DBT + Metabase Self-Serve Analytics

Problem and constraints

15 analysts spent 2+ hours daily running manual reports from Google Sheets. Stakeholders couldn't access data until afternoon reports were ready.

Approach

Migrated 50+ reports to BigQuery with dbt transformations (staging, intermediate, marts). Built automated data quality checks. Deployed Metabase for self-serve exploration.

Architecture

Diagram of the reporting path: BigQuery source tables into dbt staging, intermediate and mart models, then Metabase dashboards for self-serve exploration. — Report path from source data through the dbt layers into Metabase.

Measured result

Reports run hourly instead of daily. 40+ stakeholders self-serve. 30% overall pipeline efficiency gain.

What	Value	Scope and source
reports migrated	50+	from manual Google Sheets workflows to scheduled BigQuery models with dbt transformations
stakeholders self-serving	40+	querying dbt models and Metabase dashboards without an analyst
analysts previously running manual reports	15	2+ hours a day each before the migration
report refresh cadence	daily to hourly	the same reports after moving off hand-assembled Google Sheets

Stack

dbt BigQuery Metabase SQL