eBay’s pickup-point direction reflects a broader shift toward orchestrated delivery nodes that lower last-mile cost by concentrating volume, widening customer choice and using existing capacity more effectively.
In Brief
- Delivery networks are moving from uniform home service toward distributed access points and multiple service speeds.
- Pickup points, partner stores, lockers, cross-docks and leased facilities are being connected through common routing and cost controls.
- The model improves density but introduces difficult decisions about inventory placement, partner terms, capacity utilization and service consistency.
The Underlying Pattern and Stakes
The strategic value of a large pickup network lies in orchestration rather than the number of locations. A pickup point converts dispersed residential stops into consolidated flows. That can reduce failed deliveries, increase drops per route and give transport capacity a more predictable destination. The same operating logic is visible in Shopee’s collection points, MercadoLibre’s multi-speed network and Pattern’s partner-led fulfillment model.
This is a redesign of node roles across the delivery network. Stores, lockers and third-party facilities become operational endpoints rather than passive commercial locations. Slower orders can fill spare cross-dock and linehaul capacity, urgent orders can draw from inventory close to demand, and categories with difficult delivery economics can move through specialist partner nodes. The opportunity is lower cost with wider service choice. The requirement is a control layer capable of matching each order to the right path.
How Companies Are Converging
Distributed access points are becoming a formal part of network capacity. Shopee expanded its Taiwan collection-point network to more than 3,100 locations, nearly 50% more than a year earlier. It also began shipping directly to lockers without additional packaging, reducing handling while helping improve buyer waiting time by 12%. The significance extends beyond convenience. Each collection point creates a repeatable delivery destination where parcels can be consolidated and handoffs standardized.
Existing physical networks are also being recruited to support faster flows. Shopee connected around 7,000 convenience and pharmacy stores to instant delivery in Indonesia, giving its network local inventory sources without reproducing that store footprint. Pattern uses a comparable partner-capacity principle for oversized products, combining four Chewy nodes with one of its own and deploying its fulfillment technology inside the partner facilities. Both configurations extend service coverage by assigning new roles to assets that already exist.
Service segmentation supplies the volume needed to keep those nodes productive. MercadoLibre directs flexible orders through a slow-shipping network when spare capacity appears in fulfillment centers and cross-docks. Shopee runs instant, same-day, next-day and collection-point options across the same broad network. In Indonesia, instant-delivery volume grew more than 35% while cost per order fell around 20%. The shared mechanism is controlled differentiation: orders receive the service level their economics and customer promise require instead of entering one default flow.
Physical expansion is being kept selective. Shopee leases and operates fulfillment centers rather than owning the buildings, adds capacity against forecast volume and maintains relatively high utilization. Pattern uses light cross-docks, warehouse automation and external logistics infrastructure, keeping capital investment near 1% of revenue. MercadoLibre is investing more heavily in its own fulfillment capability, but it also relies on density and spare-capacity routing to improve returns from each node. Ownership varies; the common priority is control over flow without unnecessary fixed assets.
Operating Model Mechanics
The first requirement is a clear hierarchy of node roles. A fulfillment center holds inventory and controls dispatch. A cross-dock consolidates flow without becoming a storage point. A partner store provides local stock, customer access or both. A locker offers unattended handoff. Treating these locations as interchangeable would obscure their different cut-off times, handling costs, storage limits and failure modes.
Order routing then becomes a cost and service decision made at transaction level. MercadoLibre’s system can place flexible orders into unused capacity, while its fast network protects more urgent promises. Shopee dynamically optimizes for speed, cost and customer preference. Pattern assesses variable economics by SKU, marketplace and country, then back-solves the inventory purchase price required to preserve unit economics. These approaches connect routing decisions to commercial value rather than treating freight as a pooled overhead.
Inventory positioning determines whether distributed delivery points actually improve performance. Shopee’s fulfillment orders arrive faster than its platform average, with more than one-third of its fulfilled parcels in Asia delivered the next day during March. That advantage depends on sellers moving inventory into platform-controlled facilities before orders arrive. Pattern assumes inventory ownership from brand partners, giving it direct control over availability across more than 60 marketplaces. MercadoLibre’s locally fulfilled cross-border operation is profitable, while international fulfillment remains scale-sensitive, illustrating the economic value of placing stock near demand.
Commercial rules must reinforce the physical design. MercadoLibre moved merchant shipping charges toward actual dimensions and weight, aligning fees more closely with handling and transport cost. Pattern uses marketplace-specific fulfillment and commission costs when setting the price it pays suppliers. Shopee reinvests part of its marketplace monetization into fulfillment and delivery while monitoring whether seller pricing remains competitive. In each model, the network cannot be separated from the rules determining who pays for size, speed, inventory and failed utilization.
To make this configuration work in daily planning, node capacity, order eligibility and service promises need to sit in the same decision layer. A pickup location that appears available to customers but is missing cut-off, capacity or exception data creates additional handling rather than savings.
Risk, Constraints and Trade-offs
Density is the central dependency. Pickup points and lockers lower cost when enough compatible orders can be consolidated into each route and location. Thin volume creates another network of underused nodes. Shopee’s collection-point expansion is supported by 29% order growth, while MercadoLibre reduced shipping cost per unit by 17% as Brazil item volume rose 56%. Those outcomes show why demand density and network efficiency must advance together.
Service choice also creates operational complexity. A network offering home delivery, pickup, lockers, same-day service and a slower economy lane must maintain distinct routing rules and customer promises. Misclassification can send an urgent order into a capacity-filling lane or move a low-value order through an uneconomic premium route. The cost advantage depends on disciplined eligibility logic, not simply adding more options.
The capital trade-off differs by configuration. Shopee and Pattern limit fixed investment through leased facilities and partner nodes, but accept greater dependence on external operating standards and contracts. MercadoLibre exercises more direct control and captures stronger density benefits, while absorbing short-term margin pressure from fulfillment expansion and shipping subsidies. Both routes require governance over service quality, capacity rights and cost pass-through when fuel or labor prices change.
Operational Self-check
Questions that expose whether a distributed pickup model is reducing total cost or merely adding another layer of complexity include:
- Are pickup, locker and home-delivery orders assigned through one cost and service engine, or through separate channel rules?
- Can the network identify which nodes create genuine route consolidation after failed handoffs and extra handling are included?
- Do partner agreements define capacity access, exception ownership and service data at the level required for real-time routing?
What This Pattern Signals
The visible shift is from last-mile ownership toward last-mile orchestration. Competitive advantage increasingly comes from controlling the decision about where an order flows, even when the physical node belongs to a store partner, landlord, carrier or specialist operator. Pickup points are one expression of that broader model.
If the pattern persists, network design will place greater weight on node interoperability, transaction-level cost-to-serve and service-tier governance. Fulfillment centers, cross-docks, stores and lockers will operate as a coordinated portfolio rather than separate channels. The evidence from Shopee, MercadoLibre and Pattern suggests a durable structural change, supported by measurable gains in delivery cost, waiting time and capacity utilization.
This article is based on recent earnings reports and public disclosures from the companies referenced.
