How Coaching Actually Helps Executives
Most of what gets called leadership development treats the leader as the problem to be optimised. That is not how I work, and it is not what the evidence actually shows changes when coaching works. Coaching does not fix anyone. It interrupts an over-functioning survival script, the exhausting, always-on operating mode where a leader absorbs every risk, controls every output and mistakes exhaustion for competence, and it hands leaders back the executive capacity that script was quietly spending.
A new peer-reviewed study puts numbers behind what that interruption actually looks like in practice. Kirschner and Terblanche (2026), published in the European Journal of Training and Development, analysed 1,130 open-ended survey responses from 686 executives across more than 300 global organisations, predominantly Fortune 500 companies. Leaders were asked two things: how coaching changed their leadership, and what colleagues told them about the change. Here is what that tells us about how coaching actually works.
Retiring the hero script
The single most common change leaders reported was in team leadership: 24.2% of responses, more than any other theme. Not vision. Not strategy. How they ran their teams.
One executive described it exactly as I would: "Has made me transition from being a hero and controlling the output of the team, to empowering and guiding the team to be owners of their work and hold them accountable." That is the over-functioning survival script in one sentence, named, and then set down. Delegation was not a productivity tactic here. It was a leader recognising that carrying everything alone had never actually been leadership; it had been self-protection dressed up as diligence.
Reclaiming the bandwidth the Identity Tax spends
Executive presence showed up as the fourth most common outcome, at 15.7%, described by leaders as increased confidence, visibility and comfort taking up space in senior rooms.
For many of the LGBTQ+ executives and senior leaders I work with, presence has never been free. There is a real cost, an Identity Tax, paid in code-switching, hypervigilance and the constant background labour of managing how you are read in rooms that were not built with you in mind. That cost is invisible on an org chart and very visible on an energy bill. When leaders in this study report becoming more confident and more visible, what they are describing, in less clinical language, is bandwidth returning. Capacity that used to go into monitoring perception becomes available for actual leadership.
It does not stay private
Here is the finding that matters most. When leaders described what colleagues actually said back to them, the pattern echoed their own account almost exactly: team engagement (27.4%), executive presence (17.5%) and communication (13.8%) topped the list colleagues raised, unprompted.
One leader put it this way: "My direct reports have commented on my delegation adjustments and have been more open to receiving new assignments, some of which I would've probably done myself prior to the coaching sessions." Another: a direct manager "commented a few times on how I am showing up at meetings in a more confident and strategic manner."
That convergence is the whole point. A leader can believe they have changed. It means something different when the people who watched the old pattern for years are the ones telling them it is gone. The high-performance illusion, the polished exterior that used to hide the strain underneath, only actually cracks when what is underneath it is visible to other people too, and what they see, in this data, is someone leading rather than surviving.
Where the study earns its credibility
A study is only as useful as its willingness to name its own limits, and this one does that unusually well. Four things are worth stating plainly.
First, this is self-report layered on self-report. Leaders described their own perceived change and separately described what they recall colleagues saying. Neither is an independent observation of actual behaviour. The overlap is meaningful, but it shows two self-report accounts agreeing with each other, not proof that behaviour changed.
Second, the sample was drawn from a post-engagement survey, which carries the usual biases: leaders with a poor experience were less likely to respond at all, and social desirability shapes what people write when a coach or coaching organisation might read it.
Third, a meaningful share of responses did not fit the coding categories: 24% for the leadership-impact question and 40% for the colleague-feedback question. The published themes describe what fit cleanly, not the full range of what leaders actually said.
Fourth, and the authors disclose this directly: the first author is CEO of the coaching organisation that supplied the data. That is a genuine conflict of interest, stated openly rather than buried, and it is exactly the kind of disclosure that should make a reader look harder at the findings rather than dismiss them.
Add a design that is descriptive rather than causal, no control group, no independent measure of behaviour before and after, and the honest summary is this: the study tells us what leaders say happened. It does not, and does not claim to, prove that coaching alone caused it.
None of that discredits the findings. A study that names its own conflicts of interest and its own coding gaps is one you can actually trust as far as it goes. And as far as it goes, across nearly 700 senior leaders and hundreds of organisations, what it shows is specific and behavioural: leaders stepping out of the hero role, reclaiming presence that used to be taxed away, and having that change confirmed by the people who work alongside them.
I coach founders, leaders and LGBTQ+ executives to dismantle their survival scripts and reclaim their executive capacity. If you are ready to operate at depth, schedule a strategy session with me.
Reference: Kirschner, J. and Terblanche, N. (2026), "Executive leaders' perspectives on the impact of executive coaching on their leadership development," European Journal of Training and Development, Vol. 50 No. 10, pp. 127-145.
