“We’re not going to automate for the sake of automating. We’re going to automate and add data into the process to create value.” — Wouter van Peteghem, Managing Partner at Alluvion.
Everyone says it: data quality is important. Yet in many organizations, it remains little more than an intention. The real question, therefore, is not whether you want better data, but whether your organization is ready to truly organize, embed, and sustain master data excellence.
In a recent podcast, we spoke with Terumo as the end customer and Alluvion as the integration partner about their journey toward First Time Right in SAP. The most valuable insight wasn’t in the technology, but in the approach: start small, make it measurable, organize ownership and only then accelerate.
Why master data excellence must become a priority now
In many organizations, we see the same patterns recurring: duplicates and inconsistencies. Manual corrections. Processes that slow down. Teams maintaining their own version of the truth. That’s frustrating. It becomes risky once you start automating. Then you spread errors faster than ever before.
At Terumo, this became very tangible in supplier onboarding. Forms, email chains, and manual data entry made the process slow and error-prone. The goal was not a more elegant data model. The goal was a faster, more reliable, and better-governed process.
Watch the webinar as well: First Time Right in SAP.
Step 1. Don’t start with tooling, start with maturity
Implementing a tool to solve a data problem achieves little if the surrounding organization doesn’t change with it. That’s why the approach started with a data maturity assessment, not to assign a score, but to determine the first realistic step forward.
Who owns master data? How do business and IT work together? Are definitions of correct and complete shared? Can you measure bottlenecks, or does everything rely on gut feeling? Without this foundation, a system quickly becomes an additional layer of complexity.
A critical note: many organizations believe they are already far along because they have an MDM tool in place. In practice, ownership often turns out to be unclear or insufficiently embedded in the organization. diffuser dan gedacht.
Interesting read: Who is steering your decisions?
Step 2. Choose one process with real pain
The most pragmatic lesson: start where it actually hurts. Not with “we want better data,” but with a specific process that is costing time or quality today.
At Terumo, that was supplier onboarding. One KPI became leading: the lead time from request to creation and approval in SAP. Through digitization and further automation, that lead time was demonstrably reduced. It only becomes truly relevant when you factor in the volume..
Importantly, this did not happen all at once. The improvement came through iterations over several years. That requires discipline. in communicatie en verwachtingmanagement.
Step 3. Design for First Time Right instead of fixing issues afterward
At Terumo, data quality became the starting point, not an afterthought. That means verification and enrichment during creation and change, instead of corrections afterward.
A key mechanism in this approach is unambiguous identification. By using a unique identifier, entities could be accurately recognized and managed consistently across systems. betrouwbaarder herkend worden en werden duplicaten teruggedrongen. Dat legt de basis voor stabielere processen in de hele keten.
Step 4. Integrate reliable external data, but do not underestimate the complexity
External data enrichment may seem straightforward. In practice, you encounter country-specific address structures, postal code formats, and variations in spelling. Data mapping was explicitly mentioned in the case as a key point of attention.
The solution lay in adding country-specific logic to ensure that source data and the SAP context aligned properly. This shows that master data excellence is not just a data challenge, but also an integration and process challenge.
An important counterpoint to keep in mind: external data is not an absolute truth. It is often the best available source, but your process determines how exceptions are handled and who makes the decision in case of doubt.
Step 5. Organize buy-in through demonstrable value
Adoption does not happen by telling people that data is important. Adoption happens when teams experience that their work becomes easier, faster, or more reliable.
That is why value was consistently linked to tangible outcomes: less manual work, shorter lead times, greater insight into bottlenecks, and improved auditability and traceability. At the same time, ownership at the management level remains crucial. It is not a one-time action, but a journey in its own right.
Step 6. Build sustainably and scalably
A common pitfall is quickly building something that works for one process. After a few years, you are left with custom solutions that are difficult to maintain. That is why a clear principle was upheld: keep the core as standard as possible and create value through integrations around it.
This prevents your solution from becoming dependent on specific individuals or exceptional configurations.
Step 7. Think in phases, not in an end state
Master data excellence is never truly finished. You can, however, structure it into logical phases. First, get your processes and master data automation in order. Then integrate reliable sources to minimize manual work. Next, scale more intelligently with techniques such as AI for, for example, duplicate detection and identifying inconsistencies.
AI without a solid foundation mainly accelerates errors. With a strong foundation, it strengthens your processes.
