Two years ago, the Y Combinator co-founder Paul Graham coined the term founder mode, the idea that great leaders should impose their vision directly on the organisation and resist the temptation to delegate too much. The essay went viral. It also became, for a year, an easy excuse for executives who wanted to talk over their teams.
This week, Stanford psychologist Jamil Zaki pushed back in a Harvard Business Review piece titled The Best Leaders Embrace the Role of Supporting Character (HBR, 5 May 2026). His argument is short and uncomfortable: in the organisations that perform best, the leader is not the protagonist. The leader is the supporting character whose job is to make the people doing the work look, and become, better. The behaviours Zaki names, practising intellectual humility, asking better questions, helping people connect work to their own values, sound mild. They are not. They are a learned skill set that most senior managers do not have.
Why This Lands In May 2026
Zaki's piece is not landing in a vacuum. Three weeks ago, Edelman's Trust Barometer 2026 reported that 42% of UK and global employees would rather switch departments than report to a manager whose values they do not share. Two weeks ago, McKinsey's State of Organizations 2026 survey of more than 10,000 senior leaders across 15 countries found that 86% of leaders believe their organisation is unprepared to adapt AI into day-to-day operations, and that the differentiator is not headcount or budget but leadership behaviour, what they call human-centric leadership. Gallup's most recent global workplace data shows manager engagement at 22%, down nine points since 2022.
The pattern across those numbers is the same. The job description for a manager has moved from direct the work to create the conditions in which the work can be done well. The leaders running organisations now were largely promoted under the older job description. Most have never been taught the second one.
What Intellectual Humility Looks Like, In Behavioural Terms
Intellectual humility is fashionable as a phrase, less so as an observable behaviour. In our own work, building on academic behaviour-change research from UCL, Cambridge and Bocconi, we measure it three ways:
- Time-to-question. In the first two minutes of a meeting, did the leader make a statement or ask a question? Most fall on the statement side, and the team takes the cue. The supporting-character version inverts it.
- Quality of follow-up. After an answer, did the leader respond with a position, or with a second question that pressure-tested the answer? Edmondson (1999) in her foundational Administrative Science Quarterly paper on psychological safety identified this as the single most reliable behavioural marker of a learning team.
- Public correction. Did the leader, at any point in a 60-minute meeting, openly change their position based on what a junior person said? In our observational work, this happens in fewer than one in five executive meetings. It is the behaviour that creates safety the fastest.
None of this is mysterious. All of it is hard to do, because the underlying habits are wired in by twenty years of being promoted for having strong views and stating them clearly.
Why Awareness Training Will Not Fix This
Most organisations respond to the Zaki problem with a workshop. A psychological safety module, a 90-minute session on inclusive leadership, a deck of frameworks. The retention curve here is well-known. Roediger and Karpicke (2006) in Psychological Science showed that active retrieval beats re-reading by around 50% on long-term retention. Sitting in a room while someone explains intellectual humility is, in retention terms, the cognitive equivalent of re-reading. The behaviour does not transfer.
What does transfer is rehearsal. In our own UCL-anchored research on communication skills, participants who learned a behaviour through immersive role-play with professional actors scored roughly 20% higher on observed behaviour than those who learned the same content through video or slide-show training. Self-rated confidence did not predict actual performance, a Dunning-Kruger pattern we then designed out of subsequent studies. The mechanism is straightforward: a leadership behaviour is a motor skill before it is a cognitive one, and motor skills require reps with feedback.
What We Do About It
Our leadership workshops and our immersive simulations are built for exactly this. A small group of senior leaders enters a designed scenario. Professional actors play the team members. The leader has to run the meeting, ask the questions, hold the silence, change their mind out loud. Then the same scene is replayed with different choices. By the third repetition, the supporting-character behaviour is no longer counter-intuitive, it is the default. That is what behaviour change actually requires.
Whether or not you work with us, Zaki's piece is worth carrying as a single question into your next executive meeting: if my team replayed the tape of this hour, would they say I was the protagonist of it, or the person who made them sharper?
Book a free 30-minute diagnostic call → or read about our research-backed approach.