Running custom SQL queries directly against BigQuery tables requires users to have query execution permissions in BigQuery, which is provided by the predefined IAM role roles/bigquery.user (BigQuery User). This role grants the ability to run jobs, including queries, within the project, and can be combined with dataset-level read access to the underlying tables. Assigning this role to a Google Group containing the BI team is the recommended IAM best practice for managing access to a group of users rather than granting permissions individually or via service account keys, which is discouraged for interactive human access and poses security risks. A Data Studio dashboard only supports pre-built reports/visualizations and does not allow arbitrary custom SQL querying, and copying data via a scheduled Dataflow batch job introduces staleness and unnecessary complexity, conflicting with the requirement to query the latest streamed data directly in BigQuery.
Community Discussion