Career Tests Are Astrology for Professionals

Somewhere in the last decade of your career, you probably took a personality test and received a four-letter code. Maybe you were told you’re an INTJ, or an ENFP, or an ISFJ. Maybe you took the Enneagram and discovered you’re a Type 4 with a 3 wing. Maybe a career coach walked you through a strengths assessment and you emerged knowing you’re an “Achiever” and a “Learner.”

And then nothing much happened, because knowing your personality type doesn’t tell you whether you’d be better compensated as a data analyst or a product manager, which city pays 40% more for your specific skill set, or whether the occupation you’re currently in is projected to shrink by 15% over the next decade.

The career testing industry has a fundamental problem. It is measuring the wrong thing, with the wrong tools, and producing outputs that feel insightful but carry almost no actionable information about the labor market.

What personality tests actually measure

The Myers-Briggs Type Indicator, the most widely administered personality assessment in the world, was developed in the 1940s by a mother-daughter pair with no formal psychology training, working from their reading of Carl Jung. It assigns people to one of 16 types based on four binary dimensions: introvert or extravert, sensing or intuition, thinking or feeling, judging or perceiving.

The test has been studied extensively by occupational psychologists over the subsequent 80 years. The research findings are fairly consistent. MBTI has low test-retest reliability, meaning a substantial portion of people get a different result if they take the test again a few weeks later. Its predictive validity for job performance is weak. The personality dimensions it measures don’t map cleanly onto the skill dimensions that determine occupational fit or compensation.

None of this stopped it from becoming a $2 billion industry with installations at the majority of Fortune 500 companies and virtually every major university career center.

The Enneagram is in a similar position. Unlike MBTI, it has almost no peer-reviewed validity research supporting its use at all. It’s a typology system with origins in spiritual philosophy, not occupational research. The number of corporate HR departments that have adopted it enthusiastically is, from a methodological standpoint, baffling.

StrengthsFinder is somewhat better grounded, but it still shares the core problem: it tells you things about your psychological tendencies, not about the structure of the labor market, what roles actually pay, what skills are transferable to what occupations, or where the compensation ceiling is in your field.

The self-report problem

Beyond the specific instruments, career quizzes share a structural flaw that no amount of clever question design can fully overcome: they ask you to describe yourself, and people are not reliable narrators of their own capabilities.

There’s a large body of research on this. People systematically overestimate their abilities in domains they find personally important and underestimate them in domains they don’t identify with. They describe their work preferences based on what they think they should prefer rather than what their actual behavior suggests. They conflate what they enjoy with what they’re good at, and what they’re good at with what the market will pay for.

A career quiz asks you whether you prefer working with people or with data. You answer based on your current job, your current mood, your self-image, and whatever framing effect the question introduced. The quiz takes your answer as ground truth and generates a recommendation from it.

This is not measurement. It’s a structured conversation with your self-perception, which may or may not be accurate, and it has no connection to the actual information that would help you make a better career decision.

What “10 careers” means as an output

Most career assessments, when they do produce job recommendations, return somewhere between 5 and 15 roles. The selection is constrained by what the test developers thought to include, filtered by whatever personality or preference categories you fell into, and presented as a meaningful set of options.

Consider what’s been excluded. The U.S. Department of Labor’s O*NET database covers more than 1,000 distinct occupations. Within each occupation, compensation varies by experience level, by geography, by industry sector, by company size, and by the specific skill mix the individual brings. The actual decision space facing any given person trying to optimize their career is enormous.

A career quiz that returns 10 options isn’t giving you 10 options from that full space. It’s giving you 10 options from the much smaller set the quiz designers thought to include, pre-filtered by your self-reported personality. The other 990-plus occupations, including potentially the highest-paying roles where your skills represent a strong match, never appeared in the analysis at all.

This is a bit like asking a travel agent to recommend vacation destinations, and the travel agent only knows about places they’ve personally visited. The recommendations might be fine. But they’re bounded by the agent’s experience, not by the full set of places that might be right for you.

The coach problem

Career coaches have a different version of the same issue. A good career coach brings genuine value: emotional support through a difficult transition, accountability for following through on a job search, industry knowledge in domains they know well, and experience with interview preparation and negotiation.

What they cannot do is systematically evaluate your skill profile against 1,000 occupations, cross-reference that against current BLS salary data for every metro area, assess automation risk across occupation categories, and return a ranked list of opportunities ordered by how well they match your specific mix of capabilities and priorities.

A career coach works from their own mental model of the labor market. That model is inevitably partial. It reflects the industries they’ve worked in, the clients they’ve seen, the job market conditions that were current when they built their expertise. A coach who knows enterprise software sales very well will tend to see enterprise software sales as an option for a wide range of clients, because that’s the domain where their pattern-matching works.

This isn’t a criticism of individual coaches. It’s a structural limitation. The labor market is too large and too dynamic for any individual to hold a complete model of it in their head. Expertise in a subset of it is the best anyone can achieve.

The result is that career coaching, at $150 to $500 per session, often produces recommendations that reflect the coach’s experience as much as the client’s situation. Which is useful, but not the same as a systematic analysis of where your particular skill profile has the highest value.

Opinions versus data

The distinction worth drawing is between systems that generate recommendations from opinions and systems that generate them from data.

Career quizzes generate recommendations from your opinions about yourself, filtered through the quiz designer’s opinions about which personality types suit which roles. Career coaches generate recommendations from their professional opinions, informed by their experience and limited by its scope.

Data-driven career matching works from a different starting point entirely. The O*NET database provides standardized skill profiles for more than 1,000 occupations, built from systematic research into what each role actually requires. Bureau of Labor Statistics data provides salary information by occupation, by geography, and by experience level, derived from actual employment records rather than surveys or estimates. Automation exposure research provides forward-looking risk scores based on task-type analysis across occupations.

None of that requires anyone’s opinion. It’s measurement data about the structure of the labor market, collected and maintained by federal agencies whose methodological documentation is public.

The question is whether your specific skill profile, mapped into the same framework those occupations are described in, matches well or poorly against each of them. That’s a calculation, not a judgment call.

What PathScorer actually compares

The site lays out the contrast directly, and it’s worth looking at it clearly.

These are not variations on the same approach. They’re different approaches to a different version of the question.

Why the astrology comparison holds up

Astrology is interesting as a cultural practice for a specific reason: it generates descriptions of people that feel accurate because they’re constructed to feel accurate. The descriptions are general enough to apply to most people, framed positively enough that people recognize themselves in them, and specific enough in detail to feel like real insight.

Career personality tests work the same way. Being told you’re a “big-picture thinker who thrives in collaborative environments” is pleasant, recognizable, and nearly content-free. Most people could describe themselves that way. The 16 MBTI types cover all of humanity; every human fits somewhere within them. Being categorized isn’t the same as being analyzed.

The deeper problem is that people make real decisions based on these instruments. They stay in fields that pay them below their market value because a quiz told them they’re suited for it. They avoid career pivots because a personality type profile doesn’t mention the destination occupation. They spend money on coaching that returns the same options they already knew about.

The career quiz industry is not malicious. Most people running it believe in the tools they’re using. But belief and validity are different things, and the gap between them costs real people real money over the course of their working lives.

A different question to ask

The right question isn’t “what type of person am I and what does that say about my career?” It’s “given everything I can actually do, which specific roles in the current labor market offer the best combination of compensation, fit, stability, and growth, and what would it take to get there?”

The first question produces a personality profile. The second one produces a ranked list with salary data, gap analysis, geographic optimization, and concrete paths forward.

Those are different questions and they have different answers. The first answer feels like self-knowledge. The second one is actionable.