The figure has circulated in every steering committee for years: a large share of ERP programs miss the targets they originally promised — on budget, on time, or on realized value. With S/4HANA migrations the stakes are even higher, because the end of mainstream maintenance for legacy systems sets a hard deadline. And yet the outcome regularly surprises everyone — at go-live.

The green status that means nothing

Classic project reporting measures what is easy to measure: completed work packages, tested processes, training rates. All of it can show green while the real risks sit in the blind spot. The status report describes the plan — not the reality in the business units.

That's exactly where the gap opens. A process counts as "tested," but three departments already work around it. A training is "complete," but half the attendees still operate in the old logic. The reporting sees none of it.

Where the value actually leaks

The promised value of a transformation isn't created at cutover, but in the weeks after — when teams actually use the new standard process. If adoption collapses, so does the business case. And adoption is a human variable, not a technical one.

A rollout rarely fails on the technology. It fails because no one measures, in time, what people actually do.

The five early-warning signals

Across hundreds of projects the same patterns appear — months before they surface in the quarterly report:

  • Workaround drift: teams quietly build their own spreadsheets because the new process is slower than the old one.
  • Key-user overload: four or five people carry half the load — and are the single biggest risk.
  • Status–reality gap: official progress and what the front line experiences drift apart.
  • Data distrust: users don't trust the migrated data and keep shadow records.
  • Unit divergence: one site carries the rollout, another has long fallen behind — the average hides both.

Why dashboards don't show this

These signals live in the heads of the people who carry the rollout every day. They sit in no system, because no one asks for them systematically. One-off consultant interviews capture 20 to 30 voices after six weeks — too few, too late, too filtered.

The difference: surveying hundreds of stakeholders in a structured way — stream leads, key users, the business, external consultants — reveals the patterns while there's still time to course-correct. That's the core of Project Intelligence.

What has to change

The good news: every one of these signals is measurable weeks or months before go-live. It doesn't take another consulting loop — just a look at the human side of the transformation, data-based, anonymous, and fast enough to feed the next steering meeting. Do that, and you walk into the decision with facts instead of a gut feeling that gets expensive at go-live.