The reason most personal development plans go nowhere is not weak discipline. It is that they were written in a way that made them impossible to finish, and nobody noticed, because a plan full of admirable intentions is very hard to object to in the meeting where it is agreed.
A plan that works is smaller and more specific than people expect, and slightly uncomfortable to write, because it commits somebody to a result rather than to an effort. Here is how to write one, step by step, and why each step is the way it is.
Step 1: Start from a gap that is costing you something
Not from what you would like to be good at. From what is currently getting in your way.
This inverts how most plans begin. Asked what they want to develop, people reach for something aspirational and slightly distant, because that is what the question invites. Asked what is costing them, they answer immediately and concretely, and the answer is usually something they have known for months.
Three questions find it faster than any framework:
- What work do you get left out of? Not because anyone is against you, but because somebody else is the obvious choice. That is a capability with your name not on it.
- What do you hand to somebody else? The task you route around, or wait for a particular colleague to be free for, is a gap you are already paying for in delay.
- What feedback keeps coming back? Not the one-off comment. The one that has arrived from three different people in two years, which is the definition of a pattern rather than an opinion.
If nothing comes up, that itself is worth knowing: either the role is a comfortable fit right now, in which case the honest plan is about the next role rather than this one, or nobody has told you the truth in a while.
Step 2: Write a capability, not a quality
This single change does more than the rest of the steps combined, and almost every weak plan fails here.
A quality is a verdict somebody else passes on you: strategic, senior, proactive, commercial, a strong communicator. You cannot practise a verdict. You cannot tell on a Wednesday whether you did it. Two reasonable people can disagree forever about whether it has happened, which is exactly what makes it comfortable to write down and useless to act on.
A capability is a piece of work. You can pick it up, do it, and everybody can see that it happened.
The translation is mechanical once you see it. Not "become more strategic" but "own the quarterly roadmap trade-offs and present them to the leadership team". Not "improve stakeholder management" but "run the monthly review with the finance team without my manager in the room". Not "be more senior" but "take the next incident as lead and write the follow-up".
The test: could somebody who does not know you look at the finished thing and agree it was done? If the answer needs a discussion about your growth, you have written a quality.
Step 3: Pick one goal, not six
A plan with six goals is a wish list, and the reliable outcome of a wish list is that none of it happens.
The reason is arithmetic rather than willpower. Development competes with delivery for the same hours, and delivery arrives every week with a deadline attached while development arrives once and asks quietly. One goal can survive a busy quarter by taking the small amount of room that exists. Six goals need a quiet quarter, and there is no quiet quarter.
Six also destroys the thing that makes a plan useful: when everything is a priority, the plan stops telling you what to do when the week gets full. One goal answers that question by itself.
Two is defensible if one of them is small. Three is where people are quietly negotiating with themselves.
Step 4: Decide what evidence would settle it, before you start
This is the step almost everybody skips, and skipping it is why so many plans end in an awkward conversation about whether the goal was met.
Write down now, in advance, what a person who does not know you would accept as proof. Not a feeling of progress, and not a list of things you attended. Something produced, something run, something reviewed by somebody with standards.
"Attended a course on facilitation" is not evidence. "Ran four cross-team sessions, and the fourth needed no help from my manager" is. The first proves attendance; only the second proves capability, and the difference matters because attendance is what plans quietly drift toward whenever the evidence was not agreed up front.
Agreeing this before starting also protects the person doing the work. When the standard is written down in advance, finishing is a matter of record. When it is not, finishing depends on whether whoever is judging happens to feel generous that quarter.
Step 5: Break it into activities that fit inside a normal week
A goal is the destination. Activities are what actually get done, and their size decides whether anything happens at all.
The unit to aim for is something that could be finished between other work in an ordinary week, not a week cleared for it. If the first activity needs an empty calendar, the plan is a small project, and small projects get deferred the moment delivery gets loud. Which it will.
Three to five activities under one goal is usually right. Each should end in something you could show somebody: a document, a session run, a decision made, a piece of work shipped. An activity that ends in "read about it" has no finish line, so it never finishes.
Order them so the first one is the smallest. The first activity has a job beyond its own content, which is to prove the plan is real by being completed in the first two weeks. A plan whose first concrete step is a month away is one that will be overtaken before it starts.
Step 6: Name what has to happen around you
Almost every development plan is written as though the person is the only variable. That is what makes so many of them undoable.
If the goal is to run the vendor evaluation, somebody currently runs it and has to give it up. If it is to present to the leadership team, somebody has to put you on the agenda. If it is to lead an incident, somebody has to accept that the first one will go less smoothly than it would have. None of that is your decision alone, and none of it happens because a document mentions it.
So write it in: what the manager gives up, who makes the introduction, what the first attempt is allowed to cost. A plan with nothing asked of anybody else is not a modest plan. It is one that has quietly moved the hardest part out of scope and will stall on it later, usually while looking like a motivation problem.
Step 7: Set a review triggered by an event, not by the calendar
Quarterly reviews are the wrong instrument for this. They arrive when the calendar says so, which is almost never the moment anything can be learned, and by then the details have gone.
The useful review happens straight after the first real attempt, while everybody still remembers what actually occurred. That is when you find out whether the activity was the right size, whether the evidence you agreed on is gettable, and whether the gap you named in step one was the real one.
Expect to change the plan at that point. A plan that survives its first contact with real work completely unaltered is usually one that was too vague to be contradicted.
A worked example, start to finish
All of the above in one plan, for a mid-level engineer whose feedback keeps saying they should be more senior.
- The gap, from step 1: they are never asked to lead incidents, and the same two colleagues always are. That is the work they get left out of, and it is the actual content of "more senior".
- The goal, from step 2: lead two production incidents end to end, including the follow-up write-up and the actions arising from it. Not "be more senior", which was the original phrasing and could not have been started on any particular day.
- The evidence, from step 4: the second incident is run without the manager stepping in, and the write-up is accepted by the team without a rewrite.
- The activities, from step 5: shadow one incident and write the notes; run one low-severity incident with the current lead observing; take the next incident in the rotation; write and circulate the follow-up.
- What has to happen around them, from step 6: the current incident lead agrees to be shadowed and then to step back, and the manager accepts that the first solo one will be slower and says so out loud beforehand.
- The review, from step 7: the week after the first solo incident, not at the end of the quarter.
The whole plan is about six lines. That is the correct length. A development plan is not a document about a person, it is a change to what happens to them, and the change fits on half a page.
Four ways a good plan still dies
Written well, a plan can still fail, and it usually fails in one of four recognisable ways. Each has a cheap defence.
- It is written once and never opened. The defence is the event-triggered review in step 7, and putting the plan somewhere it is seen in the normal run of work rather than in a folder named after the quarter.
- The first activity never starts. Almost always because it was too big. Cut it in half, then in half again. A first step that takes two hours is not too small to matter, it is the thing that gets the rest moving.
- The manager changes and the plan evaporates. Extremely common, and the reason years of this work gets thrown away at every reorganisation. It has to be written somewhere a successor will find it, not held in one person's notes.
- It quietly becomes a list of courses. This is what happens when step 4 was skipped, because attendance is the easiest thing to produce when nobody agreed what proof looks like. Go back and write the evidence.
The uncomfortable bit
A plan written this way can fail visibly, and that is the point of it. If the evidence is agreed in advance and the activities have finish lines, then at the review it is clear whether the thing happened.
That is exactly why the vague version is so persistent. "Become more strategic" cannot embarrass anybody. It cannot be failed, which feels safe, and it also cannot be passed, which is the cost nobody notices they are paying: a year later there is nothing to point at, and the promotion conversation is somebody's sincere opinion against somebody else's.
A plan that can be finished is also a plan that can be shown. That is the whole return on writing it properly.
