The Collapse of Trust in Modern Hiring
The foundation of modern hiring is broken.
Resumes are unreliable (46% embellishment rate) and generative AI has made them worse. Recruiters waste 20+ hours per week on manual screening with low-confidence results.
For decades, recruiters relied on resumes and credentials as a proxy for ability. Today, those signals have collapsed.
We call this "Signal Collapse" — a high-volume, low-trust crisis fuelled by two forces:
- Systemic Unreliability — The market has always been built on claims, not proof. Global studies show as many as 46% of job seekers admit to resume embellishments.
- The AI Tsunami — Generative AI has weaponised this unreliability. Candidates now use technology to artificially enhance resumes at scale, making it nearly impossible to separate signal from noise.
This collapse traps recruiters in what we call the "Hiring Vortex" — a painful loop of manual work and low-confidence decisions. Our research quantifies the pain: recruiters report losing an average of 20 hours per week on manual screening.
The result is an expensive guessing game. Companies rely on gut feel and subjective interviews — poor predictors of success. This leads directly to costly mis-hires.
“We just wasted 3 weeks. The person can’t do the job.”
To build a high-performing organisation, you must fix the signal. You must move from judging unverifiable claims to verifying Proof of Work.
Hiring Methods — Predictive Validity Comparison
| Method | Predictive Validity |
|---|---|
| Resumes | r = 0.18 |
| Unstructured Interviews | r = 0.18 |
| Structured Interviews | r = 0.51 |
| Work-Sample Tests | r = 0.54 |
Source: Schmidt & Hunter (1998), Psychological Bulletin
Verifiable Proof of Work
What if there was a method proven by 85 years of research to be 3x more predictive?
Work-sample tests are the most predictive hiring method ever studied (r=0.54). We built our entire platform on this foundation.
If the problem is Signal Collapse, the only solution is to re-introduce trust through verifiable proof.
Foundational meta-analyses in Industrial-Organisational psychology have already proven there is a better way. The seminal work of Schmidt & Hunter (1998), which synthesised 85 years of research, found that work-sample tests are nearly 3x more predictive of on-the-job performance than resumes or unstructured interviews.
Schmidt, F. L., & Hunter, J. E. (1998). "The Validity and Utility of Selection Methods in Personnel Psychology: Practical and Theoretical Implications of 85 Years of Research Findings."
Psychological Bulletin, 124(2), 262–274.
Our platform is designed to do one thing: help you hire using the most predictive, fair, and defensible method — at scale. We call this Verifiable Proof of Work.
Our mechanism is the "Role DNA" — a unique, scientific model of what great looks like for any specific role. This DNA powers every assessment.
We are not a black box that replaces recruiters. Our core philosophy is "Empower, Don’t Replace." We provide a transparent, scientifically-grounded framework to help you make high-confidence decisions, faster.
Predictive Accuracy
Based on 85+ years of I-O psychology research
Source: Schmidt & Hunter (1998), "The Validity and Utility of Selection Methods in Personnel Psychology"
The 6 Core Dimensions — The "What" We Measure
A predictive method is only as good as what it measures.
We measure the two "gold-standard" predictors — General Mental Ability and Conscientiousness — through 6 observable, on-the-job behaviours.
Our framework is built on a second layer of foundational science. The seminal meta-analyses by Schmidt & Hunter (1998) and subsequent reviews (Sackett et al., 2023) identify two gold-standard predictors above all others:
- General Mental Ability (GMA) — The strongest single predictor of job performance (r=0.51).
- Conscientiousness — The most robust Big Five personality trait for predicting success (r=0.31).
We’ve deconstructed these abstract constructs into 6 measurable, on-the-job behaviours:
Cognitive & Analytical
How they find signal in noise.
GMAExecution & Reliability
Bias-to-action, sequencing, and completeness.
ConscientiousnessCommunication & Collaboration
How they align humans and ideas.
GMA/ConscientiousnessAdaptability & Learning Agility
Response to new information and change.
GMAEmotional Intelligence
Tone control, empathy, and conflict regulation.
ConscientiousnessJudgment & Ethics
Principled reasoning and risk calibration.
GMAThe 24+ Archetypes — The "How" We Measure
Dimensions are too broad to measure directly. You need the right questions.
We use 24+ Question Archetypes — scalable structured interviews that are one of the most powerful predictors of job performance (r up to 0.51).
A dimension like "Cognitive Ability" is too broad to measure fairly. You must observe it through specific, real-world applications.
To solve this, our framework deconstructs each dimension into a bank of 24+ Question Archetypes.
These are not generic brain-teasers. They are scalable, structured, situational interviews. Meta-analyses show that structured interviews are one of the most powerful predictors of job performance, with validity (r up to 0.51) that rivals GMA itself.
Each Archetype is a repeatable, role-specific "mini-simulation" designed to isolate and observe one of our 6 Core Dimensions:
Intent: Detects structured thinking and causal inference — how the candidate finds signal in noise.
“Given [metric shift], identify likely causes and the data you’d check first.”
“You face [N] conflicting inputs; rank and justify the top 2.”
The Role DNA — The "Context" That Matters
A one-size-fits-all test is unfair. The science agrees.
Our "Role DNA" adapts assessment weights to each role using Job Analysis-Based Weighting — a best practice endorsed by selection psychology.
A one-size-fits-all test is fundamentally unfair and, as I-O psychology has proven, less predictive. The scientific consensus is that the predictive power of various skills shifts with job demands and context — a principle known as Differential Prediction.
Our "Role DNA" framework calibrates the assessment to each role’s specific context. This is a direct application of Job Analysis-Based Weighting — a best practice widely endorsed by selection psychology to enhance predictive validity.
Our Adaptive Complexity model works on two layers:
How weights adapt to the role
Base Presets (by Role Family)
Our system applies Specific Validity Theory by beginning with a scientifically-backed preset for each of the 6 Dimensions based on the Role Family. A "Sales / Customer" role automatically receives higher default weights for Emotional Intelligence and Communication, while a "Finance / Data" role prioritises Cognitive and Execution.
Contextual Modifiers (by Role Specifics)
The AI Co-pilot fine-tunes these weights based on the specifics of the job description and the recruiter’s input. A role tagged as "Customer-Facing" or "Cross-Functional" automatically increases the value of Communication and EI, while "Time-Sensitive" increases the weight of Execution.
The Scoring Matrix — The "Proof > Polish" Principle
We score proof, not polish.
Our scoring uses Behaviorally Anchored Rating Scales (BARS), proven to reduce bias and increase reliability. We score reasoning, not presentation.
A predictive assessment is only as good as its scoring system. A great test can be rendered useless — and biased — by subjective, gut-feel evaluation.
Our framework eliminates this. Our scoring is built on a simple, powerful philosophy: Proof > Polish.
“We don’t score a candidate’s accent, their grammar, or how polished they sound. We score only the observable evidence of their reasoning and judgment.”
We use a Universal Scoring Matrix — a direct application of Behaviorally Anchored Rating Scales (BARS). Unlike simple 1–5 scales, BARS anchors every score to a concrete, observable behaviour. This method is proven to improve fairness and accuracy by forcing all evaluators to use the same defensible standard.
Universal Scoring Matrix (BARS)
Off-topic, generic, or clearly incorrect. No evidence of the target skill.
Partially relevant; thin logic or vague evidence. Acknowledges the problem but doesn’t structure a solution.
Structured, clear, and mostly complete reasoning. Provides a feasible solution.
Concise, evidence-based, and anticipates 2nd-order effects. Defines a solution that is also principled and risk-aware.
Based on Behaviorally Anchored Rating Scales (BARS) — validated in I-O psychology to increase rater reliability and reduce subjective bias.
From Guesswork to Certainty
Every layer validated. Every step defensible.
From method to scoring, each layer is backed by decades of validated research. The result: The Talent Intelligence Matrix — a new, verifiable standard of assessment.
This entire, multi-layered framework — from Signal Collapse to Proof > Polish scoring — is a complete, scientifically-defensible system. We’ve systematically validated every step:
This rigorous process creates our proprietary data asset: The Talent Intelligence Matrix (TI Matrix).
We think of the TI Matrix as "The New CV" — a verifiable, multi-dimensional, and portable record of a candidate’s true ability. The new source of trust in the hiring market.
By synthesising the proven predictive power of work-samples, General Mental Ability, structured interviews, and behaviourally-anchored scoring, we’ve built an engine for predictive, fair, and AI-resistant hiring.
This methodology moves hiring from an art to a science. It empowers recruiters and hiring managers to stop guessing and start knowing.
Validation at every layer
Most predictive (r=0.54)
Most predictive (r=0.51 & r=0.31)
One of the most predictive (r up to 0.51)
Enhances validity and fairness
Reduces bias, increases reliability