*Why the 70% failure rate and the burning platform don't hold up, and which levers do: effective sponsors, real involvement, measured success.*

Prefer to listen in English? You can find the English version on Spotify here:
https://open.spotify.com/episode/6OWWsIfy1DK6666SMmQMZ7?si=SL0oxwY3S-qc6FidhirJtA

Change Folklore vs. Evidence: What Really Works in Transformation

Seventy per cent of all change initiatives fail. People only move when the platform is burning. The right bonus will get everyone on board. These ideas show up in strategy papers, steering committee decks and kickoff speeches. Measured against the research, several of them don’t hold up.

Prof. Dr. Renate Osterchrist has looked at change from three sides: in-house in HR development at Heraeus, for ten years as a consultant at McKinsey, and today as Professor of Leadership, Change Management and Human Resources at Munich University of Applied Sciences. Her starting point was a 2024 review by Hagl and colleagues that brings together 30 years of peer-reviewed research on change interventions, 119 studies in total. Practitioner data was missing, because firms like Prosci, McKinsey and BCG don’t publish in peer-reviewed journals. Osterchrist added it, together with her own interviews in companies.

The result is a toolkit for anyone who has to explain to a board why a change intervention is worth the money, and why the executive team has to step up. Episode 20 of A Change in Conversation goes through what survives the test and what doesn’t.

The 70 per cent nobody ever measured

The best-known number in change management goes back to 1993. In their book on re-engineering, Hammer and Champy estimated that up to 70 per cent of efforts miss their intended results. It was a guess from experience, and two years later Hammer himself complained that it was being misread. The number stayed anyway, and it is often wrongly attributed to John Kotter.

Osterchrist knows the consulting side well and has a simple explanation for its staying power: it sells. Look how many fail, and here is how we do it better.

The real issue is how success gets defined. BCG reported in 2024 that only 30 per cent of companies fully meet timeline, budget and scope in large tech programmes. Hitting all three corners of the project triangle at once is rarely realistic, so Osterchrist reads 30 per cent as a remarkably good result. Another 35 per cent or so get most of the way there and are doing fine. Nobody mentions them. In Prosci’s public data, 40 per cent of organisations say they met or exceeded their goals. Leave out those who say it is too early to tell, and the figure rises to 55 per cent.

So decide what success means before the programme starts, and treat “too early to tell” as an honest answer.

Fear is a weak motivator

The burning platform, made famous by John Kotter’s eight steps from 1995, is still the go-to opening in many change plans. Studies have not been able to confirm that it works, and Kotter has since moved on from it.

Complacency is real. But people need a light at the end of the tunnel more than a fire behind them, a reason why the tunnel is worth going through. Medicine shows the limits of fear: doctors warn patients about stroke risk every day, and many still don’t change how they live. What moves people is a clear answer to what they gain.

Not every change comes with good news for everyone, and leaders have to be honest about that. What helps is explaining the why and then making progress visible early. Quick wins build momentum and give people the sense that their effort pays off.

Why bonuses are a weak lever

The clearest gap between research and consulting practice is pay. Strategy consultants often treat incentive systems as a driver of change. Research dismissed that long ago. Bonuses narrow attention and create tunnel vision rather than creative ideas. In 30 years, Osterchrist has not come across a compensation system that rewards individual high performers and encourages real collaboration at the same time. Good communication from leaders has a stronger effect on motivation during change than financial rewards.

A mediocre sponsor is not good enough

Prosci’s data shows a clear pattern. Successful transformations have far more sponsors who are seen as effective. Failed ones have more sponsors who are only somewhat effective or not effective at all. Being visible is not enough.

Effective does not mean doing everything. The executive team explains the strategy and where the business is going. Middle managers, who in Osterchrist’s view are blamed far too often and unfairly, translate that into what it means for their own team, because employees want personal messages from their direct supervisor. Programme leads make that as easy as possible for leadership.

And when leadership keeps putting other things first? Osterchrist’s advice from experience is to say so openly and stop the initiative, rather than pushing it through half-baked.

How many people need to be involved

Two McKinsey studies from 2021 and 2024 suggest that around 7 per cent of the workforce needs an active role for a transformation to reach critical mass. In a company of 10,000, that is 700 people. Sitting in a town hall doesn’t count. With leaders on board and a network of change agents, the number is within reach.

Osterchrist treats it as a rough guide. In large software rollouts, her interviews point to quality over quantity: a smaller group of well-chosen key users who know the content beats a large network of change agents who don’t.

Engagement theatre and the playing field

Involvement can backfire when it is staged. People are invited to shape everything and later learn that the outcome was fixed all along. The answer is to define the playing field up front. Everyone knows how long and how wide the pitch is, and nobody plays in the stands. How the game is played inside those lines is open for discussion.

A useful tool here is the delegation continuum, an old leadership framework going back to Tannenbaum. In one case, an automotive supplier with 15,000 employees used it to prepare a workshop for its top 50 leaders, including topics such as plant closures. It made four things explicit: what is decided and only needs explaining, where input is wanted before the CEO decides, where a compromise is worked out together, and where leaders get a direction and decide on the implementation themselves. It also forced the CEO to get clear for himself where he actually wanted debate.

In international programmes, local leaders play a double role. They see themselves at the top of their country, but from a global view they are part of the cascade. They need to be involved, and they need a feedback loop that keeps the cornerstones consistent while leaving room for local variation.

AI: the same levers under more pressure

For AI transformation, Osterchrist sees the same six areas of intervention at work: communication, support, involvement, reinforcement, social influence and coercion. A few things are sharper. Companies introduce AI to save time but rarely say what that time is for. The pace leaves many people breathless. And trust is fragile: inviting people to experiment and then replacing them with an agent works once or twice. After that, nobody experiments anymore. Clarity, such as switching off old tools, is accepted. Manipulation is not.

The blind spot: measuring success

One gap runs through the studies and the interviews alike. Change practitioners want to measure more, but leadership often doesn’t sign off. The top team isn’t aligned on what it wants to achieve, and measurement takes effort. That leaves programmes without the data they need to work iteratively.

The comparison with medicine fits here too. A doctor tries one treatment, checks the blood test two weeks later, and adjusts the dose. Change work too often uses a watering can instead: more of the same for everyone, with little idea of what is working. Without a data foundation, the discipline risks losing the backing it needs.

Key Takeaways: What Leaders Should Remember

Define success before kickoff, and allow for “too early to tell”.

Replace the burning platform with a credible upside and early quick wins.

Don’t rely on bonuses. Leaders who communicate well move people more.

Insist on an effective sponsor. If priorities lie elsewhere, say so early.

Give enough people a real role, choose them well, and be clear about what is open for debate.

Measure as you go, so you can do more of what works and less of what doesn’t.

Listen to the full episode

Renate Osterchrist and Arne Kötting recorded this conversation in German. An AI-dubbed English version is also available. Renate’s summary of her findings and the underlying study are linked in the show notes.