I am working on a SQL query will lots of joins and it is currently taking around a min to complete which as per the amount of records in all tables and keeping the complexity in mind is fine. Though the current time taken is already good but anyways I was checking the query plan to see if there is any possibility for enhancing it. I saw that there is a particular point which shows around 70% in RID lookup for the main table. It's a heap table but the lookup seems to be because of an amount column which acts as output in aggregate function but is not used in any of the NC index. Rest all of the used or output columns are in NC indexes. My question is if it's beneficial OR best practice to add amount column in indexes (nc of course) AND what are the implications of adding a column directly to index vs adding a column as covered. Please share ur views.
Note: I have no authority currently to share the screenshot so please forgive me for that.