Google Analytics Integration
Please use javascript or React SDK version 11.20.0 or higher.
GA4 is the latest version of Google Analytics, a widely used web/app data analytics tool. By integrating Hackle with GA, you can view the results of experiments run in Hackle directly in GA data reports.
You can use the data already sent to GA as-is to review experiment results in the familiar GA reports. If you use BigQuery, even deeper analysis is also possible.
Policy
Since GA data reports are used, GA's experiment-related limitations apply. For detailed GA limitations, refer to the link.
GA Version
GA4 integration is supported (UA not supported)
Supported development environments
Web/WebApp environments are supported.
Number of connectable GA Property IDs
One GA Property ID can be connected per Hackle Workspace. Development Environment is not supported.
Supported experiment types
A/B Test only (MAB Test not supported)
Number of A/B Test groups (Audiences)
Up to 4 test groups can be created per experiment. Audiences are created one per experiment number - experiment version - test group.
Audience retention period
365 days
Setting Up Integration
Before integrating, both the Hackle SDK and GA SDK must be installed in your service.
Go to Workspace Settings > External Integration in the Hackle Dashboard. (Link)

Select Google Analytics and start the configuration.
After logging in with your Google account, allow access to the Google account.
The Google account you log in with must be an editor (or higher) of the Google Analytics property.

In the popup that appears, enter the GA Property ID to connect with the Hackle Workspace. The GA Property ID can be found in the GA console under Admin > Property Settings > Property Details.

Enter the Property ID and click the SAVE button to save the changes and complete the integration. Once the integration is complete, the integration status will display as "Integration Complete" and the connected Property ID value will be visible.

Starting an Experiment
When you create an A/B Test in Hackle and start the experiment, audiences are automatically created in GA. Audiences are created per experiment version and per test group.

In the GA console, you can verify this under Configuration > Audiences as shown below. Audience names follow the format: HACKLE-EXPERIMENT-{experiment number}-{experiment version}-{test group}.

When actual user traffic enters the experiment, the user count in the GA audience will also increase. Click "Add new comparison" in the desired GA report and select the audience you want to filter, allowing you to view data filtered to that audience only. Add comparisons per experiment group to compare performance across groups.

After the experiment ends, if you no longer need data analysis for that audience, it is recommended to delete the audience.
FAQ
Q. The audience appears correctly in GA4, but data is not accumulating for that audience.
A. To view user information in GA4, you need gtag function insertion and domain allowlisting.
For the gtag function, refer to the gtag installation guide below.
For the tag to work, the domain also needs to be allowed. To allowlist your domain, check whether the domain has been added under Settings - Data Streams.
Last updated