Data Mesh

Data Mesh: Why This Approach Makes Your Data More Productive

Data Mesh is an innovative concept that enables organizations to handle their data more productively. It takes a decentralized approach to data architecture and data management. The aim is to increase efficiency, flexibility, and scalability in data management in your company. The basic idea is that the team that produced the data, such as marketing, sales, or service, also understands it best. This makes data mesh something of an alternative to the conventional, centralized data warehouse approach.

Data Product: Your Domain Teams Help Themselves and Others

With Data Mesh, the focus is on your domains and the associated teams. The domain team creates so-called data products from its own operational data and analytical data models. Your domain team can use them to carry out analyses independently. These data products are also made available to other domains and teams based on a data contract, an internal agreement, to meet their data requirements and promote innovation. A data mesh architecture, therefore, gives the domains and teams in your company more autonomy and, at the same time, enables your various teams to perform cross-domain data analyses.

As a decentralized, domain-oriented approach, Data Mesh is an interesting alternative to traditional data environments. We would be happy to show you how you can use it to break down silos in your company, promote collaboration and make your data the basis for well-founded business decisions.

Sabrina Barko Sherif | Consultant Data & AI

Arvato Systems

Our Data Mesh Services at a Glance

Technology & tool consulting based on your existing system landscape & hyperscalers (Azure, AWS, SAP & GCP)

Advice on the implementation and success factors of a data mesh strategy in your company

Creation of a roadmap for the implementation of data mesh

Support in identifying, prioritizing, and managing domains and data products to increase efficiency

Support in establishing a federated data governance team

Data engineering services for setting up and maintaining the technical infrastructure & creating data products

Performance reviews to check the implementation to identify bottlenecks and make continuous improvements

The 4 Principles of the Data Mesh Strategy

1. data as a product

Data Mesh treats data as an important asset — as a data product with a clear value proposition and defined stakeholders. The focus is on providing high-quality, usable data, including the necessary analysis tools, in order to fulfill specific specialist business requirements in your company with these data products. Data as a product is a central principle of Data Mesh.

2 Domain-driven Data Decentralization

Data Mesh assigns responsibility for the data to specific, specialized business domains within your company. This domain-driven data decentralization ensures that those who are most familiar with the data from their department also take care of the context, quality, and use of this data. Promoting responsibilities and expertise within the respective domain is also one of their tasks.

3. Self-service in data matters

Data Mesh enables your users to access the data relevant to them independently and to edit and analyze this data using intuitive tools and interfaces. This allows users in a specialist domain to derive efficient insights from their data. Another important principle of Data Mesh is that each of your teams has access to a self-service data platform.

4. federated governance

Federated governance of your data mesh ecosystem allows for highly automated management of shared data, creating transparency and increasing user efficiency. The federated governance group in your company consists of representatives from your specialist domains and platforms and representatives from other relevant areas, such as compliance, legal, and security departments. This group defines your data mesh, platform guidelines, product development, and operating standards and monitors the functioning of your data mesh ecosystem.

What the 3 Main Data Mesh Components Do

The data product - productivity as a program

A data product is a logical unit that combines all components for processing and storing domain data, thus supporting data-intensive or analytical use cases. These data products can be made available to other teams. However, your respective domain teams are the sole owners of their data products. They are responsible for the operation of the data product throughout its entire life cycle. They also have the task of continuously monitoring data quality and costs and ensuring availability. If a data product is also to be made available to other domains in your organization, a data contract is defined for this purpose.

Data contracts - benefits for other departments

A data contract is a formal agreement between two parties on the use of a data product that one party supplies and the other uses. Such a contract defines guarantees for the data set to be provided and sets out the expectations on the part of the users of the data product users. It specifies the endpoint, syntax, semantics, and data quality to be supplied. Data contracts ensure stability, trust, and quality within the network of your company's various data domains; data contracts ensure stability, trust, and quality. They can also be used to visualize the data flow across data products. Your organization's various data contracts are usually defined and collected in a central data catalog, together with other metadata.

The self-service data platform - your technical basis

What a self-service data platform looks like can vary from organization to organization. However, your chosen solution must allow each domain team to have its own self-service area. In this way, the platform supports the domain team in creating its data product, using it for its intended purpose, and publishing it. The platform also allows cross-domain access to monitors and documents. A data catalog is a central entry point for the individual data products and makes them findable in your other domains. In addition, your self-service data platform is used to automatically enforce the guidelines set by your company's Federated Governance Group throughout the entire organization.

The 4 Crucial Roles in Everyday Data Mesh Life

1. domain team

2 Federated Governance Group

3. data platform team

4. enabling team

Domain team - the decentralized data specialists

Each of your domain teams is focused on a specific business domain. The domain team includes the specialists for their own data. They have a deep understanding of the data requirements and processes within their own domain or department. Therefore, they are responsible for the collection, storage, preparation, quality assurance, and provision of data, including for other teams. The teams in your various domains act autonomously from one another and make independent decisions in the context of their domain, from using certain technologies for their data product to scaling the data and measuring their performance. Even though all domain teams are autonomous, they work closely together. In particular, they need to help create appropriate APIs to exchange data between the different domains in your organization.

Federated Governance Group - guardians of the data mesh

The Federated Governance Group is an association of managers from the individual domain teams. This group brings together representatives of your domain teams and, where appropriate, other experts, moderators, and managers to establish guidelines and standards for data management throughout the organization. These policies are recorded in the Self-Service Data Platform and automatically enforced throughout your organization. Through its oversight, the Federated Governance Group ensures product and operational consistency and maintains interoperability across your data mesh ecosystem.

Data Platform Team - Infrastructure Specialists

The Data Platform Team is responsible for developing, maintaining, and optimizing the data infrastructure and platform that underpins your organization's data mesh and is used by all your domain teams. In addition, the Data Platform team establishes a central data catalog for the various data products within your company, optimally designs data integration through smooth processes, and distributes the necessary access rights regarding data security on the self-service data platform. It continuously monitors the platform's status, solves any problems that arise, and thus ensures availability and performance.

Enabling team - facilitator of the data mesh idea

The enabling team spreads the idea of the data mesh within your organization. Especially in the initial introduction phase, there is a need for explanation. The enabling team, therefore, consists of specialists with extensive knowledge of data analytics, data technology, and self-service data platforms. It supports your domain teams on their way to becoming fully-fledged members of the data mesh.

Our test: Is your company fit for data mesh?

We support you even before the planning and introduction of your data mesh ecosystem. Because the first question you need to answer is: Is your organization ready for data mesh? And would data mesh even be a recommendable data strategy in your situation? First of all, we should clarify what your ideal path to becoming a modern, data-driven company looks like.

We would be happy to support you on your way to a data mesh ecosystem - just get in touch!

Your Contacts for Data Mesh

Volker Greaves

Expert for Data and Automation

Sabrina Barko Sherif

Data and automation expert

Make data more productive with Data Mesh