It is one of the most quoted statistics in management literature, and one of the most unmoving. McKinsey's transformation research, repeated across multiple cohorts of major change programmes, finds that around 70% of large-scale change programmes miss their stated goals. Different sectors, different decades, different consultancies running the change, the headline number stays roughly the same.
This is the part that should be uncomfortable. It is not a problem of knowing. The frameworks have been mature for thirty years. It is not a problem of investment: UK organisations now spend an average of £1,068 per employee per year on training (CIPD Learning at Work, 2024). It is also not a problem of intent: nobody launches a transformation hoping for failure.
It is a problem of behaviour. And behaviour is the layer most transformation programmes never seriously try to reach.
What McKinsey Actually Found
The McKinsey transformation work, including the survey research summarised in their Why Transformations Fail series, repeatedly identifies the same set of distinguishing features in the minority of programmes that succeed. Two are particularly relevant for our purposes:
- Front-line ownership of the change, not just executive sponsorship, but visible behaviour shifts in the layer of managers who actually deliver it.
- Capability building targeted at specific behaviours, not generic leadership courses, but a defined list of "this is what we now do differently" practised in realistic conditions.
Both of these are behaviour-layer interventions. Both are routinely missed in favour of communication, governance and process design, the visible parts of a transformation that look reassuring on a steering committee slide.
Why Behaviour Is The Hardest Layer
Two findings from the wider behavioural-science literature explain why the behaviour layer is so consistently skipped.
The first is the retention gap. Roediger and Karpicke (2006), in Psychological Science, showed that being tested on material increases long-term retention by around 50% compared with re-reading the same content. Active retrieval beats passive review, by a lot. Most transformation training is the second kind: presentations, slides, e-learning. The retention curve is brutal, and managers cannot enact behaviours they have already forgotten.
The second is the practice gap. Even when behaviours are clearly defined, very few programmes give managers structured opportunity to practise them in realistic conditions with feedback. In our own research, building on academic behaviour-change work from UCL, Cambridge and Bocconi, participants who learned a feedback skill via role-play practice scored approximately 20% higher on observed behaviour than those who learned the same content via passive modalities (slides, video). Self-rated confidence did not predict observed performance, a Dunning-Kruger pattern that we designed out of subsequent studies. The mechanism is straightforward: behaviour change is a motor skill before it is a cognitive one, and motor skills require reps.
What This Means For An Active Transformation
If you are running a programme right now, the practical implication is uncomfortable but useful: the parts of the programme most likely to be cut under deadline pressure are also the parts most strongly correlated with success. The communication push and the new policy document will happen. The behavioural rehearsal, the live, feedback-rich, friction-creating bit, almost always slips, because it is the hardest to schedule and the easiest to deprioritise.
Three questions worth asking your sponsors this week:
- Have we defined the specific behaviours we are asking the front-line manager layer to do differently? (If the answer is "lead better" or "be more inclusive," the answer is no.)
- Have those managers had structured practice of those behaviours in realistic, feedback-rich conditions, not just attended a session?
- What is our 90-day measurement, not of training delivery, but of behaviour observed?
What We Do About It
Our Change Resilience Lab exists for exactly this reason. We build a bespoke simulation around the actual transformation our clients are launching, staged with professional actors playing the affected stakeholder groups. The leadership team lives through the change in a safe environment first. Resistance, hidden alliances and political risks surface before the real launch, while there is still time to act.
Whether or not you work with us, the question worth carrying is simple: in your transformation, what is the actual mechanism by which any individual manager's day-to-day behaviour will change? If you cannot answer that, the 70% number is not a sector trend. It is a forecast.
Book a free 30-minute diagnostic call → or read about our research-backed approach.