How to Learn New Skills While Working Full Time

The plan is always beautiful on Sunday.

An hour before work. An hour after. Three hours each weekend day. Twelve hours a week, which means in three months you will have done something serious. You write it in a calendar and feel, briefly, like the problem is solved.

Monday you hit two of three sessions. Tuesday the meeting runs long. Wednesday you are so tired that you open the material, read the same paragraph twice, and close it. By Friday you are four sessions behind, and being behind is worse than not starting, because now the plan is a document that proves something about you. Week two is where almost every self-directed skill build dies, and it dies of arithmetic rather than character.

You did not lack discipline. You wrote a schedule for a person who does not exist — someone with your obligations and none of your fatigue.

This page is the version that survives contact with a real job. It is the execution layer under any career switch, which is why it sits beneath how to change careers: the framework there tells you what to learn, and this one is the only reason you will still be learning it in March.

Two errors, and the second one is the one nobody names.

The first is volume. You budgeted the hours that exist on a calendar rather than the hours that exist in a life. The calendar does not show the friend who calls, the appointment that runs over, the evening you are genuinely too flattened to think. A twelve-hour plan has no slack, and a plan with no slack converts one bad Tuesday into total failure.

The second error is that hours are not interchangeable. An hour at 6am on a clear Saturday and an hour at 9:40pm after a day of meetings are the same unit on a spreadsheet and completely different instruments. One of them can hold a new abstract concept. The other cannot hold anything that has not been held before. Most failed study plans are not short of hours — they are spending their best hours on the easiest tasks and their worst hours on the hardest ones.

Fix both and the required weekly commitment drops to something you can defend for a year. And a year of five real hours a week beats three weeks of twelve, which is not a motivational claim — it is what the research on how memory works predicts.

The convenient thing about being time-poor is that the constraint pushes you toward the schedule that works better regardless.

Hermann Ebbinghaus's work on the forgetting curve established the basic shape more than a century ago: newly learned material decays quickly, and the steepest loss happens soon after learning. A single long session does nothing about that decay. Sessions spread across days keep intercepting it, which is the whole basis of the spacing effect — the finding, replicated across a very wide range of material, that distributed practice produces more durable retention than the same total time massed together.

What you do inside a session matters as much as when it happens. Jeffrey Karpicke and Janell Blunt compared students who practised retrieving what they had studied against students using a richer-looking technique, and the retrieval group retained substantially more a week later. The detail worth carrying: the students who studied the comfortable way expected to do better. Robert and Elizabeth Bjork call the general principle desirable difficulties — conditions that feel harder in the moment reliably produce more durable skill than conditions that feel smooth.

Put those together and a working adult's constraint stops looking like a handicap. Short, frequent, effortful sessions with self-testing in them are the method. You were going to be forced into it anyway.

The entire system is three commitments and about five hours:

• Two blocks of ninety minutes, same slots every week, protected like a doctor's appointment.
• Fifteen minutes of maintenance on the days between them.
• One rule about what is allowed inside each.

That is smaller than the plan you wrote on Sunday, and it is roughly four hundred hours a year if you keep it, which is more than enough to become genuinely employable in a bounded skill. The reason to write it small is that the plan's job is not to be impressive. Its job is to still exist in week nine.

Protected means three specific things. It is on the calendar with a name. It is in the same two slots every week, so it becomes a location rather than a decision. And it has a defence — a sentence you say to whoever asks, out loud, in advance, to the people who share your evenings.

Pick the slots by honestly ranking your week's cognitive quality rather than its convenience. For most people that is early morning, one weekend morning, or the ninety minutes immediately after work before you sit down. The worst possible choice is "after everything else is done," because everything else is never done and that slot exists on no real day.

Only one category of work is allowed in a block: material that requires a fresh brain. First encounter with a new concept. The first attempt at a problem you do not know how to solve. Debugging something genuinely broken. Writing the hard part. If you find yourself organising notes, choosing a course, watching an introductory video or updating your tracker inside a block, you have just spent your most expensive ninety minutes of the week on the cheapest available task. That is the single most common leak in self-directed learning, and it feels productive the entire time it is happening.

Fifteen minutes on the other days, and the content is fixed: retrieve, do not re-read.

Close the notes. Write down what you remember about the last block. Answer three questions you wrote for yourself. Redo one problem you already solved, from scratch, without looking. It will feel worse than re-reading and it is doing several times more work, which is exactly the Bjork point.

The margins are also where passive input belongs — a talk on the commute, documentation over lunch, a chapter before bed. Passive input is genuinely useful and it is not learning; it is context that makes the next block faster. Filing it in the margins rather than the blocks is most of what the system does. If fifteen minutes is all some weeks will give you, that is a viable floor rather than a failure, and the case for it is in learning any skill in fifteen minutes a day.

Match the task to the brain you actually have, not to the position on your list. Working adults lose more progress to this mismatch than to missed sessions.

Fresh brain only: new abstractions, the first attempt at an unfamiliar problem type, debugging, designing something from nothing, writing an explanation of a concept you just met.

Tired brain is fine: retrieval on old material, redoing a solved problem, reading documentation on something you already understand, mechanical practice, watching a talk, tidying yesterday's notes.

Never in either: choosing which course to take, comparing tools, reorganising your system. Those are decisions, not learning, and they expand to fill whatever time you give them. Make them once, on paper, outside your five hours.

Two failure modes follow directly from ignoring this. Attempting a hard new concept at 10pm produces the experience of reading the same paragraph four times, which most people interpret as being bad at the subject rather than as using the wrong hour. And spending a Saturday morning on notes produces the sensation of a productive weekend with no new capability in it. If the focus itself is the thing collapsing rather than the schedule, that has its own diagnosis in why you cannot focus while studying, and the question of which hours are actually your best is covered in the optimal time to learn.

Five hours a week does not appear. It gets taken from something, and a plan that does not name the source is a plan that fails quietly while you feel guilty about it.

So name it. Realistic candidates: one show you watch, one weekend morning, the scrolling hour, one recurring social commitment paused for a season, one hobby explicitly on hold. Write the actual sentence — "I am not watching anything on Tuesday or Thursday until March" — because a named trade is a decision and an unnamed one is an intention.

Two things are not on the list. Sleep is not available, and not for wholesome reasons: memory consolidation happens during it, so hours taken from sleep degrade the value of every hour you kept. Trading eight hours of sleep for one of study is a net loss on the thing you are trying to buy. And the people who depend on you are not on the list either, which is precisely why the number is five hours and not twelve.

The other thing to drop is breadth. One skill at a time, for a defined stretch. Two half-learned skills is the most reliable way to spend a year and have nothing you can demonstrate — and demonstrable is the only currency in a career switch.

The mechanics matter more than the intention, and there are only six of them.

1. Name the two slots and put them in the calendar for twelve weeks. Not "most evenings." Tuesday 7:00 and Sunday 8:30, with a title.
2. Decide the next action before the block starts. The last thing you do in a session is write the first thing you will do in the next one. This removes the ten-minute orientation tax that otherwise eats a ninth of your week.
3. Start with three minutes of retrieval. What did I do last time, what was hard, what was I about to try. It is a warm-up and a spaced review at the same time.
4. Log one line at the end. What you did and what is still confusing. This log becomes your evidence in month three, when your sense of progress is unreliable and the log is not.
5. Apply the never-two rule. Missing one block is data. Missing two consecutively is where the whole thing quietly ends, so a missed block gets rescheduled inside the same week even if it shrinks to forty minutes.
6. Review the log every four weeks. Read the confusion column. If the same item has appeared three times, it is a real gap and it belongs at the front of the next block rather than at the back of your mind.

Twelve weeks of that is around sixty hours of high-quality, spaced, retrieval-heavy practice on one bounded skill. That is enough to produce an artifact. It is not enough to become an expert, and expecting the former rather than the latter is most of why people quit. If you want a more granular scaffold for building the calendar itself, there is a walkthrough in making a study schedule.

None of this needs a product. Two calendar entries, a text file for the log, and free material will run the entire twelve weeks.

What hurts most when your hours are scarce is doubling back. Learning something in the wrong order costs a week, and a week is a fifth of your quarter. Mochivia's roadmaps are prerequisite-ordered specifically to prevent that, the lessons are shaped for retrieval rather than recognition, and they are sized to fit a block instead of a semester. If your available hours are fragmented rather than merely few — which is the standard situation later in a career — the strategy shifts, and that version is in career change at 40.

Everything above is one idea in different clothing: your constraint is not hours, it is the mismatch between the plan you wrote and the person who has to execute it on a Wednesday.

So write the smaller plan. Two blocks, fifteen minutes between them, hard things in the blocks, review in the margins, one named sacrifice, one log. It will feel insufficient in week one, which is the correct feeling and also the reason it will still be running in week thirty.

Open your calendar now and put two ninety-minute entries in next week with real names on them. That single action is worth more than the rest of this page, because it is the only part that survives closing the tab.