Master data can be defined as the data that has been cleansed, rationalized, and integrated into an enterprise-wide “system of record” for core business activities.

Berson & Dubov (2007)

Master Data Management (MDM) is the framework of processes and technologies aimed at creating and maintaining an authoritative, reliable, sustainable, accurate, and secure data environment that represents a “single version of truth,” an accepted system of record used both intra- and inter-enterprise across a diverse set of application systems, lines of business, and user communities.

MDM System provides mechanisms for consistency in using master data across organizations. Also it provides the consistency in understanding and builds the trust of master data entities. MDM System is designed to accommodate and manage the change.

Many organizations in current business trends they have multiple, often inconsistent repositories of data. The reason is due to the LOB (Line of Business) divisions, different channels of industry, cross domain distribution of information and packaged systems. Sometimes the reason may be due to Mergers and Acquisitions within.

Before we think about MDM and its practices. Let us step back and raise the questions to understand the dimensions of the system requirements

  1. How are we going to use the master data.?
  2. How do we architect the solution?
  3. What master data do we need to manage?

Considering methods to use the MDM system through collaborative authoring helps the system to coordinate a group of uses and systems in order to reach agreement to set the master data. MDM is mainly focused for the operational transactions and business processes of the enterprise systems where there is much interacting with other application systems and people, also it would be a source of authoritative information for downstream analytical systems and sometimes it is a source of insight itself.

The best way to architect such solutions are through consolidation which is an analytical approach to store the physical data which matches by consolidated view of master data. As this is not a pub-sub system, it is also not used for transactions but could be used for reference. Summarizing my statements. We can bring together master data from a variety of existing systems into a single managed MDM Repository Hub. The data is then transformed, cleansed, matched and integrated to provide the complete fine-tuned record for one or more master data domains. Such data is easily helpful to use the latest analytical and reporting systems as a trusted source to downstream and becomes reference to other operational applications.

Data redundancy is always a solution through MDM to provide a read-only source of master data as a reference to downstream systems.

The following methods can help us to choose a better solution to architect the MDM system and manage it.

  1. Consolidation Implementation: To bring data together by data cleaning and find the matched records. Benefits for reporting purposes and preparing data for downstream systems. Draw-back is more of data is not always latest with operational systems. This implementation would be mainly used for Analytical approaches.
  2. Registry Implementation: Providing cleaned and matched Read-Only Information for identifying the cross references of the information to source systems. This solution is fastest to build but may me more complex to manage. This implementation would be mainly used for Operational approaches.
  3. Co-existence Implementation: storing of master data in one or more locations considering the existing systems doesn’t change and always provide the read-only information. The issues might occur in this implementation are mainly through the systems that are not always consistent. This implementation would be mainly used in collaborative, operational and Analytical approaches.
  4. Centralized Transaction Implementation: Data Centric approach of centralizing complete master data for one or more domains. The benefits for this implementation supports new and existing transactional applications and the issues that duly encounter needs changes to existing systems to exploit. This implementation is also used in collaborative, operational and Analytical approaches.

 

Conceptualizations of MDM are used in various industries like

  • Customer Data Integration Systems
  • Product based Systems and PIM (Product Information Management)
  • IFX (Interactive Financial Exchange)
  • SID (Shared Information/Data Model) of Telecom
  • HL7 ( Health Level 7) of Healthcare