GROW is the framework a large share of professional coaches run on. It was set out by Sir John Whitmore with Graham Alexander and Alan Fine in the early nineties, it is four stages, and there is nothing proprietary about it.
Goal. Reality. Options. Will.
Written down it looks almost too simple, which is why people read it and move on without using it. The value is not in the acronym. It is in being made to finish each stage before starting the next, which is the thing that does not happen when you think about a goal on your own.
Here is the whole thing as a session you can run in about forty minutes with a notebook.
Goal: what specifically, by when
The mistake at this stage is stopping at the topic.
"Get fitter" is a topic. A goal has an end state you could hold up against reality and get a yes or a no.
Ask yourself:
- What do I want to be true that is not true now?
- How would I know it had happened. What would I be able to point at?
- By when, and is that date real or is it decoration?
- Is this actually mine, or is it something I think I should want?
That last one is worth sitting with. A goal inherited from someone else's expectations behaves exactly like a goal you want, right up until the first hard week, at which point it evaporates and you conclude you are undisciplined.
Write one sentence. If it takes more than one sentence it is more than one goal.
Worth knowing: if the goal is one you have been carrying for years without starting, the wanting itself may be the obstacle rather than the plan. See why you procrastinate on the thing you actually want before you spend forty minutes on a session.
Reality: where you are, without the story
This is where solo sessions fall apart, because it is the boring stage and it involves saying unflattering things with nobody present to hear them.
- What have I actually done about this in the last three months? Not intended. Done.
- What have I tried before, and what specifically happened when it stopped?
- What resources do I already have that I am not using?
- On a scale where ten is finished, where am I, and what is the evidence for that number?
The last question is the most useful one in the framework. People give a number that feels right and then discover they cannot justify it, and the gap between the felt number and the defensible one is usually where the goal has been living for years.
Be specific about the last failed attempt. "I lost momentum" is not a finding. "I stopped the week my mother was in hospital and never rebooked the slot" is a finding, and it has an obvious fix.
That specificity requirement is the same one a weekly review runs on, and for the same reason: an abstraction cannot be acted on.
Options: generate before you evaluate
The rule is that nothing gets judged until the list is long.
Most people produce three options, dismiss two on practical grounds, and take the one they were already going to take. The generating and the evaluating have to be separate passes or the first pass never happens.
- What could I do, if resources were not a constraint?
- What would I advise a friend in this exact position?
- What is the option I have already thought of and rejected. Why did I reject it?
- What would the version of this be that takes an hour a week rather than ten?
- What have people who have done this actually done? Not what should work. What worked.
Ten options, then evaluate. The tenth is frequently better than the first three and it does not exist when the list stops at five.
Go further: the hour-a-week question is not a consolation prize. Fifty hours a year is a draft, a language to conversational, or a business with its first customers. Finding an hour that survives a bad month is the harder half of it.
Will: what happens this week
This is where the session either produces something or does not, and it is the stage most self-directed sessions skip entirely.
By this point you feel clearer, and clarity feels like completion. It is not. Clarity is the setup.
- What is the specific first action, and when this week does it happen?
- What is in the calendar slot that action needs. What moves?
- On a scale of one to ten, how likely am I to do it?
- What could stop it, and what is the plan if that happens?
- Who else knows this is happening?
The eight-out-of-ten test is the one to keep if you keep nothing else. A seven means you already know you probably will not do it, and the honest response is to cut the action in half and ask again rather than to promise harder.
The fourth question is a one-line pre-mortem, and it is worth doing properly rather than in passing. The full version takes twenty minutes and produces three or four if-then lines you can keep beside the plan.
The two stages that get skipped, and what happens
Run alone, this collapses in a predictable way, and it is always the same two stages.
Reality gets skipped because it is uncomfortable, so the options are generated for a situation you are not actually in. Will gets skipped because thinking about a goal produces a feeling of progress that closing the notebook does not disturb.
That is the specific job another person does in a coaching session: not supplying insight, but refusing to move on. It is one of the four things a coach actually does that never appear in the marketing. Someone asks the scale question and then waits through the silence. Someone comes back the following week and asks about the action, which means Reality next time is built on evidence instead of memory.
You can approximate the first part with discipline and a notebook. The second part is genuinely hard to arrange for yourself, and the arrangements that work are worth reading before you decide the framework alone is enough.
Running it on yourself: four practical rules
The framework assumes a second person. These are the substitutions that hold up.
- Write everything. Held in your head, Reality compresses to a vibe. On paper it has to be sentences, and sentences are checkable.
- Timebox each stage. Ten, ten, ten, ten. Without a clock, Options expands to fill the session because it is the enjoyable one.
- Do not skip forward when a stage is uncomfortable. Discomfort at Reality is the signal that the stage is working.
- Book the next session before you close the notebook. Otherwise Reality next time is memory, and memory is generous.
The session, in five lines
- One sentence for the goal, with a real date
- Reality in evidence, including a number out of ten you can defend
- Ten options before any evaluation
- One action, this week, with a day and a slot
- Under eight out of ten, halve it and ask again
Common questions
How long does a GROW session take?
About forty minutes alone, and it can run to an hour the first time on a goal you have been carrying for years. Sessions on the same goal get shorter, because Reality becomes a delta rather than a full account.
How is GROW different from OSCAR?
OSCAR is Outcome, Situation, Choices, Actions, Review, and the first four map closely onto GROW. The difference is the fifth: it builds the review into the model rather than leaving it to chance. For a goal that runs over months, that turns out to matter more than anything in the front half. A coach runs that fifth stage as a scheduled step rather than a hope, which is one of several things that change when a goal is set with somebody else in the room.
Can I run GROW on someone else without training?
You can run a useful conversation with it, and the thing to resist is answering. The value a second person adds is the question and the silence after it, not their view of your situation, and the strong temptation is to supply the view.
How often should I run one?
Quarterly on a goal, roughly, with a fifteen-minute weekly review in between. Running GROW weekly turns it into a ritual with no new Reality to report, and that is how people conclude the framework does not work.
Does it work on goals that are not personal?
It was built for workplace coaching and it moves across cleanly, which is part of why it spread. The stage that changes most is Reality, because at work other people hold half the evidence and you have to go and ask them for it.
The takeaway
The framework is four questions in an order, and the order is the whole product. Skip Reality and you plan for a situation you are not in. Skip Will and you have had a nice think.
Take two things if you take nothing else: the number out of ten you have to justify, and the eight-out-of-ten test on the action before you write it down.
One thing to do today. Run the last stage only. One action, this week, with a day and a slot, and the likelihood out of ten written next to it. If it is under eight, halve it and write the smaller one instead.
