My smartwatch found thousands of steps inside a shopping cart
TechnologyMy smartwatch celebrated ten thousand steps before lunch. I had spent most of the morning pushing a shopping cart through a warehouse store.
The number looked impressive until I remembered that my hand had been fixed to the cart handle. On another day, the watch recorded very little during a stroller walk. The device was not counting an objective quantity called “movement.” It was estimating steps from the signals available at my wrist.
I am not using the watch for medical decisions or a formal training plan. I use it as a gentle reason to move. Even at that low level, a confident number can turn into a judgment: good day, lazy day, streak preserved, streak broken.
I checked the watch fit and profile settings, then compared three ordinary situations: free walking, pushing a cart, and carrying a bag. I noted the pattern rather than trying to force an exact ground truth from one afternoon. If the device offers workout modes or phone-assisted tracking, I can test whether they make the estimate more useful for those situations.
The goal is not to prove the watch wrong. It is to decide what the number is good for. A rough trend may motivate me even if it is not an audit-grade step count. A competition or health conversation would require more caution about what is actually measured.
How do you keep wearable estimates from becoming moral scores? Which activities make your tracker obviously overcount or undercount, and do you correct them or simply interpret the trend differently?