Inside-out: what the research says about lasting leadership change

The inconvenient gap between knowing and being

There is a type of leader every experienced HR director recognises. They complete the 360. They attend the programme. They practise the behaviours in the workshop role-play. Six months later, under real pressure, they are indistinguishable from the person they were before the intervention started. The skill was acquired. The person did not change.

This is not a failure of willpower or intelligence. It is a structural problem in how most leadership development is designed. The intervention targets what a leader does; the part that needs to change is who a leader is. These are not the same target, and conflating them is the source of enormous waste in organisational development budgets.

A decade of doctoral research at the University of Johannesburg, which underpins the CapabilityFX approach, examined this gap directly: what distinguishes leaders who sustain behaviour change from those who revert? The answer is not the quality of the training content, the seniority of the coach, or the frequency of the feedback. It is whether new behaviour becomes integrated into the leader's sense of self.

What the research actually finds

The distinction the research draws is precise, and worth sitting with. There are two possible outcomes when a leader engages with development:

Skill addition. The leader learns a new behaviour, practises it, and can demonstrate it in conditions where they feel psychologically safe and externally supported. The skill sits alongside the self. It is learnt, not internalised.

Identity integration. The leader comes to see the new behaviour as an expression of who they are. Being a certain kind of leader, someone who holds accountability without micromanaging, or who sits with ambiguity without rushing to premature resolution, becomes central to their self-concept. The behaviour no longer requires external reinforcement.

The behavioural outputs of these two outcomes look identical in the short term. This is why organisations are deceived by post-programme surveys and 30-day follow-up assessments. The divergence appears later, specifically under conditions of stress, cognitive load, and reduced external observation. Under those conditions, the leader whose new behaviour is skill-additive reverts. The leader whose new behaviour is identity-integrated holds.

The mechanism is not complicated. When a person is under pressure, available cognitive and emotional resources contract. The self falls back on its most deeply held orientations: what it believes, how it interprets threat, what it considers its role to be, and what it feels entitled or obligated to do. If the new behaviour has been integrated into those orientations, it survives the pressure. If it has not, it is the first thing that goes.

This is the inside-out principle. Lasting change starts with who the leader is. External behaviour is the downstream expression of internal state. Development that works on the inside, on values, on narrative, on the leader's relationship to authority and challenge, produces behaviour that holds. Development that works only on the outside produces behaviour that lasts until the next hard quarter.

Why organisations design interventions that cannot deliver this

The inside-out principle is not controversial in the academic literature on adult development, identity formation, or leadership effectiveness. The more interesting question is: why do organisations persistently commission development that cannot produce it?

Three pressures explain most of it.

The packaging problem. Identity development does not fit neatly into a two-day offsite or a six-session coaching contract. It requires repeated engagement over time, conditions of genuine psychological safety, and skilled facilitation capable of working with what surfaces when a leader's assumptions are actually examined. This is harder to sell, harder to schedule, and harder to evaluate against quarterly OKRs.

The measurability problem. Pre- and post-programme competency surveys are good at detecting skill acquisition. They are not designed to measure whether a leader's self-concept has shifted. So organisations measure what they can measure, invest in what they can measure, and conclude that what the measurement shows is what matters. This is a category error, not a measurement problem.

The discomfort problem. Development that addresses who a leader is, rather than what they can do, will at some point require a leader to sit with the gap between their current self and the self the role demands. That is uncomfortable. Senior leaders are not accustomed to discomfort in a professional setting, and some organisations treat any sign of developmental discomfort as a facilitation failure rather than as evidence that real work is happening.

The result is a market saturated with interventions that are easy to commission, easy to measure, and insufficient to produce lasting change.

What reversion actually looks like

The pattern is observable enough that practitioners learn to recognise it quickly.

Consider a senior operations director at a regional logistics company who completed an executive leadership programme focussed on coaching skills. In the weeks following, her direct reports noted a real shift: she was asking more questions, holding back her opinion in team discussions, and visibly making space for others to solve problems. The 360 scores moved. The development investment appeared to have worked.

Four months later, during a period of serious supply chain pressure on her division, the shift unwound almost entirely. She took back control of decisions she had started to delegate. She reverted to telling rather than asking. She reassumed the role of problem-solver-in-chief.

When this was explored in a subsequent session, her account was revealing. She had not experienced the coaching behaviours as an expression of how she led. She had experienced them as a technique, something she applied when conditions were stable enough to afford it. When conditions stopped being stable, she stopped affording it. The skill was real; it just was not hers. It had not become part of how she understood her own leadership identity.

Contrast this with a second case: a divisional general manager in a professional services firm who had come to see his tendency to rescue his team from difficulty as not just a management habit but as a distortion of what good leadership meant to him. The development process had involved sustained reflection on why rescuing felt necessary, what it said about his relationship to competence, and what kind of leader he was trying to be for his people. The new behaviour, staying out of problems he had previously dived into, was not a technique. It was a conclusion.

When commercial pressure arrived in his division, his team noted that he held his new orientation. He still felt the pull to intervene, he said so directly. But the pull did not override a deeply considered position about who he was and what his people needed from him. Identity-integrated behaviour held under the exact conditions where skill-additive behaviour fails.

What designing for the inside actually involves

Development that works inside-out is not softer or more indulgent than skills-based development. In many respects it is more demanding. It asks leaders to examine assumptions they have held unchecked, to sit with uncertainty about whether their current identity serves their role, and to do that work in a professional context where most signals reward confidence over inquiry.

Practically, it involves several conditions that most standard programmes do not create:

Time and repetition. Identity does not shift in a single session. Effective inside-out development is structured over months, not days, with intervals that allow real-world application and reflection.

Exploration of values and narrative. What does this leader believe leadership is for? What story do they tell about who they are and where they came from? What does that story make it impossible for them to do? These are not philosophical indulgences; they are diagnostics for where development needs to work.

Attention to the leader's relationship to authority and challenge. How a leader responds when their authority is questioned, or when the situation exceeds their established competence, is a reliable indicator of where their identity needs to grow. Development that does not surface this material is working on the surface.

Skill transfer situated within identity development. This is the key structural move. Skill transfer is necessary, it is just not sufficient on its own. When a coaching conversation or a workshop exercise is framed not as "here is a technique" but as "here is what leaders who are committed to developing their people actually do, and is that who you are?", the integration probability rises sharply.

The DUAL model, which structures CapabilityFX's development approach across four stages, Discover, Understand, Accept, and Lead, is designed precisely around this sequencing. The early stages of the model are not about skill transfer at all; they are about creating the conditions under which a leader's current identity can be examined honestly. Skill transfer comes after, and within, that work. The 4D method operationalises this sequence in practice, ensuring that each engagement phase is doing the right developmental work at the right depth.

This matters for how engagements are structured. A development programme that skips straight to skill transfer, without investing in the Discover and Understand phases, is likely to produce the kind of short-term gains and long-term reversion that disappoints both participant and sponsor. The phases are not decorative; they are load-bearing.

The questions worth asking before you commission anything

If you are responsible for leadership development in your organisation, the inside-out principle suggests a useful set of diagnostic questions before signing off on the next programme.

What is this intervention actually designed to change? If the answer is primarily behavioural, with no structured attention to the assumptions and identity dimensions that drive behaviour, the conditions for reversion are already present.

How will we know if identity integration has occurred, not just skill acquisition? Post-survey data that measures demonstrated behaviour in the 30 days after a programme is not the answer. Observation under genuine pressure, over six to 12 months, is.

Is there space in this design for leaders to examine assumptions, not just practise techniques? If the answer is no, the design is working on the outside only.

If these questions surface gaps in your current approach, the use-cases section of this site includes applied examples of inside-out development in practice, and the services page describes how CapabilityFX structures engagements to address the whole picture. For organisations using assessments to anchor development, it is worth noting that both the Five Lens Development Platform (Ennea International's assessment) and the future-readiness assessment (Tomorrows Compass's tool, for which CapabilityFX is a licensed distributor) are designed to surface identity-level material, not just competency gaps. They are starting points for inside-out work, not replacements for it.

The person two years later

The standard pitch for leadership development is improvement. Better decisions, stronger teams, higher engagement scores. Those are legitimate outcomes, and none of this argues against pursuing them.

What the research adds is a time horizon. Outside-in development produces visible results quickly. Six months in, the scores have moved, the behaviours are present, the investment looks justified. Two years in, the question is whether the person has changed, or whether they have learnt a set of techniques that returned to dormancy when conditions demanded the real self.

Inside-out development produces leaders who are genuinely different people two years later. Not because they have acquired more techniques, but because who they are and how they understand their own leadership has shifted in a direction that holds. That is a different kind of result, on a different timeline, and it requires a different kind of development investment.

If that distinction matters to your organisation, we are worth talking to.

The leaders described here are representative composites drawn from patterns we observe in practice, not identifiable individuals.