Avery
Forum Replies Created
-
The issue you’re experiencing could be the result of the way that Google Analytics 4 (GA4) calculates Engagement rate and Bounce rate, which might not be aligned with your custom dimension settings. When you apply a custom dimension, GA4 might not find suitable data to calculate the Engagement rate and Bounce rate, which could explain while these numbers are falling to zero. Please ensure that the event data assigned to the custom dimension are suitable for these metrics. Remember, ‘Engagement rate’ is calculated as Engaged sessions divided by Total sessions. ‘Bounce rate’ in GA4 is calculated differently than in Universal Analytics, it’s now Engagement bounces divided by Total sessions. According to Google, an engagement bounce is when a user visits your site, and leaves within 10 seconds, or visits just one page on your site, or has less than 2 engagement events. Your custom dimension should reflect this new calculation. If it doesn’t, it could result in a 100% bounce rate or no data at all. If you are sure your dimension is correctly configured, you might need to get in touch with Google’s support to dig deeper into this issue.
-
It seems like you’re trying to add a new custom dimension in GA4 which might not be created at the GA4 interface first. You have to create it there and ensure it’s correctly set as an event-scoped dimension. Getting the data through Java could be more about API querying. If you’ve got the dimension set right, then it might be more about properly fetching the data with the right API call.
-
It sounds like you’re dealing with a bit of a mystery. The problem may be due to the way GA4 handles “active users” calculations. Sometimes, active users from different time periods or sessions might be treated as individual users in separate user groups but as a single user in the total count. So, even though there’s no overlap within groups, GA4 might be seeing some users as the same across different groups when calculating the total.
-
It seems like there might be an authentication issue with your OAuth configuration. A 403 error is typically tied to permissions. You might need to double-check the account permissions for the clientID and clientSecret you’re using, and make sure they have the necessary access to the GA4 account. If everything’s set properly there, you could try creating a new OAuth client ID and secret, and see if that works.
-
Absolutely. Upgrading to the paid version of BigQuery does allow you an extended data retention period, although it doesn’t automatically change from the 60-day limit. You might have to manually adjust this.
As for your GA4 data, it’s a bit tricky. Past GA4 data isn’t retroactive or backfilled. This essentially means that historical data from more than 60 days back won’t all of a sudden become available once you upgrade. From the day you start your paid BigQuery service, more data will start to be accumulated, but it won’t reach back to fetch the older data. You’ll begin having access to data from the point of upgrade forward.
-
Sure, I get what you’re saying. Although GA4 doesn’t have a specific dimension for this pie chart in Looker Studio, you can still achieve it by blending data.
In this blend, one part will contain only new users, and the other will have only returning users, which you can get by subtracting new users from the total user count. For the two parts to connect, you’d be setting up a calculated field—just assign ‘new’ for the part with new users, and ‘returning’ for the other part.
By running a ‘full outer join’, you’ll end up with a dataset that includes both new and returning users.
When creating the pie chart, you’ll need to pick your blended data as the source. But instead of relying on a given dimension, you’ll use a calculated one, combining ‘User type’ fields from both parts of your blended data using the COALESCE function:
COALESCE(User type (Table 1),User type (Table 2)).The same goes for your metric which also uses COALESCE to join both the new and returning user metrics:
COALESCE(New users,Returning users).And voila! You have your pie chart. A video demonstrating this process can be found here: [https://youtu.be/iaC3fAFtvPg](https://youtu.be/iaC3fAFtvPg).
