Leadership lessons for setting up a data governance framework from scratch
I believe every enterprise today should be working to become data-driven. But I also caution those same institutions not to become data-determined.
This is the paradox in our moment of growing mountains of raw data, ubiquitous digital technologies, and unfolding artificial intelligence.. Yes, automation, autonomous systems, and robotics are transforming our world. But judgment, reasoned decision-making, and even leadership intuition have never been more important.
The difference between data-driven and data-determined
An analogy I’ll make is from Competition in the Age of AI. This book suggests that today’s leader is not so much the commander of a ship, amplifying his or her sea-faring skills with radar, sonar, GPS, digital weather monitoring, and more. Today’s executives and managers are more akin to fleet commanders.
They blend the use of all these tools with traditional human management and guidance tactics. My imagined data-driven navy includes everything from sailboats to hovercrafts — a complex system that can’t simply run on autopilot.
“In contrast with the current wealth of data, analytics, and AI, we still appear to be suffering from a shortage of managerial wisdom,” write Iansiti and Lakhani.
Data and AI provide powerful insights, but it is the wisdom of leaders — their ability to interpret, contextualize, and act upon this information — that truly drives success. Data-driven leaders use data to inform their decisions, but they never allow data alone to dictate the course of the business.
Data leadership as the first building block for data governance
In the data-driven enterprise, knowledge graphs emerge as crucial tools for cultivating wisdom. While generative AI, LLMs, and ML dominate headlines, it’s the often-overlooked “semantic AI” of knowledge graphs that truly paves the way for effective data governance.
The knowledge graph is the digital tool that connects the disparate stores of valuable but isolated data, compounding precise knowledge while identifying information that is incomplete. It delivers higher-level reasoning and decision-making. In a word, wisdom.
This is not, of course, to suggest heavy-handed, top-down management of data assets. Quite the contrary. As I argued in an article here last week on Agile Data Governance, an iterative, bottom-up approach to building data understanding and competence among teams is essential. Leverage of our data catalog and knowledge graph tools, what I’ve called the respective “nervous system” and “brain” of the data-driven enterprise, is key. And, while bottom-up data governance is critical, it’s the top-down vision provided by strong data leadership that truly enables an organization to navigate the complex waters of reaching the data-driven shores.
The top of the organization cannot effectively function without a vision of the full picture. The fleet commander, to return to my analogy, needs to see the entire armada, to understand what is working and what is not, to enable its interaction, to direct and coordinate.
Key components of agile data leadership when starting from scratch
Agile data leadership is about:
A focus on meaning: A key function of knowledge graphs is to link information through structured frameworks known as ontologies. These define the relationships between the concepts emerging from data. Beyond the pattern recognition of most AI systems, knowledge graphs atop a data catalog deliver complex insights into the significance and implications of data assets. This will define relationships between data concepts from day one.
Knowledge mapping: These graphs allow data to be transformed into more intelligent inferences, to answer complex questions, and by integrating data from multiple sources, knowledge graphs enable enhanced understanding.
Contextual awareness: Combined with other tools like our AI Context Engine, semantic AI enhances the contextual awareness of business goals, challenges, and competitive threats from the very beginning.
Encourage data literacy: As you build your framework, ensure other leaders understand data principles, just as they understand accounting basics.
Agile governance means agile leadership: Create a governance structure that can handle both structured and unstructured data, making all data assets accessible from the start. Data then becomes a utility for various and diverse use cases, and it allows enterprises to handle vast, diverse datasets through a deeper layer of understanding.
If you’re a leader looking to enhance your organization’s data-driven capabilities, demo data.world’s data governance platform today.
