When we hear about data management these days, the terms “data quality” and “data governance” come up quite a bit—and they should because these are important functions for ensuring that organizations leverage their information in the best ways possible.
The problem is, these terms are often used interchangeably. Even worse, they’re used to suggest a hierarchy within the data operations that might not actually exist. For example, some see data quality as being a mere component of data governance, which fails to give quality the prominence it deserves.
For sure, both data quality and governance play vital roles in data-driven organizations. Their roles and responsibilities are quite different, however.
Data quality is about making sure that all data owned by an organization is complete, accurate, and ready for business users to analyze, share, turn into decision-making insights, etc. The quality of data has always been important. But the strategic value of data quality has risen dramatically as companies gather ever-growing volumes of data from more and more sources, and in various formats.
Organizations today collect data from multiple enterprise applications, Web sites, mobile devices, and social networks. The volume of data is likely to increase even more with the growth of the Internet of Things (IoT) and its countless connected objects all generating and sharing information. A recent study predicted that there will be 20.4 billion connected devices by 2020, making it imperative that organizations have strong data quality processes in place to ensure the tsunami of data is kept clean and up-to-date.
Data governance has to do with creating the framework and rules by which organizations will use data. As such, its purpose is quite different from that of data quality. Although data governance is still considered an IT-based function at some enterprises, its main role today is to make sure that the necessary data informs critical business functions.
An easy way to tell the difference between the two is to see whether either could be a standalone function. Data quality has one basic purpose: to collect and cleanse data and make sure that it’s complete, timely, and accurate. Without quality data to build a framework around, there’s no reason to have a data governance process. In essence, data governance becomes meaningless without a steady supply of quality data.
Data quality and governance are complementary functions that have fundamentally different responsibilities. There’s really no reason to confuse the two. To be sure, data governance is extremely important for organizations, especially as volumes of data and data sources continue to grow and information resources play an increasingly vital role in business operations and success. To ignore data governance is to welcome problems such as lack of control over data resources, misuse of data, security and privacy vulnerabilities, integration issues, failure to comply with regulations, etc.
In addition to understanding the differences between the two data-focused disciplines, it’s important for organizations to truly grasp just how important data quality is to their success.
Data quality should be a strategic priority for any enterprise. Much of the data that’s coming into organizations is incomplete or inaccurate, so it’s vital that companies have in place processes and tools to clean data before using it for analytics, particularly if it ends up being leveraged for customer engagements. There’s a huge downside to neglecting data quality. It can result in a substantial drop in revenue, poor customer service, loss of competitive edge, damage to the brand, and other negative impacts.
A solid data quality management strategy should include three main components: Gaining organizational alignment on business rules; putting the right technology in place to manage data quality; and committing the required time and staff resources to maintain data quality.
Each of these is highly important for companies looking to leverage their customer data as a strategic advantage. Organizational alignment is needed because high-quality data means different things to different parts of the organization – whether it be sales, human resources, product development, or some other area. The entire organization should agree on which business rules to use in determining data quality.
Having the right technology in place can make data quality management a lot easier and less costly. Surprisingly, a lot of organizations continue to manage data quality manually, using spreadsheets and data stewardship processes that require extensive manual input and are therefore labor intensive and costly. To effectively conduct data quality management at the level needed by many companies today requires automation, and that comes from the use of the right technology.
Even though many of the data quality tasks can be automated, organizations still must commit the time and resources needed to manage the process successfully. People are needed to define business rules, select and use the right software, and oversee the entire process.
Managing data quality is one of the most challenging processes IT organizations face today—and it’s not getting any easier given the growing volume and sources of data. Many enterprises are facing data quality problems characterized by data redundancy, incomplete or missing data, out-of-date data, lack of data standards, and the improper parsing of record fields from disparate systems.
These are significant challenges. Poor quality data can derail business initiatives as well as drain productivity throughout the organization. If data quality issues aren’t addressed, analysts, executives and other business users won’t be able to generate optimal value from data and the insights they gain from the data.
Fortunately, with the right processes, technologies, and resources in place, organizations can enhance their data quality efforts. Given the major investments they’ve made in data management, mining, and analytics tools, enhancing data quality is a sound business decision that should provide a solid return.