Option B is the most robust and efficient for handling potential duplicate timestamps. Here's why: Inner Query & QUALIFY: The inner query filters for the specific event type ('form_submission') and form ID ('contact_us'). The 'QUALIFY' clause, along with OVER (PARTITION BY USER ID, EVENT TIMESTAMP ORDER BY EVENT TIMESTAMP) = 1', efficiently de-duplicates the data. It assigns a row number to each combination of USER_ID and 'EVENT_TIMESTAMP' , and only keeps the first occurrence (ROW_NUMBER = 1), thus removing duplicate timestamp entries for the same user. Outer Query & Aggregation: The outer query then groups the de-duplicated data by 'USER_ID' and uses 'HAVING COUNT( ) > to identify users who have more than 3 distinct form submissions. Here's why the other options are less suitable: A: This option doesn't handle duplicate timestamps. It will count each duplicate timestamp as a separate form submission, leading to inaccurate results. C: Option C has similar function as option B, however it's slightly less elegant using 'WHERE rn = 1 ' than using Snowflake's Qualify command D: This Option aggregates to just Date Level and thus doesn't provide a robust solution within a 24 hour Period. E: Option E uses a 'COUNT( ) OVER clause that doesn't provide de-duplication of the underlying data, thus a duplicate Event timestamp would be counted twice.