Today in the BigQuery FinOps Dashboard, identical queries are counted as distinct if they query a sharded/partitioned table but with a different date suffix.
This provides an incorrect picture as far as most frequently queried queries, potentially hiding potential optimization opportunities. It's not always the most expensive queries you want to optimize, but the ones that are executed at the highest frequency.
Currently it's hard to identify those queries, but aggregating identical queries on the same table (regardless of the date suffix) would help identify those.
Please do not upload confidential information