THE PROBLEMS ENCOUNTERED WITHOUT DATA MANAGEMENT

At its simplest, data management is the management of data, information represented in a formalised manner suitable for communication, interpretation or processing. This far-reaching but simple statement implies that all aspects of the management of data, including the storage of data in a file or a database to support a limited set of business processes, are ‘data management’. However, data management as a formal term is normally associated with the provision of an enterprise-wide service. The definition that has been in use within the BCS Data Management Specialist Group for some time is:

From this definition it can be seen that data management is a far-reaching function. It is involved with the definition of data, to enable that data to be shared between information systems and become a corporate resource. It is also involved with the management of the data in active information systems to ensure that it is reliable (that is, of good quality) and that the relevant data is available to the users that need it.

Data management, and its ‘big cousins’ information management and information resource management, can be compared to other corporate business functions, for example, personnel and finance. The data management function looks after the data resource in the same way as the personnel department looks after the personnel in the organisation and the finance department looks after the organisation’s money. I have used the term ‘data management function’ here because the data management responsibility may not reside in a single organisational department. The management structures needed for data management are discussed in Chapter 11.

It is important to recognise that the majority of data does not ‘belong’ to the data management function, as indeed the majority of personnel within a company do not ‘belong’ to the personnel department. The management of a large proportion of the data is normally the responsibility of the different functional departments in the organisation. The data management function should, however, provide the essential quality control of the enterprise’s data and be recognised as an authoritative source of information about the organisation’s data.

Only very few organisations have managed to implement an effective corporate data management function. The remainder are counting the costs of ineffective or non-existent data management. These costs come from the development of information systems to meet narrow departmental or business function and process needs without recognising that each information system in an organisation should be a subsystem of a larger integrated enterprise-wide federation of information systems designed so that the appropriate information is provided to the appropriate user in the appropriate place at the appropriate time. At the heart of such a federation of information systems is the requirement to share data; for this to happen, there has to be common enterprise-wide definitions of data and a co-ordinated enterprise-wide control of the actual live data. In other words, there has to be data management as defined by the Data Management Specialist Group and this data management has to be properly resourced, supported and managed in order to be effective.

Without effective enterprise-wide data management in place within the organisation:

• The information systems within the enterprise cannot be interfaced. Because they have not been defined and developed to work together, the information systems cannot be ‘joined’ other than at the most rudimentary, technical level.

• Data is not shared between the information systems. Even if it is possible to technically connect the information systems, it is still usually impossible to share data between those systems because of the incompatible data definitions in use in the disparate information systems.

• Communication breaks down and information gets lost. Without true data sharing, it becomes impossible for departments to obtain the information they need to carry out their jobs effectively within the necessary time frame.

• Data is unnecessarily transcribed and rekeyed. If there is a need for information to be shared between incompatible information systems the only option that is often available is to have the information from one system rekeyed into the other system or systems. This is not only very resource-intensive but there is a danger of the data being incorrectly transcribed if the definitions are ambiguous or the concepts underlying those data definitions are incompatible.

• The wheel keeps being reinvented. A substantial portion of the time and cost of the development of any information system is taken up with the analysis of the information and data requirements and the subsequent development of the database to meet those requirements. Without effective data management, this effort is required for each new information system irrespective of whether the same or similar information or data requirements have been analysed before. With enterprise-wide data management in place, the new information system can reuse existing data definitions, producing savings in both time and cost of the development of the new system. Implementing data management specifically to save system-development costs is unlikely to lead to an increase in information sharing across the enterprise, whereas if data management is implemented to improve information sharing there is almost certainly going to be a reduction in system-development costs in the long term.

• The competitive edge of the organisation is reduced. If information is delayed or lost, for example, because the mechanisms for the sharing of data between information systems are inefficient or are not even provided, the ability of the organisation to provide an efficient service to its customers, and thus to compete with its rivals, will be impeded.

• Frustration sets in. Users of information systems who consistently find that they do not have the right information at the right time to carry out their job effectively soon become frustrated. The staff of the IT or IS department who are providing support for those users find that they are constantly being criticised and themselves become frustrated at their inability to provide an effective service. The IT or IS department staff involved in the development of new and replacement systems find themselves having to develop from scratch systems whose requirements overlap systems developed before either by themselves or their colleagues. They also become frustrated.

The requirement for information is changing all the time. Where data is poorly or ambiguously defined, it may be difficult to respond to requests for new information or information presented in a different manner. Real business opportunities may be lost as a result of the inability to respond quickly to changing requirements. There are many situations where information exists in systems as data but is not accessible to the users within the right time frame.

Such costs impact on the business, the users of the information, as well as on the IT or IS department, who provide the technical infrastructure to provide the information to the users.