Master data is a term from computer science ( data management ) and business administration (in connection with the use of business application software ) that is used to describe data that contains basic information about business-relevant objects (such as products, suppliers, customers, employees) (e.g. B. Name and price, discount agreements, names and addresses, entry date, ...) that are required for ongoing (e.g. periodic) processing. They are therefore also called static data or basic data , also known as “master data” or “core data”.
Related categories of data:
In the context of operational quantity and value considerations, there are in addition to the master data and the term inventory data (English inventory data or quantity on hand data ) which extends over so-called transaction data form (such as data about orders, orders, reminders, payments) and change. ( Ref : Hansen) data generated in connection with the change of master data, change data mentioned. They are processed in master data management (often not as part of operational processes).
There are no mandatory dependencies on these data categories with regard to the organizational and storage form of the data stocks . However, there are certain preferences; For example, master data is often stored and processed in databases (which makes it easy to evaluate and its quality can be protected), operational transaction data (such as 'received payments') are often in a common, operating system-specific file format .
Master data is used in a wide variety of areas, mostly in applications from companies and other organizations.
The following types of master data (and their identification terms in brackets) can serve as examples:
- Customer master record (customer number of a customer at a department store or hotel)
- Personnel master record (personnel number of the employee in the personnel administration of a company)
- Material master record ( material number of an inventory item in materials management )
- Vendor master record (vendor number in accounts payable , same as customer master record / number)
- G / L account master record (G / L account number in the accounting )
- Asset master record ( asset number in asset accounting )
- Patient master record (patient number in a doctor's office or clinic)
Master data can be defined using three criteria:
- on the existential dependency that can exist between individual entities (as well as between the associated entity types ; synonymous object types): B. an order anticipates the existence of a customer; therefore: no order without a customer; see also Referential Integrity
- about the different volume of data over time (there will be more orders than customers over time) and
- about the content or processing and about temporal aspects.
In the example above, master data is the customer data, while the order data falls under transaction data.
1. Definition by means of existential dependency
In an entity relationship diagram (ER diagram), the entity types can be arranged in such a way that the original, i.e. H. existentially independent entity types on one side (assumed left) and the existentially dependent entity types to the right. Data of the entity types on the far left are called master data, while data of the entity types on the far right with the strongest dependencies are called transaction data. The data of the object types in between can also be gradually assigned to the master and movement data depending on the location. For every two existentially dependent entity types it then applies that the existentially independent entity type contains the more basic and the dependent entity type the more flexible data. The two terms master data and movement data are thus gradually antonymous. Since the more flexible data refer to the more robust data, the latter are also called reference data. ( Lit .: Sinz)
For example, customer and article data fall under master data, while data on order and invoice items are movement data.
In this context, a distinction is sometimes made between master data and so-called main master data (e.g. customer) and auxiliary master data (e.g. industry, region, shipping type, delivery condition or payment target).
2. Definition by means of change volume (statics vs. dynamics)
Data that changes frequently (creation, change, deletion) are often referred to as "movement data " - although the change volume (change frequency) alone is not an exclusive criterion for this classification. In spite of this, movement data is generated dynamically as events occur over time and usually much more frequently than the master or inventory data on which they act. Their existence (from creation to deletion ) is - apart from data backup and historization - often only limited to the time until the process it represents has found its expression in the respective master or inventory data. "Master data is data that remains unchanged over a longer period of time ." ( Lit .: SAP)
3. Definition according to content or process-oriented and temporal aspects
Master data are status-oriented data that are used to identify, classify and characterize issues and are available unchanged over a longer period of time. They are also known as fixed dates.
Movement data are processing-oriented data that are constantly being created by the operational performance processes, are continuously incorporated into the data processing operations and usually cause a change in the inventory data. The motions are as transactions (Engl. Trans action ) denotes their data forms usually contain u. a. a time specification (date, time, ...).
Inventory data is - like the master data - condition-oriented data, which characterize the operational volume and value structure, and therefore not unchanged. In contrast to the master data, they are subject to a systematic change as a result of operations, which is caused by the processing of transaction data. ( Lit .: Scheithauer)
Example: A 'cash withdrawal' (= movement) occurs - in the business transaction or business process of the same name - at a certain point in time and leads to a change in the 'account balance' (= inventory data). Without such and similar movements, the account would remain unchanged for a long time, like the account master data (e.g. the credit limit ), i.e. static.
The three definitions do not always lead to the same scope of the terms master and transaction data; the definitions of terms are therefore not equivalent. So is z. B. the stock level of an article in a supermarket according to definition 1 a master date; since it is a property of the existentially independent entity type article. Since the inventory changes relatively frequently, it is a movement date according to definition 2. According to definition 3, it is an inventory date, because the inventory describes the quantitative status of an article.
The property of a date to be classified as a master date or a movement date depends on the environment (context). What is master data in one application or database (e.g. article data in a warehouse management system) can be movement data in another database (e.g. article data in a database for creating a group-wide product catalog).
Definition 1 on the basis of existential dependency is mainly used for data structuring and data storage (e.g. by data managers and database administrators).
Definition 2 on the basis of the change volume is often used in connection with data-changing processes (change service) (e.g. by application developers), while
definition 3, based on business process-oriented and temporal aspects, is more used when considering business processes.
Master data history
Since everything changes in the long term that were thought to be constant and invariable in the medium term, master data often also has a validity date (valid from, valid to). These slow changes are then recorded in the master data history . The master data history plays an important role where master data - e.g. B. Cost centers or profit centers after a reorganization of the company - change from one business period to the next. Transaction data from different business periods are often available in an ERP system at the same time, e.g. B. for evaluation purposes . Now it may z. B. In the document display it does not happen that a document from the previous period with a cost center account assignment '1000' is displayed in terms of responsibility and organizational assignment to the current, current definition status of cost center '1000'; rather, the display of these cost center attributes must correspond to the time at which the receipt was posted.
Properties of the master and transaction data
Despite all the differences in the definitions above, there are several similarities:
- are characterized by a certain statics (invariant over time) and usually have no time reference,
- are often used by several applications or business areas, e.g. B. partial master records (purchasing, construction, disposition, accounting, sales, work preparation),
- in analytical evaluations are often the criteria according to which the evaluation is carried out (e.g. product, branch, customer). Thus they are candidates for the dimensions in Online Analytical Processing (OLAP) and
- are mostly held for the long term.
- are characterized by a certain dynamic (time variant) and usually have a time reference (validity date),
- are often used by a few applications,
- often provide the facts in analytical evaluations, with which they are candidates for the cell contents in OLAP,
- are usually only required for a limited period and therefore
- held short or medium term.
- are characterized by a certain dynamic (temporally variant),
- often provide the facts in analytical evaluations. Thus, they are candidates for the cell contents in OLAP and
- are mostly held for the long term.
Master data maintenance
The maintenance of master data is important for companies for various reasons: A high quality of the master data is a prerequisite for efficient processes, satisfied customers and reliable analyzes. If, on the other hand, the master data is incorrect, this can impair and delay the processes in the company (for example when issuing invoices), cause costs and even result in a company's reputation loss.
In practice, master data maintenance arises from the need to use the data stored in the master data record several times or over the long term. The aim is to prevent master data (e.g. the bank details or address of a customer) from having to be re-entered for each business transaction . Rather, the data is entered into a customer master record once. This master record can then be used as an internal data reference source for all subsequent orders from this customer with the help of his customer number.
The master data maintenance thus prevents data redundancy and effort in data acquisition. The data sets essentially serve to uniquely identify the unit concerned. Using an identification term, master records, for example in an ERP system, can be called up, displayed, checked, changed or transferred to a further process as a data record at any time by authorized persons and other system functions .
According to a study by BearingPoint, around 50 percent of companies in Germany in 2016 had a concept for integrating master data into the company's digital structure. Such a concept was being implemented in 38 percent. Around a third of the companies pursued a comprehensive master data management approach and an integration of customers and suppliers.
Further examples of master and transaction data
- In an application for materials management , master data are e.g. B. the article or part master data, movement data are z. B. Inventory receipt and issue data.
- In a PPS system , parts lists and work plans are part of the master data. Movement data is data on purchase orders, orders and deliveries.
- In the case of a product configurator , the product features form the master data; the individual configuration that a customer makes when he puts together a product according to his wishes is part of the movement data.
- In a library system, the catalog and user data (e.g. name and address) are master data, while the data relating to the borrowed books of a user and the notification of a return are movement data.
History of terms
The division of the data world into master and movement data happened in the early days of electronic data processing . When there were no computer workstations for the direct input of data, the master data was available as a sorted pile of punched cards , as magnetic tape - or on magnetic disk files , which were still relatively expensive at the time . Changes were collected or stacked in movement files over a certain period of time in order to be processed against the master data in one pass.
With the beginning of dialog processing and the keeping of almost all data on direct access memories and the use of database management systems , the focus shifted in the 1980s away from change management to semantics - the existential (in) dependence of entity types or their characteristics in the tables of relational database management systems . The importance of original data, i.e. data from existentially independent entity types, is also shown in the fact that an instance for the management of master data has been set up for this data in some companies .
In various laws ( Information and Communication Services Act IuKDG, Teleservice Data Protection Act TDDSG) master data deviating from the derivation of stocks also referred to as inventory data, if they are necessary for the establishment, implementation or modification of a contract for the transport of information. Examples of inventory data there are customer data (name, address, booking account, etc.) and associated permanent system data (telephone number, connection ID, passwords, etc.). In this context, the term inventory data is used as a synonym for master data, whereby the term master data would be appropriate to avoid confusion and misunderstandings.
- Elmar J. Sinz: Conceptual data modeling in the structured entity relationship model (SER model). In: Günter Müller-Ettrich (Ed.): Effective data design: practical experience. Cologne 1989, ISBN 3-481-00003-0 .
- Erik Scheithauer: University of Applied Sciences Frankfurt / Main (PDF; 1.2 MB).
- Hans Robert Hansen , Gustaf Neumann : Business Informatics. Volume 1, Stuttgart 2005, ISBN 3-8252-2669-7 .
- Knut Hildebrand, Boris Otto, Anette Weisbecker (eds.): Master data management. HMD 279. dpunkt.verlag, Heidelberg 2011, ISBN 978-3-89864-750-2 .
- Accounting with SAP R 3 , by Paul Wenzel
- Knapp Consult love your data touchconsult.de data integration by Knapp: Consult
- See section grammar in: master data . In: Duden online dictionary . Bibliographisches Institut GmbH - Dudenverlag.
- Customer: Create customer master record (sold-to party). ( Memento from March 4, 2016 in the Internet Archive ) help.sap.com.
- ENT tumor documentation for all clinics, installation and use. (PDF; 427 kB) Stefan Langenberg, accessed on May 2, 2020 .
- o. V .: Why does a company have to maintain its master data? Retrieved August 20, 2020 .
- o. V .: Maturity and relevance of master data management. In: https://www.bearingpoint.com . 2016, accessed August 20, 2020 .