I hold onto phones. The one in my pocket is on its fourth spring, and it is the third device in a row I have kept past the point where the carrier assumes I will be shopping again. That habit changes what I am actually buying. A phone that feels excellent for eleven months and a phone that is still tolerable in year four are two different products, and the spec sheet does not separate them. Nothing on the box describes year four. So I keep my own short list of things to check, and none of them are about the camera.
Battery capacity is the first one, though not in the way it gets advertised. The milliamp-hour figure is easy to read off the box. What happens when it drops to 70 percent of that figure, somewhere around cycle 800, is the part that decides year four. At that point the question is whether a replacement is a walk-in service or a small surgery. Some manufacturers price an official swap at a level that works as a nudge toward a new handset. Others keep it modest and stock parts for years. The spread between the two is not small, and it is fixed on the day I choose the model. I look up the official replacement price before I buy, not after the battery has gone soft, because the number is public and it tells me how the maker thinks about year four.
Update support is the second, and it is the one that has shifted most in the last few years. The headline promises now stretch to six or seven years on flagship lines, but the promise has two layers and they expire on different schedules. Full OS upgrades are one thing. The patch layer decides whether my banking app keeps opening. A device can sit two major versions behind and stay perfectly usable, while a device that has fallen off the patch list starts getting refused by the software I actually need. I read the published support policy for the exact model. A marketing line written for the family covers whichever member drew the longest window.
Those two facts collide hardest at the bottom of the price range. A budget handset with three years of patches and a sealed, expensive battery works out to roughly thirty months of real service before the replacement cost lands, a year short of the window I am buying for. The discount at purchase runs out before the phone does. A mid-tier model at a higher sticker, carrying five or six years of support and a serviceable battery, covers the same window with room left over. The arithmetic is dull and it keeps landing in the same place. Cost per year of ownership stops resembling the sticker almost immediately.
My pre-purchase routine now runs in a fixed order. Support policy page first, because it is the hardest fact to reverse after the money is gone. Then the battery service price. Parts availability comes third, which usually means checking whether independent repair shops carry the screen and the cell or whether every fix routes back through one channel. A model with parts on every corner in its launch year can become a mail-in-only device once the volume drops, and a screen quote from a local counter is a two-minute phone call. Only after those three do I look at the chip, the panel, the camera stack. The order feels backwards to anyone shopping for the first six months of ownership. It is built for the last six.
Buying guides make this awkward, since most of them are written for the first year of ownership. A Korean rundown of current self-purchase options takes a different cut, sorting its picks by plan tier and by how long the buyer expects to hold the device. The holding-period axis is the one I care about, because a four-year buyer and a fourteen-month buyer should be walking out with different boxes. The piece works through it model by model. My own version of that check has settled into three lines:
- Official battery replacement price for the exact model, published before purchase.
- End date for security patches, which usually arrives earlier than the end date for OS upgrades.
- Whether the battery and screen are sold to independent shops or handled through a single channel.
A second life is worth accounting for as well. A phone that leaves daily rotation in year four rarely goes in a drawer. It becomes a household spare, a car unit, a first device for a kid, a music player wired into a speaker. Every one of those roles depends on the same two variables. A handset still receiving patches carries resale value and a plausible second owner. Once it drops off the list, it is a paperweight with a nice screen. The gap shows up in the used market within weeks of a support cutoff, and it is wide enough to swallow whatever the sticker discount was at purchase.
The decision gets shorter this way. Two or three published facts, all checkable in about ten minutes, remove most of the lineup before I have opened a single review. What remains is a small set of devices that will still be doing their job long after the purchase has stopped feeling like a purchase. The rest of the comparison is preference, and preference is cheap to change. A support calendar is the one term I cannot renegotiate later.