The problem did not sit within one individual function.
Microsoft Navision was established as the company's central ERP system. Many commercial processes worked reliably within it. At the same time, significant parts of international procurement, production and goods management were also being handled through Excel, email and direct coordination between people.
Sales, procurement, customer service, production control and logistics all held information – but they did not always have the same picture at the same time.
This created additional queries, duplicate data maintenance, unnecessary goods movements and a lack of prioritisation. When incoming shipments arrived, there was not enough early transparency regarding which goods they contained, when those goods were required and which customer orders were behind them.
Goods were received, temporarily stored, moved again and later prepared once more for onward shipment. In some cases, loading units were handled several times, even though a more direct flow of goods would have been possible.
Before discussing digitalisation, we first had to make visible where time, information, movement and cost were being lost.
Not every existing process was a bad process.
That distinction was critical to everything that followed. Many ways of working had developed over years and worked well in their basic logic. Employees understood the processes, the results were sound and the routines were fundamentally practical in day-to-day operations.
The task was therefore not to reinvent the entire organisation or to change functioning processes simply because a system preferred a different logic.
We first examined the actual current-state processes: What works? Where is work duplicated? Which information is maintained more than once? Where do unnecessary movements, waiting times, intermediate storage or repeated queries occur?
From this, the target process was developed. Good working practices were to remain. Unnecessary effort, information gaps and media breaks were to disappear.
Putting process before software does not mean preserving old routines without questioning them. It means first distinguishing what creates value from what merely creates effort.
Digitalisation should not merely represent work. It should create operational control.
Another point was at least as important: even before development began, it had to be clear which information would later be required for leadership and decision-making.
At the beginning, the company had only a limited number of reliable operational KPIs. The design process therefore did not ask only which inputs a process required, but also which data the process needed to generate.
The transport costs of individual transport units, for example, could be broken down across the quantities of goods actually contained within them. This also created a better basis for pricing calculations in sales.
A digital process should generate the data with which the company can subsequently be managed.
Why not simply put everything into the existing ERP?
Technically, much of the functionality could probably have been implemented within Navision. Operationally, however, that would not automatically have been the better solution.
International suppliers, quality management, external logistics partners and other participants needed to receive information and feed their own data back into the overall process.
Directly integrating all of these external companies into the internal ERP would have created additional licence costs, complex permission structures and data-protection questions. At the same time, the system would have been significantly more extensive than many of those users actually required.
Together with the internal IT team, we therefore developed an independent web application. It was not intended to become a second ERP system, but a lean process and information platform precisely for the areas in which internal and external participants needed to work together.
Navision remained in place where it made sense. The web application handled the international operational process logic. Relevant data was subsequently transferred back into the core system.
At the same time, information that previously had to be collected from different screens and positions could now be presented together in the context of the relevant transaction.
The right system decision is not automatically the greatest possible degree of integration. What matters is which architecture makes the process simpler, more controllable and usable for everyone involved.
The application was not developed for employees. It was developed with them.
As COO, I initiated the project and led it as Project Lead. Following alignment with the executive management, sales, procurement, customer service and IT were involved early in the development process.
The basic logic was developed together with the operational functions. New development stages were presented regularly, tested and adapted on the basis of feedback from the people who would later use the application.
As the project progressed, international locations, logistics partners and suppliers were also brought into the process. Requirements were discussed on site and the rollout was built step by step across countries and partners.
I personally visited suppliers and logistics partners in Europe and Asia, presented the system, conducted training and explained why individual working practices needed to change and how the additional transparency would also benefit the respective partners.
COO and Project Lead · process design · alignment across functions · coordination with IT · international implementation · supplier and partner discussions · training and operational support during rollout.
Then live operations revealed what no concept can fully anticipate.
The implementation worked – but it also exposed friction in several places.
Actual production volumes did not always match the plan. Quality inspections meant that individual batches could not be released. Packaging structures had to be changed at short notice. At some partners, the necessary buffer space was also missing, making it impossible to implement the planned process exactly as designed.
At the same time, the digital maturity of international suppliers varied considerably. Many wanted to maintain their existing Excel-based ways of working in parallel. This initially created parallel data environments internally as well.
The training effort was significantly greater than originally expected. Particularly where leadership did not consistently support the change, a technical implementation quickly became an organisational issue.
The pandemic subsequently introduced a situation that nobody could have anticipated during the original process design. Procurement routes changed at short notice, production was shifted between regions and suppliers, logistics partners changed and goods originally planned for ocean freight sometimes had to be moved by air freight at short notice as priorities changed.
A digital process must not only handle the planned normal case. It must be able to cope with an operational reality that changes.
Transparency became operational speed.
Not every effect was consistently quantified financially at the time. The operational impact, however, was clearly visible.
Digitalisation begins before the software.
Putting process before software does not mean preserving existing ways of working at any cost. It means making a clear distinction before a technical decision: What works? What creates unnecessary effort? Where is information missing? Which data will be required to manage the company better afterwards?
Only then can a meaningful decision be made about whether a standard solution, a modification, an interface or an individual application is the right answer.
Execution is equally important. A technically good system does not change an organisation by itself. The people who work with it need to understand the logic, be able to provide feedback and actually support the change in day-to-day operations.
And even then, no process model remains final. Markets, value chains, customer requirements and operational reality change. Good systems therefore need to be not only efficient, but adaptable.
An initial confidential conversation is enough to assess the starting situation, responsibility and potential assignment scope.