How do you decide which goal belongs in the Goal step of a GROW conversation?

What each one is actually for

Regret Minimization Framework

The regret minimization framework is for the moment when comparing the options has stopped producing anything. You have listed the pros and cons more than once, the list keeps coming out even, and the deadlock is not about information. So you change where you are standing rather than what you are weighing: project forward to eighty, put that person in front of this exact fork, and ask which version they would still be sore about. What makes it work, when it works, is that the pressures making the choice feel impossible right now — how it looks this year, the cost of the first uncomfortable month, the fear of visibly trying — do not survive being viewed across forty years, while what you genuinely care about does. It returns a verdict, usually fast, and then it is finished. It is not a planning tool and it does not know what Tuesday looks like.

GROW Model

The GROW model is a way of running a conversation, usually with someone else asking, so that a vague wish becomes something with a date on it. Developed in coaching practice associated with Graham Alexander, Alan Fine and John Whitmore, it moves through four stages in order and refuses to skip. Goal: what do you want, stated concretely enough that you would know if you had it. Reality: what is actually happening now — what you have already tried, what the evidence is, who else is involved — kept scrupulously separate from what you wish were happening. Options: everything you could do, generated before any of it is evaluated, because judging as you go kills the unusual ideas first; the coach's job here is mostly to keep asking what else after you think you are done. Will: what you will actually do, by when. The last stage has a piece of technique in it that most people skip and that does much of the model's real work — you rate, on a scale of one to ten, how likely you are to do the thing you just committed to, and if the number is below about eight you change the action until it is not.

What you get out of pairing them

What the age-eighty question buys a GROW conversation is protection against the most expensive failure a coaching session can have, which is being excellent about the wrong thing. A session where Reality was honest, the Options were genuinely creative, and the Will step produced a dated action can still be a waste if the Goal at the top was picked because it was the one that was easiest to articulate at ten past the hour. Nothing later in the sequence checks it. GROW has no step called are you sure this is the one, and everything after Goal quietly assumes the answer is yes. What GROW buys someone who uses the regret question is the half of the process that framework openly does not have. A regret verdict is very easy to agree with — yes, I would regret not doing this — and just as easy to still be agreeing with in eleven months, unchanged, because agreeing with a verdict and having a plan are different events that feel similar from inside. The concrete gain from the pair is narrow and worth having: the goal at the top is one you would defend at eighty, and the bottom of the page has a specific action with a date and a realistic chance of happening.

What Regret Minimization Framework rests on

• Self distancing improves judgment randomised trials
  Evaluating a situation from a distanced, third-person vantage point instead of an immersed first-person one improves the quality of judgment about it.
• Long term framing surfaces values practitioner report
  Framing a decision against a long time horizon surfaces which option better matches a person's deeper values and cuts through short-term ambiguity.

What GROW Model rests on

• Implementation intention closes intention action gap randomised trials
  Pre-committing to a specific situational cue paired with a specific response lets the behavior fire automatically when the cue occurs, without depending on in-the-moment willpower.
• Self generated options increase commitment observational studies
  Options a person generates themselves produce more ownership and commitment than options handed to them.
• Structured reflection clarifies goals observational studies
  Walking through Goal and Reality in a fixed sequence clarifies what a person actually wants and where they currently stand.

Where they actually meet

Both practices keep the person as the source of the answer and change only the position the question is asked from — that is the shape they have in common, and it is unusual enough to be worth naming. A GROW coach who supplies the answer has broken the model, because the whole bet is that a plan you produced yourself is one you will still be following in March, while a good plan handed to you tends not to survive first contact with a bad Tuesday. The regret framework makes the same bet in a different medium: it hands you no conclusion, only a vantage point, and whatever the eighty-year-old says is your own material arriving from an angle you cannot reach while standing in this month. There is a second thing they share, which is a discipline about not letting stages collapse into each other. GROW insists that generating options and judging them are separate activities, done in that order, because judging as you generate reliably kills the unusual idea before it is finished. The regret framework insists on the same separation one level up: decide what matters first, and only then decide what to do about it, because working out how to do something is a very effective way of never asking whether it is the thing. The seam between them is simply where each one starts and stops. One answers is this the right goal and goes no further. The other picks up at now what and assumes the first question is closed.

The principle underneath both

A question asked from a different position than the one you are stuck in gets further than more thinking from inside it — and the answer still has to be yours, or it will not survive contact with an ordinary week. Both practices supply a structure for the asking and refuse to supply the content. What differs is only the relocation: the coaching sequence moves you between stages that you would otherwise let blur together, and the age-eighty question moves you along time so that this season's pressures are too far away to be heard.

What that makes the real goal

Ending up with an answer you will actually act on — which requires two things that come apart more often than people expect. It has to be about something you would still defend in forty years, and it has to be something you generated yourself, in a form specific enough to happen on a particular day. Stated that way the goal has nothing to do with coaching or with decision heuristics. It is what a good therapist, a good manager and a serious personal review are all trying to produce, and you can look anywhere for the pieces.

Other practices that serve the same goal

WOOP is the closest single practice that already does both halves and it is the best-evidenced thing on this page. Developed by Gabriele Oettingen out of research on what she calls mental contrasting, it runs four steps in order: name the wish, picture the best outcome vividly, then deliberately name the obstacle inside yourself that will actually get in the way, and finally write an if-then plan for that specific obstacle. It has randomized-trial support across several domains, and the reason it works is thought to be the pairing rather than either half — imagining the outcome alone tends to sap effort rather than build it, and it is the collision with the obstacle that converts the wish into energy. Motivational interviewing supplies a sharper version of GROW's commitment number: the confidence and importance rulers, where you are asked how important this is out of ten, then why that number and not a lower one, which reliably produces the person's own case for change rather than yours. It has trial support as a counselling style. And a decision journal serves the same objective from the other end, replacing an imagined eighty-year-old with a real record — you write the choice, the reasoning and what you expect, and reading it back a year later tells you things about your own judgment that no thought experiment can.

A tradition that already does this

The Spiritual Exercises of Ignatius of Loyola, written in the sixteenth century, already contained both moves in one procedure, guided by another person in a fixed sequence — which is to say, a structured coaching conversation four hundred years early. The section on making a decision offers several methods to be used when the choice will not settle. One asks you to imagine a stranger in exactly your position and consider what advice you would give them. Another asks you to picture yourself at the point of death and ask what you would then wish you had chosen. And the process does not end at the verdict: the method requires that the choice be confirmed, made concrete, and acted on, precisely because the tradition knew that a decision reached in a moment of clarity has a habit of dissolving on the walk home. That is the same worry the commitment number at the end of a GROW session is built to catch.

Using the Regret Minimization Framework to enhance the GROW Model

The place to spend this is the first ten minutes, when there are two or three plausible goals on the table and no way to choose between them — the point where sessions either pick one arbitrarily or spend the whole hour litigating which one is real, and neither outcome is any good. Write each candidate as a single sentence on its own line. Then ask the question about each in turn: at eighty, looking back, which of these left undone would still sting, and which of them, done, would that person not remember happening. Ask it in that pair, because the second half does as much work as the first — the goal that survives is usually not the most impressive one, it is the one whose absence keeps registering. Two rules make this usable rather than disruptive. First, if you are the one coaching, the question is yours to ask and the answer is not yours to have; you offer the vantage point and then you are quiet, because a goal you supplied is exactly the kind that gets agreed to warmly and abandoned in a fortnight. Second, timebox it hard and close it. Five minutes, before Goal is written down, and then the goal is locked for the rest of the session. Reopening the goal during Options is the classic way one of these conversations dissolves into a pleasant hour that produced nothing, and it happens because Options is where the discomfort starts and revisiting the goal is a comfortable place to retreat to. This is a bounded tiebreaker at one specific point in the sequence, not a replacement for the Goal step and not a standard opener for every session.

Using the GROW Model to enhance the Regret Minimization Framework

Once the age-eighty question has produced a verdict, run it straight into GROW starting at Reality, because Goal has just been settled and going over it again is procrastination in good clothes. Reality first, and the discipline here is to write only what is true today rather than what is true in general: what you have actually done about this so far, what evidence you have rather than what you assume, who else is affected and whether you have asked them. Most people discover their Reality section is much thinner than expected, and that thinness is itself the finding. Then Options, and the rule that makes this stage worth anything is that you generate before you judge — six at minimum, written down, including the ones that are obviously bad, and then ask what else three more times after you feel finished, because the interesting option is almost never in the first four. If you are stuck, use the unlock questions coaches use: what would you do if it could not fail, what would you do if money were not part of it, what would you tell a friend in this exact position. Then Will, which is where this stops being a conversation. One action, with a date and a time, small enough to be real — not begin the transition but book the conversation with my manager for Thursday at nine. Then do the thing almost everyone skips: rate out of ten how likely you actually are to do it. If the number is below eight, do not motivate yourself, change the action. Shrink it until the number is an eight, then write it as a specific if-then line — when Thursday's calendar reminder goes off at eight forty-five, I open the message and send it. That single number is the closest thing either of these practices has to a working lie detector, and a regret verdict that never gets one is the kind that is still true and undone next November.

The offer letter, open in another tab since Monday

It is Thursday and the tab is still there. You have been toggling between take it and stay since the weekend, and every argument you construct gets answered by the next thought about an hour later, which is how you know more thinking is not going to finish this. So you stop arguing and ask the other question instead: at eighty, looking back at this exact fork, which version of not-choosing actually stings. The answer arrives faster than the last four days would have predicted and it is not close — you would regret staying. Not because the new job is obviously better on paper. Because you can already see the shape of the next three years in the current one, and the eighty-year-old is unmoved by the reasons that have been holding you, which are mostly about how it will look to leave now. Fine. That is the Goal, and it is closed. The letter is still unanswered, though, because I would regret staying is not a plan and does not tell you what to say to your manager. So: Reality. You have told nobody. You have not read the contract's notice clause. You do not know whether the handover you are worrying about is genuinely three months or whether you invented that. Options, and you make yourself get to six before judging any: accept as offered, accept and ask for a later start, negotiate the start date and the salary, decline, tell your current manager and see what happens, ask the new firm for a week. Then Will, one action with a time on it: send the acceptance Friday at noon, after reading the notice clause tonight. How likely, out of ten, that you send it at noon? Six. So it changes — not send the acceptance, but draft the acceptance tonight and put it in the folder, then press send at noon. Nine. That is the version that happens.

When this is worth reaching for

Three sessions in, and the goal written at the top of the page has been different each time. Nothing in the structure catches that, because each individual conversation was competent; the drift only shows up if you put the three pages side by side, and when it does it usually means the real question has never been asked out loud. That is one condition. Another: a decision has been genuinely settled — you have said it to someone, you are not arguing with yourself about it any more — and eleven months later nothing has happened, not because you changed your mind but because there has never been a specific action with a date attached. A third, and it belongs to whoever is coaching rather than being coached: the Options stage keeps collapsing back into a debate about whether the goal is right, which is nearly always a sign the goal was locked before it was settled. Not, though, when the goal is clear and agreed. Then adding this is a step the conversation does not need, and it will feel like one.

Where this does not apply

The framework is out of scale for most of what gets coached. It works on choices with real weight and a long horizon — the kind you might still be turning over at eighty — and a GROW conversation about restructuring a weekly meeting, or getting better at delegating, is not that. Asked about a smaller reversible goal, the age-eighty question produces a shrug rather than clarity, and the shrug is contagious: it can make a perfectly good, modest goal feel too trivial to work on, which is a way of doing nothing that feels like having standards. Where the Goal is already clear and agreed, skip this entirely; the pairing is for the specific case where Goal is contested or vague. There is also a limit that matters if you are coaching someone else rather than yourself. A regret question aimed at another person's life is a powerful thing to hand them and it can open material that a performance conversation is not the right container for — a marriage, a bereavement, a sense that the last decade went somewhere they did not choose. If that opens, stop coaching toward the original goal, say plainly that this seems bigger than the session, and do not press on. And for someone in a depressive episode the question is unreliable in a particular direction: low mood changes what an imagined future self appears to say, and a verdict reached from inside it is not the neutral reading it feels like.

What the evidence does and does not support

The pieces here are unevenly evidenced, and the strongest support sits at the very end of the sequence rather than at the front. The Will step's if-then commitment — deciding in advance that at a specific moment you will take a specific action — is among the most replicated findings in behaviour-change research, tested across health, study and work behaviours in randomized trials. The distancing move at the front, judging a situation as an observer rather than from inside it, also has randomized experimental support in the judgment literature. Between those two poles the ground is softer. GROW's other components — that options you generate yourself produce more commitment than options handed to you, and that walking Goal and Reality in a fixed order clarifies what you want — rest on observational work and on decades of practitioner report rather than on controlled comparison against other conversation structures; the model as a whole has never been isolated in a trial, partly because it is hard to blind a coaching conversation. The age-eighty framing has no trial behind it at all and is a heuristic that spread because it is memorable. Nobody has tested using a regret check to resolve a contested Goal step. What would have to be run is a comparison of coaching conversations where the goal was tie-broken this way against ones where it was not, followed long enough to see which goals were still being pursued a year on — and no such study exists. Until it does, this sequencing is a reasoned application of each practice's documented strength to the other's documented gap.