A few reporting tables can be refreshed separately, but that becomes harder to manage as finance requests more data. At that point, you need one pipeline that selects the approved fields, transfers only new and changed records, and updates all destination tables after the source import is complete. That’s the type of BigQuery ETL process covered here: http://datrise.com/en/pipeline/bigquery-to-bigquery . It shows how an incremental sync can keep the second project current without repeatedly moving the full dataset. You can use the same pipeline for new finance tables as reporting requirements expand.