DM Review would like to welcome Vicki Raeburn as our new data quality columnist. She has more than 30 years of data experience to share with business and IT executives alike. The point of arrival (POA) for any customer information solution (master data management, customer data integration [CDI] or data warehouse) is a complete view of all customers and how they interact with your company. The benefits of achieving this POA are many: cross-selling opportunities, tailored customer service options, appropriate sales force allocation, compliance with know-your-customer regulations and so on. At the POA, your company will maximize its opportunities for profitable revenue growth within regulatory boundaries. The key to creating a complete customer view is high quality customer data. Creating high quality customer data depends on making the right software and technology choices. It also requires implementing a robust data governance process that combines software components, enterprise-wide executive commitment, data quality processes and metrics, and behavioral change management processes. This column is the first in a series on creating high quality customer data by establishing the right data governance processes. I will start the series by taking a look at a common data quality conundrum: Why does everyone who originates data think their group is creating good quality data, but end users of data that has been combined across the company think nobody is doing anything about data quality? In short, how does good data go bad? If you have ever worked on a customer data project, you have probably heard data originators and data stewards around the company say, “But my data works fine for my applications.” And, it probably does. The business users of data within a function or a division generally strive to optimize the data that they need for their business activities. As you create a company-wide customer information solution, you face a different task. You are trying to round up all the data, rationalize it and maintain it for current and new applications. But, the source data was originally created at one time in the past for very specific purposes. The owners of the data, whether purchased or input, have managed its quality to meet the data needs of the applications that are crucial to their activities. They are defining data quality in the traditional way: to meet the requirements of the intended use. If they work in fulfillment and are focused on mailings to customers, they won’t be too careful about telephone numbers. If they are salespeople entering their own customers’ data, they will be casual about legal names and addresses. If all the U.S. billing department’s customers are primarily in the U.S., they won’t be too particular about filling in country codes. If the data definitions have a Humpty Dumpty-ish quality - “When I use a field name, it means just what I choose it to mean, neither less nor more” (Humpty didn’t really say “field name,” but he probably won’t mind the adaptation) - when you try to consolidate across the organization, they work locally because everyone in the originator’s part of the company uses them the same way. The problems you face are simple to describe. The current generation of CDI software is excellent for locating the problems, too. But, locating the data problems and even the source of the data problems will not guarantee higher quality data going forward. You need more than software. You need the data originators and data stewards across your company to change the way they collect and maintain their view of customer data to support the business needs of others outside their immediate department or function. Changing how data is collected or acquired to meet the needs of other departments or functions is almost certainly not in the originators’ performance goals today. And, you may need departments or functions across the company to help pay for the implementation of the new customer information system - something that is definitely not in their budgets today. The data and systems they have today meet the requirements of the intended use. The solution is enterprise-wide executive support combined with a data governance program that is focused on changing behavior as well as on the software solutions available for cleansing and maintaining data. Any change that affects the whole company or enterprise is significant. Change requires planning, communication and training. Your data governance program should include a clear statement of the benefits of the POA for the entire company and for the individual teams contributing data, resources and budget dollars. You will need a well-planned change management program focused on transitioning from how data is collected and maintained today to how it should be collected and maintained for the benefit of the entire company. This planning and training will be needed for all departments and functions that contribute data. Finally, you will need a regular communications program that emphasizes revenue growth success stories that are created by the customer information solution that you are pursuing. Data governance requires good software tools to locate and manage quality issues. Equally important, it depends on the commitment of enterprise-wide senior leadership and of all data originators and data stewards. Data governance is a never-ending process. The main work to establish enterprise-wide governance processes can take three to five years. The people element will be a key to your success.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access