Sampling is the practice of analyzing a portion of data to draw conclusions about the full dataset. Analytics platforms use sampling to handle large data volumes quickly, trading complete accuracy for faster report generation.
How Sampling Works in GA4
Standard reports in GA4 are always unsampled. You receive 100% of available data regardless of how much traffic your site generates.
Sampling occurs in exploration reports when queries exceed 10 million events for standard properties or 1 billion events for GA4 360. Custom reports with high-cardinality dimensions like user IDs trigger sampling more often.
Identifying Sampled Data
GA4 shows a green checkmark icon when reports use complete data. A yellow percentage indicator appears when sampling is active, showing what portion of data the report represents. Higher percentages mean more accurate results.
Accuracy Implications
Sampled reports provide estimates rather than exact figures. Small sample sizes can produce unreliable results, especially for low-traffic segments. Large sites may see samples that poorly represent overall behavior.
Metrics from sampled data should guide directional decisions rather than precise calculations. Revenue figures and conversion counts in sampled reports are approximations.
Reducing Sampling
Shorter date ranges: Narrower time periods mean fewer events to query, often avoiding sampling thresholds.
Simpler reports: Fewer dimensions and segments reduce query complexity. Remove unnecessary breakdowns when sampling occurs.
BigQuery export: GA4's free BigQuery integration provides access to raw, unsampled event data. Run queries directly on complete datasets for precise analysis.
GA4 360: The paid version increases sampling thresholds significantly, enabling larger queries without data loss.
