Trick question. The answer is "don't". I can't find a direct citation for external HDDs being the lowest parts-bin of drives manufactured, but the pricing tells the story enough for me.
My data is valuable enough to me that I usually spend ~twice as much on a high-end drive and put it in my own enclosure. I usually go for IronWolf drives, and I'm a big fan of ICY BOX enclosures: https://icybox.de/en/product.php?id=297
smartctl knows how to parse it, it just never does by default (a long-standing bug that isn't high priority enough to fix).
Do `smartctl -a -v 1,raw48:54 /dev/xxx`, or `-v 7,raw48:54` for raw seek.
However, these values have never been useful to diagnose a drive as failing before data corruption appears (and it gets kicked of a raid), in my experience.
Meanwhile, they're cheap and Seagate never pulled that SMR-in-NAS-disks shit that WD and a few others did.
WD's recent weirdness with the Red drives has pushed me to just spend the extra money for SSDs. I've been happy with Samsung's high-end products but I'm sure there are caveats.
I use Samsung for solid state - their Evo series are very good. The m2 Pro 980s are seriously fast.
https://hexus.net/tech/news/storage/148295-samsung-latest-ss...
This WD dumbassery seems like its mostly on the WD side. I'm not aware of any dumb decisions being made on the HGST-side of the WD company yet. It should also be noted that Sandisk is also owned by WD.
https://www.tomshardware.com/reviews/sabrent-rocket-4-plus-m...
I RMA'd the drive and the replacement had the exact same issue.
AKA: The issue could very well not be the hard drive manufacturer, but instead the company that delivered your hard drives to you. I prefer to pick up hard drives physically from Microcenter for this reason, it means that I can properly "baby" those fragile drives all the way home.