GROW is a simple four-step structure for a coaching conversation, and its simplicity is exactly why it's held up for decades of use in management practice: Goal, Reality, Options, Will. Each letter is a phase of the conversation, not a rigid script - the point is making sure all four actually happen, in roughly that order, rather than jumping straight from "here's my problem" to "here's what you should do," which is the default most people fall into without a structure.
Goal: what does this conversation need to accomplish?
Before diving into the problem, get explicit about what a useful outcome looks like - for this specific conversation, not necessarily the person's whole career or project. "What would make this conversation worth having?" is often enough. Skipping this step is the most common failure mode: without it, a coaching conversation drifts into venting or problem-description with no clear endpoint, and both people leave unsure whether anything was actually resolved.
Reality: what's actually true right now?
A clear-eyed look at the current situation, separate from both the goal and the eventual plan. This is where a coach's real value often shows up - not by providing answers, but by asking the questions the person hasn't asked themselves: What have you already tried? What's actually gotten in the way? What does this look like from someone else's perspective? The goal is an accurate picture, not a flattering or catastrophizing one.
Options: what could actually be done?
Generate real possibilities before evaluating any of them - a common mistake is jumping to evaluating the first option that comes up, which quietly forecloses better ones that hadn't been named yet. "What else could you try?" asked two or three times past the first answer usually surfaces options neither person had considered at first. This is also where a coach's own suggestions belong, if at all - offered late, as one option among several, not as the answer the whole conversation was steering toward.
Will: what will you actually do?
The step most often skipped or rushed, and the one that turns a good conversation into an actual change. Not "what should you do" in the abstract, but a specific, checkable commitment: what exactly, by when, and how will both people know it happened. A conversation that ends at Options without a real Will step is pleasant and often forgotten within the week.
Where GROW commonly breaks down in practice
The most common failure isn't skipping a step outright - it's rushing through Goal and Reality to get to Options, because generating solutions feels more productive than the two steps that precede it. The result is a set of options generated against an unclear goal and an incomplete picture of reality, which tends to produce a plan that doesn't actually fit the real problem. The fix is almost always slowing down, not adding more structure - genuinely staying in Reality a beat longer than feels natural before moving on.
The second common failure is a coach who nominally asks GROW-shaped questions while already having a specific answer in mind, steering the Options step toward it rather than genuinely opening the space. The tell is usually subtle - a slight impatience when the other person's suggestions diverge from what the "coach" was expecting - and the fix is mostly self-awareness: if you already know what you think they should do, that's advice, not coaching, and it's more honest to just say so directly.
Want a structure for actually preparing a real one-on-one, not just this one model in isolation? The Compassionate Accountability Planner's one-on-one planning mode builds a full agenda around exactly this kind of conversation.