A data product is more than a dataset. This template captures the information a consumer needs to discover, understand, trust, and use a product, and the commitments the producer makes. Complete it during design and publish it with the product.
| Section | Questions to answer |
|---|---|
| Identity | Product name, domain, version, and a one-sentence description of the business purpose |
| Ownership | Business owner, technical owner, data steward, and support route |
| Consumers | Intended consumers and use cases; known unsuitable uses |
| Business meaning | Key concepts and definitions, with links to the enterprise glossary; important business rules |
| Contents and schema | Entities, fields, identifiers, and relationships; reference data used |
| Sources and lineage | Authoritative sources, transformations, and upstream dependencies |
| Interfaces | Access mechanisms (tables, APIs, events, retrieval services) and usage examples |
| Quality expectations | Material quality dimensions, measurable thresholds, and how status is published |
| Service objectives | Freshness, availability, and support expectations; dependencies and exclusions |
| Security and privacy | Classification, access policy, masking, permitted purposes, and retention |
| Contract and change policy | Compatibility rules, versioning, deprecation notice, and consumer notification |
| AI readiness | Suitability for retrieval, training, or agent use; provenance and permission handling |
| Observability | Signals monitored, alert routing, and incident procedure |
| Lifecycle | Creation date, review cadence, and retirement criteria |
Quality test. If a new consumer cannot answer What is this? Who owns it? Can I use it? How fresh and reliable is it? How do I access it? What does each important field mean? What happens if it changes? from the published template, the product is not yet mature.