Data governance, known in Italian as governance dei dati, is the set of roles, rules, processes and controls through which an organisation decides who is accountable for data, how it may be used, how reliable it must be and how it demonstrates that all this actually happens. It is not software, nor a project that comes to an end: it is a system of accountability that follows data throughout its entire life within the company.
The issue concerns those who lead the business as much as those who run the systems. A report showing different figures depending on the department, personal data kept for no reason, an artificial intelligence model fed with information whose origin nobody knows: these are governance problems before they are technology problems. If you would like to know more about governance, you can read this article.
Data governance in 4 key points
Assigns responsibilities
For each data domain, it is defined who decides, who oversees quality and who applies the controls.
Sets the rules
Shared definitions and criteria guide the collection, use, access and retention of data.
Provides for controls
Documented checks make it possible to identify anomalies and confirm that the rules are being followed.
Serves decision-making
Management and business functions can use data whose meaning, provenance and limitations they understand.
Data governance: what it is and how it translates into Italian
In Italian, data governance is rendered as “governo dei dati” or “governance dei dati”. In a company, it refers to the system that assigns decision rights over data and establishes how their use is verified.
It is a branch of corporate governance: it applies accountability criteria, shared rules and documented controls to a specific asset, data.
Data governance and data management: what is the difference?
Data management comprises the activities through which data are collected, stored, integrated, transformed and made available. Data governance sets the responsibilities, priorities and criteria that guide these activities and oversees their application. The distinction concerns the function performed: in practice, data governance and data management work together.
Data governance and data management
| Data governance | Data management | |
|---|---|---|
| Question it answers | Who decides, under which rules and with which controls? | How are data collected, stored and made available? |
| Nature | Direction, accountability and control | Management and operational implementation |
| Who leads | Sponsor, governance committee and data owners, supported by stewards and control functions | Process owners, IT and the teams that manage and use the data |
| Result | Policies, roles, standards, evidence of controls | Databases, integration flows, catalogues, reports |
| Horizon | Continuous, tied to objectives and risks | Ongoing, tied to the life cycle of data, systems and processes |
Without management activities, rules remain on paper. Without shared governance, individual processes may use different definitions, permissions and quality criteria, making it difficult to compare results and assign responsibility.
Benefits of data governance
The benefits of data governance can be seen in business processes: in the consistency of reports, in the handling of anomalies and in the ability to reconstruct how a piece of data has been used.
- More reliable decisions. Shared definitions allow management and business functions to compare the same indicators without attributing different meanings to them.
- Compliance support. Documenting data, responsibilities, access and retention periods helps verify compliance with applicable obligations.
- Less reconciliation work. Common rules on master data and calculations reduce the need to reconstruct the differences between reports every time.
- Verifiable foundations for analytics and AI. Documented provenance, quality and conditions of use make it possible to assess which data are suitable for a specific purpose.
- Anomalies assigned to an owner. Every issue detected follows a defined path of reporting, assessment and resolution.
To verify results, you can look, within the chosen domain, at discrepancies between reports, anomaly resolution times and whether definitions and responsibilities are documented. Indicators should be chosen according to the problem the programme is meant to solve.
Who is responsible for data governance: roles and responsibilities
Responsibility for data governance is distributed across roles with different tasks and authority. For each data domain it must be clear who makes the decisions, who oversees their implementation and who manages the systems.
- Executive sponsor. Champions the programme within senior management, secures its mandate and steps in on obstacles that require organisational authority.
- Governance committee. Coordinates cross-functional decisions, approves policies within its mandate and resolves conflicts between functions.
- Data owner. Is accountable for decisions concerning a domain, such as customers, suppliers or products. Approves definitions and quality requirements and oversees access and usage criteria within company policies and applicable constraints.
- Data steward. Maintains operational definitions, checks that quality rules are applied and oversees the reporting and resolution of anomalies.
- Data producer. Generates or enters data into processes. A salesperson filling in a customer record, for example, must know the required fields and the rules for completing them.
- Data custodian. Oversees the infrastructure that hosts the data and applies the agreed technical measures for access, protection and management.
- Data Protection Officer, or DPO. Where designated, advises on and monitors the protection of personal data. The role is not the same as overall responsibility for data governance.
A data governance specialist helps build and coordinate this system: they define processes, support owners and stewards, organise the glossary and catalogue and prepare evidence of controls. Responsibility for individual domains remains with the relevant roles.
Principles of data governance
The principles of data governance guide decisions on data and must be applied through company rules.
- Accountability. Every relevant data asset has a named owner.
- Transparency. Definitions, rules and decisions are documented and accessible.
- Quality. Data must be fit for use: accurate, complete, consistent and up to date.
- Security and confidentiality. Access is proportionate to role, and the processing of personal data complies with the law.
- Standardisation. Formats, naming conventions and definitions are consistent across functions and systems.
- Traceability. The origin of a piece of data and the transformations it has undergone can be reconstructed.
- Ethical use. Data is used for legitimate purposes consistent with the trust of customers, employees and partners.
Every principle must translate into a verifiable rule. For the completeness of a master record, for example, you need to establish which fields are required, who checks they are filled in and how anomalies are handled. Thresholds and check frequencies depend on how the data is used and how critical it is.
Data governance framework: where to start
A data governance framework connects objectives, roles, rules, processes and controls. It defines how decisions on data are made, how they are applied and what evidence allows them to be verified. It should be built around the company’s size, risks and organisation.
Methodological references include DAMA-DMBOK 2 by DAMA International, updated with the 2024 revision. The guide frames governance within the wider data management system and provides a reference to be adapted to the company’s context.
The methodological reference should be distinguished from the organisational model, which distributes authority between the central structure and the domains. To choose a model, you need to establish which decisions require a common criterion and which can be left to individual functions.
Three organisational models
Centralised
A central body retains authority over the main governance decisions and sets common rules for the domains.
When it makes senseWhen uniform criteria are needed and decisions can be effectively coordinated by a single oversight body.
Federated
Domains have explicit decision-making responsibilities and take part in common coordination on shared standards and cross-cutting issues.
When it makes senseWhen functions or companies have different needs and must agree on how to govern shared data.
Hybrid
The company combines centralised decisions with local autonomy, specifying for each area who approves the rules and who oversees their application.
When it makes senseWhen some data require uniform oversight and others can be governed with greater autonomy by the domains.
Role names can overlap across different frameworks. That is why the chosen model must state concretely who decides on definitions, quality, access and exceptions, which decisions require agreement between domains and how conflicts are resolved.
The operational components of data governance
Policies work when they rest on concrete components. These are the main ones.
Metadata, catalogue and glossary
Metadata are the data that describe data: name, meaning, format, owner, origin, confidentiality level. A data catalogue gathers them in a single point of reference; the business glossary sets the shared definitions of business terms.
Data lineage
Data lineage traces the path of a piece of data: where it originates, which systems and transformations it passes through, and where it is used. It helps assess the impact of a change, find the source of an error and prove where a figure in a report comes from.
Data classification
Classification assigns data to internal categories according to their criticality and protection requirements. Public, internal and confidential are examples of levels a company may adopt. Each level must have corresponding rules for access, sharing and protection; retention periods should also be defined according to purpose and applicable obligations.
Data quality
Data quality indicates how fit data is for its intended use. Accuracy, completeness, consistency, timeliness, uniqueness and validity are dimensions to be assessed in relation to the process. Control requires explicit rules, periodic measurement and an owner for handling anomalies.
Master data management
Shared master records (customers, suppliers, products, sites) are used by many processes and systems: if different versions exist, problems multiply. Master data management defines a reference version for each master record, along with the rules for creating and maintaining it.
Data governance policies
Data governance policies set out the approved rules on data access, use, retention and deletion. They must assign responsibilities, refer to the enforcement procedures and specify how compliance with the rules is verified and how exceptions are handled.
How to launch a data governance programme
A realistic programme moves in short steps, within a clearly defined scope.
Define the mandate
Objectives, sponsors and boundaries: why governance is needed and what must change.
Choose a priority domain
Customers, suppliers, financial data or ESG data: an area where the problem is felt and measurable.
Appoint the roles
For that domain, data owners and data stewards with real time and a real mandate.
Map the current state
Where the data sits, who uses it, which definitions are in circulation, which rules already exist.
Write the minimum rules
Definitions, classification, quality and access rules for critical data.
Build the tools
Glossary, catalogue, lineage, quality controls: with the complexity the domain requires, no more.
Measure and report
Present the domain indicators at an agreed frequency: critical data with an assigned owner, open and resolved issues, resolution times and the results of quality checks.
Extend
Assess the results of the first domain, correct the rules and responsibilities that did not work and use the lessons learnt to launch the next domain.
An example of data governance for customer master data
In a hypothetical case, sales and finance use two different definitions of “active customer”. Sales considers a customer active if there is an open negotiation; finance, if they have made purchases in the reference period. The reports therefore return different figures under the same name.
The customer domain data owner coordinates the discussion and approves two distinct definitions, “customer with an open deal” and “customer with purchases in the period”, specifying the intended use of each. The glossary records the definitions, the period to be considered and the sources to be used.
The data steward checks that fields and reports comply with these rules and logs any issues. IT implements the agreed logic in the systems and configures access according to the approved criteria.
Control leaves evidence that can be consulted: the approved glossary, the issue log and the results of periodic checks on reports. When a definition changes, the change is approved and documented before it is applied.
Data governance, GDPR and the Data Governance Act
A data governance system helps organise the information, responsibilities and checks needed to meet applicable obligations. Its existence alone does not demonstrate regulatory compliance.
GDPR. Regulation (EU) 2016/679 governs the processing of personal data. The accountability principle requires the controller to comply with the principles it lays down and to be able to demonstrate that compliance. Documented roles, mapped data flows and controls on access and retention can support this work. The record of processing activities and the designation of a DPO are subject to the conditions and requirements set out in Articles 30 and 37 respectively.
Data Governance Act. Regulation (EU) 2022/868 governs, among other things, the re-use of certain categories of protected data held by public sector bodies, data intermediation services and data altruism. It is distinct from corporate data governance: it concerns a European framework for sharing and re-use, whereas internal data governance establishes responsibilities, rules and controls within the organisation.
Data governance and ESG
For sustainability indicators, sources must be documented, an owner assigned and calculation methods made traceable. Data governance links each indicator to the source data, the transformations applied and the checks carried out before reporting. In this way, the company can explain how a result was obtained and manage any corrections.
Data governance and artificial intelligence
Data governance applied to artificial intelligence establishes which data may be used and under what conditions. The controls cover the data used to train and evaluate models, the information entered during use and the documents the system can consult. For each use, provenance, quality, confidentiality, authorisations and, for personal data, the lawful basis for processing must be verified.
This oversight falls within the broader governance of AI, which also covers risk assessment, human oversight and the allocation of responsibility for systems. Governing data makes it possible to document part of the conditions on which their use is based.
Three organisational models
Centralised
A central body retains authority over the main governance decisions and sets common rules for the domains.
When it makes senseWhen uniform criteria are needed and decisions can be effectively coordinated by a single oversight body.
Federated
Domains have explicit decision-making responsibilities and take part in common coordination on shared standards and cross-cutting issues.
When it makes senseWhen functions or companies have different needs and must agree on how to govern shared data.
Hybrid
The company combines centralised decisions with local autonomy, specifying for each area who approves the rules and who oversees their application.
When it makes senseWhen some data require uniform oversight and others can be governed with greater autonomy by the domains.
The most common mistakes
- Leaving the programme to IT alone. The functions that produce and use data must take part in decisions on their meaning, quality and use.
- Buying tools before defining responsibilities. A platform does not make up for the absence of a mandate, shared definitions and approved rules.
- Extending the programme to all data straight away. An initial scope that is too broad can spread resources thin and slow down checks. The first domain must have explicit objectives and boundaries.
- Appointing owners without authority or time. The role must carry real decision-making power, resources and a path for resolving conflicts.
- Approving policies without checking how they are applied. Every rule must have someone responsible for verifying it and a procedure for handling deviations.
- Leaving people without instructions. Those who produce or use data must know the rules that apply to their work and how to report a problem.
The Bliss angle: the real value of Data Governance
At Bliss, we believe the quality of a decision also depends on the ability to reconstruct the information on which it is based. When two functions present different figures, clarity is needed on definitions, responsibilities and verification criteria.
Data governance makes these elements explicit and allows management to assess how they are applied over time.
The Bliss Agency perspective
Domande frequenti
What is meant by data governance?
Data governance is the system of roles, rules, processes and controls through which an organisation governs its data. It establishes who decides on its use, who is accountable for its quality and how the application of rules is verified.
What is a data governance model?
It is how the organisation allocates decision-making authority and responsibility for data. It may involve central oversight, responsibilities distributed across domains, or a combination of the two.
Who is responsible for data governance?
Responsibility is shared across several roles. The sponsor backs the programme, the committee coordinates cross-functional decisions, data owners are accountable for their domains and stewards oversee the operational application of the rules.
Who is the data owner?
This is the role accountable for decisions concerning a data domain, such as customers or suppliers. It approves definitions and quality requirements and oversees usage criteria within company policies and applicable constraints.
What role does the data producer play?
The data producer generates or enters data in business processes. They must apply the rules set for collection and data entry and flag any errors they cannot fix directly.
What is AI data governance?
It is the governance of the data used by artificial intelligence systems, from training to operational use. It defines responsibilities and controls over the provenance, quality, confidentiality and conditions of use of the data.
Fonti e riferimenti
- Parlamento europeo e Consiglio, Regolamento (UE) 2022/868 del 30 maggio 2022 relativo alla governance europea dei dati (Regolamento sulla governance dei dati)
- Commissione europea, Data Governance Act
- Parlamento europeo e Consiglio, Regolamento (UE) 2016/679 (GDPR)
- Garante per la protezione dei dati personali, Regolamento UE 2016/679, sintesi dei principi, del registro dei trattamenti e del DPO
- DAMA International, DAMA-DMBOK

