Did That Coaching Commitment Actually Work, or Did It Just Feel Good?

What each one is actually for

PDSA Cycle

Plan-Do-Study-Act comes out of quality-improvement work, and the whole point is to stop trusting a change because it felt like it worked. You write down, before you do anything, exactly what you expect to happen if you make one small change — not a vague hope, an actual prediction you could be wrong about. You make that change for real, but on a small, contained scale: one week, one habit, one room, not your whole life at once. Then you look at what actually happened and hold it up against what you predicted, on purpose, before you decide anything. Only after that comparison do you decide whether to adopt the change for good, adjust it and run it again, or drop it because it didn't do what you thought.

GROW Model

GROW is a sequence for a coaching conversation, whether you're running it with a professional coach, a mentor, or by yourself with a notebook. Goal asks what you actually want out of this — from the conversation, the month, or the situation. Reality asks what is actually true right now, stripped of the story you tell about it: what have you tried, what's actually in your way, what resources do you already have. Options asks you to generate more than one path forward, on purpose, even the ones that seem unlikely, before picking. Will is the step most people skip in their head but the structure forces onto the page: what will you actually do, starting when, specifically enough that there's no ambiguity about whether you did it.

What you get out of pairing them

PDSA buys you an actual answer to a question most people never ask about their own habits: did that change work, or did it just feel like it worked because you wanted it to? Without a written prediction to check against, you'll misremember — you'll credit a change that did nothing, or drop one that was working before you'd given it long enough. GROW buys you the thing a coaching conversation usually loses on the walk back to your car: a specific action, not a feeling of clarity. Feeling clear about your goal is not the same as knowing what you're doing on Tuesday. Put together you get a third thing neither one supplies alone — a commitment that gets checked against reality instead of just made and forgotten, so the next conversation starts from what actually happened instead of another lap through the same unresolved thing.

What PDSA Cycle rests on

• Feedback against prediction corrects belief plausible mechanism
  Comparing the actual result of a change against the predicted result before adopting it corrects mistaken beliefs faster than adopting or rejecting changes based on intuition alone.
• Small scale testing precedes scaling observational studies
  Running a change through a small planned test cycle before wider adoption reveals real-world effects while limiting the risk of being wrong.

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

PDSA's Do step and GROW's Will step are doing the same job with different vocabulary: both refuse to let the process end at a decision. PDSA won't let 'we should try eating dinner earlier' count until there's an actual, dated, small-scale test of it. GROW won't let 'I want to feel less behind at work' count until there's an actual, dated action attached to it. Neither structure trusts a good intention to survive on its own once the conversation ends — both build the specific next step into the process itself, not left as assumed follow-through.

The principle underneath both

A plan works better when it's tied to a specific moment and a specific action, rather than left as a general intention. 'I will exercise more' rarely survives contact with a real Tuesday. 'When I get home from work on Tuesday, I will put my shoes on before I sit down' has a much better shot, because the cue itself — walking in the door — triggers the behavior, instead of you having to notice the cue, remember the goal, and decide to act, all in the moment, on willpower alone. GROW's Will step is built to produce exactly this kind of if-then commitment. PDSA's Plan step does something adjacent from the other direction: it forces you to write down, in advance, exactly what you expect the specific action to produce, which is what makes it possible to check afterward whether the commitment was actually worth keeping.

What that makes the real goal

If tying a plan to a specific cue and a specific action is what makes it survive a normal week, the goal both structures are chasing is the same one: closing the gap between deciding to do something and actually doing it, without depending on remembering to feel motivated in the moment it needs to happen.

Other practices that serve the same goal

This gap gets closed by several practices that have nothing to do with quality-improvement charts or coaching sessions. WOOP — Wish, Outcome, Obstacle, Plan, a method developed by the psychologist Gabriele Oettingen — pairs a vivid picture of a wish with an honest obstacle and then an if-then plan, and has randomized trial support for producing more follow-through than positive thinking alone. Habit stacking ties a new action to an existing routine you already do without thinking — 'after I pour my coffee, I will write one sentence' — borrowing a cue that already fires reliably instead of building a new one from scratch. Commitment contracts, where you put money or a public statement on the line if you don't follow through, work on the same gap from the outside: they don't make the action easier, they make not doing it more costly.

A tradition that already does this

Field trials in agriculture were running plan-test-compare-decide cycles before Walter Shewhart and later W. Edwards Deming formalized the four-step version for factories in the mid-twentieth century — test a change on a small plot before trusting it on the whole field is older than either the name PDSA or the discipline of quality management. GROW's own history runs the other way: Sir John Whitmore built it out of coaching techniques used with athletes, where 'try the adjustment in practice, then decide if it's real' was already how a coach and an athlete worked together, long before it became a four-letter word for a conversation in an office.

Using PDSA to Check a GROW Commitment

A GROW conversation usually ends at Will — you've named the action, you know when you're doing it. What it doesn't build in is a way to tell, next time you sit down, whether that action was actually worth keeping. Borrow PDSA's Study step: before you close the conversation, write down what you predict will happen if you do the Will action for a week — not a hope, a number or a specific observable change you'd expect to see. 'If I block two hours on Tuesday mornings for the proposal, I predict I'll have a full draft by Friday instead of a half-finished one.' Then when you come back to Reality next time, you're not asking 'how did it go' from memory — you're comparing what you predicted against what actually happened, a completely different, much more honest conversation. It also gives you three real options instead of one vague verdict: the prediction held, so keep it; it partly held, so adjust the specific action and test again; or it didn't hold at all, so the Will you committed to wasn't the right lever, and Options needs another pass.

Using GROW to Set Up a PDSA Cycle

PDSA's Plan step asks you to predict what a change will do, but it doesn't ask why you're bothering to test that particular change — and a test with no real stake in the outcome is easy to abandon halfway through. Run a GROW pass before you plan the cycle. Goal: what do you actually want that this specific change might get you — not 'eat dinner earlier' but 'stop being wired at 11pm and getting five hours of bad sleep.' Reality: what's actually happening now, specifically — what time dinner lands, what you eat, what you do in the two hours after. Options: more than one candidate change to test, not just the first idea — earlier dinner, less caffeine after 2pm, a walk after eating — so the one you pick is the one you'd actually bet on, not the only one you thought of. Will becomes the PDSA Plan itself: the specific, dated, small-scale test, with your prediction attached. The GROW pass doesn't replace PDSA's rigor — it makes sure the thing you're rigorously testing is a change you actually have a reason to want.

Wednesday night, deciding whether the new bedtime actually did anything

You moved your last meal to 6:30 instead of 8:00 four days ago, and it's Wednesday night and you're trying to figure out if it's working. If you'd only run GROW, you have a feeling — you think you're sleeping better, maybe, though it's hard to say, and there's a real chance you're just telling yourself that because you wanted the change to work and don't want to admit four days of earlier dinners for nothing. But you wrote a prediction on Sunday when you planned this: you expected to fall asleep within twenty minutes instead of your usual forty-five, based on three nights of data, not one. You check your notes. Two of the three nights, you were out in under twenty-five minutes — close, real, worth another week. One night you were up past midnight, but that was also the night you had wine with dinner, which you'd noted as a possible confound before you even started. You're not guessing anymore. You have something to compare against what you actually expected, which is the difference between deciding you're 'basically sleeping better' and deciding, specifically, that earlier dinner is worth one more week before you call it a real change.

When this is worth reaching for

Three different "this year will be different" resets by March, or a manager whose weekly check-ins always end in a new goal and never in "did last week's actually work" — that's the shape of over-committing to too much and under-checking whether any of it moved. The same gap shows up anywhere being wrong about a change costs something real enough to be worth a week of data first — a new work routine, a training change before a race — rather than something you can just try and immediately feel the difference from.

Where this does not apply

If you already know for certain what works — you've tried the thing before, multiple times, with a clear result — running a formal prediction-and-check cycle on it is just friction, not rigor. Sixty seconds of writing down a prediction is not the bottleneck for putting your shoes by the door; it's the bottleneck for something you're genuinely unsure about. And GROW's structure can feel heavy-handed for a decision that's actually simple and immediate — you don't need Goal, Reality, Options, and Will to decide whether to call the plumber today. Save the pairing for changes complicated or uncertain enough that a plain gut check has actually failed you before.

What the evidence does and does not support

GROW's Will step and PDSA's Plan-Study structure each rest on different kinds of support. Implementation intentions — pairing a specific cue with a specific response — carry some of the strongest randomized-trial evidence in behavior-change research, replicated across dozens of studies on everything from exercise to medication adherence, so GROW's structure is reaching for a genuinely well-tested mechanism when it forces a concrete Will. PDSA's own evidence base is different in kind: it's a proven method for organizational and clinical process improvement with decades of documented use, but applying the logic of comparing prediction to result to personal-habit decisions is the same reasoning applied to a much smaller, less-controlled setting than a factory floor. No study has run GROW-plus-PDSA as a combined protocol against either one alone; what's above is two well-supported mechanisms reasoned into a single conversation, not a tested pairing.