How to Learn a New Skill While Working Full Time

The standard advice is to enrol in something. A ten-week evening course, a certification programme, a subscription to a video platform. Completion rates for self-paced online courses are notoriously low — the widely quoted figures sit in single digits — and the people who drop out are not lazy. They are tired, and the course was designed for someone with a different week.

What works instead is unglamorous: smaller time budgets than you think you need, and learning attached to work you were going to do anyway.

Be honest about the budget

Count the hours you actually have, not the hours you would have if everything went well. For most people in a full-time job with ordinary commitments, the genuine, repeatable figure is somewhere between three and six hours a week. Not twelve. Not "an hour every evening".

Then shape the hours to match your energy rather than your intentions. Three patterns hold up:

• Two 90-minute blocks at a fixed time. Usually weekend mornings. Long enough to get past the setup phase of anything.
• Thirty minutes before work, five days. Suits memorisation-heavy material — languages, regulation, terminology.
• One three-hour block, once a week. Suits project-shaped learning where you need to hold context.

What almost never survives contact with a real week is thirty minutes after dinner. That slot is where the day's overflow lands.

Why evening courses fail

They fail for structural reasons, not motivational ones.

The content is sequenced for a cohort, so a missed week compounds. The deadlines are soft, so they lose to any hard deadline at work. And crucially, nothing you produce in the course is used by anyone. Work that nobody consumes is the first thing to be dropped when a week gets difficult.

The exception is a course with real assessment, real cohort pressure and a fee large enough to hurt. Those finish at much higher rates. If you are paying for structure rather than content, pay enough for the structure to work.

Learning inside the job you already have

This is the single highest-leverage move available and it is under-used because it feels like it does not count.

Look at your current role for tasks that are adjacent to the skill you want. If you want to learn SQL, ask for the reporting job nobody enjoys. If you want to move into project coordination, volunteer to run the schedule for a small workstream. If you want to write, offer to own the release notes or the internal newsletter.

Three things make this superior to a course:

1. The work has a real deadline and a real audience, so it gets done.
2. You get feedback from people who know what good looks like in your industry.
3. It becomes an evidenced line on your CV rather than a certificate.

Say it plainly to your manager: "I want to develop X. Is there work coming up where I could take that on, even if I am slower at first?" Managers are usually short of volunteers.

The 20-hour and 200-hour distinction

Skills sort into two categories, and confusing them causes most of the disappointment.

Twenty-hour skills are those where twenty focused hours takes you from nothing to genuinely useful. Spreadsheet modelling, a specific software tool, basic SQL, reading a balance sheet, the fundamentals of a project management method, presenting. You will not be expert. You will be able to do the task without supervision, which is what most jobs need.

Two-hundred-hour skills are those where twenty hours only buys you vocabulary. Programming, a foreign language to working level, statistics, design, accounting, anything with professional licensing behind it. Twenty hours here produces the ability to talk about the field, which is worse than useless in an interview because it invites questions you cannot answer.

Before you start, decide which one you are attempting. If it is a 200-hour skill and you have four hours a week, you are looking at a year. That is a fine plan. It is a bad surprise.

How to prove it afterwards

A certificate proves attendance. Employers discount it accordingly. What survives scrutiny is an artefact plus a number.

• An artefact. A dashboard, a script, a document, a redesigned process, a translated set of materials. Something a stranger can look at.
• A number. How long the task used to take and how long it takes now. How many people use the thing. How much it reduced errors or rework.
• A sentence that fits on a CV. "Built the weekly sales report in SQL, replacing a four-hour manual process" is worth more than three certifications.

If your learning happened entirely outside work, build something small and public instead: a repository, a portfolio piece, a written analysis of something in your industry. It does not need an audience. It needs to exist so you can point at it.

Nobody has ever been hired for finishing a course. People are routinely hired for something they built while finishing one.

One caution on market specifics. In regulated fields, and in parts of continental Europe where formal qualifications carry more weight than portfolio evidence, the certificate genuinely does matter — sometimes more than the artefact. Check the convention in your market before deciding that evidence beats paperwork. In the UK and North America, for most non-regulated roles, it does.