting only 10 million events and performing a rule of three on the total.
Another less obvious but very critical issue: data retention. We can configure it in the GA4 admin, but free accounts only have up to 14 months of retention. Retention is the length of time granular data is stored. That is, in this system, if your retention is 2 months (the default for all accounts, which you need to change), you will never be able to request a report for data from 3 months ago.However, it also has certain disadvantages:
Processing times. You won’t notice this in small accounts, but when you have a lot of data, you’ll find that using this system can be exhausting, waiting for it to show you data.
Lack of a comprehensive API. While we can use this system, we often can’t do so outside of the analytics interface, which is very limiting.
Where we can use this system:
Especially in Explorer reports. These reports are designed to take full advantage of GA4’s granular query system, so when using these reports, we usually work with this system (unless our queries are very simple).
In the Funnels API. For some reason, GA4 has only created APIs for this type of explorer report (presumably there will be more one day, but we’ve only had this API for a long time). So if you want to take advantage of these reports outside of GA4, you can only do so by creating funnels.
And nowhere else. As you can see, this system is powerful, but it’s used in fewer places.
Twisting the twist: A GA4 feature that confuses us all.
We’ve explained two reporting systems and roughly where we’ll use each of them. But you’ll see that we’ve clarified that this is “in most cases,” “many times,” “almost always.” This is because in GA4, the report you use isn’t 100% tied to the reporting system that GA4 uses behind the scenes.
What? Didn’t the standard reports use summary tables and the drill-downs use granular tables? Yes, that’s what it would be like, but in practice it’s not that simple.
When you query a report, GA4 checks which of the two systems will provide the best data. (Of course, what it considers to be the best data can be a bit biased, as it focuses more on resource usage than on data quality.) In many cases, when using standard library reports, it will decide that the best system to provide you with data will be the aggregated one, but this will not always be the case.