The GROW model was introduced in the coaching block as a structured way to keep a conversation focused: Goal, Reality, Options, Way forward. I’ve been running one-on-ones in something close to this shape since the day I took this role, mostly because the best managers I worked under as an engineer ran their conversations this way and I’d been copying the shape without naming it. The formal version is letting me notice where the conversations get stuck — and gives me a way to teach the shape to my senior engineers.
The model, briefly
| Stage | Question | What it sets up |
|---|---|---|
| G — Goal | What do you want? | A specific, ownable outcome to point the rest of the conversation at |
| R — Reality | What’s happening now? | An honest picture of the current state, including the parts that are uncomfortable |
| O — Options | What are the possibilities? | A wider set of paths than the one already in the person’s head |
| W — Way forward | What will you do next? | A concrete commitment small enough to act on and big enough to matter |
The four stages aren’t equal in difficulty. Three of them are usually quick. The fourth is where the conversation either works or doesn’t. Which one is the bottleneck tells me what’s actually going on for the person — and that’s the diagnostic the framework gave me.
Where conversations get stuck, and what it means
Stuck on Goal. When someone can’t articulate what they want, the conversation isn’t about the work — it’s about something earlier. They might not have permission, internally, to want the thing. They might be confusing their goal with their manager’s goal, or their team’s. They might not have given themselves the time to figure out what they want at all. Pushing into Reality before Goal is clean is one of the most common things I notice myself wanting to do; the rest of the conversation drifts because there’s no anchor.
When this is the stuck stage, my move is to stay with the question longer, with variations. If nothing constrained you, what would you want? What would the “good” version of this look like in six months? Whose answer is the one you’ve been using until now? The point isn’t to extract a goal under pressure — it’s to give the person room to find the one they actually own.
Stuck on Reality. When someone can’t describe what’s happening clearly, two things are usually true. Either they don’t have the data they need, or they have it and are flinching from it. Both require a different next step. If it’s data, the coaching move is to help them go get it — and to pause the conversation rather than improvise around the gap. If it’s the second — and it often is — the coaching move is patience. The version of Reality the person can name in five minutes is rarely the version they need to face. Letting them sit with the question, without rescuing, is the work.
I’m a storyteller — my instinct is to help them articulate Reality by offering my read of it. The discipline I’m running is to wait until they say it, even if mine would be a tidier description. Their version is the one they can act on.
Stuck on Options. This stage is usually the easy one to power through, which is exactly why I have to be careful about missing what it’s telling me. When someone can only name one or two options, the conversation is happening too narrowly. They’ve already decided, and we’re reverse-engineering the justification. The coaching move is to widen, sometimes uncomfortably: what would you do if budget were doubled? what would you do if you couldn’t have any more headcount? what would the most cautious version of this look like? the most aggressive? The point isn’t that they’ll choose one of these. It’s that the contrast usually reveals what’s actually in play.
If this stage keeps being thin, it’s often a signal that the Reality stage didn’t go deep enough. People who haven’t fully named the reality can’t see the options that reality permits. The fix is upstream.
Stuck on Way forward. This is the stage I find hardest as a coach. The person has a goal, a clear picture of reality, several options — and then can’t commit to one. The temptation is to push for a commitment, and the version that comes out under that push is usually a polite, hollow one that the person will not actually carry out.
When the Way forward is the bottleneck, the question is almost always: what’s the smallest version of this you’d be willing to try by [date]? Smallness is the key. A commitment that’s safe to honour is one the person will honour. A commitment that’s too ambitious is one they will quietly let slide. Helping them find a version they can actually do, this week, is what the W stage is really for.
A worked example
A senior on my team came to me on Tuesday with a question about how to handle a colleague who’d been making decisions in his area without consulting him.
Goal: “I want the working relationship with him to be better.” That’s vague. I asked again. “I want him to bring me in before making decisions in my area.” Better. Specific, ownable.
Reality: “He’s been doing this for about three months. I haven’t said anything about it. I think it’s because he doesn’t realise it’s a problem.” We sat with that. I asked what evidence he had for the doesn’t realise read. He paused. “Actually, I’m not sure. I haven’t checked.” That was the real Reality, and naming it changed the rest of the conversation.
Options: First pass: “I could talk to him, or I could escalate to our manager.” That’s two. I asked for more. After some pushing: also possible — write a Slack message rather than a meeting, raise it in the next 1:1 instead of separately, ask a peer for their read first. Five options now. Different costs, different risks.
Way forward: First answer: “I’ll talk to him next week.” I asked what he’d do this week. He thought. “I’ll write what I want to say, by Wednesday, and decide on Thursday whether to send a Slack or schedule the conversation.” Concrete. Small. His.
That conversation took twenty-five minutes and changed something. The version of the conversation that was just answer-giving would have taken five minutes and changed nothing, because the answer would have been mine and not his.
Why this is a team effectiveness lever
The GROW move is a slow conversation that produces fast outcomes. A coached conversation takes twenty-five minutes; the answer-giving version takes five. The trade looks bad on a daily ledger and pays back enormously over a quarter, because the person who walked through GROW once with you can run it on themselves the next time. The five-minute answer doesn’t compound. The twenty-five-minute coaching session does.
Multiply that across the team. A team where each member can run GROW on themselves is a team where most of the routing-through-the-manager stops happening. The manager’s calendar opens up. The team’s velocity goes up. The throughput effect is exactly what coaching as a daily behaviour is meant to produce.
What the session added
Two things.
Names for the stages. I’d been running the shape by feel; now I have explicit labels. The labels let me notice in real time where in GROW are we right now, which tells me whether the conversation is moving or stuck.
A teachable shape. I’m going to walk my senior engineers through GROW so they can run it on the people they mentor. That’s the multiplier I keep coming back to in this series — the moves I’m making land for one person at a time; the moves I can teach land for many.
What I keep working on
I rush. The model invites silence, and silence makes me twitchy. I have to actively suppress the impulse to fill the pauses in someone’s Reality with my read of their Reality, or to push them to a Way forward when they’re still in Options. The rule I’m settling on: if I feel the urge to speak, count to five first. By the third second, the person usually starts speaking again, and they say something neither of us would have arrived at if I’d jumped in.
GROW isn’t four questions. It’s a structure that gives me a place to put the silence. That’s, almost entirely, what makes it work.