Coaching
Developing people by helping them think — not by handing them the answer. The skill that turns a smart individual contributor into someone who makes everyone around them better.
🎯 By the end you'll understand
Why the shift from “expert who tells” to “coach who asks” is the hardest move for a capable technical person, how to run powerful questions and the GROW model, where the boundary sits between coaching and mentoring, managing, and therapy — and the honest cases where coaching is the wrong tool and you should just answer.
KEY INSIGHTThe one idea to take away
When you give the answer, you solve one problem. When you help them find it, you grow a person who will solve a thousand.
Telling is faster — almost always, in the moment. That's exactly why it's a trap. Every time you supply the answer to someone who could have reached it, you get a quicker fix today and a more dependent, less capable person tomorrow. Coaching is the deliberate choice to trade a little speed now for capability that compounds. It is the single highest-leverage thing a leader does, and it feels, to a smart person who already knows the answer, almost physically uncomfortable.
Coaching is not being soft, withholding, or playing guessing games. It's helping someone do their own thinking so the insight — and the ownership — is theirs. The measure of good coaching isn't how wise you sounded; it's how much clearer they got.
DEFINITIONWhat coaching actually is
A working definition: coaching is helping someone improve their own thinking and performance by asking, listening, and reflecting — rather than by directing or instructing. The coach assumes the other person is capable and resourceful, and that the job is to unlock their thinking, not to transplant yours.
That assumption is the whole game. A manager solving a problem treats the person as a channel to a fix. A coach treats the problem as a vehicle for developing the person. Same conversation on the surface; opposite intent underneath — and people feel the difference instantly.
Coaching — a developmental conversation in which you help another person reach their own clarity, decision, or solution, primarily through questions and listening, holding the belief that they are capable of the answer. Contrast with telling, where you transfer your conclusion to them ready-made.
CORE CONCEPTThe hardest shift: from expert to coach
If you got here by being good at things — solving the hard problem, debugging the gnarly system, seeing the answer faster than the room — then your whole identity and reward history is built on knowing. You were promoted for having answers. Now leadership asks you to deliberately not use that strength, and it feels like driving with the handbrake on.
Here's the maths of why you have to. As an individual contributor, your output is your output. As a leader, your output is the output of everyone you develop. If you keep being the answer-vending machine, you stay the bottleneck: every hard call routes through you, your people stay junior, and your team's ceiling is your own bandwidth. The moment you start growing thinkers instead of dispensing answers, the ceiling lifts — your leverage becomes the capability of the whole team, not the hours in your day. This is the same logic as delegation (3.8): you scale by building capacity in others, and coaching is how that capacity actually gets built.
That itch to jump in and just say the answer is your expertise begging to be used. Treat it as a cue to pause, not a green light. The most useful sentence in a leader's vocabulary is often: “What do you think we should do?”
CORE CONCEPTAsk, don't tell — the power of a good question
A powerful question is open, short, and aimed at their thinking rather than at extracting facts for yours. It opens up the problem instead of narrowing toward your pre-formed answer. A weak question is closed or leading (“Don't you think we should just roll it back?”). A powerful one hands the steering wheel back:
- “What have you tried so far?” — surfaces their thinking and usually reveals they're further along than they let on.
- “What's really going on here?” — gets under the presenting problem to the actual one (often it's not technical at all).
- “What would you do if I weren't here?” — the most clarifying question in coaching. It removes you as the crutch and almost always produces a real answer.
- “What else?” — the deceptively powerful follow-up. The first idea is rarely the best; “what else?” mines the second and third, which are usually better.
- “What would make this easier?” — shifts from stuck to actionable without you supplying the action.
Powerful questions — open, concise, non-leading questions that prompt the other person to do their own thinking, rather than questions designed to gather information for the asker or to nudge toward a predetermined answer.
And then the hard part: actually listen (1.15) to the answer. Coaching collapses the instant the question is a formality and you're just waiting for your turn to talk. The question only works if you genuinely don't know what they'll say and you genuinely care.
SEE ITTell vs coach — see the two timelines
A report comes to you with a problem. You have two doors. Door one: give the answer. Door two: ask a coaching question. Toggle between them and watch both the short-term effect (speed) and the long-term effect (capability and ownership). Then meet the case where telling is actually the right call.
The pattern: for genuine judgment-and-growth situations, coaching costs minutes now and buys a more capable person forever. For a pure-knowledge gap or a real emergency, coaching is self-indulgent — just tell them. The skill isn't always coaching; it's knowing which door you're standing in front of.
FRAMEWORKThe GROW model
If you want a structure for a coaching conversation, the most durable one is GROW (popularised by Sir John Whitmore in Coaching for Performance). It's four stages, and the discipline is to move through them in order rather than leaping to options before you've understood reality.
Goal
What do you actually want here? What does “done” look like? Pin the objective before touching the problem — people often realise their stated goal isn't their real one.
Reality
What's actually happening now? What have you tried? What's in the way? This is where “what else?” earns its keep — most people under-describe reality.
Options
What could you do? Generate several before judging any. The job here is breadth — “what else could work?” — not picking the winner yet.
Will / Way forward
What will you actually do, by when? Commitment, not intention. End with a concrete next step they chose, so the ownership is theirs.
GROW model — a four-stage structure for a coaching conversation: Goal (what they want), Reality (what's true now), Options (what they could do), Will/Way-forward (what they'll commit to). Its value is preventing the rush to solutions before the goal and reality are clear.
GROW is a map, not a script. Real conversations loop — new reality surfaces while exploring options, the goal shifts once reality is clear. Use the stages to notice “we're picking options before we agree on the goal”, then steer. Reciting the letters at someone is just telling in a costume.
CORE CONCEPTThe coaching habit: stay curious a little longer
Michael Bungay Stanier's The Coaching Habit compresses the whole skill into one line worth memorising: stay curious a little longer, rush to action and advice a little more slowly. You don't need to become a professional coach. You need to delay the moment you start solving by about thirty seconds — long enough for one more real question.
The enemy of this is what Stanier calls the advice monster — the compulsive urge to jump in with your idea the instant you half-understand the situation. The advice monster feeds on three illusions: that you have the answer (you usually only have an answer), that your job is to add value by talking, and that giving advice makes you look capable. It's fast, it feels great, and it quietly trains your people to stop thinking and just bring you problems to vend on.
Advice monster (Stanier) — the reflexive impulse to offer solutions, opinions, and answers before fully understanding the situation or letting the other person think. Feeding it feels productive but starves people of the chance to develop their own judgment.
The fix isn't to never give advice — it's to tame the monster: ask one more question before you reach for it. The cheapest, most repeatable coaching move in existence is the three-word question “And what else?” — it costs you nothing and consistently produces the better idea that was sitting behind the first one.
COMPARISONCoaching vs mentoring vs managing vs therapy
“Coaching” gets used loosely for any one-on-one. Keeping the lanes distinct matters, because using the wrong one quietly does harm — coaching when someone needs expertise wastes their time; mentoring when they need to think for themselves makes them dependent; wandering into therapy is both unhelpful and out of bounds.
| Mode | Core question | Source of the answer | Best when |
|---|---|---|---|
| Coaching | “What do you think?” | them | they're capable and the win is growth + ownership |
| Mentoring | “Here's what I'd do, and why” | you, from experience | they lack experience you have and want it |
| Managing | “Here's what needs to happen” | the role / the goal | direction, standards, and accountability are needed |
| Therapy | “What's underneath this?” | a trained professional | genuine psychological distress — refer out |
Mentoring vs coaching — mentoring transfers your knowledge and experience (“here's how I'd handle it”); coaching draws out their own thinking (“how are you thinking about it?”). Mentoring is telling with a warm relationship; coaching is asking. Good leaders fluidly switch between them — and know which one they're in.
Coaching touches feelings, but it isn't therapy. If a conversation moves into trauma, clinical anxiety or depression, or anything beyond work performance and ordinary stress, your job is to be humane and to point them to a professional — not to “coach” them through a mental-health issue you aren't trained for.
IN PRACTICEWhen to coach, and when to just tell
Coaching everything is a recipe for being the most exhausting person in the building. The honest rule is a quick triage on what the situation actually needs:
- True emergency? Prod is on fire, a deal closes in ten minutes, safety is at stake → tell. Coach later, in the retro. Asking “and what else?” while the building burns is malpractice.
- Pure knowledge gap? They literally don't know the API, the policy, the fact → tell (or point them to where it's written). Making them “discover” a fact only you have is theatre.
- Judgment, decision, or growth? There's no single right answer, or the point is to build their capability → coach. This is most of the interesting work, and it's where coaching pays.
And give the autonomy real teeth: when they reach their own answer, let them run it — even if yours was marginally better. That ownership is the entire point. A solution someone owns gets executed with energy; a solution imposed on them gets executed with compliance. This is autonomy and intrinsic motivation (2.4) in action: people commit to what they author. Helping them find their answer isn't just nicer — it produces better follow-through and faster learning than the right answer handed down.
Report: “Should I refactor the auth module now or ship the feature first?”
Advice-monster reply: “Ship first, refactor later.” (Done in three seconds; they learned nothing.)
Coaching reply: “What happens if you don't refactor before shipping?” → they reason aloud, spot the real risk, and decide — and next time they don't need you for this class of call at all.
WEAKNESSWhere coaching goes wrong
The default failure: you slip back into expert mode and just answer, because it's faster and feels good. It works once and costs you forever — your people stay dependent, you stay the bottleneck. If every problem still routes through you, you're managing a queue, not developing a team.
You already have the answer and you're asking questions to make them arrive at it. People see through this instantly, and it's worse than just telling — it's telling plus condescension. If you genuinely have the answer and it's time-sensitive, say it. Only ask if you're truly open to a different answer than yours.
Coaching a pure-knowledge gap or coaching during a real fire is slow and maddening. “What do you think you should do?” when the server is down and they need a command is not leadership — it's abdication wearing a halo. And remember: not everyone wants to be coached. Read the person and the moment.
KEY INSIGHTWhere this connects in your course
- Delegation (3.8) — coaching is how you build the capability that makes delegation safe; together they're how leadership scales beyond your own hours.
- Feedback (3.10) — feedback tells someone what to adjust; coaching helps them work out how. The strongest leaders weave the two: name the gap, then coach the path.
- Listening (1.15) — coaching is impossible without real listening. The question only works if you actually hear the answer instead of waiting to talk.
- Autonomy & motivation (2.4) — people commit to what they author; coaching produces ownership, which drives execution and learning that telling never does.
Coaching is the quiet force multiplier underneath everything else in this stage. Up next, you'll zoom out from the individual to the group — how the people you're developing actually work together (Team Dynamics).
GLOSSARYKey terms introduced
- Coaching
- Helping someone improve their own thinking and performance through questions and listening, rather than by directing or instructing.
- GROW model
- A coaching structure: Goal, Reality, Options, Will/Way-forward — run roughly in order to avoid jumping to solutions too early.
- Powerful questions
- Open, short, non-leading questions that prompt the other person's thinking rather than gathering facts for yours or steering to your answer.
- Advice monster
- (Stanier) The reflexive urge to offer solutions before understanding the situation or letting the other person think it through.
- Ownership
- The sense that a decision or solution is genuinely one's own — the source of energetic follow-through and real learning.
- Mentoring vs coaching
- Mentoring transfers your experience (“here's what I'd do”); coaching draws out theirs (“what do you think?”).
ACTIVE RECALLCheck your understanding
Answer in your own words first, then reveal.
1Why is the shift from “expert who tells” to “coach who asks” so hard for a strong technical person — and why is it worth it?›
Their whole reward history is built on knowing the answer, so deliberately not using that strength feels like driving with the handbrake on, and telling is genuinely faster in the moment. It's worth it because a leader's output is everyone they develop: keep vending answers and you stay the bottleneck; start growing thinkers and your leverage becomes the whole team's capability.
2What are the four GROW stages, and what's the common failure they guard against?›
Goal, Reality, Options, Will/Way-forward. The failure they prevent is rushing to options and solutions before the goal and the actual reality are understood — picking answers to the wrong, under-described problem.
3What's the “advice monster,” and what's the single cheapest move to tame it?›
The reflexive urge (Stanier) to jump in with solutions before understanding the situation or letting the person think. The cheapest counter-move is the three-word question “And what else?” — it costs nothing and reliably surfaces the better idea sitting behind the first one. The mantra: stay curious a little longer, rush to advice a little more slowly.
4Give two situations where coaching is the wrong tool and you should just tell.›
(1) A true emergency — prod is down, a deal is closing — where you tell now and coach in the retro. (2) A pure-knowledge gap — they simply don't know a fact, API, or policy only you have — where “discovering” it is pointless theatre. Bonus: fake Socratic “guess my answer” questioning, which is transparent and worse than telling, and the reality that not everyone wants to be coached.