Case Study: DoorDash achieves 10–100x faster delivery matching and multi-order routing with FICO® Xpress Optimization

DoorDash optimizes complex deliveries up to 100x faster with FICO® Xpress Optimization

DoorDash, a technology company that connects consumers with local businesses, faced a real‑time dispatch challenge: its delivery‑matching system needed to solve larger, more complex routing trade‑offs faster to improve delivery times and Dasher efficiency. The prior bipartite/Hungarian approach was too slow at scale and could not support multi‑delivery routes, limiting the platform’s ability to optimize multiple overlapping pickups and dropoffs.

DoorDash reformulated dispatch as a mixed‑integer program and deployed FICO® Xpress Optimization as the production solver. The change delivered 10–100x faster solve times, enabled multi‑order routes and larger problem instances, improved decision quality (reducing edge effects), sped model iteration, and increased Dasher satisfaction through better matching.

DoorDash

Sifeng Lin

Operations Research Scientist