They have the data to infer reliability rates based on things like returns, time to purchase similar item, frequency of repeat purchase, etc. Over the volume o

They have the data to infer reliability rates based on things like returns, time to purchase similar item, frequency of repeat purchase, etc. Over the volume of say Amazon, The noisiness of varied intent will normalize itself out across a sector. They could surface this as a reliability metric on some kind of relative scale. They could...