1

This is a question I couldn't quite explain to my boss' satisfaction.

I've made a data studio with GA4 that filters for certain blog pages. I used an individual metric to show the overall engagement rate for the pages (39%) and then a table that used page title as a break down dimension. She is confused because the engagement rates of the individual pages are between 22 and 35 so eyeballing it 39 doesn't seem like it could be the average. (I've checked and both entities have the same filter and time period set on them)

I said that you can't really average percentages and that it is likely that the denominator for the overall number is lower because sessions would only be counted once there but then counted on each of the pages. (lower denominator of users means a higher rate). This would depend on the engagement rate counting unique sessions but counting multiple engaged sessions? - like this:

Grouping sessions engaged sessions Engagement Rate
All 3 100 41 41%
Page 1 60 20 33%
Page 2 50 15 30%
Page 3 30 6 20%

I also said that regardless, it is the comparison of the trend over time and between the pages that is useful for analysis. But she would really like a stronger answer. Anyone else have a better idea?

Thank you very much.

1 Answer 1

1

One engaged session is likely to cover more than one page, whereas one unengaged session won't. Unless you were using landing pages as a dimension (in which case each would be unique), the sum of engaged sessions will always being higher than the average engaged sessions for all pages.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.