How to choose a career track: pick ten real job postings for the role, list the skills they repeat, check whether you want to do that work daily, then test it for four weeks before paying for long training. The decision is made with evidence from the live job market, not from job titles or salary lists.
Most people do this backwards. They choose a label that sounds good, buy training, and only then discover what the day actually involves. The sequence below is slower at the start and much faster overall.
Why Job Titles Are the Wrong Starting Point
A job title compresses a wildly variable reality. Two people with the same title at two companies can do almost nothing in common. One “analyst” builds dashboards all day; another sits in stakeholder meetings and writes documents.
So titles are useful only as a search term. The real information is one level down: the tasks, the tools, and who you work with. That information is sitting in public job postings right now, which makes the research cheap.
Salary data has a related problem. Published ranges blend seniority levels, locations and company sizes, and they describe people already doing the job, not people entering it. Use ranges to rule out tracks that cannot possibly work for your situation, not to rank the ones that can.
How to Choose a Career Track Using Real Job Postings
This is the core exercise and it takes an evening or two per track.
- Shortlist two or three candidate tracks. More than three and you will not do the work properly. Fewer than two and you are not comparing.
- Pull ten current postings per track, filtered to the experience level you would realistically enter at and the locations or work arrangements you would accept.
- Strip each posting to three lists: required skills and tools, day-to-day responsibilities, and stated qualifications.
- Tally the repeats. Anything appearing in seven or more of ten is the real requirement. Anything in two of ten is one company’s wish list.
- Separate hard gates from soft preferences. A specific licence or degree stated as mandatory across most postings is a gate. “Certification preferred” is not.
- Mark your own status against each repeated item: have it, partly have it, do not have it.
What you end up with is a ranked gap list for each track, built from current demand. That is the document the training decision should be made against.
Read the responsibilities, not the requirements
The requirements section tells you what to learn. The responsibilities section tells you what your Tuesday looks like. People leave jobs over the responsibilities, not the requirements.
Read those paragraphs slowly and ask whether you would be willing to do that specific activity for most of a working day, repeatedly, for years. Writing documentation. Handling escalating complaints. Reconciling numbers. Running meetings. Sitting with ambiguous problems and no clear answer.
Run the Day-in-the-Life Test
Before any money moves, check the track against four things about yourself. Write your answers down, because vague answers hide bad fits.
- Interaction load. Do you want most of your day with people, or mostly alone with work? Be honest rather than aspirational.
- Structure. Do you want defined problems with known methods, or open problems where you decide the approach?
- Feedback speed. Some roles show results in hours, others in quarters. Slow-feedback work frustrates some people badly.
- Pressure shape. Steady predictable load, or quiet stretches punctuated by intense deadlines and incidents?
Then check the practical constraints that quietly disqualify tracks: hours and shift patterns, on-call expectations, travel, physical demands, and whether remote work is genuinely available at entry level in that field rather than only at senior level.
A track that fits your skills and fails on interaction load or on-call expectations is still the wrong track.
Talk to Two People Who Do the Job
Ten postings tell you what employers advertise. Two practitioners tell you what the job is. This step is the highest-value hour in the whole process and most people skip it because it feels awkward.
Find people in the role through professional networks, local meetups, alumni groups, or community forums for that field. Ask for fifteen minutes and come with specific questions:
- What does a normal Tuesday look like, hour by hour?
- What part of the job surprised you?
- What do people in this role complain about most?
- What did you learn in the first six months that no course covered?
- If you were entering today, what would you learn first and what would you skip?
That last question is the one that saves money. Practitioners will frequently name a tool or topic that courses emphasise and employers do not care about, and a dull unglamorous skill that matters enormously.
Test the Track for Four Weeks Before Committing
A short trial is the difference between a decision and a guess. Give each finalist track four weeks of light contact before committing to a long programme.
| Week | What to do |
|---|---|
| 1 | Learn the fundamentals from free material. Follow the field’s main publications and communities. |
| 2 | Do the smallest real version of the work. A tiny project, a sample analysis, a mock task from a posting. |
| 3 | Finish that piece of work to a standard you would show someone, and show it to someone. |
| 4 | Read ten more postings with four weeks of context and see whether they read as interesting or as a chore. |
Then score the trial honestly on three questions. Did you look forward to the sessions or avoid them? Did the difficulty feel interesting or demoralising? After four weeks, do you want more of this or are you relieved it is over?
Enthusiasm in week four is a far better signal than enthusiasm in week one, because week one enthusiasm is about the idea of the career and week four is about the work.
Then Choose the Training to Fit the Gap
Once the track is chosen, the training decision gets easy, because you have a gap list taken from real postings. Work down it in order of frequency.
Three things to keep in mind:
Match training to the gates. If postings consistently require a specific third-party certification, prepare for that exam specifically. Remember that the credential is issued by the organisation that owns the exam. Coursework prepares you for it and can give you a certificate of completion for the course itself, which is a different thing from the certification.
Build evidence as you learn. Finished work that someone can look at does more in an application than a list of completed modules. Keep what you produce during training.
Sequence, do not stack. One track at a time, in the order the postings suggest. Studying three tracks simultaneously produces three shallow starts.
Watch for the sunk-cost trap
If the four-week trial tells you the track is wrong, that is the trial succeeding, not failing. Four weeks is a cheap price for not spending a year on the wrong path. Change the track and keep the research method.
Decide With Evidence, Then Commit
Choosing a career track well is a research task with a known method: postings for demand, practitioners for reality, a short trial for fit, and a gap list to direct the training. It takes a few weeks. It replaces a guess you would otherwise spend a year testing at full cost.
When you have a track and a gap list, the Career Track learning paths page lays out what each path covers in what order, so you can line it up against the skills your ten postings kept repeating.