"Hong Kong's Largest Last-Mile Delivery Network"
Built and scaled a smart locker ecosystem from zero to 1,000+ locations handling 11 million parcels annually, transforming Hong Kong's last-mile delivery infrastructure.
Operational data, customer complaints, and field observations revealed a fundamentally broken last-mile delivery model in Hong Kong.
Mismatch between delivery windows and customer availability caused repeated attempts, inflating per-order costs and frustrating customers.
Manual door-to-door delivery in dense high-rises required multiple elevator trips and long walking distances, making human-only delivery unviable at scale.
Package theft, damage, and loss at unattended doorsteps were persistent complaints, particularly for high-value items from premium brands.
No unified network connected logistics providers, merchants, and consumers. Each carrier operated independently, duplicating effort with coverage gaps.
Hong Kong residents in high-rise apartments who shop online frequently but are rarely home during delivery hours. Failed first-attempt deliveries exceeded industry benchmarks.
Courier companies (S.F. Express, UPS) handling high-volume daily deliveries. Drivers spent 30-40% of time on failed delivery reattempts.
Online retailers and DTC brands shipping hundreds of orders daily, seeking reliable last-mile options that reduce failed deliveries.
By owning the physical last-mile infrastructure (smart lockers + micro-warehouses) rather than competing on delivery labor, we can build a defensible network effect where more locations drive more logistics partners, which drives more consumers, which attracts more merchants.
They optimize driver routes; we eliminate the need for door-to-door delivery entirely.
We partnered with 800 convenience stores rather than competing against them, converting them into extended network nodes while our smart lockers handled the high-volume, 24/7 automated overflow that stores couldn't support.
We combined lockers with micro-warehouse sorting, in-house delivery teams, and an open API platform — a full-stack logistics solution, not just hardware.
Hong Kong parcel volume growing 15-20% year-over-year
High-rise environment made door-to-door delivery increasingly uneconomical
Government infrastructure initiatives created regulatory tailwinds
Build an in-house delivery team alongside third-party partnerships to control quality and cost
Pure asset-light marketplace model — we needed to own the physical infrastructure to guarantee service levels
Expand to 2,000+ points via convenience store partnerships for coverage
Owning all locations exclusively — capital-intensive and slower to scale
Open API platform for partner integration
Closed ecosystem — would have limited network growth and partner adoption
Deploy core smart locker hardware and validate unit economics with an in-house delivery team before external scaling.
Automate sorting to handle growing volume — full and semi-automated systems serving 1,000+ lockers, plus logistics provider partnerships to extend reach.
Elevate the platform from utility to brand differentiator — AI facial recognition lockers with YSL Beauté, gift redemption programs, express delivery service.
Saturate the market — 800 convenience store partnerships expanding to 2,000+ points, 11M annual parcels, largest last-mile network in Hong Kong.
Tableau-based visualizations to forecast capacity demand by district, prioritizing locker placement in high-density, high-traffic areas rather than uniform distribution.
Designed lockers as active logistics nodes — parcels sorted overnight at transit hubs, transferred to micro-warehouses, then distributed to lockers. Eliminated the "hub to door" model.
Integrated in-house delivery teams with third-party logistics (S.F. Express, UPS). Built a cost model optimizing per-order spend while maintaining service quality.
IVR voice navigation for self-service resolution and cloud CCTV for remote anomaly detection — replacing on-site staff with scalable, tech-enabled support.
Prioritized reliability and cost efficiency over speed promises that would strain the network.
Scoped out refrigerated storage to avoid hardware complexity and regulatory burden.
Deprioritized reverse logistics to keep locker turnover high for outbound deliveries.
Established Hong Kong's largest last-mile delivery network, becoming the default infrastructure for major logistics partners.
Secured collaborations with YSL Beauté for AI-powered facial recognition smart lockers, proving platform value beyond utility.
Built an open API platform adopted by S.F. Express, UPS, and other carriers, transforming from single operator to logistics ecosystem.
Smart customer service (IVR + cloud CCTV) measurably reduced complaints and increased customer satisfaction scores.
Automated sorting systems and capacity alerts reduced failed deliveries and improved locker turnover through data-driven reallocation.
The decision to build modular infrastructure — separating locker hardware, sorting systems, and API layers — meant scaling from 100 to 1,000+ locations didn't require re-engineering the core platform.
Cash-back pickup incentives and capacity alerts both emerged from observing real user behavior — pickup times, utilization patterns, complaint trends — not pre-launch assumptions.
Every product decision — from micro-warehouse design to convenience store partnerships — required alignment across engineering, operations, and commercial teams. The strongest outcomes came when all three moved together.
Tableau dashboards became the shared language across teams — from locker placement decisions to cost model optimization to partner performance tracking. Metrics guided priorities at every stage.
This project shaped how I approach subsequent product initiatives — with stronger strategic foresight, deeper user empathy, and a relentless focus on long-term value creation. The smart locker platform remains a cornerstone of my career, exemplifying my commitment to building products that drive efficiency, trust, and meaningful impact at scale.