Cross-system metric mismatches are common in analytics environments, especially when reporting in platforms such as AbbVie customer insight dashboards depends on multiple upstream data sources. Interviewers want to see whether you can move from “the numbers do not match” to a structured SQL-based diagnosis.
Describe how you would identify the root cause of a data discrepancy between two systems that should report the same customer metric. Explain how you would use SQL to compare row counts, key coverage, duplicates, null handling, transformation logic, and timing differences. You should also discuss how you would narrow the issue from aggregate mismatch to specific records and how you would validate whether the discrepancy comes from joins, filters, late-arriving data, or business-rule differences.
Answer at the level of a senior analyst: focus on a practical, repeatable debugging workflow, the SQL patterns you would use, and how you would communicate findings once you isolate the issue.