The normal state
Where are trade orders planned and steered in a steel trading business? In a surprising number of operations, the honest answer is: in a spreadsheet. Maintained by hand, mastered by a few people, disconnected from the ERP - and business-critical.
That is not a failure of individual companies. It is the result of a system landscape grown over decades, in which every new requirement was answered faster with a table than with a system project. ERP, warehouse management, processing and shipping ran side by side - and Excel filled the gaps between them.
The cost of those gaps is well known: duplicate data maintenance, knowledge held in heads rather than systems, no real-time view of order status. The more interesting question is what getting out actually looks like.
The project
The company replaced its fully hand-maintained spreadsheet for planning and steering trade orders with the sector solution from Zirkel Technologies - connected to the SAP system through an interface that transfers all order data automatically.
The results reported in the interview, by the company's own account: around 30 percent of working time saved in work preparation and steering, picking without waiting times, faster and more fact-based coordination between work preparation and the warehouse. A side observation worth noting: the intuitive interface particularly helped temporary workers who are not native German speakers to understand orders reliably - digitalization as an answer to the skills shortage, not only to process cost.
One detail deserves particular attention. Asked about long-term wishes, the managing director names three: fully paperless operations, a bidirectional interface - and doing away with Excel files. Because for the monthly on-time delivery statistics, the company still needs Excel.
What the case reveals about legacy modernization
Three patterns from this case apply far beyond the individual company.
First: entry works one-way - the value sits two-way. An interface that pushes data from the ERP toward the shop floor is the right first step: contained risk, fast benefit. But as long as confirmations do not flow back into the core system automatically, the ERP stays blind to what actually happens in the hall. Analyses - on-time delivery, for instance - still have to be compiled by hand. Which is precisely why the spreadsheet ends up back on the table, just in a different place.
Second: point solutions don't disappear, they relocate. Closing one gap makes the next one visible. That is not failure - it is the normal course of modernization in stages. What matters is whether the stages follow an architecture, or whether each stage creates a new island.
Third: the target is not any single system, but a single truth. Order status, stock, processing progress and delivery performance belong in one continuous data flow - regardless of how many systems run behind it. Whether the leading system is SAP, Microsoft Dynamics 365 or a sector solution is decided by the trading house's process, not by the label.
The step after the first step
For trading houses that have their first digitalization step behind them, the real architecture questions begin:
- Which confirmations is the core system missing? Work steps, actual times, material consumption, remnants - everything that today is captured after the fact, or not at all.
- Which analyses still live in Excel? Every one of those tables marks a missing integration - and a KPI risk, because manually maintained statistics are only as reliable as their last update.
- Can the existing interface architecture carry the expansion? Point-to-point interfaces are quick to build and hard to extend. An integration layer - on a MuleSoft basis, for instance - turns individual connections into an architecture in which new systems dock on instead of forming new islands.
Zirkel Technologies supports exactly this step: assessing grown system landscapes and modernizing them in stages - vendor-neutral, with no license interest, and with a project team that does not need processing, batch control and delivery performance explained first. The output of such an assessment is not an off-the-shelf system recommendation but a sequence: what gets integrated, what gets replaced, what stays - and in which stage.
Because the goal is not to ban Excel. The goal is that nobody needs Excel to know where their orders stand.
← Back to Insights