Question

A large retailer is reorganizing its analytics platform around a data mesh. Each business domain (Sales, Inventory, Marketing) must own its own data products, control access, and publish curated datasets that other domains can discover and subscribe to without copying raw data. The central platform team wants to enable self-service data sharing across domains and organizations while keeping each domain autonomous in BigQuery. Which approach best supports this data mesh model on Google Cloud?

Accepted Answer

Give each domain its own BigQuery project to own and curate its data products, and use Analytics Hub to publish and subscribe to shared datasets across domains as listings in data exchanges.

Answer

Consolidate all domain data into a single centrally-owned BigQuery project and grant each domain team dataset-level IAM roles managed by the platform team.

Answer

Export each domain's curated data to Cloud Storage as Avro files and have consuming domains load copies into their own BigQuery projects on a nightly schedule.

Answer

Create one BigQuery dataset per domain inside the platform team's project and share tables by emailing scheduled query result exports to consumer teams.