Run the utilities to keep an eye on the wear rates - but of course backups are essential.
We're putting read intensive (= cheap) enterprise SSD with 1 DWPD specs in moderately high transactional server environments and for secondary storage behind traditional backup systems (Commvault, Netbackup, IBM's ProtectTier [which is EOL] etc.) and not seeing wear rates that are cause for concern and fall well outside typical replacement cycles. These are not used in mission critical/healthcare/Fortune 100 of course; those get high end (IBM) FlashCore modules anyway (no drive interface to slow things down) with petabytes of <100us latency systems.
That said, nothing is unbreakable. We had a higher end system go sideways due to crap Samsung RI drive firmware and some half-assed DRAID6 code and spent 2 weeks making the server environment whole again. DR site to the rescue.