Restore test
Actually recovering data from backup to verify the backup works and the process is understood.
Backup software reports success on the job, not on the restore. Corrupt sets, missing dependencies, and broken chains all report as green until someone tries to use them.
Testing should include a full system recovery rather than only single-file restores, and it should be timed, because the elapsed duration is the only honest measure of your real recovery time.
Where this comes up
Restore test sits inside our backup and disaster recovery work. If you are trying to work out what this means for your own environment rather than in the abstract, that is the page worth reading next, and a short call will get you a straight answer faster than either.