It all comes down to what’s called master data — “entity” data about customers, suppliers, partners, products, materials and accounts that is used with each transaction, application, report and decision. Effectively managed and integrated, this data can deliver the insights necessary to improve decision making — driving measurable business success. Unmanaged, it represents a wasted opportunity.
- Create business, technical and architectural requirements
- Identify operating and engagement models to facilitate effective decision-making and efficient operational processes
- Scope a realistic MDM implementation plan and effort
- Prioritize data domains and supporting processes based on business value
- Implement a market-leading single, or multi-domain MDM capability
- Identify and implement value-added technologies such as data quality and metadata management
Reference data is used to represent a vast array of business concepts that are important in such diverse activities as segmentation, driving business rules, and reporting to customers and regulators.
In many enterprises, the number of business concepts represented by reference data far exceeds that represented by tables and columns implemented in databases. Problems occur when reference data is not synchronized across the enterprise, and are not understood clearly — or not even known about — by the people who have to use them.
- Dealing with the unique requirements of “external” reference data (which comes from outside the enterprise) versus the unique requirements of “internal” reference data (which is produced wholly within the enterprise)
- Understanding the many different roles involved in RDM and making sure the people playing these roles are well-trained and have all the tools and techniques they need to be successful
- Recognizing that producing reference data is quite different to distributing it, and that an integrated supply chain for reference data is put in place
- Making sure the content of reference data is as high quality as possible; not only must there be zero defects, but content like definitions must be complete, accurate, and consumable
- Handling complexities like hierarchies, crosswalks, splits, merges and standard conversions