Enterprise Master Data Management: An SOA Approach to Managing Core Information

This is a great book that merges two hot topics in data management today: Service Oriented Architecture and Master Data Management. We highly recommend looking into this book as it can certainly bring value to your organization. Check it out here.

Customer Data Integration: Reaching a Single Version of the Truth

Often times MDM is brought into an organization because of a data warehousing project or to assist with identity resolution type work. This book provides great insight into how to reach that single version of the truth. Get more info.

Data Driven: Profiting from Your Most Important Business Asset

This is a fantastic book that digs deep into the value of Master Data Management and the value that it can bring to your organization. Perfect whether you are pitching MDM or trying to sell its value. Read more about it.

Master Data Management Tools

 

The problem of managing master data can be addressed using a variety of solutions, but among these solutions are a common set of important tools to be used in the processes of managing important information. Some of the physical tools include data networks, data warehouses, data marts and an operational data store.

Data networks provide a method for the transmission of data within an enterprise, allowing for the transfer and storage of information. Data can be held on a common server, accessed by various departments or subsidiaries of an enterprise, and used in business operations. Utilizing a data network, it becomes possible to efficiently collect important master data from different sources, allowing for its consolidation and proper management. Data networks also create an easy method for sharing critical business information to be used in daily transactions and business functions, as data can be accessed from many different locations in an enterprise, allowing for corporate data to be quickly shared between departments or even companies.

Data warehouses serve as storage location for the master data, providing an electronic repository where data can be analyzed and reported. In addition to the storage of data, data warehouses involve processes for the retrieval, analysis and management of data and information needed to properly access it. Data warehouses are generally optimized for data retrieval, and lack data normalization to preserve data integrity, which is instead handled by the operational systems that feed information to the warehouse. Some of the advantages of data warehousing are the use of a common data model to increase the ease of analysis and reporting, independent yet cooperative work with operational systems to allow for data to be accessed without slowing down operational systems while still enhancing their ability, and long-term storage for master data that may be destroyed on the source systems and still remain intact in the data warehouse.

A subset of the data warehouse is the data mart. Data marts contain data relating to a more specific purpose or subject, and are used by groups within an enterprise to access relevant information to their business needs. Data marts can exist independently of a data warehouse, but generally data marts are either consolidated to form a data warehouse, or subsets of data are extracted to form data marts. Due to the more limited scope of data marts, they provide easier access to frequently-used, relevant data for specific entities in an enterprise as well as creating a data access point with a more defined set of end-users to facilitate operations.

In contrast to the more long-term storage intended for data warehouses, operational data stores are responsible for the more volatile “real-time” storage and use of data. Relevant and detailed data is gathered from multiple sources and stored to allow for access to current operational information needed for daily transactions and corporate functions. Data is cleaned, any redundancy is resolved, the remaining data is checked against enterprise rules for data integrity, and then this enriched and refined data is stored, with data being frequently rotated as more recent data history takes the place of older data that is then relegated to a data warehouse.