The Order That Made Him Slower
The founder had a good operations lead and a stalled decision. A supplier contract needed renegotiating, the terms were bleeding margin, and the man who owned the relationship had been circling it for five weeks. So the founder did what most leaders do under pressure. He told him what to do.
"Rewrite the terms. Take out the volume clause. Get it done by Friday."
The work arrived on Friday. It arrived exactly as ordered, and it was wrong. The clause that came out was the one holding the supplier's service level in place. The margin stayed broken, and the relationship carried a fresh wound into the next quarter.
I have watched that scene replay inside hundreds of businesses. The founder's read was correct. The instruction was clear. The execution landed, and the outcome failed anyway. What went missing sat on one side of the exchange. The operations lead carried the instruction rather than the problem, so he did what the sentence said and left the thinking with the person who had already done the thinking.
That afternoon the founder asked a different kind of question. Not whether the job got done, and not why the job was hard. He asked: "If you owned this contract and the profit that comes out of it, what would you change, and what would it cost the supplier to agree?"
The conversation ran forty minutes. The operations lead rebuilt the terms himself, flagged the volume clause himself, and closed nine days later with the service level intact. Same man. Same contract. One difference. The answer came out of his own head, so he carried it.
Here is the principle underneath the story. Telling transfers information. Asking transfers ownership. Only one of those survives the moment you leave the room.
Why Telling Fails Even When You Are Right
Instruction feels efficient. You hold the answer, you hand it across the table, and the work moves. The appeal is real, and so is the cost. Three mechanisms run underneath, and each one has been documented.
Pressure creates a push-back. Jack Brehm set out the theory of psychological reactance in 1966. When a person believes a freedom is being taken away, they act to restore it, and the strength of the restoration rises with the strength of the push. An instruction lands as a restriction on how someone is allowed to think. What you receive is surface compliance, and the counter-move sits underneath it, waiting for the first unobserved decision.
An answer you generate, you keep. Edward Aronson made the case in American Psychologist in 1999 that self-persuasion outperforms direct persuasion, because a person who has argued themselves into a position cannot quietly walk away from it. The argument belongs to them. When you supply the conclusion, the conclusion stays yours, and you become the one who has to defend it every time the situation shifts.
And you remember what you built. Norman Slamecka and Peter Graf showed in 1978 that people recall words they generated far better than words they read or heard. The generation effect is one of the more durable findings in memory research, and it reaches past word lists into plans. A plan somebody constructs holds a different place in their attention than a plan they were handed. Ask them a week later what the plan was, and you will hear the difference.
Put the three together and the case against telling stops being about manners. You are handing over a conclusion the other person has no stake in, no memory of building, and a mild pressure to resist. Then you wonder why the same conversation has to happen again next month.
There is a second cost, and it lands on you. Every answer you supply trains the room to wait for the next one. A team led by instruction becomes a team that cannot decide without you in the room, and the business gets its ceiling installed at the founder's calendar.
The Field Was Built on Questions, Not Scripts
NLP gets described as a bag of techniques. Read the primary material and something different shows up. The first instrument Richard Bandler and John Grinder produced was a question set.
The Meta Model, published in The Structure of Magic I in 1975, is a catalogue of language distortions paired with the question that recovers what got deleted. Who specifically. Compared to what. How do you know. What stops you. Bandler and Grinder extracted those questions while modelling Fritz Perls and Virginia Satir, and what came out of the modelling was not a script to recite. It was a way of asking that pulls a vague sentence back into contact with the experience behind it.
Milton Erickson worked the other direction with the same instrument. In Hypnotherapy: An Exploratory Casebook, published with Ernest Rossi in 1979, the case material shows him pacing a person's present experience and then posing a question that turns attention toward a specific possibility. He rarely issued an order. He asked a question with a direction built into it, and the person walked toward the answer under their own power.
That difference matters to a leader, so name it plainly. A clarifying question recovers information. A moving question changes where a person's attention goes. The first is diagnosis. The second is influence, and it is the one most managers never practise.
The tradition kept building on questions. Robert Dilts, in Roots of Neuro-Linguistic Programming in 1983, treats the modelling interview itself as a structured sequence of questions aimed at recovering another person's strategy. Whatever you think of the packaging, the underlying design is consistent. You do not install an answer in someone. You ask in a way that makes them find it.
What the Research Shows About Asking
The question side of influence has been studied directly, and the results are worth knowing before you build a habit on it.
Asking about a behaviour nudges the behaviour. Samantha Wilding and colleagues published a meta-analysis of the question-behaviour effect in the European Review of Social Psychology in 2016. Across dozens of studies, asking people whether they intend to do something produced a small but reliable increase in the odds that they did it. The effect is modest, and it is one of the few findings in this area that has been replicated at scale. A question does not only collect a position. It moves the person slightly toward it.
How you ask yourself matters as much as what you ask. Ibrahim Senay, Dolores Albarracin and Kenji Noguchi reported a set of experiments in Psychological Science in 2010 on the private language people use before a task. Participants who asked themselves "will I?" performed measurably better on a demanding task than participants who told themselves "I will". The interrogative form beat the declarative form. A question opens a slot the mind moves to fill, while a statement closes the subject and asks for nothing.
The strongest sellers ask, and they ask in a specific order. Neil Rackham's team analysed tens of thousands of sales calls for SPIN Selling in 1988. Their finding was blunt: the high performers asked more questions than the low performers, and the questions that separated the two groups were implication questions, the ones that make the buyer work out for themselves what the current problem costs. Chris Voss, a former FBI hostage negotiator, built the practitioner version of this around calibrated questions, and his working rule is to open them with what or how rather than why, because why reads as an accusation in almost any language.
And the honest note. The Meta Model and Erickson's question work are modelled instruments from practitioner material, not the output of controlled trials. When reviewers have tried to test NLP as a package, the results are thin. Tomasz Witkowski's 2010 review in the Polish Psychological Bulletin and the systematic review by Jackie Sturt and colleagues in the British Journal of General Practice in 2012 both found the evidence base insufficient to support NLP as an intervention. Take the tools from the modelling tradition that produced them, and take the mechanism from the separate research traditions that studied asking. Keep the two claims apart and the lesson stays honest.
Four Questions That Move a Person
Four structures cover most of the ground. Each one does a different job, and each one has a failure mode attached.
1. The specificity question
This is the Meta Model doing its original work. Somebody hands you a fog, and you ask the question that turns the fog into a target.
"They never listen to me." The question: Who specifically, in which meeting, and what did they do after you spoke?
Founder situation: a marketing lead keeps saying the sales team ignores the pipeline. Twenty minutes of specificity produces a real problem, which is that the handover email goes out on Friday afternoon when nobody reads it. Fog produces a mood. Specificity produces a task.
Failure mode: running the questions as an interrogation. Three in a row turns a conversation into a cross-examination, and the other person defends instead of thinking.
2. The criterion question
Ask what the outcome has to give the other person for it to be worth doing, and you surface the standard they are measuring against.
"We need to move faster here." The question: What would this project have to give you for you to call it worth the quarter?
Founder situation: two department heads block each other on a launch. One is protecting revenue timing, the other is protecting client trust. The criterion question names both rulers before either becomes a personal position.
Failure mode: collecting the criterion and then ignoring it. Ask and discard, and you have taught the room that your questions are decorative.
3. The counter-example question
A generalisation collapses under one honest example, and the collapse needs no argument from you.
"This team cannot execute." The question: Has there been a launch in the last year where this team executed well? What was different about that one?
Founder situation: a founder convinced his operations function is broken recounts two good quarters and describes the exact conditions under which they happened. The conditions turn out to be repeatable. The diagnosis shifts from a judgement about people to a design question about conditions.
Failure mode: hunting the counter-example to win. The moment the question carries a scoreboard, people find a way to answer it defensively.
4. The commitment question
This is the future-facing one, and it is tied to Senay's interrogative form. You move the person from what they think to what they will do when the old pattern comes back.
"I will delegate more." The question: Thursday at 4pm, the proposal is on your desk and the client is waiting — what will you do next?
Founder situation: a founder accepts every piece of work that arrives. The commitment question produces a specific plan for the specific moment, which is the shape that survives contact with a busy week.
Failure mode: accepting a vague intention as a commitment. "I will try" is not a plan. Ask for the scene, the action and the person to whom it happens, and write the answer down.
The Protocol — Five Steps in One Conversation
Run the sequence in order. The order carries as much of the effect as the questions do.
- Name the outcome, then ask for their read before you offer yours. State where the conversation needs to end, then ask what they see. Their read arrives uncontaminated while yours is still off the table, and you learn the shape of the gap in their words rather than yours.
- Ask for the obstacle in their language. "What stops this from happening?" is a different question from "what is the problem", and it points the person at a cause rather than a complaint. Write down the phrase they use, because their wording is the door you will open later.
- Use one question family, not three. Pick specificity, criterion or counter-example for this conversation. Stack all three and the other person hears an agenda instead of interest.
- Let silence do the work. The answer that matters arrives three to five seconds after the question stops. Most leaders fill that gap with a suggestion, which converts the question back into an instruction. Hold the silence and let the person reach.
- Close on the commitment question, with a scene, an action and a date. "What will you do on Thursday afternoon when the client chases the proposal?" Then write the answer down in their words and send it back to them within the hour. That message is the ownership document.
Two steps carry most of the weight. Step one protects their thinking from your answer, and step five converts a conversation into a plan. Skip step one and you will spend the rest of the conversation persuading them out of a position they took to sound decisive.
For the private version, point the same questions at yourself. Ask "will I?" before the day starts, and answer with a scene and an hour. The interrogative form works inside your own head as well as across a table.