Run Small Career Experiments

What you will be able to do

By the end of this lesson, you will be able to design small tests that tell you something real about your target role before you resign, retrain heavily, or build a whole search around guesswork. You will also be able to judge the result of an experiment clearly by deciding in advance what counts as a yes, a no, or a maybe.

The idea explained plainly

Most career changers try to solve uncertainty with imagination. They read role descriptions, watch videos, talk themselves into a story about a new field, and then mistake that story for evidence. That is how people spend months preparing for work they have never actually touched.

A small career experiment is a limited piece of contact with the target role. It is not a full commitment. It is a test with a short timeline, a narrow scope, and a specific question behind it.

The question is usually one of these:

• Do I like the actual work?
• Can I do a basic version of the work well enough to keep going?
• What parts of the role fit me, and what parts do not?
• What gap is real, and what gap only looked large from a distance?
• Does this target still make sense, or should I adjust it?

The point is not to prove that you are fully ready. The point is to replace fantasy with contact. Good experiments are small enough to finish, concrete enough to teach you something, and cheap enough that a disappointing result does not damage your finances or confidence.

Useful experiments usually fall into a few categories:

• Shadowing: observing someone in the role closely enough to see the pace, tools, constraints, and boring parts.
• Volunteer or pro bono projects: doing a small real assignment for a friend, community group, or small business.
• Freelance samples: taking on a contained paid or unpaid piece of work that resembles the target job.
• Internal stretch work: raising your hand for tasks in your current workplace that overlap with the target role.
• Short courses with deliverables: not just watching lessons, but producing a sample, memo, plan, analysis, or presentation.
• Informational conversations that probe the work: not vague networking chats, but conversations that ask what people actually do, what frustrates them, what beginners misunderstand, and what good work looks like.

Notice that some of these test interest, some test ability, and some test fit. You usually need more than one. A conversation can correct your picture of the role, but it cannot tell you how you feel when doing the work yourself. A course can help you practise a skill, but it may hide the messiness of real stakeholders, unclear requests, and time pressure.

That is why I want two experiments, not one. One should involve doing. The other can involve observing or probing. Together they give you a better picture.

Keep the experiments small. If a test takes six months, costs a lot, or requires quitting your job first, it is not a small experiment. It is an early commitment disguised as research.

A good experiment has five parts:

1. The target role you are testing.
2. The question you want answered.
3. The activity itself.
4. A deadline, ideally within 30 days.
5. A result rule: what counts as yes, no, or maybe.

That last part matters most. If you do not define the result rule in advance, you will interpret the experiment according to your mood. A hard week at your current job can make any new path look attractive. A clumsy first attempt can make a promising path look impossible. Pre-deciding the meaning of the test keeps you honest.

Worked example

Priya is 41 and has worked in accounting for fourteen years, most recently in a mid-sized company where she handles reconciliations, reporting cycles, and process clean-up. In lesson 1 terms, she knows the problem is not simply her manager or employer. Her leaving brief says she wants to leave repetitive month-end work, low ownership over process decisions, and a narrow definition of success based only on accuracy and deadlines. She wants to keep analytical work, structured thinking, and cross-functional usefulness. Her next role must include investigation, pattern-finding, and work that improves how decisions get made.

In lesson 2, Priya mapped transferable value from accounting: interviewing colleagues to understand process breakdowns, spotting inconsistencies in data, documenting workflows, and writing clear summaries for non-finance teams. In lesson 3, she compared several targets and chose UX Researcher as her primary target and Operations Analyst as her backup. UX Research appealed because it seemed to combine structured inquiry with synthesis and influence.

Now she needs to test that target before she starts an expensive certificate or announces a full switch.

Her first mistake is a common one. She starts by reading articles, following UX people online, and watching videos about research methods. Useful, but still distance. She knows more words, but she has not touched the work.

So she designs two experiments.

Experiment 1: Do a small piece of real research.

A friend runs a small meal-prep business and wants to understand why some first-time customers do not order again. Priya offers to conduct five short customer interviews, look for patterns, and write a findings summary.

She defines the scope tightly:

• Write a simple interview guide with eight questions.
• Conduct five 20-minute interviews.
• Take notes and group the findings into themes.
• Write a two-page summary with key issues and possible next questions.
• Finish within three weeks.

Before starting, she writes her result rules.

• Yes: I enjoy preparing questions, listening for patterns, and turning messy conversations into usable findings. I would willingly do more of this.
• No: I strongly dislike the core tasks, especially the interviews and synthesis, even when the project is small.
• Maybe: I like some parts but dislike others enough that I may need a related role rather than this exact one.

The project teaches her more than expected. She likes building the interview guide. She likes hearing the same friction points appear across customers. She especially likes writing the summary and clarifying what the business owner should pay attention to next.

But she dislikes recruiting participants. Chasing people for time slots drains her quickly. She also notices that she is less interested in the product-design side than in the operational patterns behind customer drop-off.

Experiment 2: Probe the day-to-day work through targeted conversations.

Priya sets up three informational conversations: one with a UX researcher at a software company, one with a service designer in a public-sector team, and one with an operations analyst who studies user complaints and process failures.

She does not ask broad questions like, “How did you get into this?” until the end. First she asks:

• What did you do last week, hour by hour?
• What part of the job takes more time than outsiders think?
• What frustrates you most?
• What kind of stakeholder conflict shows up in this role?
• What does weak work look like from a beginner?
• What adjacent backgrounds tend to do well here?

These conversations sharpen the picture. The UX researcher confirms that participant recruiting, stakeholder persuasion, and ambiguous product questions are constant parts of the role. The service designer describes broader systems work but also heavy facilitation. The operations analyst describes investigating recurring service failures, reviewing customer feedback, tracing process causes, and recommending changes.

Priya sees the pattern. What she liked in her small project was not mainly user research as a product discipline. It was structured inquiry tied to process improvement. Her original target was close, but not exact.

Her final read on the experiments looks like this:

Primary target tested: UX Researcher

Result: Maybe

Why: Strong interest in interviews and synthesis, weak interest in participant recruiting and product-centered context.

Implication: Do not commit to a full UX pivot yet.

Refined direction: Shift attention toward operations research, service operations analysis, or customer insights roles where investigation and synthesis connect more directly to process improvement.

This is a good outcome, not a failed one. Priya did not waste months preparing for the wrong version of the work. She learned that her first target was adjacent to the right answer. That is exactly what small experiments are for.

Common mistakes

• Using learning as a substitute for testing. Reading, courses, and videos can help, but if there is no deliverable or real contact with the work, you are still guessing.
• Choosing experiments that are too large. If the test needs a major payment, a formal program, or many months, it is too big for this stage.
• Asking vague networking questions. “How did you break in?” is fine, but it will not tell you what the work feels like. Ask about tasks, frustrations, tools, trade-offs, and beginner mistakes.
• Testing only what flatters you. Do not design an experiment that includes only the part you already know you enjoy. Try to include at least one ordinary or awkward part of the role.
• Ignoring a mixed result. A maybe is useful. It often points to a better adjacent role, as it did for Priya.
• Failing to write down the result. If you do not capture what happened, the lesson fades and you drift back into speculation.
• Running too many experiments at once. Two focused tests are better than six half-finished ones.

Exercise

Design two career experiments for your primary target role. If useful, one of them may also help you compare your backup role. Both experiments must be completable within the next 30 days.

Use this format for each one:

1. Target role: name the specific role you are testing.
2. Question: what are you trying to learn?
3. Experiment type: shadowing, volunteer project, freelance sample, internal stretch work, short course with deliverable, or targeted informational conversation.
4. Exact activity: describe what you will do in plain terms.
5. Output: what tangible thing will exist at the end? Examples: summary memo, sample analysis, slide deck, observation notes, reflection page.
6. Deadline: choose a date within 30 days.
7. Yes / No / Maybe rules: define the result before you begin.

If it helps, copy this structure:

Experiment 1
Target role:
Question:
Type:
Exact activity:
Output:
Deadline:
Yes means:
No means:
Maybe means:

Experiment 2
Target role:
Question:
Type:
Exact activity:
Output:
Deadline:
Yes means:
No means:
Maybe means:

A strong result will meet all of these checks:

• It is specific. Someone else could understand exactly what you plan to do.
• It is small. You can complete it without quitting your job or overhauling your life.
• It produces evidence. There is an output, not just a feeling.
• It tests the work itself. At least one experiment involves doing or closely observing real tasks.
• It has a decision rule. You already know what different outcomes will mean.

If your plan does not pass those checks, shrink it and sharpen it. The goal is not to impress anyone. The goal is to learn enough, quickly enough, to make the next decision with less fantasy and more proof.