Few things are as quietly risky as an ERP system that still works. For one high-tech manufacturer, a heavily customized on-premise system had run the business reliably for more than fifteen years. It had also become the single biggest obstacle to growth.
The company designs and builds precision instruments for semiconductor, aerospace and research customers. Its products are complex, highly configurable and sold worldwide. Its legacy ERP, by contrast, had been extended so many times that every change took months — and support for the underlying release was coming to an end.
The cost of standing still
Leadership had put off modernization for years, for understandable reasons. The system held decades of engineering and service history. Production couldn’t stop. And earlier estimates had painted a picture of a long, expensive reimplementation.
But the cost of standing still kept rising:
- Hundreds of custom programs had to be retested after every patch
- Configurable products required manual handoffs between sales, engineering and production
- Reporting depended on nightly extracts to a separate database
- New acquisitions couldn’t be integrated without a dedicated project
- Skilled support for the old platform was getting harder to find
Choosing the right path
The company weighed three options: a fresh implementation, a full system conversion, or a selective approach that carried over what still worked and redesigned what didn’t. A fresh start would have meant abandoning years of service records; a straight conversion would have carried every customization forward.
It chose the selective path. Master data, open transactions and several years of history moved to SAP Cloud ERP Private Edition, preserving continuity for service and warranty teams. Finance and order-to-cash processes were redesigned around standard practices. And most custom code was either retired or rebuilt as extensions on the SAP Business Technology Platform, keeping the new core clean.
“We didn’t want to carry fifteen years of workarounds into a new system,” the company’s CIO explained. “We wanted to bring our history and leave our habits behind.”
How the migration was run
A conversion of this kind succeeds or fails on preparation. Working with Oraya®, the program team focused on four disciplines:
- Custom code analysis. Every custom object was classified as retire, replace with standard, or rebuild as an extension. Roughly two thirds were retired.
- Data readiness. Data owners in each function cleaned and harmonized records before any migration run, and three full rehearsals were completed ahead of cutover.
- Business-led testing. Key users tested end-to-end scenarios with real orders and configurations, not just individual transactions.
- A tightly managed cutover. Go-live took place over a long weekend with an hour-by-hour plan, limiting production downtime to a single shift.
What changed after go-live
The benefits appeared quickly and have continued to grow:
- Product configuration now flows directly from quote to production order, with no manual re-entry
- Embedded analytics replaced the nightly extracts, giving managers real-time margin and backlog visibility
- The IT team spends far less time on maintenance and more on improvement projects
- A recent acquisition was brought onto the platform in weeks rather than months
Just as important, the company is now positioned to adopt new capabilities — from AI-assisted planning to predictive service — as they become available, without another major project.
Lessons for other manufacturers
Legacy modernization rarely has a perfect moment. But the manufacturers that handle it well share a few traits: they’re honest about which customizations still add value, they invest early in data and testing, and they treat the move as a chance to simplify rather than replicate.
For a look at how another high-tech company approached its transformation, read the Tolphi Manufacturing customer story. And if your own system is nearing the end of support, our team can help you weigh the options.
