Turning siloed analytics into a data product

The problem

Galery, a startup studio running merchandising sites for 30+ musical artists, had a Google Analytics 4 property for every single storefront. GA4’s own interface gives you reports, but no way to see the underlying, non-sampled event and user data, and no way to query across properties at all. That made it impossible to build any cross-artist analysis, or to package the data as something Galery could offer back to its artist partners.

What I did

Process

• Picked the export strategy. CSV exports were the obvious first option, but manual and fastidious at 30+ properties. GA4’s native BigQuery Linking integration does the same job automatically and without sampling, running daily with no manual work. I standardized on that.
• Designed the warehouse structure. One BigQuery dataset per GA4 property (analytics_<property_id>), each receiving daily events_YYYYMMDD and pseudonymous_users_YYYYMMDD tables straight from GA4. On top of that, an analytics_overview dataset with per-property views and a property_names lookup table, so anyone querying the data can just search by artist or site name.

Outcomes & artifacts

• All 30+ properties onboarded, each linked to the shared BigQuery project with its views wired up, plus parameterized SQL queries to add a property and create its events and user views. Onboarding the next artist site now takes three queries.
• A full handoff doc, with a glossary explaining terms like property, dataset, and view (not everyone touching this is a data engineer), a diagram of how data flows from Analytics into BigQuery, and a screenshot-by-screenshot tutorial for connecting the next new site. The Galery team doesn’t need to loop me back in to keep growing the roster.

Impact

• Insights that were previously invisible. Cross-property analysis across the whole artist roster, the kind GA4’s own interface simply can’t do.
• A foundation for AI-driven analytics. With the data structured and queryable in one place, Galery has been able to start building analytics models directly on top of it.
• A new product line. Galery has started packaging these insights and proposing them back to the record labels behind each artist, turning what used to be an internal reporting gap into a sellable product.

Sitting on data you can’t actually query yet? Let’s talk.