Incremental validity is whether a test adds meaningful predictive power beyond what's already captured by measures already in use.
Suppose a company already uses an IQ test to help predict job performance, and someone proposes adding a personality assessment on top of it. The personality test might correlate with performance just fine on its own, but that's not really the question that matters for whether it's worth adding. The real question is: once you already know someone's IQ score, does the personality test tell you anything more about how they'll perform? If it does, it has incremental validity. If the personality score is basically redundant with what the IQ test already predicts, adding it isn't buying anything.
This distinction matters a lot in practice, because plenty of measures can look impressively valid in isolation while contributing almost nothing once you account for what a cheaper or already-standard test already covers. Incremental validity is the check against stacking redundant assessments that all measure overlapping ground, dressed up as separate tools.
It's a more demanding bar than ordinary criterion validity, since a measure has to prove its worth relative to the competition, not just relative to chance. This is one reason serious test batteries, in hiring or clinical settings, are built and pruned carefully rather than simply piling on every test that shows some correlation with the outcome.