Audit Measurement Data Quality

What is your approach to ensuring data quality in your analyses?

Using the supplied customers and measurements tables, audit every measurement record. Flag missing or unmatched customers, missing or future dates, null or negative metric values, missing source systems, and duplicate customer-date measurements.

Output

1. One row per measurement, including valid and invalid records.
2. Return measurement_id, customer_name, issue_count, issue_details, and quality_status.
3. Order by measurement_id ascending. Use valid when no issues exist and invalid otherwise.

Output details

• Treat DATE '2026-10-26' as today's date.
• Treat measured_at equal to today's date and metric_value equal to 0 as non-future and non-negative, respectively.
• Treat NULL or whitespace-only source_system values as missing.
• Flag duplicates only when more than one record shares non-NULL customer_id and measured_at values.
• Return unmatched customer_name as NULL and issue_details as an empty string when no issues apply.