The management of the fundamental data building blocks that are shared across multiple business transactions. It is an infrastructure that is built once and then managed. it is also a program, not a one-time project.
Different types of master data examples: Customer, Employee, Product, Material, Supplier, Financial, ...
Why is MDM important ?
- Master data is a strategic asset of the business and needs to be managed with the diligence of other strategic assets.
- Developing a strategy for managing master data can help alleviate some of the business’s pain points.
- Master data management provides the business with consistent and comprehensive core information across the enterprise.
- Master data management provides better performance, process improvement, and assists with regulatory requirements.
- Master data management helps satisfy governance, regulatory, and compliance requirements
Various Terms are used for MDM :
External Databases (Service Provider): Third party suppliers and managers of domain specific master data
Examples: database marketing, government service bureaus
Persistent (Database): Master information file/database, system of record (SOR). Based mainly on Operational Data Store (ODS), active data warehouse
Relational DBMS + extract-transform-load (ETL) + data quality (DQ)
Registry (Virtual): Metadata layer + distributed query (e.g., EII)
Enterprise application integration (e.g., EAI), distributed system, Portal
Composite (Hybrid): Ability to fine-tune performance and availability by altering amount of master data persisted. Based on XML, web services, service-oriented architecture (SOA)
An approach to MDM Implementation : Start with Quick wins
There are plenty of ways to get MDM done. Some are components of large transformation programs, while others are independent, standalone programs.
In an MDM implementation, one of my guiding principles is to think big strategically but start projects in smaller phases.
MDM : Key lessons learned
In an MDM implementation, there are some key lessons learned that should be considered when initiating an MDM program.
• Joint business and IT team
• Make the case for change
• Data as a common good
• Think big but start small
• Measure and communicate success
• Processes first, technology last
• Business ownership of data
• Roles and responsibilities
• Data cleanliness and migration
• Communicate, communicate, communicate
No comments:
Post a Comment