I've collected slips in the process of doing fulfillment accurately weighing shirts before and after label and tape. The shirt is the biggest source of variability in package weight. For many sizes (notably MM), the weight distribution in practice straddles an ounce boundary. It would be nice for inventory_to_shippo_labels.py to populate weights as confident, or leave the weight if not confident, instructing the user to weigh the box (and pad the measurement with a known expected weight for the label and tape).
We'd need to do some modeling to get good distributions here. For example, it may be the case that the variance for a given shirt size is well modeled as a percentage of the mean (because what's varying is the density of the fabric).
I've collected slips in the process of doing fulfillment accurately weighing shirts before and after label and tape. The shirt is the biggest source of variability in package weight. For many sizes (notably MM), the weight distribution in practice straddles an ounce boundary. It would be nice for
inventory_to_shippo_labels.pyto populate weights as confident, or leave the weight if not confident, instructing the user to weigh the box (and pad the measurement with a known expected weight for the label and tape).We'd need to do some modeling to get good distributions here. For example, it may be the case that the variance for a given shirt size is well modeled as a percentage of the mean (because what's varying is the density of the fabric).