
I made a five-year plan once. Wrote it out, felt good about it, told a few people. It survived about eight months. Not because I got lazy — because I'd built it from a list of "good careers" instead of anything about how I actually work.
That's the trap most advice on how to choose a career walks people into: pick from a list, match a personality type to a title, done. It doesn't hold up, because the list was never about you.
Here's what I use instead. Not a quiz. Not a title. A process: gather evidence, run small experiments, compare what you find, and keep a record you can actually update.
It's not that you lack information. If anything, you have too much, and none of it is about you specifically.
Job titles that mattered ten years ago are shrinking or gone, and new ones keep appearing. Every "top careers" list contradicts the last one you read. More options doesn't feel like freedom — it feels like standing in a hallway with forty doors and no way to check what's behind any of them without opening it.
Somewhere along the way, "what career should I pick" quietly turned into "what does this say about who I am." That's a heavier question, and it's usually why people freeze. You're not just choosing a job. You're worried you're choosing a self.
Most people have close-up knowledge of maybe three or four jobs — their own, and the ones held by people close to them. Everything else is guesswork dressed up as research: a job description, a stock photo of someone smiling at a laptop. You can't compare what you haven't seen.
I stopped here for a while. Not for lack of a framework — for lack of evidence. The framework came after.

Before comparing paths, gather data on yourself. Not opinions — evidence, from things that already happened.
Track two weeks of your actual days. Not what you're supposed to enjoy — what leaves energy left over versus what empties you out, regardless of whether you were good at it. I used to assume I liked deep solo work because I was decent at it. Turned out I was decent at it and quietly drained by it, every time.
Look at your own procrastination. What do you keep putting off, even when it's small and easy? Avoidance is data. Consistently ducking a certain kind of task — cold outreach, detailed editing, public presenting — tells you something a personality label won't.
What do people come to you for, unprompted? Not what's on your resume — what colleagues or friends pull you into. I'd never have called myself "the person who fixes messy plans," but three people in three different years asked me to do exactly that. I hadn't noticed the pattern until I wrote it down.

Once you have real evidence, compare specific paths against it — not against a fantasy version built from someone else's highlight reel. This comparison isn't a one-time event for most people: as of early 2026, roughly two in three recent job changers had switched occupations entirely, not just employers, so treat this as a skill you'll use again, not a puzzle you solve once.

Job titles hide enormous variance in what a Tuesday looks like. "Product manager" at a ten-person startup and at a 5,000-person company are close to different jobs. Get specific: what does an average day involve, hour by hour, not mission-statement language.
How steep is the ramp, and does it match how you learn? Some paths reward fast generalists who pick things up in weeks. Others reward people willing to sit inside narrow expertise for years before it pays off. Mismatching yourself to the wrong curve is what burns people out early.
The category most advice skips, and often the deciding one: hours, travel, location flexibility, income timeline. A path can be a perfect interest match and still be wrong if it needs constraints your actual life can't absorb right now.
Where does this path lead in five years — a direction you'd want, or just the only visible one? Some careers fork constantly. Others run a single track. Worth knowing which before you commit.

This worked for my situation: I stopped trying to think my way to certainty and started running small, cheap tests instead. Harvard's career center outlines a five-step framework for career decisions that treats the choice as a process with reflection built into each stage, not one leap.
Twenty minutes with someone actually doing the work beats an hour of reading about it. Ask what a normal week looks like, and what never makes it into the job description. Same discipline as preparing questions for a job interview — gathering real information, not making small talk.
A weekend project, a freelance gig, a volunteer task inside the field gives you your own reaction to the work, not just a description of it — and that reaction ages better than research does, since the skills a given job requires have shifted roughly 25% since 2015, per LinkedIn's 2026 workforce data. I bombed my first attempt: picked a project adjacent to the field, not inside it, and learned almost nothing. The second, closer to the real work, told me more in four hours than three months of reading had.
A half-day shadow or short volunteer stint gets you inside the room without asking anyone to bet on you. Low-cost for both sides — most people say yes more readily than you'd expect.

Evidence and experiments only compound if you keep them somewhere. Otherwise you relearn the same things every time you circle back to "so what do I actually want to do."
Keep a running, short list — two to four paths, not twenty. Add or drop as evidence changes your read. Not a permanent list. A current one.
Under each option, log what you've learned: interview notes, what the short project told you, where your energy went during the shadow day. This is where the real comparison happens — not in your head, on the page.
Write down what you still don't know. People skip this because it feels like admitting you're not done. You're not supposed to be. "Still figuring this part out" is an honest status, not a failure to reach one.
I keep this whole structure in Macaron now, re-sortable as options change. The scope is narrow — options, evidence, open questions — and inside it, it stays current instead of stale.
That's where mine stands right now: two options still open, one mostly ruled out by a project that told me more than I wanted to hear, one open question about income timeline I haven't tested yet. Not a conclusion. A current read.
Don't think your way to a tiebreaker — run the cheapest experiment for each option instead. An informational interview or short project usually breaks the tie faster than more reflection does. If both still feel equally live afterward, that's useful too: the deciding factor may be a constraint, not an interest difference.
Treat them as a starting point, never a verdict. The O*NET Interest Profiler, maintained by the U.S. Department of Labor, is a solid free option for a structured first list. Lighter tools — including casual quizzes, some originally built for teens — can help generate first ideas, but they're built on a handful of self-reported answers, not your actual behavior. Weight your own evidence above any quiz result.
Ask exactly what outcome it's designed to produce, in specific terms — not "clarity," but "you'll leave with X." Check whether it gives real exposure to the field, or only frameworks and worksheets. Frameworks don't substitute for small experiments. If a program can't point to anything you'll test against real evidence, it's teaching you to think about a decision, not helping you make one.
When it replaces your own evidence instead of prompting you to gather it. "Follow your passion" skips the actual comparison — daily work, learning curve, constraints, growth — and hands you a feeling instead of a finding. Advice that tells you what to conclude, rather than what to go check, is worth being skeptical of.
Informational interviews, short paid or volunteer projects, and single-day shadowing are the three lowest-cost tests available, and none require quitting anything or announcing a decision. Job content changes faster than a static description can keep up with, so a short, real-world test tells you more than reading ever will.
You decide if this maps to your situation. The version of me who wrote a five-year plan from a "top careers" list had no evidence behind it. The version who tracked energy, ran three small experiments, and kept a running map had something to actually compare. That's the whole difference — not more certainty, just more to compare against.
At Macaron, we built the Career Decision Map for exactly this stage — logging the options you're weighing, the evidence from each experiment, and the questions still open, updated as you go instead of rebuilt from memory. If you're stuck between two or three directions, try mapping them out and see whether the evidence actually points anywhere. It won't tell you which career is right — nothing can do that part for you — but it keeps what you've learned in one place.
Related articles