3 Data Quality Capabilities That Supercharge Data Governance
With growing regulations and greater public concern over data privacy, companies are increasingly focused on data governance. Yet there’s no single way for a business to approach data governance that will ensure perfect results. Instead, companies can adopt a number of different strategies, all of which can be made more effective when bolstered by key data quality capabilities and processes.
What capabilities are generally implemented to improve data governance?
To improve data governance, companies need to implement new capabilities across the organization. But even more so, cultural change is typically required including people and process. Data governance programs must facilitate communication and discussion across an organization about data. Efforts to establish data governance programs often start with hiring a Chief Data Officer or creating a Data Governance Council to steer the initiative, but it is more than just top down. There must be an emphasis on changing the workplace culture and generating buy-in so that everyone is focused on doing whatever they can to identify what data is important to ensure compliance, protect the data, validate quality and meet corporate goals. This requires a strategic approach and being able to answer questions like which lines of business are going to be particularly valuable and important, and what data is needed to support these lines. Business processes may also need to be adjusted to ensure that important data is captured and governed.
There’s an overarching effort needed to map data in terms of which business processes use it and for what purposes. Some organizations approach this process using a lean or agile data governance approach where the mapping is completed incrementally over time.
With this mapping in hand, which offers a high-level understanding, data quality tools and capabilities play a key role in supporting data governance programs.
A foundation of strong data governance is built by applying data quality capabilities like data profiling, data deduplication and the application of business rules.
Your company must be data-driven to succeed. That means putting more data in the hands of more business users. But – how do you do this responsibly? Learn how you can accelerate business innovation with trusted data, while also supporting and demonstrating compliance with GDPR, CCPA, Australian Privacy Act, and other regulations.
Data profiling offers a detailed view into the content of each of the data elements in a particular data source, offering statistical summaries and samples of field contents. Rather than pulling data or running queries, a data profiling tool provides a way to quickly yet deeply assess data quality at a detailed level based on the actual data content. Do numeric fields have letters in them or do name fields have email addresses? Are all of the account IDs unique? Is personally identifiable information finding its way into other fields that may not be properly masked, creating risk?
Data profiling can be applied to any type of data, whether master or operational, from a metadata management system, a governance system, or an IT operational system to quickly get a sense of the actual data versus what is expected.
Deduplication is critical to data governance
A second data quality capability critical to data governance is the ability to ensure consistent presentation of key entities, whether accounts, contacts, parts, products, vendors or even locations. Data deduplication and entity resolution are techniques where companies leverage the insights from data profiling to match, link and resolve these entities. Duplicated data whether within or across data sources can be removed or referenced as alternate or subsidiary sources. Being able to cleanse, match, resolve and consolidate data in this way is crucial so that aggregated content and analytics deliver a high degree of accuracy and trust. As these data quality techniques can be applied to any type of data, whether leads in a CRM system, parts data, product data or even metadata, they become a powerful tool in the data governance process.
Finally, companies can use data quality tools that apply business rules to assess, baseline and target the data quality levels they need. While data profiling reveals at a glance that some fields have the wrong type of data, business rules – whether simple or complex – can be applied to critical data to ensure that it meets the data quality dimensions that are important to the organization. The rules might state that particular fields can’t be blank or check relationships between fields (for example, a customer designated as “active” should have detailed contact information). Business rules can cross data sources to help ensure that events follow an expected sequence. While these types of tools are very important for compliance or anomaly detection, they can also be used to provide ongoing evaluation of key data elements that you’re tracking in your data governance program because of their high business value.
With a data governance strategy that prioritizes data quality, companies can find ways to maximize the value of their data and monetize it. It can be the foundation of better analytics and more informed decision making, rather than just a way to ensure compliance.
But leveraging this data is also important as a method to improve understanding of and trust in machine learning and AI. These emerging capabilities are dependent upon quality data. If companies do not have transparency into the data that their AI tools are using, there will naturally be a tendency for many both within and outside the organization to doubt the results. The only way AI and machine learning can provide value to the business is if people have confidence in the analyses they are generating.
Those are just some of the reasons that data quality matters so much to proper data governance.
Watch our webcast Overcome Compliance Challenges in the Age of Data Democratization to learn more.