Follow Datanami:
June 11, 2024

Unlocking the Full Potential of Data: The Crucial Role of Data Governance in Integrated Analysis

Bidish Sarkar

Organizations today are built on data. However, in the quest to become data-driven, businesses must maintain their trust, which becomes challenging with the huge volumes of data generated and collected every day. The quality of data is another critical factor to consider. Did you know that bad data quality impacts 31% of an organization’s revenue today, up from 26% from 2022 to 2023, according to the State of Data Quality Survey 2023?

With GenAI likely driving automation in decision-making or providing suggestions to end users, the impact of data quality (or lack thereof) is expected to become much more significant in the coming years. That brings us “Data Governance” and its essential role in ensuring data quality, reliability, and consistency. Data governance is a well-defined approach to managing the data in your organization from the point of acquisition throughout its entire life cycle (while being shared internally and externally), to when it is archived or permanently deleted. There is a growing acceptance within the data and analytics community in most enterprises about the importance of data governance. The numbers indicate this trend – with the data governance market growing at a whopping rate of nearly 21%, estimating a value of about $5.3 billion by 2026.

Data governance is crucial in helping businesses improve their integrated analysis capabilities by ensuring data quality, reliability, and consistency. It starts by fostering collaboration and alignment among stakeholders in an enterprise involved in integrated analysis, such as data analysts, data scientists, IT professionals, and business leaders. By establishing clear roles, responsibilities, and communication channels, data governance promotes cross-functional teamwork and ensures that integrated analysis efforts align with organizational goals and priorities. For a global major in retail banking, the impact of better data governance led to faster and easier availability of real-time and personalized offers tailored to customized requirements, resulting in 30% higher efficiency for the marketing team when launching new product offerings.

(LeoWolfert/Shutterstock)

How Data Governance Balances Data Accessibility and Security

Balancing data accessibility with security needs careful consideration. While allowing authorized users to access data for decision-making is vital, it’s equally important to protect it from unauthorized access and breaches. Achieving this balance poses challenges but implementing access controls helps ensure that sensitive data remains secure while still being accessible to those who need it.

Moreover, managing the increasing volume and complexity of data adds to the difficulty of maintaining data security. A robust data governance framework addresses this by setting clear rules for managing data access and usage. This involves classifying data based on its sensitivity and prioritizing security measures accordingly. Furthermore, data governance fosters transparency and accountability by keeping track of accessibility and usability through audit trails and logs. This enables organizations to monitor data usage, identify unauthorized activities, and take necessary action to mitigate risks.

As part of the data governance process, it is important to establish frameworks and guidelines within organizations to define policies and standards, access control protocols, data classification guidelines, and overall monitoring and enforcement policies. While the chief data and analytics officer (CDAO) often leads such an initiative, it is imperative to bring together the Chief Information Security Officer (CISO) and the office of risk and compliance management to help define the different policies and guidelines. Often, successful organizations have dedicated data stewards assigned by each business unit to manage and protect various data sets.

Addressing Challenges in Data Governance Implementation

The biggest challenge in implementing data governance is resistance to change and cultural barriers within organizations. This, coupled with data ownership and accountability issues, often creates significant challenges for organizations in implementing an effective data governance solution. The first step in a successful organization-wide implementation of data governance is to garner executive sponsorship. It invariably needs backing from the C-suite to ensure that different divisions and business units support the effort. A crucial second step is to socialize the benefits with the end user community and showcase the advantages of such an approach. In this regard, carrying out proof of concept with a specific business unit or user group and using that group as a business champion helps immensely.

(metamorworks/Shutterstock)

As part of the proof of concept, it is essential to decide on a specific platform or a set of tools to implement data governance within the organization. Two approaches are currently in vogue within enterprises for implementing a robust data governance solution.

  1. Identify an end-to-end commercially available platform and customize it according to your organization’s purpose. Multiple platforms are available in the market that can help effectively implement data governance. Some providers of data governance platforms are Collibra, Alation, Informatica, IBM, and Ataccama, among others.
  2. Leverage a combination of tools available within current the enterprise data architecture to drive effective data governance. For example, if you use Databricks, you can effectively leverage components within their Unified Data and Analytics platform to drive data governance. If you are leveraging any of the hyperscalers, then they provide a set of native solutions to help drive data governance as part of their analytics ecosystem:
    1. In the case of AWS, you can leverage a combination of AWS Lake Formation, AWS Glue Data Catalog, AWS Identity and Access Management (IAM), and Amazon Macie to achieve this.
    2. For Microsoft Azure, this can be achieved through a combination of Azure Purview, Azure Data Catalog, Azure Active Directory (AAD), and Azure Information Protection (AIP).
    3. In the case of Google Cloud Platform the tools that need to be leveraged are Cloud Data Catalog, Cloud Identity and Access Management (IAM), Cloud Data Loss Prevention (DLP), and Cloud Security Command Center (Cloud SCC).

Once you have selected a platform or a set of tools and carried out a successful proof of concept with such a toolset to establish that it meets your needs, you will need to plan the actual implementation. Planning a phased roll-out across the organization is paramount for iterating and improving. The success of the overall data governance initiative often depends on a strong change management plan that incorporates continuous training and adoption drives.

Last but not least, centralized ownership is critical when defining the boundaries and setting guidelines. A centralized ownership structure ensures that data governance policies and standards are established and followed consistently across the organization. Such a centralized approach avoids confusion, ensures alignment with organizational objectives, and maintains the integrity and security of data assets.

An Increasing Need for Federated Data Governance

(Cherdchai101/Shutterstock)

The growing complexity of data sources is further pushing businesses to turn to federated data governance. Take one of the leading health insurance companies, for instance, which uses federated data governance to improve the quality and efficiency of patient care.

This approach enables businesses to achieve a balance between centralized and decentralized models. Clear ownership at a centralized level establishes standards, while individual divisions oversee specific data sources. This encourages uniform data management across departments, as seen in modern architectures like a data mesh. Centralized management of privacy and security ensures compliance with regulations, irrespective of the data’s origin or utilization within the organization.

Impact of Leveraging Data Governance in Integrated Analysis

The specialty risk division of a leading provider of insurance solutions brought down their average time for generating a new analytical report from a week to less than three days by driving data governance across all their data assets. In addition, this resulted in all compliance requirements, such as GDPR and CCPA, being addressed upfront. The program also improved the organization’s data culture by helping business users easily find the data they can trust. It enhanced collaboration between the sales and marketing division with the risk division by enabling business users to connect with data owners and providing easily understandable data lineage.

Enhancing Data Governance in Integrated Analysis in the Age of AI

In the last couple of years, Generative AI (GenAI) has emerged as a transformative force in empowering both data governance frameworks and integrated analysis. Through its capabilities in augmenting datasets, detecting anomalies, preserving data privacy, and enabling predictive analytics, GenAI significantly enhances data governance practices and facilitates informed decision-making processes. Its ability to create synthetic data supplements existing datasets, while its anomaly detection algorithms contribute to maintaining data quality. Furthermore, GenAI techniques such as differential privacy ensure the protection of sensitive information and enable organizations to make proactive decisions based on data patterns through predictive analytics.

About the author: Bidish Sarkar is the Senior Vice President of Data & Analytics at Persistent Systems. With over two decades of experience in the IT industry, he has a proven track record in managing executive relationships and delivering high-value strategic solutions. Bidish has previously held key roles at Genpact, HCL Technologies, and Cognizant.

Related Items:

Making the Leap From Data Governance to AI Governance

Data Governance: Simple and Practical

The Rise and Fall of Data Governance (Again)

 

 

Datanami