Why many roadmaps are too vague from the outset

In many data programs, the roadmap is the point at which all good intentions are written onto one slide simultaneously. Improve governance. Increase data quality. Modernise the platform. Expand self-service. Trial AI. Sharpen the operating model. Everything is understandable – just not prioritized.

This turns the roadmap into a wishlist. And wishlists don't help in real-world conditions. Budgets are limited. Teams are limited. Management attention is limited. This is precisely why a good roadmap must enforce sequencing.

What a robust roadmap must deliver

The right order dictates the perfect target architecture

Many roadmaps fail not because the target states are wrong. They fail because the sequencing is unsustainable. Then platform topics are clarified before ownership. Self-service is rolled out before terms and quality logic are stable. Or AI initiatives are scaled even though the necessary data foundation is still selective, inconsistent or organizationally disputed.

A resilient plan acknowledges these tensions. It is built not for the ideal state, but for the reality of the organization: available teams, limited attention spans, competing programs, politically sensitive responsibilities.

In 2026, external time pressure will be added: With the end of maintenance for SAP ERP 6.0 at the end of 2027, the data architecture will become a scheduled obligation for many companies – another reason to consciously set the order now rather than later under migration pressure.

How the target image and maturity level become a genuine sequence

If Step 1 delivers the target image and Step 2 the gap image, then the roadmap answers the operational consequence: Which initiatives go into implementation first, which run in parallel, and which are deliberately set for later?

This is precisely where the real management quality of a data strategy emerges. For now, an analytical picture is transformed into a decision-making program. Not everything that makes sense will automatically be included in the first wave. And not everything that is quickly visible automatically deserves priority.

• Wave 1: Measures with high impact and viable prerequisites
• Wave 2: Topics that build on Wave 1 or close gaps emerging there
• Later: Initiatives with strategic potential, but currently insufficient basis

How to tell if the roadmap is viable

• The initial priorities are limited and clearly justified.
• Every measure has a discernible relationship to the target image and the gap image.
• Dependencies are visible and not implicitly hidden.
• There is clear ownership, not just programme rhetoric.
• The organization can explain why certain topics are deliberately brought up later.
• Exit plans for critical data services are documented and, where DORA applies, robustly tested.

That's precisely what separates roadmaps that only generate buy-in from those that enable actual control.

How AdEx Partners is proceeding in this step

We build roadmaps as decision programs, not as slide decks. From the gap analysis, we identify three to five initiatives with the highest potential impact for Wave 1, and in parallel define the foundational work packages for later scaling.

Every initiative undergoes a sequencing test against dependencies and absorption capacity. What doesn't pass is deliberately postponed – the result is a controllable sequence, not a consent slide.

Why the real discipline begins with Wave 1

Most data programs don't lose their effectiveness at the start, but after the initial wave. Early successes are visible, attention moves on – and suddenly the roadmap is competing with the next trend being pushed through the company. It's precisely there that it's decided whether activity turns into control.

Staying the course means: Wave 2 will not be left to chance. It requires fixed review points at which progress, knowledge gaps, and priorities are reassessed, and visible management commitment beyond initial success. „Not everything at once“ is only a strength if what is consciously coming later actually arrives – and doesn't quietly disappear as soon as the next topic gets louder.

The series as a whole

This closes the logic of the series. Firstly, it clarifies what data initiatives are intended to contribute to. Secondly, the real starting situation is honestly assessed. Thirdly, this builds a sequence that brings together scarce resources, prerequisites, and impact.

That's precisely the real strength of a good data strategy: it connects ambition and feasibility in such a way that data programs do not become a portfolio of activities, but rather a manageable path to transformation.