Data Governance has gained prominence over the last decade due to the reputational impact of data breaches. This has led to enterprises limiting the scope of Data Governance to only protecting the data managed by them. This approach severely hinders their Digital Transformation initiatives.
Digital initiatives create a log of the entire transaction thread and results in a manifold increase in data management efforts. Enterprises can realize business value by exploring and exploiting the data generated as part of the digital transactions in a compliant manner through Data & Analytical initiatives.
Modern approach to Data Governance:
Enterprises need to transform their Data Governance approach to enable them to benefit from their Digital Transformation exercise. Instead of hindering innovations through a “Command and Control” approach, it is imperative that Data Governance efforts aid value realization from data.
Enlarge the scope of Data Governance: Data Governance should also encompass Analytical governance. The scope should address the why, the what, the where, the when, the who and the how of securing data across the end-to-end Lifecyle.
Bias towards hyper automation: Enterprise data workloads have different velocity, variety, veracity. The quality of data plays a critical role in the analytical value of the data. Data Governance tools should enable automated data classification as per policies, continuous data quality monitoring and data lineage creation. Through dynamic application of policies across the data value chain, the right people should be provided with the right data with the right level of access. The evolving field of automated DataOps is increasing looking to be “Data Governance on Steroids.”
Expanding the boundaries of data governance: Most enterprises today are in an advanced stage of modernizing their data landscape. The adoption of Cloud Data Warehouses enables them to explore analytical use cases which involve the mixing and merging of internal, partner and external datasets. It is imperative that Data Governance is also applied on not just internal data, but also the external and partner datasets they mix and match.
Mindset shift: Cloud-based Modern Data Warehouse technology enables secure sharing of data sets with partner and external entities, often through Application Programming Interfaces (APIs). There is an inherent need to ensure that the sharing of data complies with regulatory requirements. Thus, Data Governance initiatives should also be expanded to enable the organization to securely share and monetize data. There will increasingly be an overlap of Data Governance activities with API Governance as well as Analytical (Artificial Intelligence) Governance.
Embedding Data Governance into all Data & Analytical initiatives: Often Data Governance is an afterthought and is driven as a technical initiative. Given the criticality of Data and Analytical initiatives, there is a need to embed Data Governance as part of all the business initiatives. We thus see regulated industry sectors like Banking and Financial Services looking at Data Governance as part of their Risk Management Function.
Data Governance Strategy
Any Greenfield Data Governance initiative needs to start with Data Inventory collection (knowing what data resides where and in what format). This needs to be followed by Sensitivity analysis (identifying sensitive data and classifying them). We then need to decide how sensitive data needs to be protected to ensure regulatory compliance and who needs to have access to what data and at what level. Policies need to be defined to enforce Data Governance before selecting the Data Governance Technology.
A brownfield Data Governance initiative would need to fill in any gaps and ensure that the scope expands to all the data consumed or generated by the company.
Implementing Data Governance
Today’s enterprises need to acquire the following capabilities to address the governance needs of data:
- Ability to understand how the data has been transformed across its lifecycle.
- Ability for the user to discover and search data.
- Ability to search data sets across the enterprise as per permissions.
- Ability to understand the attributes associated with the data.
- Ability to monitor the accuracy, completeness, consistency, duplication, validity, and freshness of data.
- Ability to monitor and manage internal, external, partner transactional data.
- Ability to monitor and manage master and reference data.
- Ability to provide the right data to the right user at the right level.
- Ability to enforce policies for data privacy protection and compliance.
- Ability to enable secure sharing of data with adequate guardrails.
- Ability to view reports on data access and consumption.
The above can be achieved by implementing Data Governance platforms supporting Data Quality Management, Data Profiling, Metadata Management, Data Lineage, Master and Reference Data Management, Data Cataloging and Data Dictionary. Increasing AI is being embedded in Data Governance platforms to address the “unknown” unknowns of data.
Operating Model for Data Governance:
Data Governance can be Centralized, decentralized, or federated across business units and functions. With the evolution of Modern Data Architecture to support Data Mesh architectures, a Centralized Data Governance team may not be enough for mature organizations and would need a federated governance model where the business governs the data, and the IT teams govern only the data platform and infrastructure.
Conclusion: A Capability-driven approach towards Data Governance:
Instead of Data Governance being reduced to just a selection of toolsets and technologies, we need to take a “Capability-based” approach for this i.e., addressing the people, process and technology needs of data governance in an integrated manner to achieve true democratization of data and aid the data-driven decision-making journey of enterprises.