Go-live for Dragonfly: Our upgrade to SAP S/4HANA at six locations
On October 4, we at RENOLIT simultaneously upgraded our central ERP system to SAP S/4HANA at six locations across three countries. The announcement of this success could be summed up in a single sentence. Behind the scenes, however, lay a massive undertaking spanning several years, involving an incredible number of challenges, a few setbacks, and major successes. One Friday, we held our breath for just a moment. This is a report on one of the most demanding projects RENOLIT has implemented in recent years.
Why we had to upgrade our SAP system
Our previous SAP system had run reliably for many years. When SAP announced some time ago that it would be ending support for SAP R/3, it was clear that we would have to leave this familiar world behind. Six of our locations were affected and had to make the switch simultaneously: Worms, Frankenthal, Waldkraiburg, and Munich in Germany, as well as Cramlington in England and La Porte in the U.S. Whoever touches this system touches the entire company. My respect for it was accordingly great. We named the program “Dragonfly.” It would be a long road before it could take flight.
Brownfield and the point where nothing worked anymore
At the beginning was the fundamental question: Do we just replace the system? Or do we use this opportunity to overhaul all our processes? In technical terms, the first approach is called “Brownfield,” and the second is called “Greenfield, ” meaning building from scratch on a greenfield site.
Of course, it’s tempting to let the dragonfly take flight on a fresh, green field. But we deliberately remained realistic and opted for a middle ground. I jokingly called this approach “Brownfield with green tufts.”
After all, a complete new build would have taken us many years. RENOLIT has grown so rapidly in recent years that this has resulted in a multitude of individual processes and workflows. Rebuilding these from scratch in a uniform manner would have been a mammoth task.
If we had tackled everything at once, it would likely have jeopardized the entire project. Our approach was therefore to adopt the existing processes and optimize them in a very targeted manner at a few key points. We thought we’d chosen a manageable path. By early 2025, however, we’d reached a point where nothing was working anymore. We’d underestimated how much experience and structure even this approach required.
Our dragonfly was grounded.
Help from an external partner and how much we learned in the process
In this situation, my fellow board member Thomas Sampers was a great help to me. I asked him to share responsibility for the program with me on the board, since he had already led an SAP implementation across 16 European countries himself. He knew from his own experience what we were missing. On his recommendation, we brought Deloitte on board as our new implementation partner in February 2025 and restructured the program.
And yet, the beginning wasn’t easy. At Deloitte, we were working with experts who had already successfully followed this path in many projects over the past few years.
The project management was correspondingly rigorous and demanding. Yes, there were bumps along the way, but it moved us forward. Above all, we learned to document our own processes precisely before changing them. All of that remains with us, even after the consultants have left.
Shortly before the launch, our partner even sent former Ironman World Champion Sebastian Kienle to us, who showed us the similarities between elite sports and a project like this. I never would have expected that kind of inspiration in an IT project. Our own people had to contribute their knowledge of our processes. That’s exactly where the real work began for many of them.
13,816 test cases leading up to the dress rehearsal
As early as 2025, we began restructuring the old system in several phases while it was still in operation, in order to make the eventual transition possible in the first place. In three rounds of testing from May through August 2026, we ultimately conducted over 13,000 tests across 832 different scenarios.
The testing was primarily carried out by our key users—colleagues from purchasing, planning, sales, shipping, finance, and controlling—who know their processes best. Every failed test was a minor setback and, at the same time, an error that we were spared on launch day. They did all of this alongside their day-to-day work and went above and beyond.
In September, we ran through the entire transition in a dress rehearsal. In project terminology, this is called a “dress rehearsal.” Interestingly, this is a term also used in space exploration. From that point on, the comparison to a rocket launch wouldn’t leave my mind.
A two-week countdown and a test that failed twice
Although the go-live date had been set for a long time, it wasn’t until late summer that it really hit us: The countdown was on. October 4, a Sunday, was to be our launch day.
However, because the transition was so complex, it ultimately took two launch weeks. Thousands of processes and tests were therefore initiated about 14 days before the big day. Countless checklists were worked through.
In the process, the old system was phased out step by step, from new item numbers to new orders to the final transaction. Our systems continued to run, and their data was temporarily stored.
Then came the moment everyone dreads in a project like this. A critical process failed twice in a row during the final week. On Friday morning, the team launched the third attempt. Another failure would have jeopardized the launch that weekend.
I admit that I was more nervous than the team that Friday. When I arrived at the IT floor that afternoon, which was serving as our control room during those days, everyone was working with surprising calm. Around 4:00 p.m., a cheer rang out from the other corner. At first, I didn’t even know what was going on until I realized that the process had completed successfully and everything was green. At that moment, a huge weight was lifted off my shoulders.
Sunday, 11 a.m.: The decision on “Go” or “No-Go”
The conversion then ran starting Friday evening. On Sunday at 6 a.m. German time, the key users at all six locations began checking the data in the new system. In Cramlington, it was 5 a.m. at that time, and for the team in La Porte, the second night shift in a row was just beginning.
Thomas Sampers and I were in Worms starting at 9 a.m. and were able to watch as one area after another reported “green.” By 11 a.m., just like a rocket launch, all sites had given the go-ahead.
The tests had been passed, all six site managers had given their approval, and our program management team led by Fabian Reiter and Martin Klinkert recommended the launch. We could still have rolled everything back to Friday’s state.
Now Thomas Sampers and I had to make the final decision. Are we really going live with the system?
For us, it was clear: Yes. The team had done a fantastic job; everything was ready.
At 12 p.m., the phase began that’s actually called “Point of No Return” in the plan. A short time later, we took off. The systems came back online, and starting Monday morning, all six sites were operating on the new system.
We were finally up and running. What a feeling!
The first few days on the new system and what has changed for us
The first few days went better than expected. By the way, one of the bug reports came from me, since I couldn’t log in at first. It quickly became clear: In this case, unfortunately, the problem was sitting right in front of the screen, and I first had to figure out the new login process myself.
The new software represents a major change. But we, too—as a team and as a company—have changed. This applies above all to how we work together. In the past, IT handled such changes largely on its own. In this project, however, the business units described, tested, and approved their own processes. I’ve seen how colleagues have grown into new responsibilities in the process.
We now understand many processes that we didn’t know well enough at the beginning. IT and the business units have grown closer together. We emerged from this project different than when we went in.
Dragonfly is up and running, and work on the processes begins
Dragonfly is up and running! I’m incredibly proud of everyone who made this possible.
As important as our external partner was, our own team played the most crucial role.
The Dragonfly team held this program together and drove it forward. Our IT team made it technically possible and consistently surpassed expectations. The key users, process owners, and site coordinators tested, reviewed, and took responsibility across six locations—many of them working nights and weekends.
I’d also like to thank those who, for months, took over or supported the work of our key users behind the scenes so that they could focus on testing. They weren’t listed in any project plan, yet they played just as vital a role in this project. For this, Thomas Sampers and I—who were responsible for this program on the Executive Board—would like to express our heartfelt thanks.
Now the long haul begins, which I, as an endurance athlete, am particularly looking forward to. Over the next few years, we will tackle our processes step by step until the grass is completely green.
