Does Google have a native GA4 to BigQuery export?

Yes. Every GA4 property can be linked to BigQuery for a daily or streaming export of raw event data. The link has no fee, but BigQuery storage and query costs apply, and streaming costs $0.05 per GB. Standard properties have a daily export limit of 1 million events.

What happens if my GA4 property goes over 1 million events a day?

Google's docs say the daily export is paused if a standard property consistently exceeds the limit, and previous days are not reprocessed. Streaming export has no event limit, and GA4 360 properties have higher limits.

Should I use the GA4 export or Weld?

Use the native export when you need event-level data and stay within its limits. Use Weld when you want GA4 report tables that sit next to ad spend, CRM and revenue data in the same warehouse, with lookbacks that keep recent days current. Many teams use both.

Does Weld sync raw GA4 events?

No. Weld syncs report tables through the GA4 Data API: 7 pre-built reports and up to 20 custom reports with up to 8 dimensions and 10 metrics each. For raw events, use Google's BigQuery export.

Is the GA4 sync to BigQuery incremental?

Yes. The first sync loads data from the start date you choose, and later syncs load new data and re-fetch a set of lookback days so revised figures are overwritten.

Why do my numbers differ from the GA4 interface?

Common reasons are summing distinct user counts across days, comparing different date ranges, and GA4 revising recent days after a sync. Lookbacks handle the last of these; the others come down to how the report is built.

Google Analytics 4 to BigQuery

Google gives every GA4 property a free BigQuery export of raw events, and for many teams that is the right place to start. Weld syncs something different: daily GA4 report tables, such as traffic by channel, campaign and page, next to your ad, CRM and revenue data, with lookbacks that keep recent days in line with GA4.

Start for free View setup guide

4.8 on G2

Source

Weld

Warehouse

Checked against Weld and vendor docs on 10 October 2026

Weld · New sync

SourceGoogle Analytics 4Connected

Tables to sync

channel_traffic15 cols
campaign_performance15 cols
page_path13 cols
events_overview10 cols
audience_overview11 cols
social_media_acquisitions15 cols

Extract
Load
Keep in sync

DestinationGoogle BigQueryReady

google_analytics_4 dataset

google_analytics_4.channel_trafficSynced
google_analytics_4.campaign_performanceSyncing
google_analytics_4.page_pathSynced
google_analytics_4.events_overviewSynced
google_analytics_4.audience_overviewSynced
google_analytics_4.social_media_acquisitionsSynced

Google Analytics 4 tables: 7
fields, typed: 93
agent context included: MCP

Trusted by data and AI teams

Options

Three ways to get GA4 data into BigQuery

Google has two native paths, and both are worth checking first. The GA4 export gives you event-level data; the BigQuery Data Transfer Service and Weld both load aggregated reports. Which one fits depends on whether you need raw events or report numbers that match the GA4 interface.

GA4 BigQuery export (native link)

Best for: Raw, unsampled event and user-level data
Effort: Low: link the property to a Google Cloud project
Freshness: Daily, or streaming within minutes
Good to know: No fee for the link itself; BigQuery storage and query costs apply. Standard properties are capped at 1 million events a day for the daily export, and Google pauses the export if a property keeps exceeding it. Streaming has no event limit and costs $0.05 per GB.

BigQuery Data Transfer Service for GA4

Best for: Report-level GA4 tables inside Google Cloud
Effort: Low: configure a transfer in the console
Freshness: Daily, refresh window of 4 days by default
Good to know: Google lists GA4 among the connectors with no transfer charge. One custom report per transfer, at most 9 dimensions and 10 metrics, and custom dimensions and metrics are not supported.

Weld

Best for: GA4 reports alongside ads, CRM and revenue data
Effort: Low: authorise, pick reports, choose a schedule
Freshness: Incremental syncs on your schedule, with lookbacks
Good to know: 7 pre-built report tables plus up to 20 custom reports. Each sync re-fetches recent days so figures GA4 revises later are overwritten.

How it works

How to sync Google Analytics 4 to Google BigQuery with Weld.
In the app, with the API or an agent.

Setup guide API guide MCP docs

Weld · Data warehouse

Data warehouseGoogle BigQueryConnected

Data locationChosen when you connect

CredentialsService account or key, stored encrypted

Schema

What lands in Google BigQuery
Every Google Analytics 4 table, keys intact.

Weld's GA4 schema has 7 daily report tables. Each row is one property, one day and one value of the report's dimension, so they work as building blocks for marketing reporting rather than as raw events.

Open the schema explorer

channel_traffic15 columns
- property_idstring
- session_default_channel_groupingstring
- datedate
- total_usersdouble
- sessionsdouble
Users, sessions, engaged sessions, conversions and revenue per day by default channel grouping. The usual starting point for channel reporting.
campaign_performance15 columns
- property_idstring
- session_campaign_namestring
- datedate
- total_usersdouble
- sessionsdouble
The same metrics split by session campaign name. Join to ad platform spend on campaign name and date to compare cost with GA4 outcomes.
page_path13 columns
- property_idstring
- page_pathstring
- datedate
- screen_page_viewsdouble
- total_usersdouble
Views, users, new users, conversions and revenue per page path per day. Useful for content and landing page reporting.
events_overview10 columns
- property_idstring
- event_namestring
- datedate
- event_countdouble
- total_usersdouble
Event counts, users and revenue per event name per day. A daily summary of events, not the event-level rows the native export gives you.
audience_overview11 columns
- property_idstring
- datedate
- total_usersdouble
- new_usersdouble
- sessionsdouble
Daily totals for users, new users, sessions, sessions per user and page views for the property.

All 7 tables

campaign_performance
channel_traffic
social_media_acquisitions
browser_and_operating_system_overview
page_path
audience_overview
events_overview

Example query

Monthly sessions, engagement rate and revenue by channel.
In Google BigQuery SQL.

Runs in BigQuery against the channel_traffic table Weld lands, here in a dataset named ga4. Engagement rate is recomputed from sums rather than averaged, and users are left out because daily user counts cannot be added up across days.

Google BigQuery · query

1select2  date_trunc(c.date, month) as month,3  c.session_default_channel_grouping as channel,4  sum(c.sessions) as sessions,5  round(safe_divide(sum(c.engaged_sessions), sum(c.sessions)), 3) as engagement_rate,6  sum(c.conversions) as conversions,7  round(sum(c.total_revenue), 2) as revenue8from ga4.channel_traffic as c9group by month, channel10order by month desc, sessions desc

Weld MCP server

Ask your agent.
It queries Google BigQuery for you.

Connect Claude, Cursor, ChatGPT or your own agent to Weld over MCP. It reads the Google Analytics 4 context, finds the tables it needs and runs the query in Google BigQuery.

Set up the MCP server

Claude CodeOpus 5

ConnectedWeld MCPAgent contextGoogle Analytics 4tables7fields93

How many campaign_performance rows do we have in Google BigQuery?

Reading the Google Analytics 4 context first, so the query uses the right tables and keys.

●weld·search_available_references

⎿raw.google_analytics_4.campaign_performance

+ 6 more Google Analytics 4 tables

●weld·get_reference_schema

⎿campaign_performance · 15 columns

columns and types

●weld·run_query

⎿select count(*) from {{raw.google_analytics_4.campaign_performance}}

submitted

●weld·get_query_status

⎿DONE

●weld·get_query_result

⎿rows returned

✓Answered from Google Analytics 4 data in Google BigQuery

>Ask about your Google Analytics 4 dataMCP

Under the hood

Things to know.
Before you sync Google Analytics 4.

Report tables and the native export answer different questions: Weld syncs GA4 through the Analytics Data API, which returns aggregated reports. If you need individual events, user journeys or event parameters, use Google's native BigQuery export; you can run both and keep them in separate datasets.
GA4 revises recent days: Google Analytics keeps processing data after the day ends. By default each Weld sync also overwrites the past 24 hours and the full days 7 and 30 days ago. Custom lookbacks replace those defaults with the day offsets you pick.
User counts do not add up across days: total_users is a distinct count for each row's day and dimension value, so summing it over a month counts returning users more than once. Sum sessions, events and revenue freely; treat user totals over longer periods with care.
Not every dimension and metric combination is valid: Custom reports must include date and can only use combinations the GA4 Data API accepts. Check a combination in Google's GA4 Dimensions and Metrics Explorer or Query Explorer before you build the report.
Rate limits are shared across your Google account: Google Analytics applies global rate limits across your account. If other tools use the same limits, syncs can slow down; Weld throttles its GA4 syncs once usage reaches 80% of the limit, and the initial sync can be slower than later ones.

Sources: Weld Google Analytics 4 connector docs · Weld Google Analytics 4 schema · Weld BigQuery setup guide · Google Analytics Help: BigQuery Export · Google Analytics Help: Set up BigQuery Export · Google Cloud: Google Analytics 4 transfers · Google Cloud: BigQuery pricing, Data Transfer Service