dataarchitect.studio

Glossary

Data Mesh

Data mesh is an organisational approach to analytical data that decentralizes ownership: instead of one central team running one platform, domain teams (orders, payments, logistics) own and publish their data as products, with a self-serve platform underneath and federated governance holding standards together.

Its four principles — domain ownership, data as a product, self-serve platform, federated computational governance — are a response to central-team bottlenecks in large organisations. The honest track record: the sociology is the hard part. Mesh succeeds where domains genuinely staff and own their data products (usually formalised through data contracts), and fails where it’s adopted as a technology purchase. Small organisations rarely need it — a bottleneck you don’t have doesn’t need decentralising.