BigQuery

What Stevin gets out of BigQuery

BigQuery is not an analytics tool, it is a data warehouse. It gets sold as 'for data analysts,' but the practical value for marketing sits in one thing: you can join GA4 data, CRM data, ad-spend data, and business data in queries you cannot run anywhere else. We connect GA4 to BigQuery as standard (free for most accounts, costs cents per query) and build the real attribution questions on top of that. Without BigQuery you work in separate silos; with it you work in a SQL layer.

What you pair it with

For your stack: GA4 export to BigQuery daily (free up to 1M events per day), exports from Stripe, HubSpot, or Salesforce through Fivetran or Airbyte, and visualization through Looker Studio or Metabase. Storage is cheap (cents per GB), queries only cost money on large scans. Do the math up front: complex joins on 100GB or more run toward 5 to 20 euros per query, so optimize with partitioning and clustering.

Common mistakes

• →Running SELECT * on large tables, an accidental query can cost 50 euros or more
• →No partitioning on date fields, query costs explode on historical analyses
• →Letting Looker Studio query live BigQuery directly, dashboards get slow and costly, better through materialized views
• →Not giving service accounts scoped permissions, a compromised key can export your whole warehouse