The interrelationship between data governance and data quality


Data is the new buzzword. And data quality is on every conversation board of organizations that are struggling with huge amounts of data and don’t know how to manage it.

While organizations generate and have access to tons of data, both internally and externally, few have truly figured out how to harness the data and turn it into a business asset.

Striving to stay competitive, every organization is aware of the potential value of data and the magic that can be created through it. However, most organizations face challenges related to the quality of the data they hold. And these challenges pose barriers to creating the insights needed to drive impactful business results, forcing stakeholders to seek out quick-response solutions that can be implemented to rack up quick wins.

Eureka… let’s fix data quality

The era of quick wins and quick fixes has thus entered the Data Management arena. Most stakeholders spend less time on the problem because the focus is still on the solution. The need to get to the root of the problem, find the real challenges and prioritize them before jumping into the solution seems out of place in this “instant” world in which organizations operate. And the timelines of managing transformation projects like migrating to the cloud, building a data lake, managing big data initiatives only make the situation worse, leaving less time to strategize and push stakeholders towards a solution.

One size fits all – leading to chaos and failure

Addressing data quality without really understanding data quality is the chosen path. And that turns out to be a recipe for disaster. As organizations embark on the data quality journey (more like a project than a program), without really understanding the stages and nuances of data quality, and its interconnections with other fundamental blocks and management of data management, namely, data strategy, data governance, Metadata Management and Master Data Management, the solution is doomed.

Data quality is often seen as a simple process, where a little profiling, fixing or transforming data quality rules will do wonders to improve data quality. This is the first point of failure – the lack of time given to understand the end-to-end data quality process and not creating a strategy to solve the real problem.

Data quality management involves several steps, and it is an ongoing, ever-evolving endeavor that cannot be solved by a single project or technology-based intervention. And more importantly, data quality cannot be fixed or improved in a silo. Data quality needs data governance. Even though both use different frameworks and attack data and deal with data issues in their own way.

The synergistic relationship is very important to understand

Data quality and data governance are therefore synergistic and interdependent in their relationship. And there’s no better way to understate that than with an example. If you’re sending your child to choose a pizza, chances are you’re asking the child to go to a particular restaurant and choose a pizza you like, with the toppings and bread choices that are yours. taste. There is a process that the child will follow, which you have put in place through your experience of choosing a pizza that your family enjoys. If your child doesn’t go through the process of picking the pizza the way you like it (for example, choosing a regular crust instead of a thin crust and adding olives as a topping, which is not enjoyed by anyone in the family), then you don’t approve of the same and reprimand the child. This process and practice of getting the pizza right and not approving the deviation is what data governance defines and puts in place.

Now let’s look at data quality with the same example. In the pizza that your child bought, there are leftovers, which the family would like to eat the next day for breakfast. Therefore, the pizza goes in the refrigerator. Because it can only be used the next day if it is refrigerated. If the pizza is left in the cupboard, it will spoil overnight and will not be safe to eat. It’s the quality of the data. Data quality establishes the system for keeping data usable and fit for purpose. Data quality works on the dimensions (accuracy, consistency, completeness, timeliness, and others) and thresholds set by an organization (95% accuracy or 85% completeness) to ensure data is clean, usable, and actionable. as if by purpose.

Here is a perspective on how to approach this complex interdependency

Having seen the interdependence of data governance and data quality and having gained vast experience in various data management programs, it has been observed that most organizations are willing to invest and work to improve the quality of their data in order to adapt it to their objective and thus leverage data to create business results.

However, often to improve data quality, organizations go ahead with cutting-edge technology-based solutions instead of taking the longest route to solving the problem by creating a solid governance foundation.

While governance is recognized as an important pillar in the data stewardship journey, the pursuit of quick wins and even faster results from data stewardship programs ensures that the fundamental pillar is put on the back burner, reducing efficiency of most data management programs.

As highlighted in the example above, data governance defines the boundaries and standards for managing data. If this pillar of data management is neglected, chaos will reign. Each stakeholder will follow their own principles and standards to manage the data and the lack of a process will lead to users finding and modifying the data according to their knowledge and requirements. And thus, the data quality will be compromised and difficult to manage. Even using point solutions to fix data quality issues won’t work in the long run.

Therefore, it is important that organizations first put in place a best-fit governance layer, based on a robust, well-articulated and published data strategy. Definition of roles and responsibilities, as well as minimum standards, policies, processes and procedures, and a well-balanced metadata management approach and plan, as well as data quality thresholds and KPIs, with Informative dashboards and dashboards are critical success factors in data management. journey. Ensuring these fundamental pillars are in place will result in data that meets the required standards and is fit for purpose, enabling organizations to use data to create meaningful insights and achieve desired business outcomes.

So in this chicken and egg story, if the egg or even the chicken came first, it was data governance!