How Huawei Built a Comprehensive Data Governance Framework for Digital Transformation

Huawei's Digital Transformation Vision

In 2017 Huawei announced a new corporate vision: “Bring the digital world to every person, every family, every organization, and build an intelligent world of interconnected things.” The company’s CIO, Tao Jingwen, added the goal of becoming an industry benchmark by achieving fully connected intelligence.

Huawei then defined a digital‑transformation blueprint with five initiatives, emphasizing data governance and digital operations to improve customer interaction and internal efficiency.

Data Governance Phases

Phase 1 (2007‑2016): Established a data‑management organization, created a data‑management framework, issued policies, appointed data owners, and improved data quality through unified architecture, trusted sources, and measurement mechanisms.

Continuously improve data quality and reduce error‑correction costs.

Enable end‑to‑end data flow to boost business efficiency.

Phase 2 (2017‑present): Built a data foundation that aggregates enterprise‑wide data, provides data services, maps data, and ensures security and privacy, enabling on‑demand sharing, agile self‑service, and transparent usage.

Business visibility for fast, accurate decisions.

Artificial intelligence for business automation.

Data‑driven innovation as a competitive advantage.

Data Classification Framework

Huawei classifies data by characteristics and governance methods: internal vs. external, structured vs. unstructured, and metadata. Structured data is further divided into basic, master, transaction, report, observation, and rule data.

Structured Data Management

1. Basic Data (reference data such as countries, currencies) is static and used for classification, requiring change‑management and unified standards.

2. Master Data (customers, products, suppliers, organizations, personnel) demands source‑of‑truth consistency and regular quality checks.

3. Transaction Data records business events; it heavily references master and basic data and must preserve those relationships.

4. Report Data is processed data used for decision‑making, including fact tables, metrics, dimensions, statistical and trend functions, and rule data.

5. Observation Data captures process‑level information (e.g., logs, sensor data) and requires clear linkage to business objects.

6. Rule Data structures business rules (decision tables, scoring cards) and must be managed for configurability, visibility, and traceability.

Unstructured Data Management

With growing big‑data needs, Huawei manages documents, images, audio, and video by extracting basic features (title, format, source) and enriched content (tags, similarity) using a unified‑distributed metadata approach.

External Data Management

External data (supplier certifications, consumer insights) is governed primarily by compliance, with clear responsibility, auditability, and controlled access.

Metadata Management

Metadata (business, technical, operational) bridges the gap between business and IT, supporting data discovery, data‑service management, data‑lake transparency, and compliance across the entire data value chain.

Huawei’s comprehensive data‑governance and metadata framework enables the company to turn raw data into trusted, reusable assets that drive intelligent, connected operations.