Knowledge,
made publishable.

There's a coaching talk I came across recently, the kind aimed at managers and team leads, that spends most of its time on the usual ground: closed questions versus open ones, leading questions to avoid. Then, almost as an aside, one line stands out from the rest: swap "why" for "what makes you think that," and watch what changes.

It's such a small swap that it's easy to read past it. But sit with it for a second and the logic is hard to argue with.

"Why did you do it that way?" sounds like curiosity. It shows up in leadership books filed under "understanding root causes," in performance reviews, in the kind of one-on-one that's supposed to feel supportive. But think about where else that exact question gets asked. Courtrooms. Disciplinary meetings. Any situation where someone is expected to account for themselves to a person with the power to judge the answer.

That's the trap. "Why" is grammatically a request for information, but socially it functions as a demand for justification. The person on the receiving end doesn't hear "help me understand." They hear "explain yourself," and they answer accordingly, reaching for the most defensible version of events rather than the true one.

"What made you think that" asks for the same information and does none of that work. It's still curious about the reasoning. It just doesn't put the other person on trial to get there. The spotlight moves off the person and onto the thinking itself, which turns out to be a much easier thing for someone to walk you through out loud.

It's worth noticing how ordinary this pattern is outside of work, too. A parent asking a kid why they did something. A partner asking why a decision got made without them. In almost every case, the honest answer to "why" starts with some form of self-defense, because that's the posture the word puts people in, whether or not it's the posture anyone intended.

Journalists get trained on a version of this same instinct. Ask someone why they left a job, and the answer is usually the polished one they've given a dozen times already. Ask what led them to start looking, and sometimes you get the real story instead, the one with an actual moment in it rather than a tidy explanation built for public consumption.

None of this makes "why" a word to avoid entirely. Sometimes you do need the reason, stated plainly, and there's no way around asking for it directly. But in a coaching conversation, where the goal is to help someone think more clearly rather than defend a decision they've already made, the softer phrasing tends to get further. It lowers the temperature just enough that people stop performing an answer and start actually producing one.

What makes this particular fix worth sharing isn't that it's clever. It's that it's almost embarrassingly available. No new framework to learn, no structural change to how a conversation runs. Just one word, swapped for another, in the exact moment most managers default to the version that puts people on the back foot without meaning to.

Try noticing it this week. Not changing anything yet, just noticing how often "why" comes up in conversations where the actual goal is understanding rather than accountability. Most people are surprised by the number once they start counting.

Then try the swap where it fits. "What made you think that" instead of "why did you think that." "What led you to that decision" instead of "why did you make that decision." The information you're after stays the same. The person answering tends not to.

There's a longer piece coming on the other habits worth building into how a leader coaches a team: what to do after someone fails, what active listening actually requires, the discipline behind a good coaching model. This one felt worth sending on its own, mostly because it's the easiest of the group to try in the very next conversation you have.

There is a version of leadership that mistakes presence for value: the manager who has an answer for everything, who steps in before the silence gets uncomfortable, who corrects the plan before the team has finished proposing it, who explains the lesson before anyone has had the chance to miss it on their own. All of this can look like help, and sometimes it is, but more often it's intervention dressed up as guidance, leaving the other person with less room to think than they had before the manager arrived.

The four coaching behaviors worth paying attention to, care, focus on strengths, restraint after failure, and the habit of celebrating what worked, share something underneath the surface. Each one asks the leader to do less, not by putting in less effort, but by interfering less: trading the reflex to jump in, correct, or perform competence for something quieter, and in the long run more useful, room for someone else's competence to develop.

Take care first. It sounds like the softest of the four, almost incidental, but the pattern is specific: people can tell when a manager's concern is real, and they respond to it not with gratitude but with trust, which is the actual currency that makes coaching possible at all. Care isn't a behavior you perform for effect. It shows up, or it doesn't, in what you choose to notice about someone's development and what you choose to overlook, in whether you remember what they told you weeks ago about what they wanted to improve, and in whether that memory shapes how you talk to them today.

Most performance reviews spend their word count on what's broken. Fixing what's broken feels like leadership. It isn't, always. People improve faster in areas where they already have some traction, and a coach who spends the bulk of their limited time chasing deficits is fighting the direction learning naturally wants to go. Strength, not deficiency, is where growth compounds fastest. Play to it, whatever that costs in the appearance of thoroughness.

Punish a mistake and watch what happens next: the next mistake gets hidden, not fixed. That's the cost most managers don't count. The reflex to punish, or just to visibly disapprove, comes from a reasonable place, since mistakes are costly and costly things seem to deserve consequences. But a team that hides its mistakes from you is a team you can no longer actually coach. Better to walk the person through what went wrong: choose curiosity over a verdict, and stay interested a beat longer than instinct wants to allow.

And then, the smallest gesture, easy enough to skip that it usually gets skipped: taking the time to celebrate what went right. Not a speech, not a performance of enthusiasm, just the plain acknowledgment that something worked, said directly to the person who made it work. Positive reinforcement works. It always has, whatever the latest management book claims to have discovered about it. Underneath all four behaviors is the same move: get out of the way of the person's own competence, instead of standing in front of it.

Nowhere is that clearer than in the way a leader asks questions, which can be either the most useful tool in coaching or its most disguised form of control. A closed question invites a yes or no and teaches almost nothing to either party. A leading question, dressed up as curiosity, is really just an opinion with a question mark attached to it. Even the word "why" sounds like genuine inquiry, but it tends to put people on the defensive, because it asks them to justify themselves rather than think out loud. The alternative, asking what makes someone think what they think, hands the thinking back to the person who has to live with the answer, instead of asking them to defend a decision that's already been made.

This is where restraint becomes technique rather than personality. A good question doesn't just gather information the leader happens to be missing. It forces the other person to construct an answer out loud, a different and more demanding cognitive act than simply agreeing with someone else's plan. The leader who resists the urge to supply the answer is doing the harder, less visible work of making someone else's thinking sharper, question by question.

Listening carries the same discipline, though it rarely gets treated as a skill on its own. Real listening isn't nodding while composing your next sentence, or holding still while running through your own list of problems for the day, or hearing what's said and then finishing the other person's sentences for them. The bar is being able to summarize, accurately, what the other person just said, a surprisingly high bar most people don't clear even when they believe they have. Most people, in a given day, are not truly listened to by anyone. The leader who manages it, consistently, gives something rare enough that it changes how a team behaves around them: more forthcoming, more willing to say the difficult thing out loud, simply because someone finally noticed the difference between hearing and understanding.

The GROW model brings this same logic to its most structured form, and its four steps ask as much of the coach as of the person being coached. Setting the goal belongs to the person being coached, not the leader, however tempting it is to suggest a better one. Assessing reality means setting aside your own assumptions long enough to see what's actually true, not what's convenient to believe about the situation. Generating options means resisting the pull to hand over your own answer, even when you're fairly certain you already know it. And deciding on the action, the final and most consequential step, has to belong entirely to the person who will carry it out, since a decision made for someone rarely survives contact with the work of actually doing it.

A coach who makes that decision for someone hasn't coached them. They've simply told them what to do.

What ties all of this together is less a technique than a posture: the coach's job is not to be indispensable, but to make their own presence, gradually and deliberately, less necessary. The care, the patience with failure, the discipline of asking instead of telling, the willingness to actually listen, the structure that keeps decisions where they belong: each one is a version of the same idea. Getting out of the way, so that someone else's thinking finally has room to arrive.

1. Four Behaviors That Support Coaching

Demonstrate genuine care. People can tell when a manager's concern is real. It shows up in what you notice, remember, and follow up on over time, not in what you say.

Play to strengths. Most people improve faster in areas where they already have some ability. Coaching time spent building on strengths tends to produce more progress than time spent closing gaps.

Don't punish failure. Punishing mistakes discourages risk-taking and teaches people to hide problems instead of raising them. Coach the person through what happened and what they'd do differently next time.

Celebrate success. Acknowledging what went well, privately and publicly, is usually more motivating than a general pep talk.

Reflection exercise — Think of a manager you considered genuinely good. Which of these four behaviors did they show most consistently, and which did they rarely show?

2. The Power of Asking Questions

A good question requires the other person to construct an answer, which builds understanding in a way that being told an answer doesn't.

• Closed questions (yes/no) produce less learning. Use sparingly.
• Open questions have no fixed answer and are the most useful for coaching.
• Leading questions carry an implied opinion and restrict thinking. Avoid them.
• "Why" questions can feel like an accusation and put people on the defensive. Swap "why do you think that" for "what makes you think that."

Reflection exercise — Write three open questions you could ask a team member this week instead of offering your own solution.

3. Active Listening

Active listening is not nodding while planning what to say next, staying silent while running through your own to-do list, or interrupting to finish someone's sentence.

A simple test: if you can't summarize what the other person just said, you weren't really listening. Tone of voice and body language often carry more than the words themselves.

Practical habits: eye contact, following without planning your response, and periodically paraphrasing what you've heard.

Reflection exercise — In your next one-on-one, paraphrase what the other person said before adding your own point. Notice what changes.

4. Coaching for Performance: The GROW Model

When coaching is tied to a specific goal, the GROW model offers a simple structure.

Goal. Clarify what the person is aiming for, and make sure it connects to the team's needs.

Reality. Assess the current situation on facts, not opinions or assumptions. Ask for evidence when someone states an opinion as fact.

Options. Help the person generate as many paths forward as possible, including ones that seem unlikely at first. Past experience often limits the options someone considers.

Will. The person being coached, not the manager, decides on the action. The coach's job is to test that action against the goal, not choose it.

Every session should end with at least one concrete action agreed. Review progress against it in the next session.

Reflection exercise — Pick a current challenge on your team. Write one question for each GROW step (Goal, Reality, Options, Will) you could ask the person facing it.

Ask a room full of managers if they know the GROW model, and most hands go up. Ask them to run a real coaching conversation through it, out loud, and the model tends to fall apart in the same four places every time. Not because it's a bad framework. Because each of its four steps has an easy way to fake it.

Start with Goal, the step everyone gets right on paper and wrong in practice. The failure here rarely looks like skipping the question. It looks like asking it and accepting the first answer that comes back, usually something vague enough to mean nothing: "get better at this," "improve the situation," "do more." A goal that vague isn't a starting point. It's a placeholder dressed up as one, and the coach who lets it stand has already let the whole conversation off the hook before it began.

Reality is where the model breaks down most often, and for a reason nobody likes to admit: honest reality checks are uncomfortable, and discomfort is the first thing most conversations quietly optimize away. The tell is a particular kind of politeness. Someone names a problem, softens it immediately, and the coach, wanting to keep things pleasant, lets the softened version stand. Revenue is "a bit behind," rather than down twenty percent. A key skill is "still developing," rather than absent. None of this is dishonesty exactly. It's the ordinary human preference for a version of the truth that's easier to sit with, and a coach who doesn't push past it is coaching against a fiction.

Options fails in the opposite direction: not through avoidance, but through speed. Someone names a problem, the coach already has an answer in mind, and the "options" step becomes a formality on the way to the coach's preferred solution, dressed up as a question so it feels collaborative. Real option-generating is slower and, frankly, more annoying than that. It means sitting with the discomfort of a few genuinely bad ideas on the table before a good one shows up, because the good ones rarely arrive first. Skip that discomfort and you get options in name only, a single idea wearing a plural.

Then there's Will, which fails in the most familiar way of all: the coach, sure they already know the right move, nudges the conversation toward it instead of letting the other person land on their own answer. It rarely looks like an order. It looks like a leading question, a raised eyebrow, a "have you considered," repeated until the person being coached arrives, gratefully, at the exact conclusion the coach had in mind the whole time. This isn't coaching. It's steering with extra steps, and the person on the receiving end usually knows it, even when they can't quite say why the conversation felt hollow.

What connects all four failures isn't laziness. It's that each shortcut feels, in the moment, like the reasonable thing to do. Accepting a vague goal feels efficient. Softening a hard truth feels kind. Moving fast past options feels decisive. Nudging someone toward the right answer feels helpful. None of these instincts are malicious. They're just the path of least resistance, and the model exists precisely because that path leads somewhere worse than the discomfort it avoids.

Used honestly, GROW isn't difficult to understand. It's difficult to sit through, which is a different problem entirely, and the only real fix for it is noticing, step by step, exactly where the temptation to shortcut shows up, and choosing the slower version anyway.