Poor data quality is more common than you think—and the consequences can be devastating.

Organizations today are drowning in data. It's coming from industry groups, your website, and your own employees. But is your data accurate? And if it's not, how much damage could it cause?

A surprising number of people make critical business decisions based on data that has never been checked for quality. Raw data might be as pristine as a mountain spring or as filthy as a sewer; most data sources lie somewhere in between these extremes. Even if your data isn't obviously dirty, it can still hurt your organization.

How Much Does Bad Data Cost?

Poor data quality is arguably one of the largest hidden costs of business. Numerous studies support this claim, including one by Artemis Ventures that showed poor data quality costing the U.S. economy a staggering $3.1 trillion each year!

If you're having trouble grasping the magnitude of this problem, consider some real-life examples of how simple errors can produce disastrous results.

Frightening Consequences

A mail-order company sent out tens of thousands of catalogs and waited for the phones to ring. After a couple of eerily quiet days, company executives checked with the business that mailed the catalogs on their behalf. That's when they discovered the catalogs had been sent to the wrong addresses. Because of a computer error, the catalogs were mailed to members of the mailing list who showed the lowest likelihood to place an order. The resulting revenue shortfall forced the company to halt operations for several months.

After someone hit the wrong key in a county computer system, a house valued at $121,900 skyrocketed to $400 million. The error likely occurred when a county employee accidentally gained access to a computer program without authorization. The inflated home value was used to calculate tax rates and led the county to budget for $8 million in property taxes that did not exist.

A drug manufacturer originally claimed its product was perfectly safe for anyone who used the drug for fewer than 18 months. While battling multiple lawsuits, the company uncovered a data error in a product safety study. In reality, the drug increased patients' heart attack risk after only four months of use—meaning many more people were exposed to its dangerous side effects.

Problems Caused by Bad Data Are Common

In extreme cases, bad data can have catastrophic consequences. Most people are fortunate enough not to experience these total disasters, but mini-disasters are quite common at almost all organizations and many are related to data-quality problems.

Unless you've taken steps to manage data quality, how confident can you be that the management and financial reports you produce are reliable? If you produce business intelligence reports directly from your operational data and base business decisions on these reports, you're putting your business at risk.

Fix Bad Data Once and for All

So how do you solve these data quality problems? You might be able to build additional checks and balances into your operational software—but if you're running package software or receiving data from elsewhere, you might not have this option.

In many cases, the only viable solution is to build a data warehouse. When done right, a data warehouse is the most cost-effective and successful approach to solving data quality problems.

A data warehouse enables you to address multiple issues:

• Make data more user-friendly and simplify reporting by renaming tables and columns logically and performing critical calculations so users don't have to.
• Validate data in the extract, transform, and load (ETL) process, ensuring data errors are not brought into the warehouse tables.
• Enhance data by providing metadata. Metadata is the "user manual" for the data warehouse, giving business users a complete understanding of the information it contains.
• Bring together data from disparate systems, providing the ability to see your organization as whole.
• Create a single version of the truth from which you can disseminate accurate, reliable information to all your business users.

Avoid Your Data Disaster

Yes, implementing a data warehouse will include some work and an up-front cost. If you haven't suffered a data disaster lately, it might be difficult to convince senior management that there's a problem. Show them the list of data disaster examples. Avoiding these problems—along with their attendant costs and bad publicity—is just good business.

To learn more about how SEQUEL Data Warehouse can solve your data quality problems, make life easier for your business users, and help you avoid a data disaster, contact us at This email address is being protected from spambots. You need JavaScript enabled to view it..

Alan Jordan has been involved in System 38/AS400/iSeries/IBM i development for his entire IT career and is a 21-year veteran with Coglin Mill. A native of Australia, he has been living in Rochester, Minnesota, since 1998, when the company introduced RODIN to the U.S. market.  Alan oversees product direction and development and also manages the customer support team. He has extensive knowledge and experience in data warehousing technologies on IBM i and is a regular speaker at COMMON and other events.