Data Governance — The DMBOK Definition

Data Governance — The DMBOK Definition

The exercise of authority and control (planning, monitoring, enforcement) over the management of data assets.

Why it matters

Most teams call ‘data quality dashboards’ ‘governance’ and wonder why nothing improves. DMBOK is precise: governance is authority + decision rights, not dashboards. Without authority, every steward is suggestion-only.

Going deeper

The four operational components DMBOK insists on:

1. Strategy — a published, dated, signed-off charter. Without it the programme is a hobby.
2. Policy — short, enforceable rules. Policies that read like essays don't survive contact with reality.
3. Standards & procedures — the how. Naming conventions, lifecycle gates, exception workflows.
4. Issue management — a queue for governance issues, with SLAs, owners and visible burn-down. If governance has no inbox, it doesn't exist.

Authority for each of the above is granted by a Data Governance Council (executive sponsors + business + IT), enacted by Data Owners (per-domain accountability), executed by Data Stewards (daily quality + curation), and operated by Data Custodians (the platform / DBA layer).

Data governance is the rules of the road, not the traffic dashboard.

A traffic dashboard tells you that 80 cars/min are passing through intersection X — it's interesting; it changes no behaviour. The rules of the road (red means stop, right-of-way conventions, fines for violation) plus a police force to enforce them are what makes the system work.

Most ‘data governance’ programmes are giant dashboards with no police force: lovely visualisations of how often producers break the rules, with no consequence for breaking them. DMBOK's insistence on authority is the missing police force.