CONNECTOR · GA4
GA4 — Web analytics with quirks. We pull the BigQuery export and ignore the GA4 UI.
CONNECTOR · GA4
GA4
Web analytics with quirks. We pull the BigQuery export and ignore the GA4 UI.
- categorySource · Web analytics
- freshness targetDaily / streaming (Firebase). Plan for late-arriving rows.
- auth modelBigQuery dataset link + IAM.
- sync modeBigQuery linked dataset → dbt staging.
TABLES SYNCED
events_YYYYMMDD partitioned tables.
SCHEMA · SAMPLE
schema.sqlSQL
0
CREATE TABLE analytics_xxx.events_20250414 (
event_name STRING,
user_pseudo_id STRING,
event_timestamp INT64,
event_params ARRAY<STRUCT<...>>
);QUERY · SAMPLE
query.sqlSQL
SELECT event_name, COUNT(*) AS n
FROM analytics_xxx.events_*
WHERE _TABLE_SUFFIX BETWEEN FORMAT_DATE( 0 , DATE_SUB(CURRENT_DATE, INTERVAL 7 DAY)) AND FORMAT_DATE( 1 , CURRENT_DATE)
GROUP BY 1 ORDER BY 2 DESC;OPS NOTES
GA4's UI numbers will not match BQ export numbers — sampling, late-arrival, and consent-mode redaction. Always source-of-truth from BQ.
Wire a real pipeline to your warehouse
30-min stack review. We come back with a scoped plan, an owner, and a freshness contract. No SDR, no slide deck.