It's your experience, not your expertise, that makes you a great leader.
Leadership has always been demanding. Maybe that's amplified now. When was the last time you had time to think? That your nervous system felt relaxed, not scanning for the next thing to do, to achieve. That you felt there was time and space to connect with your team, your family, properly. Not just being there but being present.
For over a decade I've been working with leaders navigating significant transitions: a new role, a strategy change, an organisation that's moved under their feet. Lately I've realised, all leaders feel in a kind of transition. The ground is moving for everyone, and I wonder if we're prepared.
The CMI surveyed over 4,500 UK workers and managers. Eighty-two per cent of those entering management positions had received no formal management or leadership training. One in four people in the UK workforce holds a management role. Twenty-seven per cent of workers describe their manager as highly effective.
These aren't figures about bad people in wrong jobs. They describe a system that promotes on technical competence and then expects something structurally different to emerge, largely unaided.
For most of the last century, expertise was enough. When AI enters that picture, it doesn't just change the technical landscape. Decades of accumulated knowledge, and a machine closes the gap in months. That's unsettling for leaders : it feels like an opportunity and a huge risk. It feels like there might be a genuine change to the value leaders bring. Hard earned expertise may no longer be enough, which feels unfair in a way worth naming.
Maybe, what remains when AI takes on the technical layer is the work leadership was always supposed to be. Reading situations accurately. Holding complexity. Making judgements that need context and consequence. Building conditions in which other people and AI can do their best work. That work was never about expertise. It was always about how you show up.
The WEF's Future of Jobs research, drawing on over 1,000 employers and more than 14 million workers, identifies the skills that will matter most through 2030. Analytical thinking is considered essential by seven in ten employers. Resilience, flexibility and agility rank second. Then leadership and social influence.
As AI takes on more of the technical and analytical layer, what determines leadership effectiveness is the human layer. These aren't soft skills. They're the primary source of leadership value in an AI-shifted world. And unlike expertise, they can't be replicated.
Technical competence gets someone promoted. The assumption, rarely stated but consistently acted on, is that leadership capability will follow. The 70:20:10 model was built on a sound insight: most development happens in the work itself, not in the classroom. But despite being quoted as established fact across the profession for decades, DeRue and Myers found there's actually no empirical evidence supporting it.
In practice, the 70, the on-the-job learning, defaults to expertise accumulation. Leaders learn what the domain requires, how to navigate the organisation's politics, how to deliver results. What they rarely develop is awareness of the conditions they're creating around them, the capacity to read the situation accurately, or an understanding of what their habituated responses do to the people they lead.
Robert Kegan's work at Harvard identifies distinct stages in how adults make meaning. The later stages aren't about abandoning what worked before. What develops alongside the expertise is the capacity to question your own assumptions, to hold ambiguity without rushing to resolution, and to lead through conditions that don't have clear answers. Most senior roles now require that capacity as a baseline. Most leaders have never been supported to develop it deliberately.
The third problem is the model of leadership itself. The hero model (the individual who knows most, decides fastest and carries most) is baked into how we select leaders, how we promote them, and how we measure their success. It produces leaders who are highly capable as individuals and significantly underdeveloped as architects of the systems they're supposed to be leading.
Go back to the late nineteenth century and the stationary steam boiler. When electricity arrived, factory owners made a logical but disastrous mistake. They replaced the boiler with a giant electric motor but kept the overhead shafts and leather belts exactly as they were. Productivity didn't move. Productivity only exploded years later when leaders moved to unit drive: a small, independent motor on every machine. The pattern repeats. Organisations plug AI into unclear strategies, siloed structures and disengaged workforces, and wonder why the transformation stalls.
McKinsey's State of AI research shows 74 per cent of organisations struggle to scale AI value, despite 65 per cent now using generative AI regularly. The RAND Corporation attributes 84 per cent of AI project failures to leadership and organisational causes, not technical ones. Bain puts transformation failure rates at 88 per cent.
The pattern across all three is the same. The obstacle isn't the technology.
Consider what leadership, done well, is actually producing. Not a set of tasks. Not a series of decisions. Something more like the conditions in which other people can do their best work. At senior level, that tends to be the job.
When strong, accelerates the right work. A clear strategy scales into faster, more integrated execution. When weak, accelerates the wrong work. AI gets applied to problems that haven't been clearly defined. A faster mess.
When strong, connects the dots. Reduces friction, spots patterns, predicts issues across the value chain. Data flows because people do. When weak, amplifies the dysfunction. Competing tools. Political battles now have a technical dimension too.
When strong, amplifies human judgement. Creates powerful human-in-the-loop systems. When weak, amplifies the resistance. The quiet sabotage. The shadow IT. People who aren't invested in the organisation aren't invested in its AI either.
When strong, removes friction from work that already flows. When weak, adds tools to a system that already obstructs. New technology layers onto existing dysfunction. People lack the permissions or data to use it effectively.
The four conditions do two things at once. They describe the sequence that anything significant has to travel through to become real. Strategy has to be clear before teams can genuinely align around it. Alignment has to exist before engagement is real rather than performed. Real engagement has to exist before tools and permissions to act actually land. You can't skip a stage.
Building these conditions isn't HR's job or IT's job. Those functions can provide expertise and resource. But building and maintaining the environment in which an organisation can execute its strategy is the work of leadership. It always has been.
The human brain under stress doesn't behave the way we need it to at senior level. Under acute pressure the prefrontal cortex (the part responsible for judgement, reasoning and reading complex situations) goes offline. The habit-driven part takes over. We stop responding to what's actually in front of us and start reacting to the pressure itself.
Under sustained pressure something else happens. The brain narrows its attention onto immediate tasks and problems. What gets lost is peripheral vision, not strategic peripheral vision but human peripheral vision. The people around you stop registering as fully as they should. This isn't callousness. It's simply that sustained pressure removes the capacity to see clearly.
Professor Amy Arnsten of Yale School of Medicine has spent decades studying what stress does to the brain. Her finding, documented in Nature Neuroscience, is direct. Under acute stress, the prefrontal cortex goes offline: the part of the brain needed most at senior level. Under pressure, the brain switches from thoughtful to automatic. The habit-driven part takes over.
A leader under sustained pressure isn't underperforming through lack of effort or capability. The thinking the role requires has been switched off by the pressure itself. Arnsten notes that awareness of this mechanism makes a practical difference. Knowing what's happening, and having the capacity to pause, can restore the function that pressure has suppressed.
Scott Dick, General Manager, Breedon Group Scotland
Cranfield General Management Programme
Scott came to a Cranfield General Management Programme with a pattern many senior leaders will recognise. His senior team was working evenings, weekends and holidays as routine. Too many decisions were escalating to the wrong level. Firefighting had crowded out strategic thinking.
The reframe that shifted things was simple: behaviour is a function of person and environment. His job was to shape both. Not with a transformation initiative. With small, visible changes and short feedback loops. Clearer decision rights. Short focused check-ins replacing sprawling meetings. Protected thinking time. Boundaries on out-of-hours contact as the default, not the exception.
Escalations to GM level fell from seven to three per week. Right-level decisions went from 52 per cent to 80 per cent. After-hours emails fell from 72 to 23 per week. By October 2025, Breedon Scotland's underlying EBIT was up 26 per cent on revenue that was down 6 per cent. Profitable growth in a difficult market, achieved not by pushing harder but by removing the friction consuming the organisation's energy before it ever reached the work.
"The team described it simply: calmer and faster."
Scott's story is an example of a leader giving himself, and his team, permission to regulate. To create the space and pace that allowed people to actually think. The results followed.
That permission doesn't always exist. When it doesn't, behaviour from the top amplifies conditions all the way down the system. Leaders under pressure create pressurised environments. People disconnect from the purpose of their work, focused instead on the immediate, the urgent, the task in front of them. Various pressure fronts create their own weather. The unintended consequences land on the front line, the end user, the customer, in ways nobody at the top ever intended or can easily see.
I was twelve years old when I first understood what systems do to people when no one in authority is paying attention. I'd snapped both the tibia and fibula in my left leg. After the surgery I was on a children's ward. Over the following days I watched what happened when the adults in charge had stopped seeing the humans in front of them.
What stayed with me wasn't the incident itself. It was the ordinariness of it. These weren't bad people. They were people under pressure, focused on time and tasks and the next thing, who had stopped noticing the impact they were having. I've seen this pattern in every industry, in every country, at every level of seniority in the decades since. It almost always starts the same way. Not with bad intent. With drift.
Three things make progress possible. Knowing your own habituated reactions under pressure, so that you can catch them before they land. Understanding the context you're operating in with enough precision to see what the existing conditions are already producing. Having a clear practice for how to show up as a senior leader in a way that develops rather than depletes the conditions around you.
By managing your own pressure, you give yourself the presence of mind to respond to the situation at hand and the people in front of you at the appropriate pace. By working on the conditions around you in this way, you in turn relieve the pressure system for others. The weather changes.
The practiceThere's substantial evidence about what determines whether senior leaders can actually execute at the level their roles require. It points consistently at five areas: clarity of intent, quality of human and political understanding, accuracy in reading the context, steadiness of presence, and the discipline to set pace rather than be set by it.
The strategic story is the foundation. Without it everything else is effort without direction. When Purpose is genuinely shared, the organisation moves with something closer to collective intelligence. When it's drifted, or was never quite real, the leader becomes the bottleneck.
Leadership is exercised through and with people. Research by Ferris and colleagues consistently finds that political skill (the ability to understand others at work and use that understanding to act effectively) is one of the strongest predictors of leadership performance.
The room you think you're in and the room you're actually in aren't always the same. Reading Place accurately requires a degree of stillness: the capacity to observe before acting, to resist the pull of the familiar response and ask what this particular situation is actually asking for.
How you show up, what you attend to, and the altitude at which you operate are signals the organisation reads continuously. A longitudinal study tracking nearly 6,000 employees found that manager stress transmits directly to employees, with measurable effects detectable a full year after the initial transmission.
Speed isn't the same as momentum. Research at Georgia Tech found that when leaders demonstrate patience it significantly boosts their reports' creativity, collaboration and productivity. Choosing not to move, when everything is pushing for movement, is one of the hardest and most underrated things a senior leader can do.
This may feel like a lot. A lot of time and attention. A lot to pull together. Most leaders have what they need to do this, if they give themselves the time and permission to do so. They've built the experienced judgement, the pattern recognition, the ability to read people and situations. Often what gets in the way is not capability. It's capacity: the headspace to pay attention effectively.
That starts with a shift in how you understand the job. Not "I am leading this team, this project, this outcome": role-defined, finishable, tied to a specific result. But "I am a leader." Identity-defined. Practice-anchored. Ongoing. The first asks: am I done? The second asks: what does this situation require of me now?
Leadership has always needed this framing. Organisations that reward heroics (the leader who swoops in, solves the crisis, carries the load) are not rewarding leadership. They are rewarding a particularly visible form of individual contribution. There is also an internal reward: the dopamine hit of putting out fires is real. The sense of urgency, the visible result, the feeling of being needed. It is just not sustainable, for the leader, or for the people around them. The conditions that produced the crisis remain untouched, producing the next one.
Leadership as a side hustle, something to attend to once the urgent stuff allows, produces the same result. Whack-a-mole. Fires extinguished, symptoms treated, the same problems returning in new forms because the conditions generating them were never the focus.
AI does not create this problem. It makes visible the cost of it. The fires become more frequent. The gap between what the situation requires and what habituated leadership provides becomes harder to paper over.
So how does it come together? Think in two halves. The first is clarity of context: reading the conditions you are operating in, understanding what they are producing, and seeing the situation as it is rather than as you expect it to be. The second is managing how you respond: the decisions you make, the actions you take, and the quality of leadership you bring to both.
The Leadership Response Cycle™ is the framework I use to underpin my work with leaders and teams. It pulls these concepts together into a structure for continuous practice. Not a one-time fix. A discipline that runs beneath everything else.
Explore what's actually happening. Analyse why. Decide on the fewest things that will make the biggest difference. Act, test, learn. Left side is clarity. Right side is response. It runs anticlockwise, and then it begins again.
The cycle is the scaffold. The bones of leadership practice. OrgBarometer™ supports the left side: a structured read of the conditions you are actually operating in across Define, Align, Engage and Enable. Autopilot™ supports the right side: awareness of your own habituated patterns under pressure, so that what you decide and how you act comes from deliberate choice rather than reflex. The 5Ps are the constant across both. They are the quality of leadership the cycle runs on: how you show up to read the situation honestly, and how you show up to act on what you find.
Not a tool to pick up when needed. The underlying infrastructure of leadership, running continuously beneath everything else.
That is why it's hard. The difficulty is rarely about capability. It's about being asked to see yourself and your contribution differently.
It's common for leaders to think they got promoted for what they do and continue to do what they did before, now in a bigger job. Professor Cliff Bowman of Cranfield University describes this mismatch. He calls it negative delegation. It conjures an image of strategic leaders sucking up the work, over-controlling and working hard to cast an unproductive shadow across their organisation.
Leaders have always needed to lead the people, not do their work. Things change too fast now. As leaders, as a leadership development community, we need to more clearly name and appreciate that each leadership promotion requires a change in focus and identity. We do ourselves and others a disservice by shying away from that.
Especially as AI is coming for your leaders' expertise. It's their experience it cannot replicate.
This can't be imported or accelerated. Someone who knows the technology without yet having the experience simply doesn't have it. Years and consequence are the only route.
The issue is rarely that leaders lack what this moment requires. They tend to have it. What gets in the way is misdirection: attention fixed on the expertise, on the visible and nameable part of the asset, while the thing actually doing the work goes unrecognised and underused. That is what this report is about.
Do you see yourself as a leader? Not as someone leading a team, a project, an outcome. As a leader, for whom this practice and this cycle is the job. That is the starting point for everything else.
How clear is your Define? Is the strategy operable, or aspirational without being something people can act on? A useful test: ask five people at different levels to describe the strategy in their own words. The range of answers tends to tell you what you need to know.
How functional is your Align? Do teams work across boundaries or defend their patch? When decisions are made in a meeting, do they stick outside it?
How honest are you being about Engage? Are people genuinely invested, or performing investment? Engagement tends to be earned through conditions rather than manufactured through communications.
How well enabled is your organisation? Do people have what they need to get the job done, or are they navigating the same obstacles repeatedly?
Can you say clearly what you're actually for in this role right now, not what the job description says, but what this particular moment requires of you? Are you reading the organisation as it is, or as you expect it to be? How much of what reaches you has been filtered before it arrives? What are you paying attention to right now, and what are you missing because of it?
There is no time. Things need to move. Moving fast on unclear foundations tends not to accelerate progress. It tends to accelerate the mess. There's a version of speed that comes from clarity rather than urgency. The two feel very different from the inside.
I've done leadership development. It hasn't made much difference. The question isn't whether you've been on a programme. It's whether the conditions you're currently creating are the conditions for good work. Working with the actual conditions, and the actual patterns running under pressure, is a different proposition.
Naming the problem feels exposing. This is the most understandable resistance and often the most costly. What goes unnamed tends to be unmanageable. The leaders who build the most capable organisations aren't usually the ones who project confidence they don't feel.
Surely this is just common sense. It is. The gap between knowing what good leadership requires and doing it consistently, under pressure, in unfamiliar territory, is where most of the work actually lives.
Leadership has always been demanding. The conditions in which it's exercised have changed, and are changing further. AI is the most visible part of that change. It isn't the cause and it isn't the point.
What senior leadership actually requires has been consistent across decades of research: the ability to hold complexity, to read the system accurately, to stay regulated under pressure, to build the conditions in which other people can contribute fully. These aren't soft skills. They're the primary source of leadership value in an environment where the expertise layer is increasingly replicated by technology.
The prize isn't complicated. Organisations where people can contribute fully and things actually move. Better places to work that are also more productive, not despite each other, but because of each other.
An honest read of where you're starting from.
The Leadership Conditions Scorecard takes around eight minutes. It will give you a read across all four conditions, where they are strong, where there is drag, and what kind of support fits where you are right now. There is no bad result. There is only an honest one.
Arnsten, A.F.T. (2015). Stress weakens prefrontal networks: molecular insults to higher cognition. Nature Neuroscience, 18, 1376–1385.
Bain & Company (2024). Transformations that work. Boston: Bain & Company.
Bowman, C. (Cranfield University). Concept of negative delegation. Referenced in Cranfield General Management Programme materials.
Chartered Management Institute (2023). Better managers: harnessing the power of people management. London: CMI.
DeRue, D.S. and Myers, C.G. (2014). Leadership development: a review and agenda for future research. In D.V. Day (ed.), The Oxford Handbook of Leadership and Organizations. Oxford: Oxford University Press.
Ferris, G.R., Treadway, D.C., Kolodinsky, R.W., Hochwarter, W.A., Kacmar, C.J., Douglas, C. and Frink, D.D. (2005). Development and validation of the Political Skill Inventory. Journal of Management, 31(1), 126–152.
Institute for Employment Studies (2016). The relationship between UK management and leadership and productivity. Brighton: IES.
Kegan, R. (1994). In Over Our Heads: The Mental Demands of Modern Life. Cambridge, MA: Harvard University Press.
Kegan, R. and Lahey, L.L. (2009). Immunity to Change. Boston: Harvard Business Review Press.
McKinsey & Company (2025). The State of AI. New York: McKinsey & Company.
RAND Corporation (2024). Why AI projects fail: leadership and organisational causes. Santa Monica: RAND.
Sluss, D. (2020). Becoming a more patient leader. Harvard Business Review, September 2020.
Torbert, W.R. and Rooke, D. (2005). Seven transformations of leadership. Harvard Business Review, 83(4), 66–76.
World Economic Forum (2025). The Future of Jobs Report 2025. Geneva: WEF.
World Economic Forum and McKinsey Health Institute (2026). The Human Advantage: Stronger Brains in the Age of AI. Geneva: WEF.
Wold, H.C., Rugulies, R. and Buch, R. (2022). The contagious leader: a panel study on occupational stress transfer in a large Danish municipality. BMC Public Health, 22, 1876.