Busting Four Data Quality Myths

Don't Fall for these Data Quality Fairy Tales

Robert CherneskyMay 15, 2019

Download E-Book

Let’s break down some common data quality myths. First off, imagine going to your next meeting and asking a simple question, “Do we believe our data should be high quality?” This is like asking, “do we believe in eating healthy?” Of course the answer is yes. We should have high quality data, but why isn’t belief turning into actionable behaviors to ensure we have accurate and trustworthy data?

The reason is rather simple. We often don’t take action until we have a fear. Either we are going to lose something or we’re unable to achieve something. Let’s dive into an example for each.

  • Fear of losing something: Take for example regulatory compliance, when we have a fear of being non-compliant and regulators place a magnifying glass on our data quality problems, our organizational behaviors will jump into action to fix it.
  • Fear of not being able to achieve something: An analytics project is a perfect example. To be effective, the data must be of high enough quality so the outcome of the analysis is trustworthy. Yet too often we fix the symptom by spending exorbitant amounts of time cleaning up the data, rather than fixing the root causes of the data quality problems.

Also problematic are organizational traps of false beliefs, or what we refer to as data quality myths that stymie the organization with either a false sense of security or denial that results in a quick fix to band-aid the data quality issues.

Let’s debunk the top four data quality myths so that your organization doesn’t repeat the bad habits of other organizations that failed to take action before a crisis occurred.

  1. Siloed Approaches: Data quality and governance require individual strategies
  2. Too Expensive: Establishing data quality is a time consuming, manual process
  3. Data Trust: Data is either good or bad quality
  4. Breaking the Silos: Businesses need a separate data quality tool

Data Quality and Governance Require Separate Strategies

Data governance is about harmonizing people, processes and technology to establish trust in data and leverage it as an enterprise-wide business asset. Essential to achieving this goal is data quality. Data is always traveling through the data supply chain and along the way it is exposed to new processes, uses and transformations, which can significantly impact data integrity. By establishing a comprehensive data governance strategy, organizations can score and monitor data quality and enact data integrity controls throughout the data supply chain for reasonability checks, timeliness checks, balancing and reconciliation and statistical controls. As a result, businesses can prevent downstream data issues and monetize data assets at the same time.

There are obvious synergies between data governance and data quality efforts. Data quality is essential to establishing trust in data but so is data governance. Governance ensures an organization’s data is both accessible and understood among business users. But without considering the quality of data before leveraging, how are business users supposed to trust it? By combining both governance and quality efforts, business users can quickly and easily find, trust and extract true insights from their data.

In addition, in combination, data governance can help automate manual data quality tasks.

Establishing Quality is Manual and Time-Consuming

 Data governance ensures users know where their data resides, what it means, and how to use it. However, by incorporating advanced controls and integrated machine learning capabilities, businesses can automatically monitor and improve data quality to generate meaningful, real-time insights.

By adding machine learning and analytics capabilities to a data governance framework, an organization can continuously monitor data integrity to ensure it is error free. When business users know data is clean and correct, they trust the information pulled to help drive better business decisions. Analytics-enabled data governance automates essential data quality tasks that would typically take large teams of people to accomplish. By cutting back human input, businesses minimize human error.

Still, understanding the quality of data isn’t always clear cut.

Data is Either Good or Bad Quality

 Data quality is a moving target and categorizing it as simply good or bad won’t work. For example, if data is just inaccurate, it is useless.

What if data is completely accurate, but it is six years old? Business users would likely consider this data low-quality because it isn’t timely but may still be useful to different departments in other ways.

A more specific example comes from the financial services industry. Business users are working to examine online banking habits to make the right financial offerings through email. The data sets are missing half of the customers’ mailing addresses.

Is this poor data integrity?

Not for their purpose. The organization only required email addresses. However, if they want to conduct a mailing, the quality is inferior because information is missing and after six years, outdated.

To avoid confusion surrounding data integrity, organizations need a data governance strategy to improve enterprise data quality with an all-inclusive data intelligence platform. That way, they can assign quality scores and users can weigh in and rate specific data assets regarding their usability and quality for different purposes.

Businesses Need a Separate Data Quality Tool

Many businesses still have disparate tools for data governance, data quality and analytics. By selecting a data intelligence platform with all three capabilities combined into one comprehensive solution, businesses can integrate data integrity monitoring and improvement directly into their data governance framework to establish trusted data throughout the organization. The platform should conduct data quality checks for completeness, conformance and validity around the data supply chain.

To enhance data transformation as it flows between and across systems, and verify that the data remains reliable, the platform should institute automated quality checks that go beyond basic quality dimensions. In addition, analytics capabilities can provide machine learning algorithms for self-learning capabilities to constantly improve data integrity without manual efforts.

Data governance capabilities should also provide transparency into an organization’s data landscape. This includes the available data, its location, the data owner/steward and data lineage. Business users across the enterprise are empowered to quickly access glossary definitions, synonyms and business attributes for data, so they can easily define, track and manage all aspects of their data assets.

By abandoning common data quality myths, organizations create understandable, high integrity data so business users trust the decisions they make.

 Are you looking for additional information about data quality? Check out the e-book below.

Get Insights

For a deeper dive into this topic, visit our resource center. Here you will find a broad selection of content that represents the compiled wisdom, experience, and advice of our seasoned data experts and thought leaders.

Download E-Book