Organisations devote enormous attention to the visible architecture of change, e.g. strategies, targets, transformation programmes. But many of the forces that determine whether change succeeds are much harder to see.
Who do people actually listen to? Where does trust sit? Which relationships allow ideas to travel across an organisation Who has influence without appearing anywhere near the top of the organisation chart?
Helen Bevan has spent more than three decades working on large-scale change, predominantly within health and care systems. She believes leaders seeking to change complex organisations need to become much better at seeing this hidden organisation.
“The theme that actually links all of this,” she says, “is about making the invisible visible.”
And the most important place to start, according to Beven, is power.
Your most influential people may not be your most senior people
Bevan defines power as “the capacity to produce a change”.
The form of power organisations most readily recognise is formal authority. In what Bevan describes as an “old power” system, authority is attached to hierarchy and influence flows downwards. Position, title and location on the org chart determine who is presumed to have the ability to make things happen.
But there is another form of power operating alongside it.
Informal power is relational. It comes from respect, credibility and connection rather than rank, with its influence travelling peer-to-peer rather than through a formal cascade.
The people occupying the two systems aren’t necessarily the same.
Bevan points to an organisational network analysis which showed that a relatively small number of people can sit at the centre of surprisingly large informal networks. She calls them “super connectors”.
Citing data from the organisational network analysis firm Innovisor, she says formal leaders typically make up around 12% of an organisational population and collectively reach somewhere around half of it. By contrast, the most influential informal leaders may constitute only around 3%, yet reach a far larger proportion of colleagues through their relationships.
“If we’re trying to make change happen in our organisations, it’s very, very helpful to find out who are the informal leaders,” she says.
Yet senior leaders often have no idea who they are.
“When you actually do the organisational network analysis and the formal leaders find out, they are really shocked because they don’t expect it to be those people.”
When a leadership team launches a new strategy or change initiative, another conversation is already happening elsewhere. Employees are testing the ‘official’ story against what trusted colleagues think. They are comparing notes on what they have been told versus previous lived experience, with some people actually become interpreters of the change for everybody around them.
While those people can accelerate a transformation, they can also quietly undermine one.
“If they agree with the change, they’re going to accelerate it,” Bevan says. “If they don’t agree with it, they’ve got their own narrative and different way of thinking about it and they’re going to block it.”
Her recommendation: find your 3%.
And don’t assume that asking for volunteers will do it.
Bevan is particularly sceptical of the familiar organisational practice of inviting people to self-nominate as “change champions”.
“I would never ask people to self-nominate to be change champions,” she says. “The people that self-nominate are the enthusiastic people that haven’t got any influence.”
Instead, ask peers who they trust, respect and turn to when they need advice. The names that recur are likely to tell you considerably more about where influence really sits.
Identifying those people is only the beginning.
Bevan argues that they need to become insiders and co-creators rather than just act another communications channel for a decision that has already been made.
For example, when an influential person is sceptical, the instinct may be to avoid them. Yet Bevan suggests taking the opposite approach.
Build the relationship. Find out what is driving their response. Perhaps most importantly, treat their scepticism as information.
And if leaders choose to involve influential people, the relationship cannot disappear as soon as the initiative has been launched.
“Informal power is all about relationships,” Bevan says. “So you have to keep the relationship.”
Bevan would ban one word from leadership
“If there’s one word that I would choose to ban in the leadership vocabulary,” says Bevan, “I would choose to ban the word empowerment.”
Which may seem an unusual position from somebody arguing for more distributed power. But her objection is to what the word can conceal.
In most organisations, empowerment still implies that somebody with greater formal authority is granting another person permission to act, with the boundaries of that authority defined by the person giving it.
And what has been given can, ultimately, be taken back.
“Very often, when we’re in a very hierarchical situation and we say to people, ‘You are empowered’, people are empowered under the terms and control of the senior leader.”
Bevan prefers agency.
Agency means feeling that you have the capacity and ability to act, make choices and influence outcomes yourself.
So what happens when organisations enter the age of agentic AI?
“We need agentic people too,” Bevan says.
As technology assumes more routine and predictable work, organisations will need people capable of sensing what is happening around them and then exercising judgement and acting without waiting for somebody further up the hierarchy to issue an instruction.
But organisations can easily undermine exactly that capability.
For example, an organisation invests heavily in improvement or change training. Employees learn new methods and tools, but then return to precisely the same working environment they left.
“They’ve got the skills, but they haven’t got a sense of agency,” she says. “They think, ‘Well, I can’t really improve anything because I haven’t got permission.’”
Agency, in Bevan’s view, cannot simply be bestowed by management.
“I don’t think as leaders that we can give people agency. Agency is something that people have to build for themselves, but we as leaders create the conditions for it.”
That requires leaders to think differently about how change begins.
Stop arriving with a cathedral
Bevan uses an image to describe one of the most common mistakes.
The cathedral.
A small group develops an idea. They refine it and construct the business case and then build the proposal and work through the detail until it resembles something complete.
By the time other people encounter it, considerable intellectual and emotional investment has already gone into the design.
The task then becomes getting everybody else to “buy in”.
Bevan thinks this is backwards.
“A cathedral [is] a complete and fully formed idea that we’ve written the business case or we’ve done the proposal, so we’re very emotionally invested in it and we’re very attached to it.”
The more finished the cathedral becomes, the less room remains for anybody else to contribute meaningfully.
Her alternative is to bring a brick.
“Create the space for every person to contribute,” she says. “If everybody can contribute right from the beginning, we help them to have an emotional connection both to the collaborative process and you help people to build their agency.”
Too often, leaders overlook that participation after a decision has largely been made isn’t the same as participation in making it. If leaders want people to feel ownership, then other people need to be able to affect what is being built.
“Brick by brick,” Bevan says, “we start to create something better than what one person would have done alone.”
Skills are not enough if people cannot connect
Bevan recalls an attempt to bring the Virginia Mason healthcare improvement approach into the NHS in England. Five hospital systems participated in a multi-year experiment. They received the same broad intervention: support, investment and access to learning.
Yet the outcomes massively diverged. Three improved, but two went backwards.
Bevan describes research led by Dr Nicola Burgess at Warwick Business School examining what separated the organisations. The standout difference, she says, was the extent of the social connections between people leading improvement locally.
One organisation showed a dense network of relationships, with people learning and sharing across the system. Another had a much more fragmented network.
“Social capital is more important than human capital,” says Bevan.
Human capital is what organisations typically invest in, i.e. skills, experience, knowledge, professional development and technical competence.
All of those matter, of course, but social capital is different. It’s the value created through relationships. It’s when people trust each another and knowledge can move across organisational boundaries? Or when somebody encounters a problem, they know who else might have solved something similar. Or groups that normally operate independently have enough connection to learn from one another.
A workforce can contain highly capable individuals and still struggle to change collectively.
Which isn’t to say individual capability is unimportant, but capability trapped inside organisational silos has limited reach.
“The strongest influence on what people do and think is what the people around them do and think,” says Bevan.
From a communications perspective, leaders often assume that better comms will change opinions. Bevan is more cautious.
“People don’t change their opinions as a result of effective communication,” she says. “People change their opinion, their mind, as a result of what their peers think.”
Leaders should invest deliberately in networks, recommends Bevan.
Create time for people to learn together. Connect change activists working in different parts of the organisation. Build bridges between groups that rarely interact. Bevan even points to “randomised coffee trials”, where colleagues are paired simply to have a conversation with somebody outside their usual network.
These interventions can sound almost trivial next to a major transformation programme, but that may be because the formal programme is visible and relationships are not.
Most change work happens above the waterline
Bevan introduced another way of looking at the same problem.
An iceberg.
Above the waterline sit the parts of organisational change leaders can readily observe, i.e. events, data, patterns, policies, processes and practices.
These attract attention because they can be measured and managed, but below the surface sit some of the less visible forces, i.e. relationships, connections, power and mental models.
Mental models are particularly important, which Bevan describes as “the assumptions that we make about how things are and why they happen”.
These assumptions influence behaviour without necessarily being consciously examined.
Bevan uses the Waters of System Change framework, developed from work by Peter Senge and others, to identify six conditions that can hold a problem in place: policies, practices, resource flows, relationships and connections, power dynamics and mental models.
The further leaders move into the less visible conditions, she argues, the greater the potential leverage for deeper change.
Yet organisations frequently concentrate their energy in the opposite direction.
“Where do we think the big levers for change are?” she asks. “Very often we think they’re above the waterline because that’s where we spend 80% of our efforts in change on the things above the waterline.”
The more powerful leverage points, she argues, are often underneath it.
Consider a process problem.
The instinctive response is to redesign the process, and sometimes that is exactly what is required.
But what if the real constraint is a relationship between two functions that do not trust one another? What if the process persists because one group holds decision-making power it is unwilling to relinquish? What if everybody is operating from an assumption that nobody has ever thought to challenge?
A technically sensible intervention can leave all those conditions untouched.
Bevan uses obesity as an example from healthcare.
A system looking only above the waterline might concentrate on treatments, programmes and clinical pathways. But underneath sit deeper assumptions about responsibility.
If obesity is understood primarily as individual failure — people simply making poor choices or failing to look after themselves — that mental model affects the policies, investment and interventions that follow.
Change the underlying assumption and a different set of possibilities becomes visible.
The same principle appeared when participants at the IMI session applied the framework to very different problems: AI transformation in biotechnology, an acquisition by private equity, declining school attendance, collective bargaining, legal-sector engagement and the introduction of aesthetics into an established wellness business.
Very different subjects exposed the same pattern.
While the first answers often concerned process, the deeper discussion concerned relationships, power and assumptions.
An uncomfortable question
There is, however, a second part to Bevan’s approach.
Once leaders have identified everything in the system that needs to change, she asks them to turn the analysis back on themselves.
“To what extent is the way that we as a team or organisation operate contributing to the conditions holding the problem in place?”
Bevan recently used the question with senior leaders in the Danish health system. They had identified numerous external obstacles. But the next task was to examine what the health authority itself might be doing to reinforce them.
She asks the same question of leadership teams, policy teams and change teams.
Perhaps the organisation wants greater collaboration while its own structures reward siloed performance. Perhaps it wants people to speak openly while senior leaders react badly to challenge. Perhaps it wants distributed decision-making while significant choices still have to travel upwards or perhaps it wants people to act with agency while designing every transformation as a finished cathedral.
“What we find time and time again,” Bevan says, “it’s not just the external world that’s got to change. We’ve got to change.”
That is a substantially more demanding form of change leadership.
It requires leaders to examine their own mental models, their own relationships and their own use of power.
To ask not only who is blocking this?, but what are we doing that allows the blockage to continue?
Not only why won’t people change?, but what conditions have we created around them?
And not only how do we get people behind our solution?, but what might they see that we cannot?
Ultimately, Bevan’s argument places less faith in the heroic change leader and more in the quality of the system around them.
Yes, formal authority still matters. And plans still matter. So does process, resources and technical expertise.
But none tell the whole story.
Because organisations run on influence that never appears on an organisation chart. They run on assumptions nobody has written down and relationships no dashboard is measuring.
And those forces are shaping change whether leaders choose to see them or not.
The challenge, as Bevan puts it, is to “hold the mirror up to ourselves”.
Because sometimes the most important part of changing a system is discovering the part of the system you have been looking through rather than looking at.
