Skip to content

Balancing Network Effects, Learning Effects, and Power in Experiments

At DoorDash, we rely on experimentation to make decisions regarding model improvements and product changes because we cannot perfectly predict the results in advance.