Latest dealsDeals See all deals→
Chapter Five

Try Before You Resign

What a test is for

People often talk about a career change as if the main act is courage. Hand in notice. Burn the ships. Commit. This is tidy advice for people who do not have rent, children, ageing parents, a reputation in one field, or a body that is already tired.

For most adults in their thirties and forties, the problem is not lack of bravery. It is lack of reliable information. A new field looks attractive from a distance because it is viewed through relief, fantasy, or a few polished examples online. The question is not only Can I get in? It is also Do I actually want the work once I am near enough to see it properly?

This is why trying before resigning matters. A good test does three things:

  • It shows what the work feels like in practice, not in theory.
  • It produces evidence you can use later if you continue.
  • It limits financial and reputational risk if you decide not to continue.

That third point matters more than people admit. Not every exploration should become a transition. Sometimes the most useful outcome of a test is a clean no. If a weekend project, a shadowing day, or a short assignment reveals that the appealing version of the field is not the real one, that is not wasted effort. It is cheap information.

This chapter sits after Three Roads Out for a reason. The kind of test you run depends on the kind of move you are making. If you are taking the first road, same skills in a new industry, your test may be mostly about context and language. If you are taking the second road, new skills in the same industry, you need to test whether you enjoy the new craft enough to study and practice it. If you are taking the third road, a deeper pivot, you need stronger evidence before you disturb income, identity, or family arrangements.

A test is not a dramatic gesture. It is a small, bounded piece of reality.

The difference between interest and fit

Nadia is 37 and works in the civil service. She has spent eleven years in policy and service delivery roles. She is good at interviews, stakeholder meetings, and making sense of conflicting information. She is less happy with the pace of policy churn and the distance between recommendations and actual user experience. She becomes interested in UX research after hearing about it from a former colleague.

From a distance, UX research seems plausible. She likes asking questions. She likes finding patterns. She likes improving services. The language feels familiar enough to be flattering and different enough to be exciting.

At this stage, many people make one of two mistakes. The first is to dismiss the idea because they are not yet qualified. The second is to overcommit before they know whether the day-to-day work suits them. Nadia does neither. She decides to test fit before making any declaration.

She starts with informational conversations, but not the vague kind where both people exchange generalities and leave feeling polite. She asks five people in UX research the same practical questions:

  • How much of your week is spent interviewing, synthesising, presenting, recruiting, and persuading?
  • Which parts are satisfying and which parts are tedious?
  • What frustrates people entering from adjacent backgrounds?
  • What does good work look like at junior, mid, and senior levels?
  • Where do researchers actually work: agencies, product teams, consultancies, government, freelance?

This gives her something useful but incomplete. She learns that she is attracted to one subset of the field more than the whole. She likes discovery interviews, service improvement, and public-interest problems. She is less drawn to fast-moving startup environments where research must constantly justify itself in commercial terms. That is already progress. She is not asking, "Could I work in UX research somewhere?" She is beginning to ask, "Under what conditions would this be a fit?"

Interest is broad. Fit is specific.

She then notices a second issue. Several researchers tell her that freelance work can be patchy, and contract roles often require comfort with uncertainty, self-marketing, and shifting scope. Nadia has tolerated bureaucracy for years, but she does not enjoy unstable income. This matters because she had quietly imagined starting with freelance projects after resigning. The conversations do not kill the idea of UX research. They do kill one version of the path into it.

This is what testing should do. It should not merely confirm attraction. It should reveal the shape of the trade-offs.

Five low-risk ways to test a field properly

Not every test suits every field, but most useful trials fall into a few types. The common feature is that they are bounded. They have a purpose, a timescale, and an output. They are not open-ended dabbling.

1. Contract or project work

This is best when the target field allows a discrete piece of work to be done without pretending to be fully established. For someone considering content design, data analysis, training, operations improvement, or research, a small project can show whether the work itself is engaging and whether the standards are manageable.

The risk is that people choose projects that are too large, too vague, or too generous. They volunteer to "help out" and end up doing free labour with no clear learning goal. A better version is narrower: conduct five user interviews and write a short findings summary; redesign one training module; analyse one process bottleneck; draft one content flow.

2. Volunteering with boundaries

Volunteering can be useful when paid entry points are scarce, but only if it is structured. The point is not to become an unpaid version of a junior employee for six months. The point is to get close enough to the work to observe standards, tools, frustrations, and your own energy.

Good boundaries include a fixed duration, a clear deliverable, and agreement about what you are not taking on. If you are exploring community programme management, for example, helping to run one event cycle may be sensible. Becoming the permanent person who rescues the organisation's admin is not a test of fit. It is just familiar overfunctioning in a new setting.

3. Short courses tied to output

A course is only a test if it requires you to make something, decide something, or show something. Passive learning often creates false confidence. It can also create false discouragement, because introductory material is sometimes detached from the actual pleasures of the work.

A useful short course ends with a sample, project, case analysis, or practical exercise. Nadia chooses a short UX research course that requires participants to plan interviews, conduct them, and synthesise findings. This matters more than a certificate. It gives her contact with the actual work.

4. Shadowing

Shadowing is underrated because it sounds too simple. But one well-observed day can reveal more than weeks of online reading. You see how much time is spent waiting, documenting, persuading, correcting, repeating, and coordinating. You also see what kinds of people and rhythms the work attracts.

Shadowing works best when you prepare for it. Do not ask only, "Can I shadow you?" Ask to observe specific activities if possible, and ask permission to follow up with questions afterwards. Otherwise you may witness a day without understanding what is typical and what is unusual.

5. Informational conversations that lead to concrete observation

Conversation alone is not enough, but it is often the door to more useful tests. The strongest conversations end with something more tangible than encouragement. A person may suggest a meetup, a project idea, a team to observe, a hiring manager to follow, or a sample task to try. If every conversation ends with "Best of luck," you are still near the surface.

Each of these methods helps answer a slightly different question. Can I tolerate the day-to-day? Can I produce work at the expected level? Do I like the environment in which this work usually happens? Would I want more of this, or only a little of it?

Nadia's two weekend projects

Nadia decides she needs more than conversations. She arranges two small weekend projects over two months while keeping her job.

The first is with a local charity that wants to improve how people sign up for a support service. Nadia offers a tightly defined piece of work: she will interview six recent users about the sign-up experience and produce a short summary of patterns and recommendations. She is explicit about scope. She is not redesigning the whole service, running implementation, or becoming the charity's standing advisor.

The second project is self-generated. Nadia chooses a public digital service she uses regularly and recruits a handful of friends and family members who fit the user group. She runs practice interviews, writes notes, groups themes, and creates a short presentation as if she were handing findings to a team.

These projects teach her more than she expected.

First, she likes the interview stage very much. She is patient, curious, and good at drawing out specifics without leading people. She notices moments where someone says one thing but means another, and she is comfortable following that thread. This is not a surprise if you look back at her civil service work, but it is useful to see it operating in a new context.

Second, she also likes synthesis more than she feared. She enjoys turning messy notes into patterns and practical recommendations. She does not love the formatting of slides, but she can do it.

Third, she dislikes the uncertainty around getting the work. The charity project was satisfying once defined, but arranging it required awkward outreach, explaining herself, and not knowing whether the effort would lead anywhere. The self-generated project was useful but lonely. No team, no manager, no shared standards. She realises that what appeals to her is not freelance independence but research as part of a stable organisation, preferably in public service or a mission-led environment.

This is the key finding. Nadia has not simply learned, "I like UX research." She has learned a more accurate sentence: I like interviewing users and making sense of findings, and I would want to do that inside a stable team rather than as an independent freelancer.

That sentence is much more valuable. It narrows the path. It rules out one route and strengthens another. It also gives her language for later chapters of this book, especially when she must think about proof, translation, and job search.

A useful test does not only say yes or no. It says yes to what, no to what, and under which conditions.

How to avoid fake tests

People are busy, so it is worth being blunt about what does not count.

Reading about a field is not a test. Listening to podcasts about a field is not a test. Following people online who seem to enjoy a field is not a test. None of these are useless, but they are still observation from a distance.

There are also more subtle fake tests.

One is choosing only tasks that flatter existing strengths. A manager exploring counselling might enjoy one deep conversation with a friend. That does not test the administrative load, emotional repetition, record-keeping, or training demands of the profession. A writer exploring product management might enjoy strategy discussion while avoiding prioritisation, conflict, and trade-offs. If the test excludes the inconvenient parts, it is not a test of fit. It is a test of fantasy.

Another fake test is overinterpreting one pleasant experience. Enjoying a workshop does not mean you should become a facilitator. Enjoying a coding tutorial does not mean you want a software engineering job. Enjoying praise from people in a new field does not mean you want their work. Adults changing careers are especially vulnerable to this because novelty itself can feel like evidence.

A third problem is running tests that are so expensive in time or money that they recreate the pressure of a full decision. If you sign up for a long programme before trying any direct exposure to the work, you may feel compelled to continue simply because you have paid for the identity. That is not exploration. That is escalation.

The antidote is to design tests with three qualities:

  1. Specificity: What exact question is this test answering?
  2. Boundaries: How long will it last, and what is the deliverable?
  3. Reflection: What will count as evidence for or against fit?

Nadia keeps notes after each project. Not emotional diary entries, just practical observations. Which parts drained her? Which parts made time pass quickly? Which frustrations felt tolerable because the work itself mattered? Which frustrations felt fundamental? This is useful because memory is easily distorted by relief, praise, or one difficult day.

She also resists the temptation to announce a full identity change too early. She does not tell everyone she is "moving into UX." She says she is exploring user research and testing fit. This may sound like a small distinction, but it protects judgment. Once people publicly declare a new future, they often become less honest with themselves about the evidence.

What to do with the result

By the end of her tests, Nadia is not ready to resign. That is not failure. It is the correct result. She now knows enough to make a better next move.

She decides on three actions. First, she will look for opportunities inside or adjacent to government where service research is already valued, rather than chasing freelance work she already knows she would dislike. Second, she will build one or two stronger samples from bounded projects, because these are beginning to look like proof rather than mere curiosity. Third, she will continue talking to people, but now with a sharper question: what entry routes exist for someone with policy and service-delivery experience who wants to move into in-house research?

This is what trying before resigning should produce: not certainty, but a narrower and more realistic plan.


Next step

Choose one target field and write down the smallest test that would expose you to the real work within the next six weeks. Define the question it is meant to answer, the boundary around it, and the output you expect to produce or observe. If you cannot design a test without quitting first, the field may be less accessible than it appears, or you may still be describing it too vaguely to explore well.

End of the book That was the last chapter Follow book Back to the book

Contents

About this book The Second First Job by Sophia Williams · 5 chapters

Changing careers in midlife rarely happens in one brave leap. This book follows people in their thirties and forties as they test options, handle money, explain the change to family, and build enough proof to be hired into work that fits …

More books

All books →