BigQuery setup for User Journey
Enable GA4 BigQuery export, connect project/dataset, and wire install → purchase events for Enterprise journey analysis.
What this is for
User Journey Intelligence reads raw GA4 events from BigQuery (events_YYYYMMDD tables) to analyze install/open → registration → sign-in → purchase behaviour.
BigQuery is required for this module. GA4 reporting API alone cannot power cohort journeys or the Campaign Journey Matrix at event level.
Available on Enterprise plans. Growth and Scale tiers see a plan gate on the User Journey Intelligence screen.
Prerequisites
A GA4 property with daily (or streaming) export to BigQuery enabled.
A Google Cloud project that receives the export, with a dataset named like analytics_123456789.
A saved client setup (edit mode) — the BigQuery tab appears after Identity is saved.
Either a service-account JSON key (recommended) or a separate BigQuery OAuth grant on your workspace Google login.
Step 1 — Turn on GA4 → BigQuery export
In Google Analytics 4: Admin → Product links → BigQuery Links → Link → pick your Cloud project.
Choose daily export (standard) or streaming if you already use it. Note the project ID and dataset ID shown after linking.
First tables appear after the next export run — allow up to 24 hours for a new link.
Step 2 — Identify project and dataset IDs
BigQuery project ID = the Google Cloud project holding the export (e.g. my-ga4-project). This is not the GA4 property number.
Dataset ID = the dataset under that project (e.g. analytics_987654321). Inside you should see events_YYYYMMDD and possibly events_intraday_YYYYMMDD tables.
Step 3 — Service account (recommended)
In Google Cloud: IAM & Admin → Service Accounts → Create. Grant this service account BigQuery Data Viewer and BigQuery Job User on the export project.
Create a JSON key and download it once. In VEOtool Client Setup → BigQuery tab, upload the key, enter project/dataset, click Test connection, then Save connection.
Service accounts avoid an extra Google OAuth verification step and are the most reliable path for production.
Step 4 — OAuth BigQuery grant (alternative)
If you prefer not to upload a key, connect Google on the workspace first, then click Grant BigQuery access on the BigQuery tab.
This requests a separate bigquery scope on top of your existing Ads/Analytics grants. It does not replace the service-account path — pick one auth method per client connection.
Step 5 — Event names
Install / first open — default first_open for Firebase mobile apps; use install if that is your canonical install event, or first_visit for web.
Purchase — default purchase; change only if your app uses a custom revenue event name.
User Journey Intelligence also tracks new_registration and sign_in for the Campaign Journey Matrix. Ensure these event names exist in your GA4/BigQuery stream.
Step 6 — Run analysis
Open Attribution → User Journey Intelligence (or the sidebar entry). Pick date range and cohort events, then Analyze.
Top KPI cards use the full matching cohort in BigQuery; acquisition flow and sample journey rows use a representative 50-user sample for readability. Use View more to load additional sample journeys.
Troubleshooting
Test connection fails — wrong project/dataset, missing IAM roles, or export not yet populated.
Zero purchasers — date range too narrow, wrong purchase event name, or filters excluding your buyers.
Campaign matrix empty — campaigns need UTM/source metadata on events; very old renames may appear as separate rows.