Cainiao merges many parcels into one and halves cross-border delivery time

A consumer ordering from several merchants on AliExpress used to receive several slow international parcels, each routed and shipped independently. Each parcel was cheap; the bundle was slow, and slow kills cross-border e-commerce.

Cainiao built a consolidation business model and the analytics to run it. A two-phase online optimization decides, as parcels arrive at the consolidation hub, which ones to merge into one package and which shipping method that package should get — fast enough to run in production, with provable bounds on how far its decisions drift from optimal.

Supporting models estimate delivery-time distributions per lane, and simulation sets service targets such as the ten-day on-time rate, so the optimizer's cost-speed tradeoffs reflect what the network can actually do.

In production since 2020 for parcels from China to more than 50 countries, the system saves tens of millions of dollars annually and cut delivery time by at least 50%, according to a 2024 INFORMS Journal on Applied Analytics paper.

Decide consolidation and shipping method in one optimizer

It couples two decisions — what to merge and how to ship — that were previously made separately and suboptimally.

Provable sublinear regret bounds let a heuristic-speed algorithm be trusted with production traffic.

Consolidation converted a pure cost center (many small parcels) into a faster product customers prefer.

When each order is optimized alone, the customer pays the seam. Consolidating across decisions — which parcels, which lane, which speed — finds savings no single decision can.

The consolidation system became core to Cainiao's cross-border offering; by 2024 it launched grouped-delivery services such as five-day delivery to the US, and the work was published in the INFORMS Journal on Applied Analytics.