A central data team can collect data from every department, yet still struggle to keep it usable. When a business unit changes a data structure without warning, downstream pipelines break. A data product changes who is accountable: the department that knows the data publishes it for other teams or external partners and takes responsibility for meeting their needs. section 18.1.3
In a data mesh, treating data as a product goes hand in hand with domain ownership. The producing department manages its data’s life cycle and quality, maintains agreed-upon schema contracts, and determines who can access it. Data should also be described, documented, standardized, and interoperable so consumers can use it within governance rules. section 18.1.2 section 18.1.4
A sales schema change
Suppose sales adds fields to support new CRM features. Rather than leaving a central team to discover broken integrations, sales documents the schema change and communicates it to teams using its data. For a significant change, the central data team can help manage schema evolution and compatibility. section 18.1.3
The platform supports the product; it does not become its owner. A central data team provides self-service infrastructure and governance support, while departments publish and consume their own data products. Kafka can serve as the shared platform for near-real-time exchange, schema enforcement, and integration with access-control and compliance tools. Those capabilities help move data reliably, but the producing department remains responsible for its quality and its agreements with consumers. section 18.1.2 section 18.1.4
