How to learn a high income skill while working

Here is the direct answer. Choosing the skill is not the bottleneck and it never was. Almost everyone who fails at this failed after choosing, in month two, with a half-watched course and no evidence of progress. What separates the people who get there is a system with four parts: a single skill held for at least six months, practice aimed at the specific things you cannot yet do rather than the things you enjoy redoing, retrieval instead of rewatching, and a shipped artifact that someone other than you can evaluate. The reading below is why each of those matters, and where the popular advice on this topic is measurably wrong.

Why the lists do not help

Search for high income skills and you will get the same eight to twenty items: software development, digital marketing, sales, data analysis, UX, copywriting, project management. The lists are not wrong. They are just answering a question that was not the hard one.

The hard one is that you have a job, perhaps forty-five hours of it including the commute, and you are proposing to acquire a professional competence in the gaps. The list gives you a destination and no vehicle. Worse, the abundance of destinations is itself a trap, because switching skill at month three feels like progress and resets you to zero. The single most predictive decision you will make is committing to one thing long enough to get past the stage where you are bad at it.

What deliberate practice actually claims, and what it does not

The foundational work is Ericsson, Krampe and Tesch-Römer, and it is worth understanding properly because the popularised version is misleading. Their claim was not that time produces expertise. It was that a specific kind of effortful, feedback-rich activity at the edge of your current ability produces it, and that ordinary experience does not. Expert violinists were not distinguished by hours holding a violin. They were distinguished by hours of solitary practice on things they could not yet play.

Now the honest counterweight, which the hustle content never includes. Macnamara, Hambrick and Oswald’s meta-analysis found deliberate practice explained about 26 percent of performance variance in games, 21 percent in music and 18 percent in sport, but only around 4 percent in education and less than 1 percent in professions. Professional domains are messy: the tasks are ill-defined, feedback is slow and ambiguous, and who you work with and what you get exposed to matter enormously.

The correct conclusion is not that practice is pointless. It is that in a professional skill, practice alone will not carry you, and the missing ingredients are real projects, real feedback and exposure to how the work is actually done. Which is precisely why the course-completion approach stalls.

Stop rewatching, start retrieving

The default self-teaching loop is watch a tutorial, follow along, feel competent, forget it, watch another. It fails for a well documented reason.

Roediger and Karpicke showed that testing yourself on material beats restudying it by a wide margin at a one week delay, and that people consistently predict the opposite because restudying feels fluent. Following along with a tutorial is the most fluent activity available to a learner and one of the least productive. You are watching someone else retrieve.

The fix is mechanical. After any lecture or chapter, close it and write down what it covered from memory before checking. Rebuild the tutorial project from a blank file without looking. Explain the concept to somebody, or to a document. Anything that makes you produce the information rather than recognise it.

And space it. Cepeda and colleagues’ review of 254 studies found the same total time distributed across separated sessions beats massing it, with the advantage widening the longer you need to retain the material. Five weekday evenings of fifty minutes will beat one Saturday of four hours, which is convenient, because you have the evenings and your Saturdays keep getting taken.

Ship something small every two weeks

This is the part that distinguishes a hobby from a skill you can charge for. Every fortnight, finish and publish a small artifact: a working feature, an audit of a real company’s funnel, a landing page, a dashboard on a public dataset, a short case study of something you fixed.

It does three jobs at once. It generates the feedback that deliberate practice requires and that watching cannot provide. It builds the portfolio that, in nearly all of these fields, is what gets assessed instead of credentials. And it gives you an external deadline, which matters because your internal ones will drift. Note that in the Ericsson framing, feedback is not optional decoration on practice; it is the mechanism.

Keep them small. A finished trivial thing beats an unfinished impressive thing every time, and the unfinished impressive thing is the standard way this ends.

A weekly structure that fits around a job

Five sessions of fifty minutes on weekday evenings, one longer block at the weekend, and one full day off. That is roughly six hours a week, which sounds insufficient and compounds to about three hundred hours in a year of genuine focused practice on a single skill. That is a lot.

Put those sessions in a calendar with times attached, not on a to-do list. A session that exists only as an intention loses to whatever is actually happening at 7pm. Alternate them: two on new material, two on retrieval and rebuilding without reference, one on the current artifact.

Review once a week. What did you ship, what did you fail to retrieve, what goes in the next fortnight’s artifact. Ten minutes.

Where a tool helps and where it does not

Being clear about the limits first. No app will teach you the skill, and no assistant will supply the taste and judgement that a mentor or a working team gives you. If you can get either of those, they beat everything in this article.

What a tool can do is hold the parts you keep dropping. It can keep the plan and rebuild it when a work crunch eats a fortnight, which is the moment most self-teaching attempts quietly end. It can turn the material you have gathered into questions so your review sessions are retrieval rather than rereading. And it can keep your notes, PDFs, documentation and course materials in one searchable place so that “what was that pattern called” takes ten seconds instead of derailing a session.

That last one is the underrated benefit. Self-taught learners accumulate an enormous, scattered pile of half-read material, and its scatter is what makes it useless.

Where Normi fits

Directly, since this is our blog. Normi holds the plan, the materials and the recall in one place. You describe what you are learning and by when, it drafts the sessions as reminders you confirm, and when you miss a week it rebuilds the schedule rather than presenting a backlog. Your notes, documentation and saved PDFs sit alongside it, so you can ask questions and get answers cited back to your own sources, and turn any of it into a quiz for the retrieval sessions.

It will not write your portfolio or tell you whether your work is good. Those remain the hard parts, and they are the parts that pay.

The one thing to take away

The people who succeed at this are rarely the ones who picked the cleverest skill. They are the ones who were still doing it in month eight, with a folder of eight or ten small finished things, having never had a particularly dramatic week. Structure is what makes month eight arrive. Motivation reliably does not, for reasons worth understanding.

If you want the plan-and-recall side handled, Normi is on iOS, Android and the web with a 3-day free trial. Set the goal, let it draft the week, and judge it on whether you are still going in six weeks.