By Elliot King

Safeguarding data quality is a continual and ongoing process. Sadly, no matter how diligent companies are, data quality problems will be created. But why is that?

  The root causes of data quality problems are deeply intertwined with daily
business operations coupled with the simple truth that businesses are not
static. Things change. Time decay is perhaps the most obvious root cause of data
quality issues. People move; they die; they get divorced; they no longer have a
need for your specific product, and so on. Updating data records to reflect
those changes in a timely fashion is difficult since you may not even be aware
that the change has occurred. Over time, data that was once right becomes wrong.

Time decay is a serious ongoing problem. But more dramatic quality issues often
emerge when organizations grow, change the application infrastructure, or must
respond to external demands. Too frequently, for example, when a company
purchases another company, there is not enough time to fully merge the
information infrastructure. Instead, the IT organization looks for a workaround
or stopgap measure to insure the consolidation does not impede ongoing
operations. These workarounds bring with them both known and unknown risks to
data quality.

A workaround is often the solution of choice in a variety of other situations as
well. A company may choose to eliminate certain applications or face a new
regulatory requirement. Branch office or remote operation personnel may feel
that the central IT staff cannot respond to their needs fast enough. In response
to the pressure to “keep things online,” an organization may opt for a
workaround.

A third root cause of data quality issues is that the company simply has not
established a sufficient data quality program. Since every piece of data cannot
be verified, the data quality system itself may be flawed. Alternatively, data
entry screens may be poorly designed and generate too many errors. And too few
companies have standard data dictionaries.

As long as business is a fluid process, data quality issues will always arise.
The key is to recognize their source and have programs in place to minimize the
number of problems and reduce their impact on business.


Leave a Reply

Your email address will not be published. Required fields are marked *