Data Products – Just another hype?

In today’s data-rich world, access alone isn’t enough. The real challenge lies in making data understandable, reusable, and valuable. At SEB, we explore how Data Mesh and Data Products can unlock smarter data sharing and drive long-term business impact.
“Knowledge is power,” a phrase first expressed by Sir Francis Bacon in 1597, remains relevant even today. However, the way in which knowledge translates into power has evolved significantly. Historically, access to information itself created a competitive advantage – for instance, an individual with access to a library had an advantage over someone without it. In today’s digital society, information is universally accessible through the internet and search engines. As a result, the comparative advantage of simply accessing information has diminished. The challenge now lies not in access, but in efficiently identifying and leveraging the right information – a challenge that solutions such as OpenAI’s ChatGPT are designed to address.
Large enterprises face a similar issue. While they hold vast amounts of data, extracting and aggregating effectively remains complex. Many organizations encountered this difficulty when working with big data and central data warehouses, which often evolved into “data swamps” instead of delivering the expected business value. A key reason for this was the inherent complexity of understanding business data. The absence of common standards resulted in diverse definitions and structures, making it impossible for centralized teams to interpret the data without deep insight into business processes and IT landscapes.
Data Mesh for smarter data sharing
Data Mesh has emerged as a modern approach to data sharing, specifically addressing the limitations of central data warehouses. The concept is built on a federated governance model, where data ownership stays within business domains but is subject to centrally defined policies and standards.
The Data Mesh framework is founded on four key principles:
- Domain-Oriented Ownership: Decentralising ownership of analytical data, placing accountability within the business domains that are closest to the data.
- Data as a Product: Ensuring that business domains treat their data as products, providing high-quality, reliable, and well-documented data to internal consumers.
- Federated Computational Governance: Establishing a governance model built on federated decision-making and shared accountability.
- Self-Service Data Platform: Enabling domain teams with self-service capabilities to manage the full lifecycle of their data products.
Data Products power smarter data sharing
We have seen that during the last couple of years, the concept of Data Products has been adopted by many companies as the new way of building data sharing solutions. Data Products represent the tangible implementation of Data Mesh principles, primarily designed to enable data sharing for analytics and reporting. The greatest impact, however, often comes from adopting the underlying principles rather than the technical implementation itself. Examples of organisational implications include:
- Data Maturity: To treat internal data assets as products, organisations must first strengthen their understanding of why and how data should be managed differently.
- Data Architecture: Current governance structures must be evaluated against the principle of domain-oriented ownership, which may require significant changes to existing models.
- Reusability and Interoperability: To maximize efficiency, data products should support multiple use cases and allow use cases to easily combine multiple products. This requires clear information models that illustrate the relationships between different data objects.
When implemented effectively, Data Products provide organisations with an efficient model for internal data sharing. By making business data understandable, reliable, and reusable, organisations can not only improve efficiency but also unlock new business opportunities. Furthermore, the emphasis on reusability reduces reliance on complex point-to-point integrations, simplifying the implementation of data management capabilities such as data quality controls.
Reflection
The question remains whether Data Mesh and Data Products represent a lasting paradigm shift or merely a temporary trend that will be replaced by new methodologies. It is my view that the foundational principles of Data Mesh are here to stay. The transition towards treating internal data sharing for analytics and reporting as a product, rather than as a purely technical integration, reflects a natural and sustainable evolution. This approach enables organisations to manage data as a strategic asset, enhancing efficiency, reducing costs, and strengthening compliance with evolving regulatory requirements.
At the same time, it is important to acknowledge that Data Mesh is not the sole framework influencing data sharing within organizations. Dependencies and intersections with other approaches –such as the use of APIs for operational data flows or established data management frameworks like DCAM –must be carefully considered. As a result, the future framework for organisational data sharing may need to extend beyond the boundaries of Data Mesh, integrating multiple methodologies in order to deliver long-term value.
For SEB, as we advance on our journey toward becoming a data-driven bank, these considerations are highly relevant and will be central to shaping our data strategy moving forward.