Leaders don’t lack answers. They lack questions.
What Hal Gregersen’s research (MIT) says about catalytic questions, and what it changes for a business leader.
For around ten years, I ran mid-sized companies, two of them under leveraged buyouts. I took one lesson from it that no dashboard shows: the higher you rise, the faster the answers come – often before the question has even been asked.
A chief executive is not isolated for lack of people to talk to. They are surrounded by an executive committee, a board, shareholders, bankers and advisers. They are isolated because those people answer before they have finished framing the problem.
That is why I placed questioning at the heart of my work with leaders. It is also why the work of Hal Gregersen, a researcher at MIT, deserves more than a passing quotation. Here is what I take from it, read through a leader’s eyes.
The CEO bubble
In 2017, Gregersen published an article in the Harvard Business Review with an unambiguous title: “Bursting the CEO Bubble”. In it, he describes a paradox. The chief executive is the person who must detect that a change of course is needed. They are also the person whose power most isolates them from the signals that would let them see it. Nobody wants to bring a problem to the boss, still less tell them they are wrong. Among the 200 or so leaders he interviewed, Gregersen found almost none who did not recognise the phenomenon.
Anyone who has held a P&L knows the mechanism. The information that reaches the top has been filtered, rounded off and brought into line with what the leader is supposed to want to hear. Gregersen’s response is not to improve reporting. It is to change stance: talk less, ask more.
The catalytic question
In Questions Are the Answer (2018), drawn from more than two hundred interviews, Gregersen identifies a particular category: the catalytic question. It removes barriers to strategic thinking and opens new, faster paths towards solutions. Its signature is paradoxical: at the moment it is asked, it seems both entirely unexpected and immediately obvious.
I would add an observation from the field. In an executive committee, you recognise it by the silence that follows. Everyone knows it is right, and nobody had asked it.
Gregersen’s earlier work on innovation (The Innovator’s DNA, co-written with Jeff Dyer and Clayton Christensen) gives a sense of the register:
- If we did not already have this process in place, would we put it in place today?
- If the law forbade us tomorrow from selling our current products to our current customers, how would we earn our living next year?
- What would happen if we did exactly the opposite of what we do?
These are not rhetorical provocations. They shift the frame, and a shifted frame releases energy. We stop debating the best answer and discover that there was another problem.
Three states nobody seeks out
Where do these questions come from? As early as 2017, Gregersen put forward a counter-intuitive hypothesis. Leaders who manage to ask them deliberately place themselves in situations where they are wrong, uncomfortable and quiet.
- Being wrong. As long as a leader is busy demonstrating mastery, they cannot ask the question that would reveal they lack it.
- Being uncomfortable. Discomfort is the reliable signal that you have reached the edge of what you already know. Questions asked on familiar ground produce nothing new. Gregersen cites the example of Uber’s CEO, who got behind the wheel as a driver in 2022. By putting himself in a position to be wrong and uncomfortable, he drew lessons from the experience that no report would have given him.
- Being quiet. This is probably the hardest state for a leader whose diary is full. Transformative questions emerge from listening and reflection. They are not born at a two-day strategy away-day, but in silence.
One older finding deserves to be put on an executive committee’s table. In their research on the innovator’s DNA, Dyer, Gregersen and Christensen found that innovative founder-CEOs spend 50% more time than others on discovery activities: questioning, observing, experimenting and widening their network. The sample consists of entrepreneurs, not mid-market chief executives, so the figure does not transfer as it stands. But it raises a genuine management question: how much of your diary is structurally protected for this?
The method: the Question Burst
Gregersen does not stop at diagnosis. He proposes a protocol, the Question Burst, published in the Harvard Business Review in 2018 and now taught at MIT. Its strength lies in its brevity and its constraint.
- State the problem in two minutes at most. It must be a problem that genuinely matters, presented at a high level: what would improve if it were solved, and why you are stuck.
- Bring together a small group that deliberately includes two or three people with no direct experience of the problem, whose way of thinking differs markedly from yours.
- Four minutes, questions only. Two rules: no answers, and no explanation of why a question is being asked. The person who owns the problem does not respond. They write everything down, word for word.
- Review alone, then commit. Identify the questions that open a new path and commit to exploring at least one of them.
The word-for-word rule is not cosmetic. A question rephrased by the person receiving it has already been digested, and therefore already neutralised.
Gregersen describes three typical outcomes. The first is a more positive view of the problem, which opens up new ideas. The second is the sense that the problem is far wider than expected. The third is the realisation that you are yourself part of the problem. All three are progress.
According to MIT Executive Education, across thousands of leaders who have taken part in the exercise over two decades, the problem is reframed and new ideas emerge in 85% of cases. When the exercise is carried out with a colleague, psychological safety improves in 89% of cases. This is practitioner feedback, not a controlled study. It remains remarkably consistent.
And artificial intelligence?
Gregersen has extended his thesis to AI, notably in a 2023 Harvard Business Review article written with Nicola Morini Bianzino, then EY’s global CTO. He has since designed a tool, the Question Burst Catalyst, available as a GPT and a Gem. One design choice sheds light on everything else: humans ask their questions before the machine does. In his view, a team that outsources its questions to AI sees its own ability to ask them wither. Language models tend towards the average answer, and it is by shifting the frame from the answer back to the question that we escape it.
AI can speed up questioning. It does not replace it.
What I take from it, as a former chief executive
First. The Question Burst is the only method I know that makes not knowing acceptable in an executive committee. The constraint of the format protects the leader: they are not exposing themselves, they are applying a rule. That is what makes it workable in a culture that does not value admitting uncertainty.
Second. Discomfort as a compass becomes a personal diagnostic tool. The question a leader avoids asking themselves is almost always the one that matters. The signal is not the intellectual interest it arouses; it is the reluctance.
Third. The bubble thesis is the strongest argument for an outside third party. Not because that person has better answers: they rarely do. It is because they are the only person in the leader’s ecosystem with no interest in answering too quickly. I was coached three times during my career as a chief executive. What I kept from it was less a set of answers than questions nobody else around me would have asked.
So, to finish: what is the question you have been avoiding for the past three months?
If an answer came to you quickly, it was probably not the right question.
Further reading
- Hal Gregersen, Questions Are the Answer: A Breakthrough Approach to Your Most Vexing Problems at Work and in Life, Harper Business, 2018.
- Hal Gregersen, “Bursting the CEO Bubble”, Harvard Business Review, March–April 2017: https://store.hbr.org/product/bursting-the-ceo-bubble/R1702D
- Hal Gregersen, “Better Brainstorming”, Harvard Business Review, March–April 2018: https://hbr.org/2018/03/better-brainstorming
- Jeff Dyer, Hal Gregersen, Clayton Christensen, “The Innovator’s DNA”, Harvard Business Review, December 2009.
- MIT Executive Education, “The leadership skill AI can’t replace: asking better questions”: https://executive.mit.edu/blog/the-leadership-skill-ai-cant-replace-asking-better-questions.html
%20(2)%20(2).png)
.jpg)
