Operating model
Make the operating
model explicit.
Understand the organisation, assign decision rights and connect practices to evidence. The KDV reference product informs this model; using it does not require a KDV platform.
BIO-INSPIRED BY DESIGN
Different roles.
A shared ecosystem.
Hives and beekeeping give us a language for cooperation, local responsibility and care for the whole. It is an organisational metaphor: people, decision rights and data products remain explicit.
The metaphor helps explain the model. It does not replace an organisation’s terminology or assign human roles based on bee biology.
EcosystemShared purpose
The boundary of collaboration: objectives, participating domains, common policies and accountable governance. Start with a use case and a named sponsor.
Typical evidencePolicy, decision log, value indicator
HiveBusiness domain
A domain with an accountable owner, its own meaning and priorities. It coordinates its Cells while respecting the ecosystem’s minimum rules.
Typical evidenceDomain charter, owner, glossary
CellBusiness subdomain
A focused subdomain inside a Hive. Stewardship and engineering work close to the data. A Cell can own several data products; the Cell is not itself a product.
Typical evidenceSteward, quality rules, product inventory
CombBusiness grouping
An optional grouping of Cells within the same Hive. A Cell may belong to several Combs. This is a business and permission grouping, not an extra infrastructure layer.
Typical evidenceGrouping criteria, scoped responsibilities
Data productAn accountable offering
A purposeful offering with a named owner, versioned contract, known consumers, quality expectations and a lifecycle. Tables, APIs and dashboards can be delivery interfaces.
Typical evidenceContract, tests, access record, usage
FarmExecution environment
Where products are cultivated and operated: production, test or development environments. The Farm is an execution boundary, distinct from business ownership.
Typical evidenceInventory, access controls, recovery evidence
THE OPERATING MODEL
Five layers. One traceable thread.
Follow a decision from intent to evidence. Each layer gives the next one something concrete to work with.
Principles & policies
What must hold truePurpose, classification, retention and the minimum governance baseline.
Organisation & roles
Who decides and deliversExplicit owners and stewards, with responsibilities assigned at the right scope.
Processes & decisions
How work movesCultivate, govern, share and activate, with review and exception paths.
Controls & enablers
How rules take effectContracts, quality checks, access policies, catalogues and operational safeguards.
Evidence & improvement
How we know and learnDecision records, test results, usage, incidents and agreed outcome measures.
01 / Operating model
Roles describe accountability.
Interfaces support the work.
A person may hold several roles. Record the role exercised for each decision. Consumer, Producer, EcOps and PlatOps are product interaction perspectives, not substitutes for business accountability.
| Role | Scope | Responsibility | Reference interface |
|---|---|---|---|
| Governance forum | Ecosystem | Sets the baseline, reviews cross-domain exceptions and resolves conflicts. | EcOps records decisions |
| Ecosystem manager | Ecosystem | Coordinates implementation, scope, policy versions and improvement follow-up. | EcOps |
| Domain owner | Hive | Accountable for the domain and escalated approval decisions. | EcOps / Producer |
| Subdomain owner | Cell | Accountable for the subdomain and its approval decisions. | Producer |
| Data product owner | Product | Owns purpose, contract, versions, publication and consumer relationships. | Producer |
| Operational steward | Cell; optionally Hive | Maintains meaning, classification and quality; reviews requests. | Producer / EcOps |
| Coordinating steward | Hive | Coordinates stewardship and resolves issues across Cells. | EcOps |
| Data engineering | Cell; optionally Hive | Builds and maintains reproducible data preparation and delivery. | Producer |
| Technical custodian | Farm | Operates environments and implements authorised controls. | PlatOps / EcOps |
| Consumer | Product | Uses data within agreed terms and provides feedback on fitness for purpose. | Consumer |
When no subdomain owner is assigned, make the domain owner’s inherited responsibility explicit. A Comb may group permissions without creating a new role. Exact segregation of duties depends on risk and context.
Original KDV interpretation
Practices connected.
References respected.
An original KDV interpretation for operational use. These connections are illustrative, not an official equivalence, assessment or reproduction of the source frameworks.
| Reference | Practice lens | KDV operational expression | Evidence |
|---|---|---|---|
| DAMA-DMBOK 2 | Governance and accountability | Named owners and scoped decision rights. | Role register and decision record. |
| DAMA-DMBOK 2 | Metadata and quality | Shared definitions, product context and quality rules. | Glossary, lineage and test results. |
| DAMA-DMBOK 2 | Security and operations | Classification, authorised access and recovery practices. | Policy record and recovery check. |
| DCAM | Capability development | Agree priorities, assign responsibilities and review improvements. | Action backlog and accountable follow-up. |
| DCAM | Evidence of practice | Demonstrate execution and outcomes over time. | Decisions, controls and indicator trends. |
| Data Mesh | Domain ownership | Hives and Cells own their products and meaning. | Domain and product charters. |
| Data Mesh | Data as a product | Versioned contracts and consumer-oriented delivery. | Contract, feedback and usage. |
| Data Mesh | Self-serve infrastructure | Reusable delivery paths with clear operating boundaries. | Service interface and operational checks. |
| Data Mesh | Federated computational governance | Common rules, local decisions and automated checks where feasible. | Baseline, policy tests and exceptions. |
Original KDV interpretation
Go to the original sources.
DAMA-DMBOK 2
The data management knowledge reference informing the governance vocabulary.
DCAM
A capability framework informing evidence-led improvement. Consult its publisher for access and licensing.
Data Mesh
The four principles informing domain ownership, data products, self-service and federation.
KDV adds an original operating interpretation. DAMA-DMBOK and DCAM material is not redistributed. References do not imply endorsement, affiliation, certification or an official mapping.
Community
Help shape the next iteration.
Bring a question, a better example or a tested improvement. Small contributions count.