GA4 Limits Nobody Tells You About

Google Analytics 4 is a powerful tool, but its technical limits are rarely documented clearly. Discovering them in production, when your reports display inconsistent data, is an experience you can avoid. Here are the real ceilings you need to know before deploying your measurement plan.

25 Parameters per Event, but Actually Fewer

The official documentation states 25 custom parameters per event. What Google does not clearly specify is that this quota includes automatically collected parameters: page_location, page_referrer, page_title, screen_resolution, language, and others. In practice, you have 20 to 22 slots for your custom parameters. Exceed the limit and additional parameters will be silently ignored, with no error message in the console.

500 Distinct Events and 50 Custom Dimensions

GA4 imposes a ceiling of 500 distinct event names per property. This seems comfortable, but poorly named events (with dynamic identifiers in the name) can exhaust this quota quickly. On the custom dimensions side, the limit is 50 event-scoped and 25 user-scoped, plus 50 custom metrics. Each registered dimension is final: even if you delete it from the interface, it continues to count toward the quota during a grace period.

30 Conversions Maximum

You can mark a maximum of 30 events as key events (formerly conversions) per property. For an advanced e-commerce site with granular micro-conversions, this limit is quickly reached. The recommended strategy: group similar events under the same name and use parameters to differentiate variants.

Sampling in Explorations

Standard GA4 reports are not sampled, but explorations are. Beyond 10 million events in the selected period, GA4 applies sampling that can strongly bias your analyses. The green icon at the top of the report turns yellow then red depending on the sampling rate. For unsampled analyses, exporting to BigQuery is the only reliable solution.

Data Retention: 2 or 14 Months

By default, GA4 retains user-level data for 2 months. You can extend this retention to 14 months in the administration settings. Beyond that, granular data is deleted. Aggregate reports remain available, but you can no longer create retroactive segments on older data. Without a BigQuery export, you lose detailed history irreversibly.

Cardinality: The “other” Values Trap

When a dimension contains too many unique values (typically more than 500 per day), GA4 groups the least frequent values under the “(other)” label. This high cardinality phenomenon affects both standard reports and explorations. Product pages with unique URLs, user identifiers, or internal search terms are the most common cases. The solution: reduce cardinality upstream (group URLs, use categories) or leverage raw data in BigQuery.

For an optimal GA4 setup that anticipates these limits from the start, or to understand in detail what GA4 is, feel free to check our dedicated resources.