LESSON - Avoid the Pitfalls of Poor Data Quality
- By Bernie Gracy
- May 8, 2007
By Bernie Gracy, Vice President, Global Strategy, Product Management and Marketing, Pitney Bowes Group 1 Software
Without clean data, there is no CRM. Poor data quality can lead to serious business problems.
Several years ago, the introduction of new customer relationship management (CRM) systems from the world’s largest software houses was heralded as the panacea for the wayward enterprise. The formula for the success of these systems seemed simple: interact with customers, collect valuable data, compile 360-degree views of individual customers, and utilize the information to build rich relationships that improve over time.
After a while, however, many enterprises found they were not getting the value from these systems that they expected, and they have arrived at similar conclusions: CRM systems are really just front-end graphical user interfaces (GUIs) and canned business processes (best practices) that rely on data. A superior CRM system does not guarantee data quality and is unable to generate return on investment on its own. It is the quality of the data that is fed into the system that makes all the difference.
Defining Data Quality
Data quality applies to more than just customer name and address data. It applies to product numbers and associated descriptions, part numbers and units of measure, medical procedure codes and patient identification numbers, telephone numbers, e-mail addresses, commodity codes, vendor numbers, and vehicle identification numbers, to name just a few.
Ensuring data quality requires:
- Understanding the nature of the data and the degree of “trusted authority” from which it is derived
- Understanding the intended use of that data
- Identifying factors that both determine and impact the data’s fitness for use
- Establishing the policies, people, processes, and technologies to manage the quality of data
Given the current emphasis on the need to maintain a 360-degree view of the customer, data quality involves being able to link all of a given customer’s records together—a task that can only be accomplished with identifiers for the records associated with each customer.
An Absolute Necessity
Poor data quality can impact an organization’s ability to increase customer retention and loyalty, limit exposure, and increase operational efficiencies. For example, the inability to eliminate redundant name and address records can result in additional mail-order campaign costs, customer dissatisfaction—even legal concerns. Imprecise data on the total business conducted with a single vendor can result in missed opportunities for better rates with suppliers.
Recent regulatory and Homeland Security initiatives such as the U.S. Department of Treasury’s Office of Foreign Assets Control (OFAC), Sarbanes-Oxley, the U.S.A. Patriot Act, and the Health Insurance Portability and Accountability Act (HIPAA) can quickly spur a company to establish a solid data foundation.
These regulations will cause even lagging organizations to recognize that an effective data quality program is quickly becoming a near-absolute requirement.
Steps for Improving Data Quality
Improving data quality and using data more strategically can be achieved by:
- Conducting a data quality assessment to help recognize the severity of database quality issues.
- Adopting a well-defined data governance plan, including a definition of who owns the data, who is authorized to access the data, and which specific standards should apply to the data.
- Developing a corporate-wide agreement on data standards for master reference data that describes common business entities like products, customers, and suppliers.
- Choosing a technology to serve as the backbone for preparation of relevant customer data that includes name and address and non-name and address cleansing, change-of-address processing, tax jurisdiction assignments, personalized messaging, tables and dictionaries, batch and real-time processing, and more.
This article originally appeared in the issue of .