What Consultations About Transformation Conceal

70 percent of all transformations fail, or so the story goes since John Kotter first said it in 1996. Since then the figure has been quoted mainly by the very firms that sell transformation programmes. Yet it has never rested on solid evidence. Anyone responsible for such a programme should know what’s missing from the pitch.

Related:Why bigger tech budgets don’t guarantee an edge  /  Deloitte study exposes CIOs as the last operational thinkers

The figure that owns the industry

In 2011 change researcher Mark Hughes examined the five most-cited sources for the 70-percent claim in the Journal of Change Management. Hammer and Champy, Beer and Nohria, a Bain article, a McKinsey paper, Kotter. In every case he found the same pattern: the number appeared either without any data or with a reference to another source that likewise cited it without evidence.

The circle closes with the senders themselves. McKinsey’s own reports cite Kotter, Kotter cites his 1996 estimate. No baseline population, no definition of failure, no survey period-none of it has ever been disclosed. The best-known metric in transformation consulting is a claim that became fact through repetition.

That’s more than pedantry. Using an unsubstantiated failure rate to justify a programme sells urgency instead of diagnosis. And urgency is expensive.

Five Gaps in the Pitch

The German Association of Management Consultants (BDU) notes a remarkable trend for 2025: consulting projects now require far more substantial justification than before. Skepticism is growing, but it rarely targets the right areas. These five points are missing from almost every proposal because their absence is tacitly encouraged by the incentive model.

First, who will operate the system in year two. The business case ends at go-live. The real burden begins afterward: operations, maintenance, second-level support, the handful of people keeping the new target vision running for three years. If you only address operations after the project, you’ve raised the ownership question far too late. These costs rarely appear in proposals because they would shatter the savings target on the cover page. The math works-until the first year of operations proves otherwise.

Second, the half-life of the target vision. A 24-month program plans for a state that is often outdated by month 18. A new AI model, a regulatory change, a competitor with a different cost profile: the target vision ages while the program is still running. You reach a goal that is no longer the right one. It’s still counted as a success because the milestone plan is green.

Third, the change budget as a token gesture. Change management appears in the deck, usually at ten percent of the volume. The real work of engaging the people who will carry or block the new system costs multiples of that. Where this line item is thin, the risk silently shifts from the consultant to the client.

Fourth, vendor lock in the architecture diagram. The recommended platform is often the one the consultancy has a partner program with. That’s not inherently wrong. It becomes a problem when the dependency is sold as a neutral architecture decision and the exit costs only surface in year four.

Fifth, the gap between pilot and scale. A successful pilot proves that something can work, not that it will pay off. Between the showcase project with the A-team and the rollout across 40 sites lies the real transformation. That’s exactly where the presentation turns vague.

The most common transformation after two years looks like the old organization with new job titles.

Where the Math Fails in Germany

Most transformation methods originate in an Anglo-American context. They assume an organization that decides from the top down. In German reality, that’s not the case.

Here, a program rarely fails because of the technology. It fails because of co-determination, the works council, a workforce that never signed off on the target vision. A rollout that hasn’t planned for interest reconciliation comes to a standstill, no matter how clean the target architecture is. Stuffing this piece into the ten-percent change line item overlooks the single most expensive line in the budget.

This DACH-specific translation doesn’t deliver a global framework. It belongs at the start of planning, not in the escalation of month 14.

Four Questions to Ask Before You Sign

This isn’t an argument against consulting. Good advisors know every one of these gaps, and many point them out openly. They just usually do so after the contract is signed, because the incentive model rewards silence rather than candor. The leverage therefore lies with the client-before the ink dries.

Four questions applied to every proposal are enough to separate the wheat from the chaff. Who bears the operating load in year two, and at what price? How long is the target vision viable before it topples at month 18? What does exit from the recommended platform cost in year four? And where does codetermination appear as a distinct work package-budgeted and scheduled?

Each of these questions has either a solid answer or a dodge. If dodging is the norm, no plan exists-only a brochure. And brochures aren’t bought for seven-figure sums.

Frequently Asked Questions

More from the MBF Media Network

Image source: AI-generated (July 2026)