How Weimeng Transformed Data Asset Governance: A Practical Blueprint for Enterprises

Introduction

In the digital economy era, data has become the most important asset for enterprises. This article shares Weimeng's data asset governance practice, showing how systematic management can improve data quality and maximize business value.

Background

Weimeng, a new‑retail service platform, processes massive user and merchant data daily. It faced problems such as lack of unified metadata standards, unclear data‑asset catalog, ambiguous ownership, frequent security incidents, low sharing efficiency, and missing value‑assessment mechanisms.

Pain Points

Metadata issues: unclear responsibility, missing business description, insufficient basic information, and unrecorded changes.

Lineage issues: incomplete or inaccurate lineage, lack of field‑level lineage.

Metric issues: non‑standard names, unclear sources, ambiguous definitions, missing thresholds and usage statistics.

Cost issues: costs not allocated to individuals or development groups, no cost‑analysis reports.

Governance System

The system follows three pillars—data assetization, quantitative operation, and value realization—covering goals, framework, implementation steps, and evaluation.

Construction Goals

Build a data‑management platform to manage tables and metrics, reducing manual effort.

Establish data lineage to trace usage and impact analysis.

Create a data dictionary and metric system to break internal data silos.

Implement data lifecycle management to lower storage costs.

Framework Design

The core is an enterprise‑level metadata management center, including asset catalog, search engine, lineage visualization, and full‑view monitoring. It addresses three metadata dimensions:

Technical Metadata

Physical details: table name, column type, constraints.

Storage details: type, location, size, partitions.

Operational details: task ID, update frequency, quality monitoring.

Business Metadata

Definitions, terminology, metric names, calculation logic, sensitivity level.

Operation Metadata

Owner, user, permissions, DDL change logs.

Asset Inventory (Metadata Service)

Data map with full‑text search and basic metadata (description, owner, business line, related tasks, access heat).

Marketplace matrix guiding users to discover required tables.

Heat analysis of table and metric access.

Storage and Cost Management

Lifecycle management based on heat analysis to keep only valuable tables.

Cost analysis separating storage cost and compute cost.

Implementation Steps (PDCA)

Phase 1: Inventory & Planning

Assess current data‑asset status and define actionable goals.

Collect requirements and create an execution plan.

Phase 2: Execution

Apply standards and platform tools to govern existing data.

Deliver a data‑governance list.

Phase 3: Continuous Inspection

Regularly check governance outcomes, analyze metrics, and improve processes.

Phase 4: Asset Operation

Track daily data flow, evaluate assets, and close the feedback loop for continuous value release.

Evaluation System

Asset analysis reports provide multi‑dimensional views (layer, database, development team, owner) of the current asset status.

Results

Key achievements include:

Metadata: over 1,000 tables now have owners; description completeness reached 91.75%; field comment completeness reached 83.76%.

Lineage: graph database enables visual traceability and impact analysis.

Metrics: standardized metric dictionary improves understanding and usage.

Cost: storage information completeness 95%; heat models support value analysis.

Conclusion & Outlook

After three phases, Weimeng improved data standardization, architecture, security, and application, lowering barriers for users and enhancing decision‑making. Future work will refine processes, adopt advanced database technologies, and further optimize data‑asset storage to boost enterprise competitiveness.