You notice that two ACME House reports show different values for the same customer metric. Explain how you would diagnose the discrepancy using SQL. Focus on metric definitions, joins, filters, date logic, aggregation grain, and how you would isolate the exact records causing the mismatch.

This is a common analytics debugging problem. Strong answers show that you can move from conflicting dashboard numbers to a structured SQL-based investigation instead of relying on intuition.