Data is the lifeblood of businesses, the glue that binds the processes. Neglecting to manage this asset is an obvious and gross mistake.
Businesses around the world aren't oblivious to this fact. Gartner forecasts that the data quality market, although not enormous in global terms, will grow to $677 million by 2011. That's big, but the growth figure is more impressive at a compound annual 17.6%.
More companies are coming around to the fact that data should be subjected to all the available rules and standards that make it trustworthy, from inception to final destination.
Data management goes beyond the straightforward process that has, until now, been the IT department's responsibility. It is an end-to-end process and is the responsibility of both IT departments, which facilitate the process to consume, transform, store and distribute data, and business people who own the data and stipulate the rules that govern it and used to manage it.
That's a shift from the traditional environment. Traditionally it was the CIO's problem, since IT departments manage systems that share and move data, such as the databases, data warehouses, other stores and ERP systems.
More companies are coming around to the fact that data should be subjected to all the available rules and standards that make it trustworthy, from inception to final destination
Mervyn Mooi is director at Knowledge Integration Dynamics.
Data quality cannot be addressed by IT departments alone. IT people don't understand how customer records should appear for business consumption and benefit. That's mainly the marketing department's responsibility. Marketers must tell IT employees how to standardise customer records and which fields are most important. IT people may also have to work closely with procurement employees, depending on the nature of the business. Whoever is at the rock-face will need to work closely with IT employees.
Data management is everyone's responsibility - even that of the data capture clerks who should report any errors or inconsistencies regardless of automated validation processes that may or may not be in place.
It's the data management function that makes or breaks data quality in any organisation as it is founded on tried and tested rules and standards in the particular industry.
For data management to succeed, it must be a component of the overall business strategy with agreed policies and procedures, because without an executive and management mandate governing data management, the risk of inaccurate and costly reporting and analyses remains high.
The right stuff
The correct structure and process would be for six groups of people to take responsibility for data management:
* CIOs and executives to drive data policies;
* Business units to lay down the data rules, structures and use;
* Data stewards to represent the data standards, usually across business units;
* IT to facilitate the rules and policies through automated processes;
* End-users and data originators to present, distribute, validate and verify the data; and
* Executives, analysts and other management to consume information for decision-making.
Based on that model, which represents a holistic approach to data management and hence with the best possible chance of success, data management is directly and indirectly the responsibility of all who use it and supply the data.
* Mervyn Mooi is director at Knowledge Integration Dynamics.
Share