On-the-go tasking systems, such as peer-to-peer food delivery or package delivery systems, help volunteers complete tasks for other users along their existing routines. In this project, we explore how we can use tasking as an opportunity to help people meet their social goals, by promoting matches with greater social benefit. For example, imagine Anna is a new member of a sorority and is looking to meet other members. She can deliver coffee to Susan, another member, as a way to create a first point of contact in their relationship. However, promoting matches with social benefit is non-trivial because we do not know where users will go, i.e. their routes are uncertain. We might miss opportunities if we pre-match users with tasks, but if we assign tasks as they arise, we might end up making sub-optimal decisions. To address these challenges, we build intelligent systems that promote social benefit without hurting other goals in the system.