How SF Express Scaled Operations: Lessons from Digital Transformation

1. Starting with Digitalization

SF Express began a digital transformation years ago, moving order placement, routing, and waybill generation online to improve efficiency and user experience. Paper waybills were replaced by QR‑code scans in 2017, and by the end of the year all orders were fully electronic.

2. Goal Alignment

Success requires a unified goal across diverse roles—operations, development, product, business, and promotion. Teams must understand business value, adopt a business‑centric language, and view operations as a value‑adding function rather than a mere execution unit.

Business focus: technical staff must grasp business objectives and communicate in business terms.

Operations mindset: shift from a pure execution team to a strategic role that evaluates infrastructure, cost, and security.

Success definition: evaluate success at the project level, breaking departmental silos to align perspectives.

3. Early Stage – Efficiency

Initial challenges included overly complex approval processes, fragmented organizational structures, and unclear collaboration between operations and development.

Process: introduce lightweight workflows and integrate tools to reduce approval bottlenecks.

Full‑stack ops team: dissolve specialized silos, give a single team end‑to‑end responsibility for incidents.

Mindset shift: treat operations and development as equal partners, fostering boundary‑less cooperation.

4. Growth Phase – Stability

During rapid growth (100 k to 1 M+ orders per day), performance issues and technical debt surfaced. The team emphasized a professional, efficient, open culture and built elastic architectures.

Application architecture principles:

Stateless services.

Single‑point of failure avoidance (the “shortest board” principle).

Distributed design for easy scaling.

Horizontal scaling, with early sharding of large tables and databases.

Infrastructure layer supports horizontal and rapid scaling, handling thousands of servers and enabling fast provisioning.

Key practices include gray‑release deployments, fine‑grained traffic routing, and service protection (rate limiting, circuit breaking, isolation).

5. Monitoring and Resilience

Monitoring covers both infrastructure (CPU, I/O, memory, APM) and business metrics (response codes, latency, thresholds). End‑to‑end tracing and comprehensive logging are essential for micro‑service fault isolation.

Disaster recovery relies on active‑active deployments, cross‑region traffic splitting, and synchronized data stores (Redis, DB) using Kafka‑based routing.

6. Continuing Value

Operations delivers value through quality (availability, MTTR, satisfaction), cost efficiency (resource utilization, capacity planning), efficiency (shifting from manual ops to platform engineering), and data‑driven insights that support business decisions.

Staying abreast of emerging technologies ensures the organization can adapt to change.

Note: This content reflects the presentation by Zeng Xiancheng at GOPS 2019 Shenzhen.