What Governance Model Should be Used for Data Projects?

Why talk about data governance?

All organizations today collect large amounts of data, covering everything from business lines and projects to human resources.

But accumulating data only makes sense if it is useful, readable, and controlled. Data governance is precisely the framework that allows everyone (management, business teams, partners) to rely on reliable and well-managed data to make decisions and take action.

Moving towards responsible data management: the 5Vs

Faced with a wealth of data, it is essential not to store everything “by default,” but to choose what is valuable to the organization. In data, we talk about the 5Vs:

1. Volume: the amount of data available

2. Velocity: the speed at which it is generated or received

3. Variety: formats, sources, complexity

4. Veracity: reliability, quality, consistency

5. Value: real usefulness for actions or decisions

In a responsible and sober approach, it is important to prioritize relevant data at the right time, rather than storing everything indiscriminately—which unnecessarily burdens systems, consumes energy, and slows down processes. So we focus on the last two Vs: Veracity and Value.

Know your processes to better frame the data

To manage your data effectively, you need to understand your business processes. Let's take a common example:

A user signs up for an offer → they are guided, supported, and then evaluated → at each stage, data is created, modified, or cross-referenced.

Mapping these flows allows you to:

• know where the data goes

• identify points of friction or transformation

• assess its quality (uniqueness, completeness, accuracy, freshness, consistency)

Security, responsibility, and legal framework

Any organization that processes data, especially when it concerns individuals, is responsible for managing it.

Data governance therefore includes:

• Securing incoming and outgoing data flows

• Access management (who sees what?)

• Documenting the purposes of use

• The data lifecycle: retention period, archiving, deletion

These elements are essential for remaining GDPR compliant, but also for maintaining a high level of trust with your stakeholders (users, funders, partners, etc.).

Architecture: working with existing structures, getting the sizing right

Information systems are not always new or harmonized, especially in the social and solidarity economy. Good data governance takes into account:

• the constraints of the existing system (heterogeneous databases, aging tools, etc.)

• real data constitution (volume, frequency, format)

• performance requirements for operational use

There is no need for a “big” system, but rather an architecture that allows for centralization, reliability, and enhancement of what matters.

Clear roles for shared responsibilities

Governance also relies on a clear distribution of roles, tailored to the size and resources of your organization. Here are the most common profiles:

Sponsor: Promotes the vision, provides resources, supports the initiative

Business representative: Gives meaning to the data, connects to the needs on the ground

Data steward: Oversees quality, documentation, and access

Analyst: Produces indicators, facilitates data interpretation

GDPR advisor: Ensures compliance and best practices for protection

In a small organization, these roles can be shared. The key is that they are identified, empowered, and work together.

In summary

Data governance is not a “technical” issue reserved for large organizations. It is a powerful lever for:

• Leveraging what we already know

• Making decisions more reliable

• Protecting beneficiaries and partners

• Avoiding wasted resources

• Making data a useful, simple, and shared tool