Skip to content

01 · Coaching Fundamentals for Leaders

Most managers, when a report brings them a problem, solve it. It feels efficient and it feels helpful, and in the short term it is both. The cost is that the report leaves with a solution but no new capability, and comes back with the next problem, and the one after that. Coaching is the discipline of helping someone find their own answer — slower in the first conversation, dramatically faster by the tenth, because the person is now solving that whole class of problem without you. This module gives you a specific structure (GROW), a question bank, and a worksheet you can put in front of yourself during a real one-on-one this week.

1. Coaching is not the four things you already do

Mode You supply Use it when
Telling The decision and the reasoning It's your call, it's urgent, or safety/compliance is involved
Advising An opinion they can take or leave They asked, and you have context they genuinely lack
Mentoring Your own experience and career pattern They want to know how someone who's done it did it
Counselling Support for a personal difficulty The issue is wellbeing, not performance — often refer on
Coaching Questions, structure, and attention They have or can get the information; the gap is thinking, confidence, or ownership

The failure mode isn't using the wrong mode occasionally — it's using telling as the default for everything, so people stop bringing you problems until they're already expensive. A useful rule: if you could honestly say "you probably know more about this than I do," coach. If you genuinely hold information they don't, give them the information first, then coach on what to do with it.

2. The GROW model

GROW (Whitmore) has four stages, and most of its value is in doing them in order. Managers who "know GROW" almost always jump from a two-sentence Goal straight to Way Forward, which is just advising with extra steps.

Stage Question it answers Signal you can move on
Goal What does this person want out of the situation, and out of this conversation? They can state it in one sentence, and it's theirs, not yours
Reality What is actually happening now — facts, not interpretation? You have heard something you did not already know
Options What could they do? All of it, including bad ideas Three or four genuinely different options exist
Way forward What will they do, by when, and what might stop them A specific action, a date, and a named obstacle

A rough time split for a 30-minute coaching conversation: 5 minutes Goal, 12 minutes Reality, 8 minutes Options, 5 minutes Way Forward. If you're spending less than a third of the time on Reality, you are very likely solving the wrong problem.

3. The GROW worksheet

Keep this open — paper or a shared doc — during the conversation, and fill it in with the person rather than about them afterwards.

GROW WORKSHEET
Person: ____________________  Date: __________

GOAL
  What they want from this conversation:
  What "good" looks like in three months:
  Whose goal is this, really? (their words / my words)

REALITY
  What is happening now (facts, examples, dates):
  What have they already tried, and what happened?
  Who else is involved, and what would they say?
  What is in their control vs. not:

OPTIONS
  1.
  2.
  3.
  4. (the one they think is silly)
  Which option is most attractive, and why?
  What would they do if resources and permission were unlimited?

WAY FORWARD
  Action:
  By when:
  What might get in the way:
  What support they want from me (specifically):
  How we will both know it worked:

4. A question bank

Questions, not statements with a question mark on the end. "Don't you think you should loop in Finance?" is advice in a disguise, and people hear the difference immediately.

Goal

  • "What would you like to be different by the end of this conversation?"
  • "If this went really well over the next quarter, what would be true?"
  • "Is that the goal, or is that a step toward something else?"

Reality

  • "What have you tried so far, and what happened?"
  • "What's the evidence for that?" / "How do you know?"
  • "What would the other person say if they were sitting here?"
  • "Which part of this is actually within your control?"

Options

  • "What options do you see?" — then, when they stop, "What else?" The best option very often arrives after the third "what else?"
  • "If budget and permission weren't a constraint, what would you do?"
  • "What would you advise a peer in exactly this situation?"

Way forward

  • "Which of those will you actually do?"
  • "By when? What's the first concrete step, and when this week?"
  • "On a scale of 1-10, how likely are you to do that?" Anything under 8, ask "what would make it a 9?" and change the action, not the person.
  • "What do you want from me between now and then?"

5. A worked coaching conversation

Maya manages a six-person customer operations team. Her report Dev has been asked to run the department's quarterly business review for the first time, and comes to her clearly hoping she'll just tell him what to put in it.

Dev: Can you send me last quarter's deck? I'll copy the structure.

Maya (Goal): I can — but before I do, what do you want to walk out of that review having achieved?

Dev: Um. Get through it without embarrassing myself?

Maya: That's a fair floor. What's the ceiling — if it went really well, what's different afterwards?

Dev: Honestly? They'd approve the second escalations analyst. We've been asking for two quarters.

Maya: So the goal for the review is a decision on headcount, not a status update.

Dev: ...Yeah. That's actually the goal.

Maya (Reality): Okay. What happened the last two times you asked?

Dev: Declined both times. First time they said "revisit next quarter." Second time nobody really said why.

Maya: What do you know about why?

Dev: I don't, really. I assumed budget.

Maya: What's the evidence for the budget theory?

Dev: ...None. I never asked. Priya in Finance would know.

Maya: What else is true about the current situation?

Dev: Escalations are up 40% year on year and our response SLA has slipped three months running. I have that data.

Maya (Options): So you have the numbers, and you don't have the reason for the previous no. What could you do about the second one?

Dev: Ask Priya before the review.

Maya: What else?

Dev: Ask the ops director directly — she was in both meetings.

Maya: What else?

Dev: I could talk to whoever did get headcount approved last quarter and find out how they framed the ask.

Maya: What else? Even a bad one.

Dev: Propose a six-month contractor instead of a permanent hire as a fallback. Honestly, that's probably the one that gets a yes.

Maya (Way forward): Which of those are you doing?

Dev: Priya and the ops director this week, and I'll build the deck around the SLA trend with the contractor option as plan B.

Maya: By when, specifically?

Dev: Message both today, meetings by Thursday, deck draft Monday.

Maya: Scale of 1-10 — how likely is the ops director conversation actually happening by Thursday?

Dev: ...Six. She's hard to get time with, and I'd be nervous.

Maya: What would make it a nine?

Dev: Asking for fifteen minutes with a specific question, instead of "can we chat." And sending it today rather than sitting on it.

Maya: Do that. What do you want from me?

Dev: Don't come to the review. If you're in the room they'll look at you for the answer.

Maya: Agreed. I'll read the draft Monday, and I won't attend.

Notice what Maya never did: she never mentioned Priya, the contractor option, or the SLA data. Dev held every piece of it. What he lacked was a goal beyond "survive," and a reason to test his own assumption about budget. Elapsed time: about eleven minutes, against the three it would have taken to forward the old deck — and Dev now runs these himself.

6. Two failure modes to watch in yourself

  • The leading question. If you have a specific idea, say it plainly and label it — "I've got a suggestion, do you want it?" — then hand control back. Disguised advice is worse than open advice, because the person learns you steer while pretending not to, which quietly costs you the trust you built in Level 1.
  • Coaching when you should tell. If someone is about to breach a policy, is missing information only you have, or is in genuine distress, "what options do you see?" is not servant leadership, it's abdication. Coaching is a mode, not a personality.

How It Actually Works

The GROW model's power comes from a specific cognitive-science finding about how insight is generated versus transmitted: self-generated solutions are encoded and retained differently than received ones. When a coach asks a question that leads someone to their own answer, that answer is produced through the person's own retrieval and reasoning pathways — which means it's stored alongside the reasoning that produced it. A solution handed to someone by an advisor is stored as an isolated fact, disconnected from any reasoning trace, which is why advice is so often forgotten or misapplied in a slightly different situation: there's no retained process to adapt, only a memorized answer that doesn't generalize.

Why the order (Goal, then Reality, then Options, then Will) matters and can't be reordered casually. Jumping to Options before Reality is fully explored means generating solutions to an incompletely specified problem — the coaching equivalent of debugging code without reproducing the actual bug first. And skipping straight to Goal-setting without first establishing Reality anchors the goal to an idealized, not actual, starting point, which routinely produces plans that don't survive contact with constraints the person didn't articulate because no one asked. The model's ordering exists specifically to prevent premature solutioning, which research on both software debugging and clinical diagnosis identifies as one of the most common and costly failure modes in complex problem-solving.

Why "the two failure modes" (rescuing and interrogating) both defeat the mechanism. Rescuing — sliding into giving the answer — reverts the interaction to advice-transmission, losing the self-generation effect described above. Interrogating — a rapid-fire sequence of closed questions — triggers the same identity-threat response covered in the feedback module (it starts to feel like being tested, not explored), which shifts the person's cognitive resources toward self-protection rather than open reflection. Effective coaching questions are deliberately open and spaced, because the goal is to give the prefrontal cortex room to actually construct an answer, not just retrieve a rehearsed one.

Exercise

Pick one real conversation coming up this week where your instinct is to give the answer — a report asking how to handle a stakeholder, a peer asking what to prioritise. Copy the GROW worksheet from section 3 and run the conversation with it in front of you. Hold two rules strictly: ask "what else?" at least three times in Options, and end with the 1-10 check on the action.

Afterwards write down three things: one fact you learned in Reality that you did not know going in; the option they produced that you would not have thought of; and how long the conversation actually took versus how long telling them would have taken. Keep this note — you will reuse it in the Module 10 project.