In my last column, I wrote about why the focus of data governance should not be on creating bureaucracy and rules, but instead on business enablement within context of use. To do this, I suggest looking at data governance not as enforcement of a discipline, but instead as a process of guiding a data expedition.
Let’s look at what a data expedition entails and how data governance will be the guide of this ongoing journey:
Observation: Expeditions do not begin without purpose. The data expedition begins with recognizing business needs or drivers. There must be a problem to be solved. Part of data governance is consistently monitoring the needs of the business and identifying ways in which data has a role in advancing key initiatives. Strong data governance programs have bi-directional communication programs to keep a pulse on data needs throughout the company. When considering whether to embark upon an expedition or not, it is necessary to consider the entire journey and all who will be involved. It does not make sense to take the entire company on a data expedition that has not been observed and noted as valuable. Exploration: Data exploration is a term widely used in the data world but often misused. The exploration phase of an expedition entails surveying the intended path or territory to note unique characteristics and attributes, identify challenges, and understand the requirements necessary for all to traverse the trails of the expedition successfully. Data exploration is very similar. It is exploring data environments—with the business drivers noted in the observation phase in mind—to find specific characteristics and attributes, identify challenges, and understand the requirements necessary to complete the data journey successfully. The documentation of these exploration notes (read: metadata galore!) is an essential element of guidance that data governance provides in the data expedition.
Discovery: Data discovery is also a commonly used term in the data world. It is the systematic process of searching for specific patterns in data. Data discovery is looking at the surveyed and documented data landscape with a question of “What could be possible?” The legwork of the data exploration phase provides the discovery teams with multiple tools and lenses to look at the data terrain from many different perspectives. The nuggets of insight that data scientists glean are the foundation for the hypotheses tested and questions asked in the next phase, investigation. The extent to which the data landscape is prepared for data discovery is part of the data governance guidance efforts. Data may be viewed in more of a“rustic” format for certified data scientists and more “pruned” for business users or citizen data scientists (Gartner).
Investigation: Data investigation is not a term that has been afforded the same popularity as data exploration and discovery. It is out there, just not as prevalent. Data investigation is the interrogation of identified datasets from the discovery phase to test specific hypotheses. Data investigators ask, “Is X possible?” or a variety of versions of “What if?” based on the insights exposed in data discovery. In the consideration of the larger data expedition, data investigation is the phase where the importance of the governing guidance really comes in to play as many more people begin to participate in the journey. Governance helps to deter- mine which “data areas” should be investigated by which parties. When implemented most effectively, a data governance program will push data decisions to the lowest level of autonomy possible. This allows interested parties to investigate the data in ways that are most effective in solving their business needs.
Navigation: This final phase of the expedition cannot be accomplished without observation, exploration, discovery, and investigation as they are necessary to generate the map and navigation tools for all to travel successfully throughout the well- planned data environment. Business users know where they want to go and what they want to do. The data expedition navigational tools help them get there and get the job done. The carefully planned guidance of governance will allow all those who traverse what was once uncharted data territory to travel with confidence to their intended destinations.
Providing structured guidance to desired destinations is much more palatable than data bureaucracy. The data expedition is by no means an easy trip—some may even argue it is never-ending. Remember—you are taking a group of people on a journey for a specific purpose. Your data governance team members should be the ones to blaze the trails.
“Do not follow where the path may lead. Go instead where there is no path and leave a trail.”
—Ralph Waldo Emerson