Aside from data, “governance” is the most used term in data governance. Duh! It’s in the name, right? But not only is it the most used, it is also the most misused and misinterpreted. The general perception of the term “governance” generates a widespread misunderstanding of the purpose of data governance. With excitement and fear rapidly growing around the use of data in businesses across all industries, data governance is becoming a hotter-than-hot topic! But, as with many corporate buzzwords, the more viral the term and concepts of data governance become, the more inappropriately used and abused they are.
Data Governance
It is common to associate the term “governance” with bureaucracy and administration which often leads to the notion of red tape, obstacles, and delays. In data governance, policies and procedures are certainly necessary, but the purpose should not be to control or rule with heavy authority. Often, companies initiate data governance with intentions of heavy control, focusing on strong, top-down authority. This is typically heard in statements such as, “We are bringing on a chief data officer to establish data governance and take control of our data issues.” While capitalizing on titles and hierarchy sounds ideal at the outset, it perpetuates the idea of bureaucracy and makes business enablement much more difficult in the long run. Data governance is not about leaving the control of data decisions in the hands of an executive or even a council for that matter. In fact, when done right, it is quite the opposite.
The purpose of data governance is to establish a framework that pushes data decisions to the lowest level of autonomy possible. Data governance establishes policies and procedures to protect and secure the core attributes of enterprise data assets (including but not limited to accuracy, accessibility, completeness, consistency, value, and reusability) and monitors the implementation and appropriate execution of these policies and procedures within defined scopes throughout the business. Rather than an executive or the data governance council taking control of data issues, data governance enables the business to take control of its own data issues and enhances the viability of the business through the more strategic use of data.
Data Ownership
A common mistake is to equate “the taking control of one’s own data issues” with data ownership. Many classic data governance “how-to lists” include the creation of data owners to establish the entities with whom data decisions and responsibilities ultimately rest within an organization. But, as technology has evolved and the proliferation of data has exploded, so has the definition and concept of data ownership. The variance in definitions has led to misuse of the term/concept of data ownership in governance programs, creating cause to evaluate whether data owners are even appropriate moving forward. It is likely that the perception of the term is also the root of the problem here. Because ownership not only implies access to/control over something tangible but also the inherent right to assign these privileges to others, the traditional data governance practice of assigning data owners/ownership is now a very slippery slope.
This, however, does not negate the expectation of the business “to take ownership” of their own data issues. See the misunderstanding of the term? What we really mean is that we expect “responsibility” and “accountability” (other commonly misused terms we will discuss next) for business outcomes when using specific data elements. If ownership is still a concept you want to keep within your governance program, consider assigning business process owners or subject/domain owners instead. The intangible aspect of processes and subject areas will make it much less likely to confuse the concept of ownership.
Accountability and Responsibility
A fundamental element to any data governance program is a solid RACI chart. In fact, almost any project/program management course will introduce the concept of (R)esponsible, (A)ccountable, (C)onsulted, and (I)nformed. The challenge is that the terms “responsibility” and “accountability” are frequently used interchangeably, making the differences between the two completely muddled. Understanding the difference is key to successful data governance. And, there is a strong relationship to authority to consider as well.
Responsibility is having the obligation to complete an assigned task. Responsibility can be given to an individual or shared by a group of individuals. Responsibilities are typically not measured other than basic check-box style completion. Unlike responsibility, accountability cannot be shared or delegated. Accountability is measured as accountable individuals fully accept the consequences for decisions, outcomes, and actions.
Responsibilities do not require authority to complete as they are assigned from those with authority. Accountability, however, requires authority. Autonomy can only be achieved when accountability and authority are in balance. This is where many data governance programs fail. When there is an imbalance of accountability and authority in either direction (accountability without authority or authority without accountability), it is a sure path to failure. If we are to remain true to the purpose of data governance—to create a framework that pushes data decisions to the lowest level of autonomy possible—then we must ensure that the level of accountability we expect from individuals matches their level of authority.
Who would have ever thought that data governance would become a trendy, hotter than hot, topic or a corporate buzzword? Most of us remember the days when it was, at best, considered a necessary evil. Now, we just have to enable an understanding about what this flashy, new phrase, “data governance” is, and what it means to be accountable and responsible for data. Tall order—but we’ve come a long way. At least, now it is cool to say “data governance.” Just wait. Soon enough, they will be asking us for metadata and data dictionaries.