How to Build a Robust Master Data Management System for Large Enterprises

Project Background

Master data management (MDM) integrates core data from multiple business systems, cleanses and enriches it, and distributes unified, accurate, authoritative master data to operational and analytical applications across the enterprise.

MDM centralizes data, ensures consistency, improves compliance, accelerates new application deployment, enhances customer insight, and speeds product launch. From an IT perspective, MDM enhances flexibility, builds enterprise‑wide data management foundations, and adapts to changing business needs.

Most enterprises face MDM challenges due to incremental business and IT development, resulting in siloed systems such as OA, HR, PLM, CRM, ERP, MES, etc.

It is impossible for a single system to cover all business processes, even for large vendors, leading to distributed deployments.

1. Project Background

A group enterprise has built numerous information systems (contract management, HR, OA, procurement, archives) over decades, accumulating data and connecting systems via APIs. Current MDM issues include:

Inconsistent naming and attributes across systems ("one object, multiple codes").

Data silos due to independent departmental maintenance.

Poor data quality and inconsistent standards.

Lack of a unified data exchange platform, causing high integration costs.

2. Project Planning

The overall goal is to create four “ones”: a universal MDM framework, an organizational plan, standardized processes, and an MDM platform for data distribution.

3. Project Implementation

The standard MDM implementation process includes consulting and planning, followed by execution, divided into four steps: current state analysis, framework planning, implementation planning, and platform deployment.

Current state analysis: assess enterprise status, identify pain points, evaluate data‑management maturity.

Framework planning: design organization, MDM policies, assessment standards, and operation methods.

Implementation: define coding, attributes, fields, approval workflows, integration, and migration strategies.

Platform deployment: realize data models, maintenance, and governance on the MDM platform.

4. Project Challenges

Challenge 1: Difficult data access – Sensitive financial and contract data require strict agreements, leading to long approval cycles. Solution: address data acquisition early in the requirements phase.

Challenge 2: Defining standards, poor quality, hard to refactor – Issues include identifier conversion, department‑code history, lack of standards, and resistance from business systems. Solution: provide professional guidance and a three‑step resolution process.

Challenge 3: Communication barriers and resistance – Conflicts between process approvals and business workflows, and resistance to system changes. Solution: differentiate technical issues (handled by the project team) from organizational coordination, requiring leadership support.