We were asked to look at an S/4HANA programme that had been running for the better part of two years and was no closer to going live than it had been twelve months in. The manufacturer behind it had done the sensible thing and stopped throwing money at it until someone honest told them what was wrong. What we found was familiar: the system had been bent so far to match the old way of working that it had become something nobody could finish, test or upgrade.
Every awkward process in the legacy system had been reproduced as a custom development. The result was hundreds of bespoke objects, thinly documented, each one a small project of its own to maintain. The team had stopped being able to see the standard product underneath.
The challenges we had to solve
- A backlog of custom code that nobody could fully account for, much of it duplicating what S/4HANA does out of the box.
- A test environment that could never be made stable, because the customisations interacted in ways no one had mapped.
- A business that had lost confidence in the date, having watched two go-lives slip.
- An honest question that had been avoided: which of these requirements were real, and which were just habit.
How we approached it
We began with an inventory rather than a plan. Every customisation was listed, traced to the process it supposedly served, and put to a simple test: does the business genuinely need this, or was it added because the old system happened to work that way? A large share fell into the second category. Sitting with the process owners, we agreed to drop them and adopt the standard flow — which is almost always easier to operate and far easier to keep current.
What remained was a much smaller set of changes that reflected things the manufacturer actually does differently, and those we kept and documented properly. With the surface area reduced, the test environment finally became stable enough to trust. We then rebuilt the cutover plan around a scope the team could believe in, rather than the one that had been quietly slipping for a year.
Where it stands
The programme reached a go-live the business recognised as real, on a system that is closer to standard and therefore something they can actually upgrade in future. The customer came away with fewer moving parts than they started with, which is usually the point of an ERP, and a clear record of why each remaining change exists.