GA4 Engagement Metrics: Session Duration, Average Engagement Time, & More

[@NiamhPhelan-k9j]

This is an excellent video, many thanks. Strangely I am seeing lots of pages with Average Engagement Time in GA4 LOWER than Average time on Page in GA3. They do tend to be the pages with higher exit rates. I would have expected the delta in such cases to be the opposite, GA4 should be higher than GA3 but I am seeing the opposite...puzzling!

[@Twooctobers]

Thanks! I would have expected the same, but it could be an indication that people tend to find your pages then not read them right away. I do this a lot when I'm researching something - I'll find multiple sites/articles on Google, then I'll go back and read them after I'm done searching. Because the pages are not in focus, they would have low engagement time in GA4, but high average time on page.

[@mari-janelle6519]

very helpful video, thanks!

[@Twooctobers]

Glad it was helpful!

[@AbdullahAlAsif]

Thank you :-) Great explanation!!! Is there any way to get these metrics in Looker Studio?

[@Twooctobers]

Yes! While there are various engagement metrics in GA4 reporting that are not in the Looker Studio connector, the User engagement metric is a sum of user engagement time and *can* be used to create them with calculated fields. For example, you can add Avg. engagement time per session by creating a field: User engagement / Sessions then set the type of the metric to Numeric > Duration (sec.) You can create a comparable Avg. engagement time per view metric by changing the denominator to Views.

[@AbdullahAlAsif]

@@Twooctobers thank you 😃

[@ultra_legend23]

Thanks for the excellent explanation, do you have any idea of how to calculate average page on time in GA4 in Big query? Thanks

[@Twooctobers]

You're welcome! I took a crack at it: gist.github.com/nicobrx/bc48e92878bfbf28bf29627906ec5339 When I compare it to the Pages and screens report in GA4, the values are similar, but not the same. That may be because events with Signals data are not exported to BigQuery, but it's also possible my query is not perfect.