You have 1000 GB of user analytics data stored in BigQuery and need to run a query that is expected to return a large number of records. Before executing the query, you wish to estimate its cost under on-demand pricing. What is the best approach to achieve this?
