Data management boosts decision-making

Regardless of industry or company size, every organisation relies on its data to produce information for business decision-making. The quality and value of these decisions is dependent on the quality of the data on which they are based.

With the flood of information available to business users today, it is common for people to attempt new and different types of information analysis so as to gain competitive advantage. But poor data management can hamper these efforts, often putting an end to them before they are even out of the starting blocks.

Few companies give data quality the attention it deserves, and even fewer are prepared to admit that their business decisions are based on inaccurate or incomplete data, often choosing to avoid the question of data quality altogether - at the expense of the bottom line.

However, data is a key strategic asset, and ensuring its quality should be an imperative for any business today.

Effective data quality management - like almost every other aspect of IT - relies on a combination of people, process and technology.

Charl Barnard, GM, Knowledge Integration Dynamics

Let`s begin by defining data quality. Simply put, it is the degree of excellence of data, implying that the data is stored according to data types, it is consistent, it is not redundant, it follows business rules, it corresponds to established domains, it is timeous, it is well understood, it satisfies the needs of the business, the user is satisfied with the validity of the data and the information derived from it, the data is complete, and there are no duplicate records.

Data quality excellence means that actions can be taken based on reliable and accurate data, and that users can be certain there are no duplicate or incorrect names, addresses, or sales figures, for example, on the system.

Data is drawn from a broad range of sources, including legacy systems, databases and the Internet, often compromising quality and making integration difficult. Before companies fork out tons of cash on expensive data collection initiatives, they need to focus on how they manage their data.

Customer data is often created and locked in multiple silos including applications, data warehouses, data marts and external sources. Significant data quality issues and conflicting semantics exist within and across data sources. While there are transactional systems of record (such as billing), companies may have no single, reliable system of record for representing customers` profiles in the organisation.

Customer profile data is thus distributed across the various IT systems in an organisation, with much duplication and a high degree of inconsistency and inaccuracy. Lack of standardisation in recording customer data, incorrect data and data that is out of date, all result in bad data - data that prohibits a company from reaching optimal performance.

Companies across all industries are recognising the need for solutions that will improve the quality of their data.

Effective data quality management - like almost every other aspect of IT - relies on a combination of people, process and technology. Only by integrating these three elements - the people who create, use and own the information; the processes that transform information; and the technology that enables information - can organisations achieve the level of data integrity and accountability required by companies today, enabling them to lower their risk, and use data as a competitive differentiator.

* The people factor - data changes and corrections have generally been left to the IT department, without taking into account the needs of business users in terms of accurate, complete and timeous information. Information users themselves need to define their particular data needs if a company`s information is to be used to strategic advantage and competitive differentiation.

* The process and technology factors - data quality depends largely on how data is collected, processed, stored and used. The key components of data quality management include: data profiling (the discovery phase in which data problems are located and understood); data quality (dependent on standardisation and validation); data integration (dependent on data linking and consolidation); and data expansion (tracking down missing data and correcting data errors).

Data quality management relies on the implementation of solutions that integrate data quality management technologies with appropriate processes. Real information quality improvement applies process improvement. Determine the root cause of the problem when you find defective data; having done that, you can define processes to prevent the recurrence of those defects, and serve the needs of the people who actually use the data. In this way, the organisation will have the ability to make better business decisions and to achieve competitive advantage.