What is the pymetrics test?

Pymetrics is a set of 12 neuroscience-based games (plus up to four optional numerical and logical games, so 16 in total) that an employer asks you to play as part of a hiring process. It takes roughly 25 to 35 minutes. Instead of asking you to describe yourself, it records how you behave while playing: how you weigh risk, how you respond to fairness, how you hold attention, how fast you learn a pattern. From that behaviour it builds a profile across 91 traits grouped into nine categories. Pymetrics has been owned by Harver since August 2022, and the assessment is delivered to candidates through the employer, not sold to individuals.

Can you fail a pymetrics test?

Not in the usual pass-or-fail sense, because pymetrics is not scored against a fixed answer key. It compares your behavioural profile to the profiles of people who already succeed at the specific company you applied to, then returns a recommendation tier to the recruiter. So whether your result helps you depends on the role and the company, not on a universal score. The same profile that matches one employer's benchmark can sit outside another's. There is no objectively correct way to play, which is also why it is harder to game than a personality quiz.

Can I take pymetrics on my own to figure out my career?

Not in a meaningful way, and this is the key thing most reviews skip. Pymetrics reaches you as an invitation from an employer you applied to. There is no consumer version you buy and take to explore your own direction, and even when you complete it, you do not receive the full trait profile the employer sees. It is a hiring instrument that answers the employer's question (does this candidate match who already succeeds here), not your question (what direction fits me right now). For self-directed career discovery you need a self-serve tool, which is the gap this review covers.

How do you prepare for the pymetrics games?

Pymetrics is deliberately built to resist preparation, because it measures behavioural tendencies rather than knowledge. You cannot revise your way to a different risk appetite. The useful preparation is practical, not academic: play in a quiet space with a stable connection, read each game's short instructions carefully since the rules differ game to game, and behave consistently rather than trying to second-guess the trait being measured. Trying to fake a profile usually backfires, because the games cross-check behaviour across tasks and inconsistency shows.

Is pymetrics accurate, and is it fair?

Pymetrics measures behaviour through performance rather than self-report, which is a structurally stronger signal than a Myers-Briggs-style type and harder to manipulate. The company also built its brand on auditing its models for adverse impact, so on the fairness question it is more rigorous than most hiring screens. The honest limit is scope and ownership: it is an accurate reading of behavioural traits for one employer's benchmark, the candidate never sees the full result, and it was never designed to tell you what work would suit you. Accurate for the employer's purpose is not the same as useful for yours.

What is the best pymetrics alternative for career discovery?

It depends on the question you are asking. If you want a self-serve assessment to understand your own direction rather than to pass an employer's screen, MyPassionAI runs a 3-minute branching quiz that returns one of 20 situational archetypes built around your current struggle and priority. If you want a different cognitive or aptitude instrument you can buy yourself, YouScience measures aptitude through timed exercises. Pymetrics answers what an employer wants to know about you. MyPassionAI answers what you want to know about your next move, and the best alternative is the one matched to your actual question.

Pymetrics Test Review (2026): What It Measures, Who It Is Built For, And The Self-Serve Alternative

If an employer just asked you to play a set of online games, or you searched pymetrics because you want to understand yourself better, the honest framing is this: pymetrics is a hiring tool, not a career-discovery tool, and the difference changes everything about whether it can answer your question. It will read your behaviour accurately. Whether that reading is useful to you depends on whose question it was built to answer, and the answer is the employer's, not yours.

What you should do next depends on where you are right now. Someone playing pymetrics this week for a consulting application needs different guidance than someone typing the word into a search bar at 11pm wondering what they should do with their career. This review covers both: what the test genuinely measures, who it serves, and what to use instead when the question is your own direction.

A note on framing first. I founded MyPassionAI, the alternative discussed later in this review, so I have a commercial reason to prefer a self-serve approach. I kept every claim to what you can verify on Harver's pages or in third-party reviews, gave pymetrics a fair and specific account of what it does well, and included the unflattering parts of my own product.

TL;DR comparison

Dimension	Pymetrics	MyPassionAI
What it is	Behavioural hiring assessment (games)	Situational career-direction quiz
Who it serves	The employer screening you	You, making a direction decision
How you get it	Invitation from an employer you applied to	Self-serve, taken whenever you want
What it measures	91 behavioural traits in 9 categories	Current struggle, priority, flow markers, values
How you answer	Behaviour recorded while playing games	Branching self-report questions
Time to complete	~25 to 35 minutes, 12 to 16 games	~3 minutes, 24 branching questions
How it scores	Trait-match against that company's top performers	One of 20 archetypes from a 5x4 matrix
What you receive	A recommendation the recruiter sees, not your full profile	Your archetype, career matches, and fit scores
Owner	Harver (acquired pymetrics August 2022)	Independent, founder-built
Best for	Passing an employer's screen	Deciding what to do next given where you are

Everything below is the work behind that table.

What pymetrics is

Pymetrics is a behavioural assessment that large employers use to screen candidates early in hiring. It is not a consumer product. Since August 2022 it has been owned by Harver, a high-volume hiring platform, and the games sit inside Harver's wider talent-screening suite.

You encounter it because you applied somewhere and the company sent you a link. That detail matters more than any other in this review, so it is worth stating plainly: you do not buy pymetrics, you do not choose to take it, and you do not get to keep a full readout of what it found. It is administered by the employer, scored for the employer, and read by the employer.

The brand was built on a serious idea. Pymetrics positioned itself as a fairer way to screen than résumé-and-gut-feel hiring, and it audited its models for adverse impact across gender and ethnicity. For an employer trying to widen a funnel and reduce bias, that is a genuine selling point, and it is one reason firms like Boston Consulting Group, Bain, Unilever, JPMorgan, Accenture, PwC and McDonald's adopted it.

How the games work

The core assessment is 12 short games, with up to four optional numerical and logical games on top, so most candidates play somewhere between 12 and 16. The whole thing runs about 25 to 35 minutes, with each game lasting one to three minutes. Names that come up across published guides include Balloons (risk), Money Exchange (trust and fairness), Digits (working memory), Keypresses (motor speed and effort), Cards, Arrows, Towers and Faces (emotion recognition).

The mechanic is the point. A personality quiz asks you to rate statements about yourself, and self-description is shaped by mood, by self-image, and by what you think the test wants to hear. Pymetrics instead records what you do under mild time pressure and infers traits from behaviour. Inflate a balloon repeatedly for more points and you reveal risk appetite. Split money a certain way and you reveal how you weigh fairness against self-interest. You are not answering questions about yourself, you are leaving behavioural fingerprints.

What it measures

From that behaviour, pymetrics builds a profile across 91 traits, grouped into nine categories.

Category	What it captures
Attention	Concentration and filtering out distraction
Focus	Sustained concentration over time
Decision making	How you handle uncertainty and choice
Risk tolerance	Comfort with potential gains and losses
Effort	Persistence when a task gets harder
Learning	Adaptability and pattern recognition
Fairness	How you weigh fair distribution
Generosity	Willingness to share or cooperate
Emotion	Reading and responding to emotional cues

Notice what is on this list and what is not. These are cognitive and behavioural tendencies an employer cares about when predicting on-the-job performance. None of them is your interest in a field, your values, your current life stage, or what kind of work would hold your attention for a decade. That omission is not a flaw in pymetrics. It is simply the line between a hiring instrument and a career-discovery one.

How scoring works, and why it is relative

This is where pymetrics differs most from the quizzes people usually compare it to. There is no universal score and no answer key. Before you ever play, the hiring company has a group of its existing high performers play the same games to set a benchmark. Your profile is then matched against that company's benchmark, and the recruiter sees a recommendation tier rather than a number.

The consequence is that the same behavioural profile can match one employer and miss another, because each company calibrates against its own people. You are not being measured against an absolute standard of who you are. You are being measured against who already succeeds at one specific place. That is reasonable for hiring, and it is exactly why the result is close to useless as a statement about your own direction: it describes your fit to someone else's team, not your fit to a life you would choose.

Pymetrics is hard to prepare for on purpose. You cannot study your way to a different risk appetite, and trying to fake a profile tends to backfire because the games cross-check behaviour across tasks. The honest preparation is practical: quiet room, stable connection, read each game's instructions, and behave consistently rather than performing a character.

The structural gap: it answers the employer's question

Strip away the neuroscience branding and pymetrics is a sorting machine pointed in one direction. It takes a pool of applicants and ranks them by similarity to an employer's existing top performers. That is a defensible thing to build, and pymetrics builds it more carefully and more fairly than most.

Pymetrics ranks you by similarity to who already succeeds at one company. That is a fact about their team, not a map of your direction.

But it cannot answer the question that brings most people to a search bar. If you are deciding whether to leave a career you are good at, or which direction fits you now that your situation has changed, pymetrics has nothing to say, because it was never pointed at you. It does not know your struggle, it does not ask what you would do if money were not a constraint, and it never shows you the full picture it built. The thing that changed in your life, your context and what you now want, is precisely what a benchmark-relative hiring screen is designed to ignore.

That is the gap a self-serve tool fills. When MyPassionAI's quiz asks Q14, "when do you completely lose track of time," and Q21, "if you didn't need money for the rest of your life, what would you wake up excited to do", it is reaching for signals pymetrics has no reason to collect: flow and values, the two things that predict whether work sustains you rather than whether you can perform it.

What MyPassionAI is, and how it differs structurally

A disclosure before this section: I founded MyPassionAI, so the operator detail here comes from owning the code, with the bias that brings. The honest counter is to publish the comparison with verifiable claims and my product's weaknesses included.

I launched MyPassionAI in 2025 after a decade optimising a career for the wrong target. The thesis is state-dependent. Career fit is not a stable function of your traits alone, it is a function of where you are right now (your current struggle), what you want to optimise for next (your priority), and how signals like flow and values interact with both.

The instrument. 24 questions, branching from the first one. Q1 asks which of four situations describes you: a student or graduate with no direction, someone in a paying career who wants change, someone with too many interests, or someone stuck or unemployed. Your answer reshapes the questions that follow, so the quiz reads your situation before anything else.

The framework. The output is one of 20 named archetypes from a 2D matrix: five struggle types (Career Switcher, Grad Explorer, Multi-Passionate, Purpose Seeker, Explorer) crossed with four priority types (Income-Focused, Lifestyle Seeker, Stability First, Experimenter). Both axes are designed to move when your life moves, which is the attribute a behavioural benchmark cannot have.

The honest scope. MyPassionAI does not measure cognition. It will not tell you your risk appetite or your working-memory span, and it cannot screen you for a consulting role. It measures situation, flow, and values, and returns a direction in about three minutes. The two tools answer opposite questions, and used together (pymetrics when an employer requires it, a self-serve quiz when you want your own read) they do not contradict each other.

What MyPassionAI gets wrong, honestly

1. It is self-report. Pymetrics records behaviour, which is harder to game than answering questions about yourself. On resistance to manipulation, pymetrics wins, and it is not close.

2. No behavioural or cognitive data. If your question is how you weigh risk or how fast you learn a pattern, MyPassionAI does not measure it. Pymetrics does.

3. Newer and not yet peer-reviewed. The matrix is informed by Self-Determination Theory and flow research, but it has not been independently validated. Pymetrics has years of deployment data and published fairness audits behind it.

4. The free tier is a teaser. Your free result reveals your archetype and a short reading. The full career matches and fit scores sit in the paid report. That is a fair criticism of the funnel.

Who should take which

Play pymetrics when:

An employer you applied to requires it (you do not have a choice here, so prepare practically and play honestly)
You want a behavioural, performance-based read rather than a self-description, in a hiring context
You care that the screening tool has been audited for fairness

Take the MyPassionAI career quiz when:

You are making a direction decision, not passing a screen: which career fits the version of you that exists today
Your situation recently changed (graduation, burnout, a layoff, a move) and you need an answer that accounts for it
You want a self-serve result you can take whenever you want, that updates when your life updates
You have three minutes and want a starting point optimised for "what should I do next"

If neither fits

Other comparisons in this silo are worth a look:

The YouScience review if you want an aptitude instrument you can buy yourself, measured by timed performance rather than self-report
The Career Dreamer AI review if you want an AI-driven career-exploration tool aimed at individuals
The RIASEC test guide if you want the Holland-code interest framework that sits underneath most traditional career assessments

If the deeper question is not which test to take but how to think about direction at all, how to find your passion is the conceptual foundation underneath every quiz in this silo.

The bottom line

Pymetrics is a carefully built behavioural hiring assessment. It measures performance rather than self-report, it audits itself for fairness more seriously than most screens, and for an employer trying to hire from a wider pool it earns its place. As a candidate told to take it, you should play it honestly and not try to study for it.

What it is not is a career-discovery tool. It reaches you through an employer, it scores you against that employer's existing team, and it never hands you the full picture. The question it answers belongs to the company on the other side of the link, not to you. According to Self-Determination Theory, what sustains people in work is autonomy, competence and relatedness, the intrinsic side of motivation, and a benchmark-relative hiring screen is structurally blind to all of it.

If your question is your own next move, take the MyPassionAI career quiz: three minutes, 24 branching questions, one of 20 situational archetypes, and career matches built around the part of you that is changing rather than the part an employer wants to sort. The quiz also flags which archetypes are most prone to staying too long in work they are good at, which is the exact trap that performs well on a screen and quietly drains the person passing it. Trusted by more than 3,000 quiz takers so far.

I wrote this review knowing I benefit from you clicking that link. I tried to earn the click anyway.

Written by Marco Kohns, founder of MyPassion.ai, former Growth Product Manager at a venture-backed consumer startup operating in 100+ countries, ex-Techstars Berlin consultant, author of a Journal of Business Research paper on generative AI for growth hacking (MSc NOVA IMS Lisbon, 18/20).