How do I change careers?

Why does the big-decision approach fail?

Most career-change advice asks you to decide first and act later: work out your purpose, pick the field, then retrain and apply. That order only works if you already know what the work feels like from the inside. If you do not, you are choosing between job titles, which are closer to marketing than to a description of the day.

People stall for years for this reason, not because they are lazy. The decision cannot be answered from where they are standing, and no amount of reflection produces the missing information. Only contact with the work does. Herminia Ibarra's research on career changers, set out in Working Identity, makes the same point: people learn who they want to be by trying things.

How do you change careers, step by step?

1. 01Explore: list directions that reuse most of what you already do well. Include one surprising option, not just sensible ones.
2. 02Try: find the smallest real version of one. Real means someone other than you cares about the output.
3. 03Learn: straight afterwards, write what you liked doing, what you avoided and what you now want to know.
4. 04Adapt: keep, drop or narrow the direction. Then run the loop again with the next one.
5. 05Commit: once a direction has survived two or three tests, invest in it, through a course, a portfolio or targeted applications.

Advice gives you an answer. Trying gives you evidence. Only one of them survives the first bad week.

How do you choose which directions to test?

Start from your skills, not from job titles. List the five things you do at work that other people come to you for. Then look for fields that are short of those things. Free tools such as the CareerOneStop Skills Matcher in the US, or the job profiles on the National Careers Service in England, let you search occupations by the skills and tasks they involve.

Aim for three directions: one obvious, one adjacent and one that feels slightly unreasonable. The unreasonable one is there because the obvious options are often the ones you have already been circling for a year.

Why change one variable, not three?

A career has three dials: the craft (what you do), the domain (who it is for) and the setting (the kind of organisation). Turning one at a time keeps most of your value and makes you easy for employers to understand. Turning all three at once usually means starting near the bottom.

What counts as a real test?

• A shadow day or a shift alongside someone doing the work.
• One small paid job in the new direction, however small the fee.
• A project with a real user who will complain if it is bad.
• A properly prepared conversation with two people who do the work now.
• An internal project at your current employer that sits next to the new direction.

Watching videos and reading job descriptions is research, not a test. Research tells you what the work is called. A test tells you what it feels like at four in the afternoon, when the interesting part is done and the admin is waiting.

How do you read the result honestly?

After each attempt, answer three questions in writing: which part did I lose time in, which part did I keep putting off, and what do I now want to find out? The middle answer is usually the honest one. A direction that survives three of these is worth spending money and months on.

Be careful with first tests that go badly. Separate disliking the work from not knowing the tool yet, an unclear brief or a tired week. If you are unsure which it was, change one condition and try again at a smaller size before you drop the direction.

What if you can't afford to take a risk?

Then the testing approach matters even more, because it is how you avoid risk. Keep your income and shrink the tests to fit: one evening task, one conversation, one Saturday shift. Small tests cost time rather than money, and they stop you making the single expensive bet that hurts if it goes wrong.

Before any real move, write down your monthly floor, how many months of savings you hold, and what the new direction would realistically pay in the first year. If the gap is too big, look for a bridge: part-time hours, a contract role, or an internal move that keeps your salary while changing the work.

How do you know when to stop testing and commit?

Commit when the same direction has pulled you back three times, when your questions have moved from whether to how, and when people in the field are starting to treat you as a peer. You will not feel certain. You will have evidence, which is the most anyone gets.

If nothing has pulled you after three rounds, that is information too. Go back to your list, change the variable you were testing, and run the loop again with a new direction.

What are the common career change mistakes?

• Paying for a long course before doing a single hour of the work.
• Resigning before testing, so money pressure picks the next job for you.
• Applying widely before testing, so every rejection teaches you nothing.
• Hiding the change on your CV instead of naming it and showing what carries over.
• Waiting to feel certain. Certainty is a result of the tests, not a condition for starting them.

Illustrative example: one change, three tests

An events coordinator wonders about project management, UX research and teaching. In one month she runs an internal project with a fixed deadline, interviews five users of her company's booking form, and teaches one evening class. The project absorbed her, the interviews were interesting, and the class drained her.

She drops teaching, keeps project management and runs a second, bigger test: managing a small freelance website launch for a friend's business. Only after that does she pay for a project management course, now with two real projects to put next to it on her CV.