DOA sucks, but it happens, go via the retailer and they’ll replace it with a new drive rather than the OEM who is likely to send a refurb.
Strangely enough before WD etc. decided the retail NAS/server market existed, those of us running multi-drive set-up’s managed just fine. I’ve used NAS drives and the failure rate isn’t vastly lower than it has been for the last 20 years (excluding notable OEM manufacturing issues eg the DeathStar’s etc.).
Cache can be added afterwards, it’s great for docker/VM/cache usage and anything that requires regular writes that would normally cause multiple drives to spin up long term (save power/noise/heat).