Leadership, Systems Thinking
The Most Dangerous Leader May Be the One with the Fastest Answer
In a multi-million-dollar deal room, one question turned a three-year ERP rollout into a process re-engineering with benefits in the first year. Why the best leaders improve the question before they supply the answer.
Insight
7 min
The room was full of very good answers.
We were in a deal room, a multi-million-dollar pursuit, and the customer's question was clear: how do we roll out an ERP upgrade that creates one centralised process across 150 systems?
The customer's thinking, and every vendor response in the room, converged on the same shape. Unify the underlying systems. Then centralise the process on top of them. The common consensus was that it would take at least three years before the business saw a real benefit.
Then I asked one question:
What if we could centralise the process without unifying the underlying systems? Is it the system we need to centralise, or the process?
That question changed the complexion of the deal.
It stopped being a technology transformation and became a process-driven re-engineering that introduced a new system, not a replacement of everything beneath it. Benefits could start within the first year. The need for a unified system disappeared, and with it much of the large change-management effort that would have followed.
Nobody in that room was wrong. Everyone had answered, quickly and competently, the question that had been put in front of them. The trouble was that the question had quietly welded two things together: the process, and the system it happened to run on.
I have thought about that moment many times since, because it breaks a rather strange mythology we have built around leadership.
The more senior the leader, the more answers they are expected to carry. A difficult transformation reaches the boardroom, uncertainty is placed on the table, and somehow the leader is expected to make it disappear.
Listen. Pause. Decide. Provide direction.
For a long time, I believed that was the job.
I no longer do.
In complex transformations, a quick answer can be premature closure wearing the clothes of confidence. It narrows the field before the problem has been understood. People stop exploring. Assumptions quietly become architecture. And the organisation becomes remarkably efficient at solving the wrong problem.
The question that usually comes second
My work sits where business meets technology, and the conversation almost always begins with technology.
Which platform should we adopt?
What should move to the cloud?
Where can we automate?
Which AI model should we use?
These are perfectly reasonable questions. They are just rarely the first ones.
The first question is closer to this:
What business outcome are we actually trying to change?
That small shift can change an entire transformation. An outsourcing problem becomes an operating-model problem. An automation opportunity exposes a broken process that should never have been automated in the first place. A modernisation programme uncovers an incentive structure that keeps recreating the very complexity the new architecture was meant to remove.
Technology does not disappear from these conversations. It finds its proper place. Sometimes it enables the outcome. Sometimes it changes the path to the outcome. Occasionally, it makes an entirely different outcome possible.
But technology can also automate an old assumption, accelerate a bad process and scale a mistake.
The leader as answer machine
That is why I have grown suspicious of the leader who always has the answer.
An answer from authority carries disproportionate weight. Once spoken, it becomes a boundary. People begin working out how to implement it rather than whether it is right. Alternative paths become distractions. Contradictory evidence becomes inconvenient.
The organisation learns to look upward for certainty instead of outward for understanding.
That can feel like alignment. It is often the opposite.
And the problem is growing. When an AI model can produce a plausible answer in seconds, answers become cheap. What stays scarce is the judgement to ask whether we are answering the right question.
Questions change the shape of the problem
So I began to do something different. Instead of asking for the solution, I tried to improve the question.
Not simply asking more questions. Endless questioning is only another way of avoiding a decision. I mean questions that change how the problem is seen:
What outcome are we really optimising for?
Whose behaviour must change for that outcome to occur?
Which constraint are we treating as permanent without ever testing it?
What do we assume to be true simply because it has always been true?
Where does the benefit appear, and where does the cost reappear elsewhere?
What would have to be true for our preferred answer to be wrong?
What happens after the first-order benefit has been achieved?
These questions do something that answers rarely do. They disturb the existing frame.
Technology problem → workflow problem → incentive → behaviour → metric → an unintended consequence somewhere else
Suddenly, the problem we started with is no longer the problem we are discussing.
That is not a loss of focus. Very often, it is the moment the real problem becomes visible.
Clarity is not the absence of uncertainty. Clarity is knowing which uncertainty matters.
This is where I think leadership matters most. Not in protecting the organisation from uncertainty, but in making uncertainty examinable.
A question-first approach
In the transformations that produced the most useful conversations, I began to see a simple pattern:
Outcome → Frame → Assumptions → Consequences → Technology
1. Start with the outcome
Before discussing technology, ask what must become measurably different for the business, the customer or the employee.
Not what are we implementing? But what are we trying to change? In that deal room, the outcome was never a unified ERP. It was one consistent process.
2. Challenge the frame
Ask whether the problem has been defined too narrowly. Is this really a technology problem? A process problem? An incentive problem? A capability problem? A trust problem? Or some interaction between them?
3. Surface the assumptions
Every transformation contains statements masquerading as facts.
Customers expect this.
Regulation requires that.
This process cannot change.
We need this platform.
AI cannot do that.
Some will be true. The important question is whether we know which ones.
4. Follow the consequences
If we change this part of the organisation, what happens elsewhere? What becomes easier, and what becomes harder? Who changes their behaviour? What new constraint might we create? And what happens when the solution succeeds at scale?
5. Only then ask what technology makes possible
This reverses a pattern I see all the time. Instead of beginning with a technology and searching for places to apply it, begin with the outcome and ask how technology changes the space of possible solutions.
Sometimes the answer will be AI. Sometimes automation. Sometimes architecture.
And sometimes the most intelligent technology decision is not to introduce more technology at all.
None of this removes the need to decide. Leadership is not an endless Socratic seminar. Eventually, someone must choose.
But there is a real difference between deciding quickly and arriving at a decision with a richer understanding of the system in which it will operate.
Where the questions lead
These questions rarely stay inside the technology.
The more I asked what outcome are we really trying to change?, the harder it became to look at technology in isolation.
The more I asked what happens next?, the more feedback loops became visible.
The more I asked where does the consequence appear?, the more organisational boundaries began to look artificial.
The more I challenged assumptions, the more I noticed that many problems were not contained in the boxes where organisations had placed them.
Technology connected to process. Process connected to incentives. Incentives shaped behaviour. Behaviour affected customers. Customer behaviour changed economics. Economics influenced investment. And investment changed the technology again.
At some point, you stop looking at individual problems.
You start looking at systems.
I learned the first version of that lesson long before I had the vocabulary for it. An anthill taught me, and I wrote about it in Systems Thinking and the Curious Case of the Anthill. The deal room was the same lesson in a very different setting: do not accept the first question as the final definition of the problem. Interestingly, this approach works even better when AI gets into the game.
Perhaps that is the leadership capability complex transformation needs most.
Not leaders who always know the answer. But leaders who can help an organisation ask a better question, widen the field of possibility, see consequences it could not see before, and then have the courage to decide.
Because sometimes the most valuable thing a leader can say is not:
"Here is the answer."
It is:
"Are we sure this is the right question?"
CONTINUE EXPLORING
More writing on technology, AI economics, systems thinking, leadership, and mentoring.
CONNECT