shopping for a continuous glucose monitor to actually test how my staple foods spike me, and the two over the counter options are the abbot one and the dexcom one. i don't care about the gap math or the monthly pricing, i care about precision, how wide a range it can read, and whether i can pull the raw data out and analyze it myself. those three criteria basically decided it.
the abbot one only records glucose between fifty five and two hundred and writes out of range for anything outside that, and for someone deliberately trying to catch big meal spikes, two hundred is exactly the ceiling you'll hit, so it hands you a blank exactly when you need the number. the dexcom one reads forty to four hundred, is a little more accurate, and exports clean csv with selectable date ranges. the only things the other one does better are one minute readings and slightly cheaper, neither of which matters for a weeks long baseline.
the failure rate question is where it got honest in the other direction. neither company publishes numbers and both sensors do die early sometimes, but the abbot one has a clean unconditional replacement promise and a money back guarantee while the dexcom one says replacement is discretionary and may not qualify. so the move is to buy the dexcom one through a big retailer instead of direct, because then you get the better sensor and a retail return policy instead of a chatbot deciding whether your failed sensor counts. easy call.