Shift-right testing
Testing and validating software in production, through canary releases, feature flags, synthetic monitoring, and observability, rather than only before release. Shift-right complements shift-left: some failure modes involving real traffic, real data, and real infrastructure only appear in production, so the discipline is to test there safely, not to skip earlier testing.
Shift-right testing accepts that some failures only appear under real traffic, real data, and real scale, so it deliberately tests in production with techniques like canary releases, feature flags, synthetic monitoring, and observability. It complements, never replaces, shift-left testing: the earlier you catch what you can, the more you can afford to learn the rest safely in the wild. The prerequisite is the ability to detect a problem and roll back fast.