We are experiencing significant costs on Snowflake due to merges being issued on tables, even though more than 98% of the DMLs on the source are inserts. Can we switch to using inserts instead of merges based on a primary key or identity value? For example, if a record has an ID greater than the maximum ID in the table, we could process it as an insert and avoid using merge operations, may be we need to mark the identity column manually for tables. This change could help reduce our Snowflake costs. Please let me know if you have any questions.