Which traits each firm benchmarks for

Firms publish very little about their Pymetrics benchmarks, so most "BCG wants X trait" guides are guesswork. Here's the honest version.

What firms have actually disclosed

BCG: Has confirmed Pymetrics is used in EMEA and parts of APAC. Emphasises learning agility and decision-making clusters in published recruitment material. The benchmark reportedly favours moderate risk and high learning rate.

JPMorgan: Uses Pymetrics across analyst programs globally. Published guidance mentions risk tolerance, fairness, and working memory. The benchmark is calibrated per program (markets vs banking vs ops).

Bain: Uses Pymetrics selectively. Less public detail; consistent reports suggest attention and fairness are weighted, with risk tolerance closer to neutral.

McKinsey: Does not use Pymetrics — uses Solve instead.

What's likely true but unconfirmed

• Consulting benchmarks favour learning agility and adaptive decision-making over high raw risk-taking.
• Banking benchmarks weight attention and working memory more heavily.
• Fairness and prosocial traits are positive across most benchmarks but rarely decisive.

What's myth

• "Always play it safe in the balloon game." Wrong — extreme safety profiles flag as risk-averse, which most consulting benchmarks penalise.
• "Click as fast as possible to show processing speed." Wrong — speed without accuracy lowers your trait estimate.
• "Match a specific score on each game." Pymetrics doesn't score games individually; it estimates traits across all games.

How to use this

Don't optimise per firm. Two reasons: (1) the benchmarks aren't fully public, so you'd be optimising blind; (2) Pymetrics detects rehearsed behaviour and discounts it. Play naturally and consistently. If your honest trait profile doesn't match a firm, no amount of game-mimicking will fix that — and the assessment isn't the only screen.

Practical takeaway

Familiarise yourself with the games. Play them in your real test mood. Trust the estimate. Spend your prep energy on the parts of the process that reward preparation — case math, structuring, video answers — not on trying to game a behavioural assessment that's designed to detect gaming.