What Is Data Mesh Governance?

Data mesh governance is a data governance-centric approach for implementing a data mesh architecture. As illustrated in Figure 1, it helps establish a balance between the decentralization of data ownership proposed by the data mesh approach and centralized enterprise data governance that advocates establishing consistency and standardization for the use of data across the organization.

With data mesh governance organizations can enable domain teams to autonomously curate, enrich and publish widely discoverable data products while providing central oversight on the use of data. This accelerates time to insights by reducing dependencies on central teams. Data mesh governance enables data users across the enterprise to quickly discover, understand and use data assets available in the data mesh by ensuring interoperability, security, documentation standards and policy enforcement for data products in a distributed setup.

Data mesh governance establishes clear roles and accountabilities amongst distributed teams, enforces the application of consistent data governance principles and sets up a federated data management model which accommodates local flexibility. By facilitating collaboration and knowledge exchange between domains, data mesh governance brings to life the spirit of adaptability and continuous improvement needed to support evolving business needs.

data mesh governance

Figure 1: The key pillars of a data mesh governance approach.

Why Does Data Mesh Governance Matter?

The inflexibility and fragility of legacy data management approaches are proving inadequate in helping digital organizations efficiently scale. As data proliferates across the organization, it becomes increasingly difficult to maintain its consistency and trustworthiness, especially when data needs to serve unique and often divergent use cases. Lack of autonomy to manage data in ways tailored for specific needs creates bottlenecks on central teams and lengthy delays in accessing and utilizing data for timely business insights. Centralized data teams with limited resources end up being responsible for optimizing data that they may not necessarily be most familiar with for different domain teams, such as sales, marketing, finance, customer service and more. This results in a loss of business context and inaccurate data being utilized for decision-making. 

How Does Data Mesh Governance Help Businesses?

Let's explore the data mesh approach, which is rooted in four key principles: domain-oriented decentralized ownership, viewing data as a product, establishing a self-serve data infrastructure and implementing federated data governance. 

By decentralizing the authority to manage data, data mesh urges the organization to share the responsibility of governing data. Not only does this offer a modular, scalable and agile model for domain teams to manage their own data but it also minimizes dependencies on centralized teams in finding, understanding and accessing data. 

This approach empowers teams to maximize the value of data by allowing business domains to treat data they understand best with processes and tools most suitable to their requirement.  requirement. By decentralizing data ownership across domains while being guided by centrally defined principles for data standards, interoperability and data quality, a unified understanding of data across the organization is created. This approach not only ensures reliability but also makes the data accessible and understandable to a broader range of users.

What Are the Key Benefits of Data Mesh Governance?

  • Reducing centralized bottlenecks and unleashing scale – Empowering distributed domain teams to autonomously curate, enrich and publish data as products to larger organizations can help alleviate the bottleneck often associated with centralized data governance. This approach allows the organization to manage large volumes of data more efficiently. Ensuring data products meet organizational standards for data quality and interoperability improves reliability and builds trust in data.

  • Bridging the business and IT (Information Technology) disconnect – Data mesh encourages domain teams that have a deep understanding of the specific business context to own and manage their data. As domain teams align data management processes to serve the unique needs of their domain, data mesh governance provides the overarching framework to align domain-specific processes with centrally defined guidelines. This factor reduces the traditional divide between IT and business domains expectations.

  • Boosting productivity by accelerating access to data – Providing governed access to the right information at the right time across all parts of the organization empowers data user to make data-driven decisions faster while ensuring data is used responsibly. Governing data at a domain level enables teams to optimize the data management processes for agility and productivity to meet dynamic business needs.

  • Improving data literacy and promoting a data-driven culture – Data mesh governance creates a shared accountability for data management. It encourages data owners to enhance their data expertise and take responsibility for data quality, governance and security. It also enables a structured approach for data stakeholders across distributed teams to collaborate, share insights and provide feedback to continuously improve the reliability and accuracy of data products. 

What Are the Challenges to Adopting a Data Mesh Governance Framework?

Data mesh does offer a promising solution for organizations aiming to be digital-first. However, its implementation requires businesses to adopt a holistic modern data governance approach capable of countering contemporary data management challenges.

Firstly, adopting a data mesh approach necessitates a cultural and mindset shift. Without well-defined roles, responsibilities and a clear understanding of the standards, best practices and guidelines for data management, organizations may encounter resistance to change. However, automation and AI-enabled solutions that support this shift can ease user adoption of new work methods and significantly shorten the learning curve.

Secondly, as distributed teams begin to curate their own data products, enterprises need to ensure that data quality, compliance and data protection principles are adequately applied. This approach helps data users determine if the data products created are reliable, trustworthy and interoperable across the larger organization. 

Thirdly, to facilitate data-driven decision-making organizations need to provide a governed platform for data exchange that makes data products available and accessible across the organization. Companies need a centralized data marketplace which can implement data access controls and security measures to prevent unauthorized access and risk exposure. 

For data mesh to scale, companies need to equip domain experts with easy-to-use, automated tools that replace manual processes capable of managing all parts of the data lifecycle. The lack of a single pane of glass to orchestrate data governance processes can limit scalability and make it difficult to realize the full potential of a data mesh.

What Are the Key Capabilities for Data Mesh Governance?

Below are the critical capabilities that ensure data mesh is supported by data that is discoverable, understandable, secure, trustworthy and valuable.

  1. Data Discovery and Classification - A fundamental principle of the data mesh approach is to enable data users to find the most relevant data assets across the organization, easily and without dependencies. However, as the number of data domains increases, the discoverability and visibility of data sets may decrease. Without data discovery and data cataloging capabilities, domain teams struggle to locate the most relevant and accurate data sets available within the organization. This struggle can lead to data silos, duplication of efforts and a lack of trust in data. 

  2. Data Lineage Tracking – In a decentralized data management setup, distributed teams bear the responsibility of managing, curating and publishing data assets. In such a context, data lineage serves as a common source of truth. It allows users to track the journey of data assets throughout their lifecycle, providing information on the origin of data, business context, procedures applied and usage terms. This information is critical for data producers as they bring together data assets from disparate sources to curate data products. Similarly, for data consumers, absence of data lineage information can cause skepticism and reduce confidence in the accuracy of data products created.

  3. Integrated View of Data Quality – In a data mesh architecture, data is managed by managed by distributed domain teams rather than centrally. This makes maintaining accuracy, completeness and reliability of data standards even more critical. Companies need to empower data producers with integrated data quality solutions. These should ensure data products adhere to centrally defined data quality rules and standards, while also accommodating domain-specific nuances that add value to the data. 

Domain teams require a solution capable of identifying anomalies, taking corrective action and monitoring the  performance of their data products as they become available to data consumers throughout the organization. Conversely, data consumers can make informed decisions on the reliability of data based on profiling details and data quality scores.

  1. Automated Data Governance – As further data domains get added to the data mesh, enforcing data governance policies and processes manually becomes unviable. To operationalize modern data governance in a federated setup, enterprises need a sophisticated technology-driven approach to manage and enforce data policies, data quality, data security and compliance across multiple autonomous domain teams. By automating data governance workflows and policy enforcement, companies can ensure the integrity of data is retained without overburdening domain teams with resource-intensive administrative tasks that are difficult to scale. 

  2. Governed Data Marketplace – Data mesh architecture is designed to make data more accessible to users across the organizations. This availability enables them to leverage it as a strategic asset for decision-making. As domain teams exercise their expertise in curating high-quality data products, it is essential for them to be empowered with a data exchange platform that accelerates and streamlines the process of data democratization. The absence of a governed data marketplace might result in poor discoverability of data products, continued dependencies on central teams and lack of oversight on the data quality, security and compliance standards of data products as they get published and consumed. The lack of this capability may eventually lead to loss of trust and slow adoption.

  3. Data Access Management – As data consumers across the enterprise grow and interact with data products through the data marketplace, managing the level of access for each user manually is not sustainable. This is where policy-based data access management plays a crucial role. Organizations can define access policies centrally based on user roles, context and sensitivity level. They can also execute them locally using data access management tools. Leveraging automation for data access governance allows data producers to control how their data products scale and establishes a foundation for sharing trusted data responsibly.

How Can Informatica Help with Data Mesh Governance?

Adopting a data governance-centric approach to data mesh provides organizations with a framework. This outlines data management roles and responsibilities, establishes data quality standards and enforces compliance and security for data sharing. It also allows flexibility and autonomy for domain teams to maximize the value of data.

Informatica Intelligent Data Management Cloud™ (IDMC) offers a comprehensive set of services to help establish adaptive, federated data governance. It provides a complete range of capabilities through a single solution for data discovery, data lineage, data profiling, data quality, data governance, data sharing and data access management. Providing a single, common platform between centralized IT teams and distributed domain teams allows for a unified yet flexible data governance model with tailored practices for each domain and centralized alignment.

Additional Resources

To learn more about Informatica’s approach to data mesh governance, listen to our demo video or visit our Data Mesh Architecture Center