Conflicting numbers for the same customer metric usually point to differences in SQL logic rather than a true business change. This matters because analysts are often expected to trace the discrepancy quickly and explain it clearly.
Two dashboards in ACME House Analytics show different values for the same customer metric, such as active customers or repeat buyers. Explain how you would investigate the mismatch. Your answer should cover how you would compare metric definitions, validate table joins, check filters and date logic, inspect aggregation grain, and use SQL to isolate where the numbers diverge.
Go beyond saying "I would check the query." The interviewer expects a structured debugging approach, the most common SQL causes of mismatches, and examples of how you would prove or disprove each hypothesis with targeted queries or CTE-based comparisons.