Track how often segments suggested by machine translation or translation memory are accepted as-is, slightly modified, or fully rewritten by editors. Useful to evaluate automation efficiency, spot recurring problem areas, and inform future improvements in MT/TM usage. Metrics can include counts, percentages, and average edits per segment.