What route optimization for parcel delivery?
Just-in-time operations, autonomous subcontractors, volumes that change every morning: the parcel delivery industry combines characteristics that are incompatible with conventional route optimization. This report shows at which level optimization becomes genuinely relevant.
The context
Parcel delivery, an industry in transition
Customer expectations on lead times keep shortening, parcel volumes keep rising with e-commerce, and the pressure on transport costs remains high. Every day, a depot splits several hundred or even several thousand parcels between its drivers, with volumes that change every morning. In such a short time window, purely manual balancing becomes hard to sustain reliably, day after day.
The goal: understand why conventional route optimization does not work in parcel delivery, and at which level it actually delivers gains.
What you will find inside
The forms of parcel delivery
Traditional, single-parcel, rapid, express: very different transport needs depending on weight and lead time.
A network organization
Line hauls between depots and delivery routes: why territory coverage drives the cost per parcel.
The two industry constraints
Just-in-time operations and subcontractors who run their own territory autonomously.
The right level of optimization
Strategic territory sectorization rather than individual driver routes.
Varied transport needs
Traditional
30 kg – 3 t · 24 h to 48 h
Heavy parcels or pallets, often several units per shipment.
Single-parcel
under 30 kg · 24 h to 48 h
One light parcel per shipment, with the same lead time as traditional.
Rapid
under 30 kg · 24 h max.
Same size as single-parcel, but with a shorter lead time.
Express
under 30 kg · under 24 h
The shortest lead time, often with a guaranteed time slot.
A network organization: to meet these lead times, carriers combine line hauls between depots, sorting on the platform and delivery routes. Poor territory coverage means parcels arriving too late to be delivered the same day, or under-filled trucks that push up the cost per parcel.
Characteristics incompatible with conventional optimization
Operations run just in time
Parcels arrive in the warehouse on a rolling basis: unloading, scanning and sorting all happen at the same time, in an operation that can last up to two hours. There is no way to wait until every parcel has been scanned before starting to sort.
Consequence: this rules out static optimization software, which would require all the data before running any calculation.
Subcontractors run their territory autonomously
Responsible for deliveries and pickups across a defined territory, subcontractors build their own routes. Contractually, there is no hierarchical relationship between the delivery drivers and the carrier.
Consequence: if routes cannot be changed every day, a global daily optimization has no purpose and can hardly be put in place.
The right level of optimization: these two constraints do not make optimization impossible, they move it up one level. The lever lies in how the territory is divided into sectors and how vehicles and drivers are sized upstream. That is what Territory Analytics & Optimization (TAO) does: audit the existing sectorization, optimize it, refine it manually, react in real time and simulate future changes.
