Data products and data mesh: how they relate, and how they are different

Data mesh, a concept introduced by Zhamak Dehghani, offers a decentralized approach to managing data, moving away from traditional centralized architectures like data lakes and warehouses. It focuses on distributing data ownership and creating reusable data products tailored to specific needs, adhering to agreed standards and schedules. The data mesh framework is built on four pillars: domain ownership, data as a product, self-service data infrastructure, and federated computational governance. While not suitable for every organization, particularly those lacking sufficient resources or complexity, the principles of treating data as a product can be beneficial across various business contexts. Challenges in implementing data products include adapting to changes in roles and processes, acquiring necessary skills, and managing potential data chaos, which can be mitigated by starting small, utilizing a data product platform, and employing automated governance standards. Emphasizing data as a product ensures that stakeholder needs and value delivery remain priorities, with the potential for organizations to transition to a data mesh approach as they evolve.