Nowadays data is one of the most important information within a company. Data gives all the important information which are needed for a company do be successful and to make good decisions concerning the long term strategy. There are essentially five types of data in corporations: unstructured (E-Mails, white papers, articles…), transactional (sales, deliveries, invoices…), metadata (report definitions, column descriptions within a database…), hierarchical (relationship information of data) and last but not least master data.
Master data is defined as the „critical“ data of a business which are shared across more than one transactional applications. In general this is divided into four groups: people, things, places and concepts. Further categorizations are subject areas, domain areas or entity types. To explain which data is affected are here some examples:
• People: customer, employee, and salesperson
• Things: product, part, store, and asset
• Concepts: things like contract, warrantee, and licenses
• Places: office locations and geographic divisions
Not all master data which was defined as such has to be managed like master data. It depends on special criteria which should be considered before handling data as master data. Some criteria are for example the value or complexity of data. Master data mostly has more value to the business than simple transaction data. Also the more complex data is the more likely it has to be managed separately and with special attention. Another example is lifetime. Master data is mostly less volatile than transactional data. For some data it is hard to define to which type of data it belongs. This can be defined on the lifetime of data to decide if it is either transactional or master data. A contract for example would be transactional data if its lifespan is very small.
The first question of Master Data Management is: “Why does Master Data have to be managed?”. Nowadays data is part of multiple applications and a lot of people and systems have access to it. So if an error appears within master data, this error appears in all applications and this can cause a lot of trouble. The second questions is: “How is Data Management defined?”. Master Data Management (MDM) contains the technology, tools and processes required to create and maintain consistent and accurate lists of master data. MDM is not always a technological problem, but mostly business processes, technical and political. Also MDM includes creating and maintaining master data. By having a good solution which includes tools and processes to keep the master data clean and consistent can save a lot of money, time and effort. All in all MDM is about “fixing poor data quality at its source and managing constant change”.
In this section I will give one example of a specific MDM Solution. This one is from Oracle. “Oracle MDM combines Operational and Analytical MDM capabilities into a full Enterprise MDM solution”. It contains a modern architecture designed to handle data quality. Additionally it provides prebuilt data models that support operational workloads and service oriented architectures (SOA). It provides tools such as “duplicate identification, elimination and prevention; data attribute survivorship; data quality rules engines; hierarchy management; data standardization; real time change management; data synchronization” and many more.