You have the skills. Rapport, listening, the right question at the right moment. Then a lesson debrief opens up, the pupil says something honest, and twenty minutes later you are three tangents deep, no closer to a decision, and the clock has beaten you both. Good coaching instinct with no structure spends itself and lands nowhere. The conversation needs a shape.
The shape is four steps: goal, reality, options, way forward. It has a name in the wider coaching world, GROW, which sounds like the sort of thing that arrives on a laminated card, and you can forget the acronym the moment you have the four words. What matters is the order. Each step sets up the next, and the whole thing keeps the thinking, and the responsibility, on the learner's side of the car.
Four steps, and none of them is telling
Read them in order before we take them apart. Agree what the pupil wants out of the time. Establish honestly where they are against it. Draw out the ways they could close the distance. Commit to what they will actually do. That is the spine. Notice what is missing from all four: you supplying the answer. The structure is built so the learner does the work, because a learner who reasons their way to a decision in the passenger seat can do it again alone at a junction, and a learner who was handed one cannot.
Start with the goal, or the hour has no owner
Getting the pupil to name what they want from the lesson is the first act of handing them responsibility. It matters far past today. Once they are driving solo they will make a hundred choices an hour with nobody to check them against, and the habit of setting their own target is where that starts.
The catch is familiar. You ask what they would like to work on and you get "I don't know, you're the instructor, you tell me." That is not a dead end, it is the exact reflex coaching exists to undo. Push gently past it with questions that give them something to reach for. What would you like to get out of today? Which part do you most want to sharpen? What would have to happen this hour for it to be worth your money? How do you want to feel about your roundabouts by the time we park up?
A goal worth agreeing is specific and checkable. "Get better at junctions" is a wish. "Approach the next six junctions without me saying a word, and mark each one myself" is a goal, because both of you will know at the end whether it happened.
Reality is where the honest gap gets measured
Once the goal is set, reality asks where the pupil actually stands against it right now. What can they already do, what will they need, what is the state of their nerves this morning. Often it is cleaner to run reality first and let the goal sharpen out of it, because a pupil rarely knows what to aim for until they have said out loud where they are.
This is where scaling earns its keep. On a scale of nought to ten, where ten is you could do that in your sleep and nought is you cannot do it yet, where are you with reversing around a corner? The number is never the point. The next question is. What puts you at a four and not a two? Now the pupil is doing their own honest read, which is the whole skill you are trying to build. Useful reality questions stay open. What do you know so far? What tends to go wrong for you here? Show me what you can already do. What is the risk if this stays as it is?
Options are theirs to generate, not yours to supply
Options is the step instructors hijack fastest, because you can see the answer and it is quicker to say it. Resist. The task here is to get the learner generating ways to close the gap, as many as they can, before any of them is judged. Quantity first, evaluation second. Run it in two passes. Get the list out, all of it, however unlikely. Then take the options one at a time and weigh them.
Questions that open the field: what could you try? What else? If I were not here, what would you do? What would a cautious driver do differently from a confident one? Then narrow: what would each of those cost you, and what would it get you? A pupil who has weighed three options and picked one has practised the exact thing that keeps them alive later, choosing under their own judgement. This is also where the linked craft of effective questioning does its heaviest lifting, because a lazy question here collapses the whole step back into you telling.
The way forward is a commitment, not a summary
The last step turns thinking into action the pupil owns. Do not summarise it for them. Make them say it. Talk me through what you are going to do. How will you make that happen at the next roundabout? What are you taking away from this? What do you need from me to pull it off? How will you know it worked?
That final question closes the loop back to the goal, and it hands the pupil their own measure of success. When they set it, own it and can check it themselves, you have built a scrap of the self-evaluation that has to run the whole time once you are no longer in the car. That is the same feedback loop the rest of the craft is aimed at, treated in full under feedback and self-evaluation.
The shape is a rail, not a script
Held too tight, the four steps turn into an interrogation, and a pupil who feels processed shuts down as fast as one who feels judged. The order is a rail to stop the conversation sliding off, not a form to fill. You move between the steps, loop back when reality shifts the goal, skip lightly through one when the pupil is already there. When a genuinely new skill is on the table, plenty of plain instruction still belongs inside the frame. Guide, prompt, then let them run it. The structure holds the coaching. It does not replace the teaching, and knowing the ratio is its own job, worked through in goal setting and in lesson planning.
The point of the shape is not tidiness. It is that a conversation with a spine actually reaches a decision the learner made, and a decision the learner made is the only kind that gets in the car with them after you climb out.