Why the People at the Top Are the Last to Hear the Truth About Themselves

There is an uncomfortable paradox at the centre of senior leadership. The higher someone rises, the more consequential their self-knowledge becomes — and the less reliable it gets. Seniority does not sharpen self-perception. It systematically degrades it.

The mechanism is not mysterious. Power changes the incentives of everyone in the room. Direct reports learn quickly which observations are welcome and which are expensive. Peers compete. Boards see a performance rather than a working week. Over time, a leader receives a version of reality that has been edited, softened and shaped by people acting entirely rationally in their own interests. Nobody lies. The information simply arrives pre-filtered.

Self-report is the wrong instrument

Most leadership development still asks the leader what they need to work on. This is a strange methodological choice, because the person being asked is the one operating with the most distorted data in the organisation.

The alternative was formalised by Marshall Goldsmith, whose work on executive behaviour change established a principle that has since become foundational in serious coaching practice: real change is measured not by the leader’s self-assessment, but by the perception of the people around them. If the stakeholders do not perceive the change, the change has not occurred in any sense that matters to the organisation.

Stated plainly, this sounds obvious. In practice it inverts almost everything about how leadership development is normally run — who sets the goals, who judges progress, and what counts as evidence.

What it looks like when done seriously

Arvid Buit, an executive coach working with chief executives and board members across Europe, applies this discipline as the opening move of every engagement rather than as a periodic review. Before any goal is set, before anything is described as a development area, he interviews between fifteen and twenty-five people surrounding the leader: board members, direct reports, assistants, occasionally a spouse.

The volume matters. A handful of conversations produces opinion. Twenty produces pattern. When the same behaviour is described independently by people who do not talk to each other and have no shared incentive, the finding stops being contestable. Buit describes the results as undeniable and extremely honest — for most senior leaders, the first genuinely unfiltered account of themselves they have received in years.

What emerges is rarely what the leader expected. One Dutch chief executive he coached had concluded that her leadership team’s flat energy was a market problem. Three weeks of structured observation showed she was speaking roughly eighty percent of the time in team meetings, filling every hesitation and reframing every difficult topic into something more comfortable. What she had labelled low engagement was learned silence. Her people had stopped bringing her bad news because she had unintentionally trained them to believe she could not absorb it.

Why the discomfort is the point

Leaders who have built careers on the strength of their own judgement find this premise genuinely difficult. Accepting that outside perception is the more accurate instrument requires setting aside the thing that got them promoted. That resistance is not a character flaw; it is the predictable defence of a self-narrative that has been rewarded for decades.

Buit frames his practice around what he calls critical friendship — a deliberate balance of care and confrontation, on the principle that leadership without accountability collapses into ego management. The stakeholder data provides the confrontation. The relationship makes it survivable.

The alternative is well documented: leaders who receive only comfortable information make progressively worse decisions while feeling progressively more confident about them.

The objection boards usually raise

A reasonable challenge surfaces whenever this method is proposed: if stakeholder perception is the measure, is a leader not simply being coached to become more popular? The distinction matters, and it turns on what gets measured.

Popularity is a global judgement about a person. Stakeholder-centred work deliberately avoids global judgements and fixes instead on a small number of specific, observable behaviours — whether the leader interrupts less, whether they follow through on commitments made in meetings, whether they ask before concluding. Those are testable. Colleagues can report on them without being asked to like anyone.

The method also builds in a second safeguard. Stakeholders are asked to contribute suggestions for the future rather than verdicts on the past, which removes most of the incentive to settle scores and keeps the conversation focused on what a leader can actually act on. A leader who becomes more accommodating without changing anything substantive will not move these measures at all.

See also: Georgian Company Formation: A Strategic Handbook for International Founders

Measuring what actually changed

The measurement discipline is what separates this from a 360-degree survey. Stakeholders are re-interviewed over the course of the engagement, and the metric is movement in their perception of specific, named behaviours — not the leader’s sense of progress, and not a satisfaction score.

This produces a distinctive pattern of results. Change becomes visible to the people in the room long before it appears in any dashboard. Meetings feel different. Questions that used to go unasked get asked. Decisions that used to be made impulsively get held a little longer. None of this registers in a quarterly report immediately, and all of it shapes how the organisation functions over the following years.

For boards evaluating whether coaching has delivered anything, this is the useful distinction. A leader reporting that they feel more self-aware is not evidence. A measurable shift in how twenty colleagues describe their behaviour is. The premise underlying stakeholder-centred coaching is simply that the people around a leader are better witnesses to that leader than the leader is.

It is an uncomfortable proposition for anyone who has spent a career trusting their own read of a room. Which is, in the end, a reasonable test of whether they mean it.