How the Skills Passport score works

A Skills Passport score is one number — between 300 and 850 — that summarizes a candidate's evidence across four weighted competency pillars, adjusted for how recently each piece of evidence was generated. This page explains exactly how that number is built.

The 300–850 scale

The composite score lives on a fixed 300–850 range. The anchoring is deliberately wide so that small differences in raw assessment performance translate to meaningful, recruiter-readable score differences without bunching at the top of the distribution.

  • 300–499 — Below the typical role-readiness threshold for the bundle the candidate took. Useful as a baseline; the candidate likely has high-leverage learning to do before the role.
  • 500–649 — Functional. The candidate clears the bar on most of the role's core competencies but has uneven coverage. Reasonable hire with growth runway.
  • 650–749 — Strong. The candidate is solidly above the role-readiness threshold and demonstrates depth in two or more pillars. The hiring decision is mostly about culture, scope, and team fit.
  • 750–850 — Exceptional. The candidate is in the top tier across the bundle's pillars and has demonstrated this on multiple recent assessments.

The bands above are the same on every Passport, regardless of which test family produced the underlying scores. That's the calibration property: a 720 on AI-Augmented SQL means the same recruiter-facing thing as a 720 on Communication.

The four pillars and their weights

Every Skills Passport composite is computed as a weighted average of four pillar scores. The pillar weights reflect predictive validity for general knowledge work; role-specific bundles can override these defaults if the role demands a different mix.

0.35

Domain

The role-specific competency the candidate is being hired for: programming language fluency, financial modeling, clinical reasoning, etc. Highest weight because role-readiness is the modal hiring decision.

0.25

Cognitive

General reasoning, working memory, pattern recognition. A reasonable predictor of how quickly someone gets up to speed on novel work.

0.25

AI fluency

Ability to collaborate with AI systems: prompt design, output evaluation, error-state handling, knowing when to override the model. Weighted equal to cognitive because this skill increasingly shapes throughput in real roles.

0.15

Communication

Written clarity, structured argument, audience adaptation, brevity. Necessary but not sufficient — kept lower-weight because most knowledge-work hires need adequate (not exceptional) communication.

Weights sum to 1.0. A candidate's composite is 0.35 × domain + 0.25 × cognitive + 0.25 × ai_fluency + 0.15 × communication, with each pillar score on the 300–850 scale.

Recency decay

Each pillar score is adjusted downward over time. Skills aren't fixed traits — coding fluency, AI familiarity, and domain knowledge all shift as tools and norms evolve. A 2-year-old programming score doesn't mean what it did when it was issued.

The decay is exponential with a half-life that depends on the underlying construct:

  • Domain (technical) — half-life ~18 months. A 700 issued 18 months ago contributes as if it were a 525 today, and the candidate is prompted to re-test.
  • AI fluency — half-life ~12 months. The faster shift in tools and norms here means stale scores decay quickly; recruiters should expect candidates to re-test annually for currency.
  • Cognitive — half-life ~5 years. Cognitive ability is more stable than domain skill, so older scores still carry meaningful signal.
  • Communication — half-life ~3 years. Slower than domain skill but faster than cognitive, since communication norms drift with industry conventions and AI authorship habits.

The Passport surfaces both the headline (decay-adjusted) score and the original assessment date, so recruiters can see whether they're looking at a fresh result or a decayed one.

What the score is not

Three things the Skills Passport composite explicitly does not claim to be:

  • A hiring decision. The score summarizes evidence on calibrated dimensions. It does not capture culture fit, motivation, ethics, scope of past work, or any of the things a structured interview is for.
  • A fixed trait. All four pillars are trainable. A 580 today can be a 720 in eighteen months with deliberate practice and re-testing.
  • A complete picture. The pillar weights are defensible defaults for general knowledge work, but real hiring loops should adjust the bundle composition (and sometimes the weights) for the specific role.

Disclaimer

Assessment results and endorsements reflect performance on standardized tests at a point in time. Hiring decisions should consider multiple factors beyond test scores.