Daniel GonzálezA “Data Threshold” in Google Analytics 4 (GA4) is a privacy protection measure that prevents individual user data from being inferred from reports and scans. When the number of users is low in a report with a specific data filter (be it audience, dates, etc.) GA4 can remove data to prevent identification of users based on demographics, interests or other data signals coming from Google Signals. These thresholds are defined by Google and cannot be changed.
You can tell that a threshold has been applied when a warning triangle appears in your reports.
Which GA4 functions are affected by the data threshold
Reports and Explorations
- Demographic Reports: Data may be removed if there are not enough users to guarantee anonymity.
- Search Query Reports: Data can be withdrawn if the total number of users for a particular query is not sufficient.
- Reports with “Google Signals”: Data may be removed if user count is low during a specific period.
- Reports with short time periods: If a report with a short time period is queried and there are low user or event counts, the data may be affected by the data thresholds.
- Exploration reports: They are also affected by this measure.
Custom Events
If you create a custom event on a property with “Google Signals” enabled, the reporting of these events may be affected and therefore you may have difficulties when viewing these events in your Google Analytics 4 reports.
BigQuery Data Export
The data exported to BigQuery does not include information from Google Signals and therefore the event counts per user may be different in Analytics and BigQuery due to how “Google Signals” handles duplicates of individual user counts. This makes BigQuery an excellent data source for extracting the “total number of events” on GA4 properties that have Google Signals enabled.
Audience Analysis
Querying reports that include demographic information or audiences defined by demographic dimensions may be affected by data thresholds, especially if the data sets are small.
GA4 API
API calls that include demographic information or audiences defined by demographic dimensions may also be affected by data thresholds.
How to remove the GA4 data threshold
Since you cannot change the way the data threshold works, the best that can be done to remove it is to look for some alternative methods of configuration or reporting:
Change user identification to “Device-based”
This solution implies that the reports will not have the information that Google Signals provides. So it is not a valid solution for those who wish to observe demographic information. The instructions to achieve this are in our article: Why do events appear only in the GA4 “real time” report?.
Generate reports from BigQuery
Again, this option is only viable for those analysts who do not need demographic information and can afford a total user count based on the user’s pseudo_id and not on Google Signals. For query construction we recommend the GA4 SQL web.
Extend the time period or reduce the filters
If it is essential to obtain user demographic information and/or perform a user count based on Google Signals, the best recommendation will be to extend the time periods of the reports or not to apply too specific filters, the important thing is always to obtain a sufficiently high number of data so that Google Analytics does not apply the data threshold.
Disable Google Signals
The last solution is to completely disable Google Signals from the GA4 property. With this the threshold will not be applied at any time but at the same time all demographic information available to Google Signals will not be available.
You can learn more about data thresholds in the official Google documentation.
