FAQ

Frequent asked questions about Google Analytics 4.

Why is Google Analytics 4 metrics higher than Universal Analytics for the same period?

Each case might be unique, but please see some hypothesis below:

Universal Analytics Hits Limit

If you are comparing data with the UA Global Property (UA-109290164-1) be aware that this property had sampled data because we went above the 10 million monthly hits limit. This might explain why Google Analytics is reporting higher numbers than UA. Google says they can "implement client sampling to reduce the amount of data being sent to Google Analytics" which might reduce the data showed on Universal Analytics reports.

Implementation in Google Tag Manager

The UA tracking was divided into specific platforms tags (EN, P4, HS) while the GA4 is firing on all pages. There might be some pages that were not tracked before by these specific triggers that are not being sent to GA4, but I'm not sure this would affect the discrepancy so much

Filters

Google's documentation points that "Your UA reports may be excluding data based on filters, and GA4 properties currently do not support these filters". I'm not sure which view you used to do this comparison, but we do have a quite good amount of filters in place.

Difference in metrics calculation

For some metrics, you'll see that GA4 has a different calculation than UA. For example the users metric: "Universal Analytics highlights Total Users (shown as Users) in most reports, whereas GA4 focuses on Active Users (also shown as Users). So, while the term Users appears the same, the calculation for this metric is different between UA and GA4 since UA is using Total Users and GA4 is using Active Users."

See more information in this post by Google: [UA→GA4] Comparing metrics: Google Analytics 4 vs Universal Analytics

Machine learning algorithms for cookieless tracking

GA4 can determine the number of all users that entered the site, not only those who accepted cookies. This is named cookieless hits. It doesn’t break the law, GA4 created machine learning algorithms that can calculate a lot of gap metrics without sending personal data. More on this topic here.