Change Management for ERP and S/4HANA Migrations

Oct 23 / Marc-André Rochefort

Version française : Conduite du changement pour les migrations ERP et S/4HANA

Key takeaways

  • ERP and S/4HANA migrations rarely fail because the software doesn't work. They fail because people work around the new system after go live. Spreadsheets, shadow systems or helpful colleagues take its place.
  • ERP adoption is a separate effort from the technical cutover. It starts before go live, continues well past it, and needs its own plan, budget and owner.
  • An ERP touches finance, procurement, warehouse, production and sales at once. The change plan has to follow the handoffs between those functions, not each function on its own.
  • ERP key users need protected time written into the project charter, a named backfill for their day job and a visible role during hypercare.
  • Neither a big bang nor a phased cutover removes the need for change management. Big bang concentrates the effort into a short window, while a phased cutover spreads it across waves.
  • Hypercare that serves adoption checks on the floor whether people are actually running the new process, not only whether bug and access tickets get closed.

When an ERP or S/4HANA migration disappoints, the software is rarely the reason. People work around it once it goes live. They go back to spreadsheets, keep shadow systems running, or ask a colleague to enter data because the new screens still feel unfamiliar. The technical cutover is one project milestone. Adoption is a separate and longer piece of work that starts before go live and carries on well after it. It needs its own plan, budget and owner.

Why does the technology usually work when adoption often doesn't?

The consultancy delivering an ERP build is judged on whether the system functions. Interfaces connect, reports run, and the data model holds. None of that tells you whether a warehouse supervisor will use the new put-away process or stick with the paper log they trust more. The technical team optimizes for a stable go live, and nobody on that team answers for the six months after. That's when a stable system meets people who haven't changed how they think about their job. Most ERP disappointment comes from there. It rarely shows up in the project plan, because nobody listed it among the change management deliverables.

Why ERP scope differs from a departmental tool rollout

A CRM or a scheduling tool touches one function at a time. An ERP touches finance, procurement, warehouse, production and sales all at once. It often changes how those functions hand work to each other. Take a change that looks small to the project team, such as a new approval step in procurement. Within the same week, it can ripple into finance close timing and into warehouse receiving. So you can't run ERP change management like a single-department rollout with one training track and one adoption metric. The change plan has to be built around the handoffs between functions. Someone also needs the job of tracking where those handoffs break.

Key users: underused by design, not by accident

Most ERP programs name key users early. They come from each business area, join the workshops, help configure the system and are meant to become first-line support after go live. In practice, they're often pulled back into their regular jobs once configuration work slows down. Nobody backfilled their operational role during the secondment. By the time training starts, the key user has drifted out of the loop. They end up relearning the system with everyone else.

The fix is structural, not motivational. Key users need protected time written into the project charter and a named backfill for their day job. They also need a visible role during hypercare, not only during design. With that structure in place, they become credible peer support, which no help desk ticket can replace. Without it, the title sits on an org chart and that's all.

What does data migration look like from the user's side?

Project teams treat data migration as a technical exercise of extraction, transformation, load and reconciliation. Users see it differently. They open the new system and find a customer record that looks incomplete. Or a part number they've used for years now carries a different code. Sometimes the history they relied on for context never came across. Whatever the technical acceptance criteria say, a user's trust in the system gets set in the first hour spent with their own data.

Change management has a narrow but specific job here. Before go live, brief users on exactly what will look different in their own records and why. Then give them a fast, visible way to report a discrepancy. One wrong field shouldn't convince them the whole system is broken. A single bad data experience in week one can undo weeks of training if nobody prepared people for it.

Big bang or phased cutover?

A big bang cutover moves every function to the new system on one date. Risk gets compressed into a single window. That forces the organization to be ready everywhere at once and gives change management a hard deadline. The catch is that there's no fallback. If adoption support is thin in one area, that area has no old system to lean on while it catches up.

A phased cutover moves business units or geographies one at a time. It spreads the risk and lets the change team carry lessons from one wave into the next. It also means running old and new systems in parallel for longer. People then wonder which process is authoritative on a given day. Neither approach removes the need for change management. Big bang packs the effort into a shorter, more intense window. A phased approach spreads it out. The change team then has to keep momentum from wave to wave without a single hard date to rally around.

Hypercare is not a technical support rotation

Most programs staff hypercare as technical support. A team fields tickets about bugs and access issues for a few weeks after go live. You need that, but it doesn't tell you whether adoption will stick. Hypercare should also answer a behavioral question: are people actually running the new process? Or have they found a workaround and stopped reporting problems because they assume nothing will change? That quiet kind of resistance to change rarely reaches the ticket queue.

Hypercare that serves adoption pairs technical support with someone who walks the floor or sits in on calls. That person doesn't ask whether the system works. They ask what people are doing differently from last month, and whether it matches what training said. That second question exposes the gap between what training covered and what people really do, while there's still time and attention to close it. Once hypercare ends and the project team disbands, that visibility goes. The habits formed during hypercare then tend to become permanent. For what to track beyond the ticket queue, see our guide to the adoption metrics that matter.

Building the capability inside the project team

ERP and S/4HANA programs are usually staffed with functional consultants, technical architects and a project manager. Change management gets added as a smaller workstream. Sometimes it's one person covering communications and training for the whole program. That's thin coverage for an ERP's scope. Programs that hold up spread change capability across several roles on the project team. One person can't be in the warehouse, at the finance close and on the sales floor in the same week. The project manager is a natural place to start. We cover what a PMP® holder already brings to change management in a separate article. The same logic applies at company level, when the goal is to build internal change capability that outlasts a single project.

Leanova Academy's CCMP® 21 hour self paced training gives one project team member a structured foundation to take this work on formally. For a full project team, a private cohort trains up to ten people together on the ERP scenarios your program will actually face. The functional leads come away with a shared vocabulary instead of relying on a single certified specialist working alone.

For change management across technology transformations more broadly, see Change Management in IT and Digital Transformations. For what happens once hypercare ends and the project team moves on, see Sustaining Change After Go-Live: Hypercare and Reinforcement.

This article is also available in French: Conduite du changement pour les migrations ERP et S/4HANA.

Reviewed September 2026. This article gives no ERP project failure rates, adoption statistics or industry research findings, because Leanova could not independently verify any of them at the time of writing. CCMP® is a registered trademark of ACMP®. PMP® is a trademark of its respective owner.