Career-change advice often begins with a dramatic instruction: discover your passion, resign and start again. That makes an exciting story. It can be terrible financial advice.

A safer career switch begins with experiments. You do not need to announce a completely new identity on day one. You need to test whether you enjoy the work, can develop the required skills and can reach the market from where you are now.

For 90 days, treat each possible career as a hypothesis. Your goal is not to guarantee a new job by day 90. Your goal is to replace assumptions with evidence.

Days 1–15: define the problem before choosing a title

Write down what you want to change. Be specific. "I hate my job" is understandable but not yet useful. Is the problem the work itself, the manager, shift pattern, commute, pay ceiling, lack of progression or unstable contract?

Next, list your constraints:

  • Minimum monthly income.
  • Hours available for learning.
  • Transport and data costs.
  • Care responsibilities.
  • Qualifications or licences you already have.
  • Tasks you can do well, even if you dislike your current title.

Choose no more than two target roles. A retail supervisor might test customer success and operations coordination. An administrator might compare bookkeeping with reporting support. A call-centre agent might explore quality assurance and workforce scheduling.

Check the roles against active adverts and the Department of Higher Education and Training's occupation data. Count recurring requirements. Do not build your plan around one unusually generous advert.

Days 16–30: investigate the real work

Job titles hide a great deal. Project coordinator can mean scheduling and reports at one employer, and chasing suppliers all day at another.

Find five people who are close enough to the work to describe an ordinary week. They do not need to be famous or senior. A former colleague, supplier, alumni contact or professional-group member may be more useful.

Ask short, concrete questions:

  • What task takes more time than outsiders expect?
  • What mistake do beginners make?
  • Which skill is genuinely used every week?
  • What evidence helped you get your first opportunity?
  • What would you test before paying for training?

The International Labour Organization includes work exposure, mentoring, networking and skills development among useful career-development activities. Use them as investigation tools, not as requests for strangers to find you a job.

If possible, arrange an authorised job shadow, workplace visit or 20-minute screen-share demonstration. Never ask someone to reveal private customer or employer information.

Days 31–60: run two low-risk experiments

Now perform a small version of the work.

An aspiring operations coordinator could build a schedule, risk log and weekly status update for a community event. A future data assistant could clean a public dataset and explain three findings. Someone considering customer success could design an onboarding checklist and respond to realistic support scenarios.

Each experiment should produce something reviewable within one or two weeks. It must be honest: label simulated work as simulated, volunteer work as volunteer work and team contributions accurately.

Choose one short learning resource only after the first experiment exposes a real gap. This prevents certificate collecting. A course should solve a defined problem, such as weak spreadsheet formulas, unfamiliar inventory terminology or poor presentation structure.

At the end of each experiment, score the role from one to five on:

  • Energy: did the work hold your attention?
  • Ability: did you improve with practice?
  • Access: can you reach entry opportunities from your location and background?
  • Evidence: can you show credible proof?
  • Cost: is the next step affordable in money and time?

Days 61–75: test your evidence with the market

Package your strongest experiment into a short work sample. Ask three people for criticism: one person in the target field, one strong communicator and one person who does not know the project.

Do not ask, "Is this good?" Ask:

  • What role does this sample suggest I can perform?
  • Which part feels credible?
  • Where would an employer still feel uncertain?
  • What should I remove?

Apply selectively to roles where you meet the essential requirements and can explain your transition. You are testing whether employers understand your evidence. Rejection is not a verdict on your identity; it is data about positioning, requirements or competition.

Days 76–90: make a measured decision

At day 90, choose among four legitimate outcomes:

  • Continue towards the role because the evidence is encouraging.
  • Adjust to an adjacent role with a shorter skills gap.
  • Run another experiment before committing.
  • Stop because the daily work, cost or access is wrong for you.

A successful experiment can save you from an expensive mistake. That is progress.

If you continue, create a transition budget and timeline. You may pursue an internal transfer, part-time study, supervised workplace learning or a staged job search while employed. Do not resign merely to make the change feel real.

Career identity is allowed to emerge after action. Test the tasks, learn from people doing the work and let repeated evidence shape your decision. Ninety days may not complete the switch, but it can turn a vague wish into a defensible next move.

Sources and limits

This framework is for exploration, not a promise that a career change can be completed in 90 days. Regulated roles may require formal qualifications, registration, licences and supervised practice.