1.3 Acceptance of Inaccurate Data

Databases have risen to the level of being one of the most, if not the most, important corporate asset, and yet corporations tolerate enormous inaccuracies in their databases. Their data quality is not managed as rigorously as are most other assets and activities. Few companies have a data quality assurance program, and many that do have such a program provide too little support to make it effective.

The fact of modern business is that the databases that drive them are of poor to miserable quality, and little is being done about it. Corporations are losing significant amounts of money and missing important opportunities all the time because they operate on information derived from inaccurate data. The cost of poor-quality data is estimated by some data quality experts as being from 15 to 25% of operating profit. In a recent survey of 599 companies conducted by PricewaterhouseCoopers, an estimate of poor data management is costing global businesses more than $1.4 billion per year in billing, accounting, and inventory snafus alone. Much of that cost is attributable to the accuracy component of data quality.

This situation is not restricted to businesses. Similar costs can be found in governmental or educational organizations as well. Poor data quality is sapping all organizations of money and opportunities. A fair characterization of the state of data quality awareness and responsiveness for the typical large organization is as follows:

They are aware of problems with data.
They consistently underestimate, by a large amount, the extent of the problem.
They have no idea of the cost to the corporation of the problem.
They have no idea of the potential value in fixing the problem.

If you can get commitment to a data quality assessment exercise, it almost always raises awareness levels very high. A typical response is "I had no idea the problem was that large." Assessment is the key to awareness, not reading books like this. Most people will believe that the other guy has a larger problem than they do and assume that this book is written for that other guy, not them. Everyone believes that the data quality problem they have is small and much less interesting to address than other initiatives. They are usually very wrong in their thinking. It takes data to change their minds.