Part 4/13:

Sesh walks through common measures of data quality: accuracy, completeness, validity, consistency, timeliness, and uniqueness. While organizations diligently monitor these metrics via dashboards, failures still occur. The key insight is that measuring data quality at the output stage is often too late. By the time issues are detected, they may cause significant damage, such as fines or project failures—exemplified by the TSB bank case, which faced a £48 million fine due to poor data testing, losing approximately $330 million in a project.

He emphasizes that focusing solely on data quality metrics at the end does not prevent problems; it only detects them after they've impacted the system. This reactive approach illustrates the need for earlier, proactive measures.