When is the right time to start managing your Data & AI initiatives as a portfolio?

The right time to start managing your Data & AI portfolio isn’t when your governance is perfect. It’s when your decisions require more visibility, consistency and accountability than your current processes can provide.

For many organizations, the decision to implement portfolio management comes with a familiar hesitation:

We see the value, but we’re not quite ready yet.

Perhaps the data strategy is still evolving. Different business units may follow different governance processes. Or more urgent AI initiatives may seem to deserve attention before investing in the structure that will help manage them.

These concerns are understandable. But they raise a more important question:

What exactly needs to be ready before you can start?

The answer may be less than you think.

When tracking initiatives is no longer enough

In the early stages of a Data & AI strategy, managing initiatives can be relatively straightforward:

• Spreadsheets capture incoming requests and initiative details.
• Project management tools track delivery and milestones.
• Presentations consolidate progress for leadership meetings.
• Steering committees align stakeholders on priorities.

This approach can work when there are relatively few initiatives, limited dependencies and a small number of decision-makers.

The challenge appears when organizations need to make decisions across initiatives, rather than simply monitor each one independently.

Consider this scenario

Three business units propose different AI projects. Each has a sponsor, an expected benefit and a delivery plan.

But funding is limited.

Leadership needs to answer four questions:

1. Which initiative should move forward?
2. Which one best supports strategic priorities?
3. Are different teams investing in similar capabilities?
4. What evidence should support the final decision?

A project tracker can tell you what each team is working on. It doesn’t necessarily provide the shared framework needed to compare competing investments.

That’s where portfolio management becomes valuable.

The objective is no longer simply to know what exists. It’s to understand what deserves investment, why it matters and whether it ultimately delivers the expected value.

Five signs your organization is ready for portfolio management

There’s no universal number of projects, employees or business units that determines when portfolio management becomes necessary.

Instead, look at the decisions your organization is struggling to make.

1. Prioritization is becoming a negotiation rather than a structured decision

Different teams often evaluate initiatives using different criteria.

One emphasizes potential revenue. Another prioritizes technical feasibility. A third focuses on operational efficiency.

All three perspectives matter. The problem arises when there’s no shared way to compare them.

Watch for these signals:

• Initiatives are approved based primarily on who sponsors them.
• Different teams use different evaluation criteria.
• Priorities change without a clear record of why.
• Leadership struggles to compare the expected impact of competing projects.

Portfolio management introduces a common framework for evaluating business impact, feasibility, strategic alignment and other relevant criteria.

It doesn’t eliminate judgment. It makes the basis for that judgment more transparent.

2. Leadership needs answers that require manual consolidation

How many AI initiatives are currently in development? Which have been approved? How much investment is committed? Which initiatives support a particular strategic objective?

If answering these questions means contacting multiple teams and reconciling spreadsheets, your organization may have outgrown its existing reporting process.

Common symptoms include:

• Multiple versions of the same initiative list.
• Significant preparation before steering committee meetings.
• Different definitions of initiative status or progress.
• Limited visibility into dependencies and ownership.

A portfolio approach creates a common view of initiatives, their status, owners and expected outcomes.

The goal isn’t another dashboard. It’s reliable information for better steering decisions.

3. Similar initiatives are emerging across different teams

As Data & AI adoption expands, multiple departments may independently identify similar opportunities.

For example, Marketing and Customer Success may both launch initiatives to predict customer churn, each with its own team, budget and data requirements.

That doesn’t automatically mean one initiative should be eliminated. Their objectives and constraints may differ.

But without portfolio-wide visibility, it’s difficult to assess whether they could:

• Reuse the same data products or technical components.
• Share development resources.
• Align their objectives and success criteria.
• Avoid unnecessary duplication of effort.

Portfolio management makes these opportunities for coordination easier to identify before investment decisions are finalized.

4. Governance responsibilities are becoming unclear

Who approves an AI use case? Who assesses its risks? Who decides when a data product should be retired?

As more stakeholders become involved, informal governance becomes harder to coordinate.

A structured portfolio helps organizations establish:

• Ownership: Who is accountable for each initiative?
• Decision rights: Who evaluates and approves it?
• Lifecycle stages: What must happen before an initiative progresses?
• Traceability: What decisions were made, by whom and why?

This becomes particularly relevant as organizations expand their AI initiatives and introduce new governance requirements.

5. Demonstrating value is becoming as important as delivering projects

Completing a project doesn’t automatically mean it created business value.

An AI application can launch successfully without achieving meaningful adoption. A data product can be delivered on time without meeting the needs of its intended consumers.

Portfolio management helps connect delivery to outcomes by establishing:

• Expected value before an initiative begins.
• Business objectives and success criteria.
• Investment and resource visibility.
• Progress and lifecycle monitoring.
• Realized value after deployment.

The conversation shifts from “What did we deliver?” to “What did that delivery actually achieve?”

The maturity trap: Waiting for perfect governance

One of the most common reasons to postpone portfolio management is the belief that the organization must first establish a mature governance model.

The imagined sequence often looks like this:

1. Standardize every process.
2. Define every role and responsibility.
3. Align all business units.
4. Establish a complete measurement framework.
5. Finally, implement portfolio management.

But this approach creates a problem.

Governance processes that look effective on paper may not reflect how teams actually make decisions. Evaluation criteria may need refinement. Different departments may require different levels of autonomy.

Governance maturity doesn’t have to be a prerequisite for portfolio management. It can develop through portfolio management.

Rather than waiting for an ideal operating model, organizations can start with their existing processes, establish a shared foundation and improve it through use.

Start with your current process. Evolve at your own pace.

A progressive approach allows organizations to introduce portfolio management without attempting to transform everything at once.

Stage	What to focus on	What you gain
1. Adopt	Bring existing initiatives, owners and processes into a shared portfolio.	Visibility into what’s happening today.
2. Structure	Establish common evaluation criteria, lifecycle stages and decision workflows.	More consistent governance and accountability.
3. Steer	Connect initiatives to strategic objectives, investments and value indicators.	Better prioritization and strategic alignment.
4. Optimize	Expand across teams and refine decisions using historical information and outcomes.	A more connected, evolving portfolio operating model.

These stages aren’t a mandatory implementation sequence or a fixed maturity score.

They’re a way to think about how portfolio management can expand as organizational needs evolve.

What could the first step look like?

Imagine an organization currently managing 40 Data & AI initiatives across several spreadsheets.

Rather than introducing an entirely new governance framework, it starts by:

• Importing its existing initiative portfolio.
• Preserving its current lifecycle stages.
• Identifying initiative owners and sponsors.
• Consolidating reporting for one steering committee.
• Introducing a shared set of prioritization criteria over time.

The organization doesn’t need to reinvent its operating model before gaining better visibility.

It begins with a process people already understand and improves it progressively.

This is an illustrative scenario, not a documented customer implementation.

What happens if you wait another year?

Postponing portfolio management doesn’t mean business decisions stop.

New initiatives will still be proposed. Budgets will still be allocated. Projects will still be prioritized, approved, delayed or discontinued.

The difference is the framework supporting those decisions.

Consider what may happen during another year of disconnected portfolio management:

• New decisions: Investment choices continue to be made using existing evaluation processes.
• New initiatives: Teams may introduce additional tools, reporting formats or definitions.
• Historical information: The reasoning behind decisions may become harder to reconstruct.
• Future consolidation: Existing initiative records may need to be reconciled when a shared framework is introduced.

Not every organization will experience these challenges. But waiting doesn’t automatically make implementation simpler.

And there’s a less obvious consequence.

The decisions made today become part of tomorrow’s portfolio.

A portfolio management approach introduced now can help establish a consistent record of why initiatives were prioritized, what outcomes were expected and how those expectations evolved.

That information can become valuable when leadership revisits investment choices.

The strongest reason to start isn’t a hypothetical cost-of-delay calculation.

It’s the opportunity to improve the decisions you’re already making.

How DataGalaxy Portfolio supports a progressive approach

DataGalaxy Portfolio helps organizations connect strategy, planning and execution across Data & AI initiatives.

Its configurable approach supports organizations as they structure and evolve their portfolio management practices.

Depending on the implementation scope, teams can:

• Centralize initiatives: Bring Data & AI use cases and products into a shared portfolio.
• Establish accountability: Document ownership, responsibilities and lifecycle stages.
• Improve prioritization: Evaluate initiatives against defined criteria and strategic objectives.
• Monitor investments: Connect initiatives to costs, resources and expected business value.
• Track outcomes: Follow progress and compare expected value with realized results.
• Expand progressively: Extend portfolio management across teams and governance processes.

Rather than requiring every business unit to operate identically, the objective is to create the shared visibility and consistency needed for strategic steering.

The starting point can be simple: one portfolio, one existing decision process and one clearly defined business objective.

From there, the operating model can evolve with the organization.

So, when should you start?

Not necessarily when you launch your first AI initiative.

Not automatically when you reach a particular number of projects.

And certainly not only when every governance process has been finalized.

The right moment is when your organization needs to:

• Compare competing Data & AI investments.
• Make prioritization decisions across teams.
• Establish clearer accountability.
• Connect initiatives to strategic objectives.
• Understand whether investments are delivering their expected value.

If those decisions are already becoming difficult, the question may no longer be whether you’re mature enough to adopt portfolio management.

It may be whether your current approach is sufficient for the decisions you need to make next.

Your Data & AI strategy will continue evolving. Your ability to steer it should evolve too.

Explore DataGalaxy Portfolio to discover how to start with your existing governance model and expand progressively.

About the author: Max Faivre

Product Marketing Manager

Share on social

X (Twitter)LinkedInFacebook