Master data management

from Wikipedia, the free encyclopedia

Master Data Management (English Master Data Management , MDM) includes all the strategic, organizational, methodological and technological activities related to the master data of a company. Master data are important basic data (data) of a company in company data processing that are not changed over a certain period of time.

Your goal is to maximize and ensure long-term data quality and data consistency across systems and applications . The data quality itself can be defined as the suitability of the data for use for a particular purpose. Master data is often stored redundantly in different databases , especially in larger companies . As a rule, this means that the necessary data synchronization is time-consuming and costly.

definition

Master data are status-oriented data that describe the core entities of a company (see. Ref : Schemm). The most important master data objects are customers, suppliers, products, capital goods, personnel and accounts ( lit .: Mertens, White). In contrast to transaction data, master data remains relatively constant in volume over time and has a low frequency of change. Reference data is a special form of master data, as it is used to classify master data objects, for example abbreviations / codes for airports or countries.

Master data management consists of a compilation of processes, guidelines, services and technologies that are used to create, maintain, standardize and manage operational data related to the core operational business of a company. A master data management system represents a comprehensive instance of master data management and thus creates a central operational reference database (system of record). It includes all master data objects of a company such as customers, suppliers and employees. The master data management does not refer to operational transaction data that arise in the context of running business processes such as orders.

Value of the master data

Master data represent a high intangible value. The better it can be used and the better it is maintained, the higher this value can be measured. Many companies recognize the master data as basic data, which must be available as a prerequisite for their business processes and activities. Real and false duplicates and errors in the master data lead to various problems for companies, e.g. delivery to wrong addresses, which lead to monetary losses.

Data quality

Assessment of the data quality

The following criteria can be used to assess the data quality.

Measurement of quality

When setting up the evaluation criteria for the quality of the master data, one encounters fundamental challenges. In order to measure quality, a uniform evaluation system must be used for all data sets. In terms of completeness, the following simple example can show the difficulty. For example, the first name attribute can only apply to natural persons and be used as a completeness criterion, but not for institutional data sets. The number of duplicates is important for assessing the quality, timeliness and value of the master data. This can be easily determined using a random sample. Only certain data sets can also be included in the measurement. For example, only active data records or a certain category / customer group / product group, region, segment, etc.

Definition of the standards for master data

Master data management can only be carried out in a targeted manner if it is clearly defined what good master data is. The standards can only be defined individually in the company. Depending on the orientation of the company, business model or department, purpose and goal, different criteria must be defined individually. Corporations have different requirements than medium-sized companies; there are no general rules. Existing data on customers, suppliers, employees, products, organizational units and charts of accounts provide a good orientation.

One should consider

  • which master data is currently available,
  • who needs which data,
  • for what purpose is the data required,
  • what is the objective for the use?
  • what conditions should be set for the future?

Master data life cycle

In principle, master data is available in the company from the first system to archiving. They go through different cycles:

  1. Plant - central plant, decentralized plant, imports. This also includes procurement and validation. With the help of IT, a minimum requirement for completeness can be implemented.
  2. Release - can be interposed to ensure functional requirements and meet schedules.
  3. Use and maintenance - is the stage in the process of actually using the data. The data is maintained during this phase. This includes updating, completing, correcting and detailing.
  4. Archiving - if the data is out of date, it can be archived. Archiving removes the data from active use, but does not necessarily delete it.

Provision of the master data

There are various IT architecture concepts for master data management ( lit .: Legner, Scheibmayer, Schemm). These can be distinguished by the parameters centrality and harmonization. Data can be kept centralized or decentralized (localized in the individual corporate entities). In addition, data can be strongly harmonized or used only partially harmonized, giving the corresponding corporate entities freedom in dealing with their master data. The different characteristics have a direct influence on the maintenance and migration efforts (e.g. the migration of data in the case of poorly harmonized databases) and the access times. Planning and security issues are also relevant for companies (especially those with several networked locations).

The literature names different characteristics which have different relevance in practice. The most important are probably:

  1. Central master data system
  2. Leading system
  3. Voting node
  4. directory

(Consolidation of the named characteristics from Lit .: Schemm, Legner, Radcliffe and Berson)

Deployment scenarios

As a rule, a distinction is made between three scenarios, which are also the common thread when introducing master data management.

Master data consolidation and harmonization

With master data consolidation, the master data of a so-called business object (e.g. supplier) is connected to the central master data repository , where duplicates are removed and, if necessary, enhanced with information (e.g. DUNS number and EAN ). The merging of master data of the same objects with possibly different identifications is called harmonization . The business case for master data consolidation is, for example, the establishment of a globally standardized reporting system across different systems.

Local master data maintenance

The master data attributes in the different systems are kept consistent via a master data server. The creation of master data and its maintenance continue to be carried out in the application systems that carry master data. If necessary, globally (company-wide) valid master data attributes are maintained on the centrally maintained master data server.

Central master data maintenance

With central master data maintenance, the process of creating and maintaining master data begins on the master data server. From here the distribution to the data management systems of the applications using this master data takes place.

Challenges for companies

Companies face various challenges in dealing with the task of master data management. These can result from very different sources.

  1. Company size
  2. Business combinations, mergers, reorganizations
  3. Increasing data volumes due to new data generation systems (collecting points, internet behavior)
  4. The data collector is not the data user
  5. Parallel systems create double data storage
  6. Different user groups in the company have to agree on a uniform common master data management, even if everyone has a different point of view and different tasks.

Master data management is a permanent process. New products, changes in the market environment, restructuring and internal process redefinitions can make it necessary to adapt the defined process steps.

Role of IT

IT has an essential function. It must provide the infrastructure, processes and interfaces with which the master data management is carried out. The actual master data management is clearly not just a matter of IT, but an independent company process.

literature

  • A. Berson, L. Dubov: Master Data Management and Customer Data Integration for a Global Enterprise. McGraw-Hill, New York 2007, ISBN 978-0-07-151089-9 .
  • Knut Hildebrand, Boris Otto, Anette Weisbecker (Eds.): Master data management, HMD 279. dpunkt.verlag, Heidelberg 2011, ISBN 978-3-89864-750-2 .
  • M. Knapp, F. Hasibether, M. Scheibmayer: Master data management lowers risk when implementing ERP. In: UdZ - Company of the Future. Volume 13, 2/2012, p. 41f. ISSN  1439-2585 (Link) (PDF; 19.2 MB)
  • H. Krcmar : Information Management. 4th edition. Springer, Berlin 2005.
  • C. Legner, B. Otto: Master data management. Whitepaper, St. Gallen 2007.
  • P. Mertens: Integrated Information Processing 1: Operational Systems in Industry. 14th edition. Gabler, Wiesbaden 2004, ISBN 3-8349-4394-0 .
  • J. Radcliffe, A. White, D Newman: How to Choose the Right Architectural Style for Master Data Management. Gartner , Stamford 2006.
  • M. Scheibmayer, E. Naß, M. Birkmeier: Master Data Management - White Paper. RWTH Aachen, Aachen 2011. (Link)
  • Jan Werner Schemm: Inter-company master data management - solutions for data synchronization between retail and the consumer goods industry. Springer Verlag, Berlin / Heidelberg 2009, ISBN 978-3-540-89029-4 .
  • A. White, D. Prior, J. Radcliffe, B. Wood, J. Holincheck: Emergence of EIM Drives Semantic Reconciliation. Gartner, Stamford 2004.

Individual evidence

  1. ↑ Master data management encyclopedia for business informatics, accessed on March 14, 2013.
  2. ^ Definition of "master data" Gabler Wirtschaftslexikon, accessed on March 14, 2013.

Web links