The New Age of Data Governance

2020 has been a transformative year for many organizations. In order to ensure companies remain compliant while also focusing on evolving regulations, governance programs are more important than ever. On September 22, data and analytics leaders from across the central U.S. came together to discuss data governance frameworks and structures, digital initiatives, innovation in a turbulent time, and approaches to maintaining flexibility between disparate government styles that influence stakeholders in a data governance journey.

The State of Data Governance

One executive, perhaps controversially, stated that data governance as it is currently known, is dead. Up until now, governance has been focused on controlling access, and defining and cataloguing data. It was activity driven; but now, data governance is focused on innovation and digitally enabling the enterprise.

Data monetization isn’t about selling data; rather, it's about deriving value from the data.

Data monetization is what is starting to turn data governance on its head, but it isn’t about selling data; rather, it's about deriving value from the data. As organizations seek to acquire more external data and create optimal models around it, how do they ensure that data is being used for the right purpose? It is no longer about how many datasets there are or metadata coverage or glossary terms mastered. Those remain important, but the conversation is altered. 

Data governance was originally about generating revenue for the company (and not about selling the data), while monetization was traditionally for financial or credit card companies to better understand the spending habits of their customers. Now, those same organizations can share and sell to targeted retailers, which is an agreed upon, secure form of monetization because the data has been sold. 

But what are organizations to do with this new wave of monetization? Value is not an objective thing, so monetization is something that everybody understands. Combining new data sets with old data sets enriches and augments the data a company can collect from its own customers with business processes. Data professionals are now creating a model set, which in turn brings in more wallet share from existing customers and revenue from new customers. It is an emerging concept but one that drives growth. 

Quantifying the Value

In terms of buy-in, ROI and innovation, perspective plays an integral role. Policies need to be put in the place about how data is governed and stewarded on behalf of clients within an organization. It isn’t only about executive buy-in, but ultimately about doing what is right for your customers and focusing on providing clarity to consumers on how data sets are handled. 

For many organizations, data governance is still looked at as overhead. What will accelerate a data governance journey, however, is not the foundation, but rather the investments into transformation. Key business partners who have the initiatives or are the catalyst for those innovations need to be shown the correlation between an action taken and results that can be replicated across the business. CDAOs have the information, the data telling how things are or what to do to improve things, and partnering with the C-suite enables data and analytics leaders to follow the dollar, quantify value, and define business cases. Otherwise, data governance by itself will not stand.

Innovative Approaches

Innovation, on the other hand, does not always have a direct correlation to governance or even buy-in. It can oftentimes come about with simplified tools and user-friendly interfaces. One executive views quality management as a sub-discipline of governance. From that perspective, discovering where data comes from, the quality and profile of the data is increasing, and innovation is emerging from where that is being applied. Innovation is in the driving of a tailored approach to a given situation.

Explainable AI is also quickly becoming an innovative governance tool to be leveraged. While businesses can continue to create models, they also need to be able to explain how any given decision was reached. It is about data lineage and showing how data was acquired, utilized and applied. Model governance and the model lifecycle are becoming bigger, more important tools for transparency. 

Regulators are still not on board with AI, however, to the extent that they need data and analytics leaders to further the cause of data science and AI. While explainable AI is a concept, CDAOs aren’t always able to explain every bit of a statistical model that the machine is able to compute, so it remains a challenge. Every data user needs to learn about data science modeling because it will be a steep learning curve for an enterprise. It’s not about writing a Python program, it’s not making a model work — it’s about the science going on behind the data science.

Data governance is nothing new, but the journey to enabling data users to understand the sensitivity of data remains immature for most. One executive stated that they don’t believe any organization in the world could claim they are very mature in their data processes. It is an ever-evolving journey, and data governance best practices are not a “good thing” to have, but rather a “must” for organizations to remain competitive. 

Sharing information across the enterprise often outweighs the cost of data governance. If a business can combine 10 use cases and identify a single solution, then it is optimizing the available data. That’s the journey a number of enterprises find themselves on. It will take time and every industry is different, but data and analytics is in a constant state of evolution. If CDAOs are emboldened to be smarter with how data is produced and used, a data governance program shows its value. 

Share information, don’t share data.

by CDAOs, for CDAOs

Join the conversation with peers in your local CDAO community.