Trust issues

Any company worth its salt knows that if it cannot be sure of the accuracy of its data, it may as well toss its future to the wind. Without a reliable, ongoing means of ensuring precise data in its systems, a company's chances of success are completely without foundation.

Data quality is hardly a new concept - it has been a corporate issue for decades. Yet still there are companies that fall into the trap of not giving the quality of their data the priority it needs, putting in temporary solutions in the belief that short-term data analysis will suffice in the long run.

Data quality is hardly a new concept - it has been a corporate issue for decades.

Mervyn Mooi is director of Knowledge Integrated Dynamics.

Companies tend not to take data quality seriously enough.

Generally, there are three approaches to ensuring data accuracy. The first is a random one, fixing data as and when needed, when mistakes become visible or obvious changes are necessary. This reactive approach is not only unsustainable, but it involves duplication of effort every time data comes into the system, resulting in an overlapping of processes and a duplication of rules: wasted time, wasted money.

Inevitably when the company realises its mistake and understands the need for its own data quality assurance, it then invests in a data quality tool. It builds rules into the software and scrubs all data in the company's various systems. While these intentions are good, this solution is still reactionary and only delivers short-term success.

Having burnt its fingers twice, the company only then opts for the only truly workable solution: the need to challenge data quality issues at the highest enterprise level requires the implementation of a data governance function.

Such a solution provides one uniform approach to addressing data quality, so all systems in the entire organisation conform to a strict data standard - a set of policies, principles, procedures and coding standards for all data. This solution is not only sustainable but workable for the long term.

There are four fundamental factors that need to be in place when a company, having learned its lessons the hard way, looks for a guaranteed solution after its short-term data quality efforts fail:

* Buy-in from top management is critical. Most management teams are not concerned with data standards - all they want is a good end result for the business. They may have heard about the importance of quality, but they know nothing of the analysis process involved in understanding data quality. A top-down approach to data quality is crucial - it should be a business standard that is implemented at the highest levels of management to drill down through the organisation.

* A culture of data quality needs to be infused throughout the whole organisation, evangelising the value of accurate data to all levels.
* Data quality needs to be recognised as an asset that is no less important than all other assets within the organisation.
* A data governance or data management function needs to be adopted by the company. This needs to be instituted just like any other IT or business function, and given the focus and budget it requires, coupled with management buy-in, ownership and accountability.

Any attempts to achieve data quality without first securing these four factors will be futile. To be sure of success, data quality needs to be seen in the wider context of the entire organisation, so a culture of data governance - embodying data quality, data management, business process management and risk management - surrounds the handling of all data in the organisation.