The hiring tools are breaking · PowerTechs

PowerTechs · a field brief for employers & hiring partners

The tools you hire with are breaking. Your own industry is saying so.

Across WEF, NACE, Harvard, SHRM and the HR-tech vendors themselves, the 2024–2026 record is consistent: the skills that decide who succeeds are the most in-demand and the least measurable — and generative AI just broke the two tools most companies still screen with.

If you hire for roles where judgment, communication, and composure decide the outcome — operations, cyber, client-facing, program management — this is about the part of those roles a resume screen was never built to see.

Sources compiled below · primary reports only · last updated 2026

0%
of US job postings request at least one durable skill (Lightcast)
1 in 700
hires actually changed after firms dropped degree requirements (Harvard/Burning Glass)
0pp
gap between how proficient students feel and how employers rate them, on leadership (NACE)
0%
of candidates showed signs of cheating across ~20,000 AI interviews (Fabric)

01 · what stopped working

The screens most companies still rely on no longer tell you much.

None of this is speculation. Each of these is a documented shift the hiring industry has already named. Open each to see who reported it.

compromisedThe résumé & the degree+
86% of US employers (89% UK) report problems with résumés as a hiring signal. Degrees were always “more proxy than direct measure” — and when firms dropped them, non-degree hiring rose just 3.5 points, about 97,000 of 77 million yearly hires. The announcement changed; the hiring didn’t.
“Not even 1 in 700 hires last year was due to this shift to skills… the root cause is we’re still crap at assessing skills.” — Matt Sigelman, Burning Glass Institute
broken by AIText & video Q&A assessments+
Candidates now use ChatGPT and purpose-built tools to game résumés, text assessments, and video interviews. ChatGPT outperforms 98.8% of humans on verbal reasoning and scores above the 70th percentile on situational-judgment and critical-thinking tests — the exact formats these screens use. One AI interview platform found 38.5% of candidates showed signs of cheating across ~20,000 interviews; 48% for technical roles.
Google’s CEO suggested returning to in-person interviews. Amazon and Anthropic added anti-AI attestations.
rhetoric > mechanics“Skills-based hiring” announcements+
85% of employers say they do skills-based hiring in 2025. But dropping degree requirements without rebuilding assessment produced a 0.14-point net effect. Only the 37% of firms that rebuilt the mechanics — mapping skills, embedding them in the ATS, training managers — drove nearly all the real change. Intent isn’t the missing piece. Measurement is.
Named leaders who rebuilt mechanics: Walmart, Apple, GM, Target, Cigna, Tyson, ExxonMobil, Koch.

Which of these do you hire with?

Check the tools in your current process. Each one opens to show what specifically broke.

Résumé screening

86% of US employers report problems with résumés as a signal. They stand in for skills they never measured — and AI-generated résumés now inflate the pile while degrading the signal. The durable layer was never on it to begin with.

Video or text Q&A interviews

ChatGPT scores above the 70th percentile on situational-judgment and critical-thinking formats — the exact ones these use. One platform found 38.5% of candidates showing signs of AI-assisted cheating across ~20,000 interviews. The format is compromised.

Self-report skills assessments

Self-report was always gameable through social desirability; GenAI industrialized it. 14% of candidates admit using AI on assessments, and 83% say they would if undetected. Faking is now the default risk, not the edge case.

SkillBridge / fellowships

The highest-fidelity method — you actually see the person work. But it’s 12 weeks, one candidate at a time, and it commits you before you can read the durable layer. It works, but it doesn’t scale, and it front-loads the risk.

02 · the measurement failure, quantified

Everyone agrees these skills matter. Almost no one can measure them.

NACE surveyed 20,000+ students and 237 employers. On the durable skills that decide who succeeds, students rate themselves proficient; employers don’t. The gap is the assessment failure — there’s no instrument reconciling the two.

students who feel proficient employers who agree
Professionalism−34.6pp
Leadership−30pp approx
Communication−25pp approx
Critical thinking−24pp approx

NACE 2024 Student Survey & Job Outlook 2025. Professionalism is exact (84.6% vs 50%); others are reported as approximate 25–30pp gaps and shown as such.

Where does this land in your industry?

The perception gap is general. The roles it shows up in aren’t. Pick your vertical to see which roles run on the durable layer — and what employers there actually screen for.

Signal: 78% of tech leaders plan to grow headcount H2 202665% say talent is harder to find

Roles where the durable six are core, not peripheral: cybersecurity analyst / incident responder (composure + analytical reasoning), IT project manager, cloud support engineer, solutions architect (clear communication to non-technical stakeholders). The technical stack gets them in the door; the durable layer decides who holds up in an incident.

Signal: one of the strongest growth areas for durable-skill demand

Roles: risk & compliance analyst, financial-operations analyst, fraud investigator. LinkedIn’s 2026 fastest-growing clusters — financial operations, risk-compliance management — are analytical-reasoning + composure-under-pressure plays. The certification is checkable; the judgment under a live risk event isn’t.

Signal: highest MOS transfer for military talent

Roles: operations manager, area / shift manager, supply-chain analyst, dispatch coordinator. LinkedIn ranks operational efficiency and process optimization among top-growing skills. This is the highest-conversion vertical for veterans — and it runs almost entirely on coordination and real-time decisions the screen can’t read.

Signal: rising expectations on adaptability & collaboration

Roles: clinical operations coordinator, care-team coordinator, non-clinical administration (team coordination + composure under pressure). The license is verifiable; whether someone holds a team together on a hard shift is exactly the layer that isn’t.

Signal: defense employs the highest % of veterans of any sector

Roles: program manager, quality-assurance lead, field service technician, production supervisor. Clearance and certification gate the door; the program’s success rides on coordination, ownership, and composure — the layer no screen scores.

03 · why it stays hard

Six reasons the durable layer resists every standard screen.

i

Faking, now industrialized

Self-report was always gameable. GenAI scaled it: 14% of candidates admit using it on assessments, 83% would if undetected.

ii

Failing proxies

Résumés and degrees stand in for skills they never actually measured — 86% of employers report problems with them.

iii

Volume without signal

The #1 skills-hiring challenge is evaluating applicant volume. AI-generated applications inflated the pile while degrading the signal.

iv

Cost & validity

Rigorous simulations are expensive to build, and 36% of employers say assessments increased time-to-fill.

v

Legal exposure

Opaque AI scoring is now litigated (Mobley v. Workday), pushing design toward transparent, audited tools.

vi

Quality-of-hire is unmeasured

QoH is the #1 recruiting metric, yet only ~25% of teams can measure it — so they can’t validate any soft-skill screen.

04 · what actually works

The methods that survive AI have one thing in common: you watch people do the work.

Filter by approach. Every entry is a real method in market, with named users and reported outcomes — not theory.

Work-sample / Virtual Job TryoutSimulation

Day-in-the-life simulations candidates work through. The closest commercial analog to a practical skills assessment — and the most validated at scale.

Walmart (400k+ hired, 95% completion) · FedEx · Delta · Humana · IKEA (49% recruiter-efficiency gain)
Game-based psychometricsGame-based

Interactive tasks, not questions — explicitly marketed as immune to ChatGPT because GenAI can’t complete them. Bias-reduced and non-self-report.

pymetrics/Harver (Unilever, JPMorgan, McDonald’s) · Arctic Shores (PwC, Deloitte, Amazon, Siemens)
Structured behavioral interviews (STAR)Structured

The standard remedy for interviewer bias and unquantifiable judgment. Employers who test skills before screening résumés make more quality hires (96% vs 87%).

SHRM-recommended · validated via TestGorilla outcomes
Skills inference from work historyGame-based

Internal talent marketplaces that infer skills from real signals rather than self-report — addressing the invisibility of skills in HR records.

Gloat (Unilever, 8–9k redeployed) · Eightfold, TechWolf (Schneider Electric)
Fellowships as extended assessmentFellowship

The highest-fidelity method: up to 180 days of on-the-job audition. Employers use them precisely because a résumé can’t show durable skills — but they’re 12-week, low-scale.

DoD SkillBridge · Hiring Our Heroes (85% offer rate, ~$110k avg start) · Booz Allen, Leidos, Microsoft
Veteran operations simulationSimulation · Veteran

A ~15-minute operations simulation as “try before you buy” for veteran candidates — the nearest existing analog to a scalable veteran skills assessment, but single-employer and narrow.

Amazon × Redeployable

05 · the frame that explains all of it

Every role has two layers. Most screens read only one.

Everything above points at the same structure. Take any posting and the requirements split cleanly in two — and only one half is the part a screen was ever able to verify.

layer one — verifiable

The technical layer

The degree, the years, the certification, the named tool. Pass or fail, and a screen checks it cleanly.

“Bachelor’s + 5 years” · “TS/SCI clearance” · “CDL” · “A&P license” · “8+ years distributed systems”
The screen can verify this.
layer two — unmeasured

The durable layer

Communication, judgment, coordination, composure under pressure. The part that decides who actually succeeds — and the part no standard screen measures.

“excellent communication” · “drive execution in ambiguity” · “real-time decisions under pressure” · “trusted advisor”
Required, but nothing measures it.

Employers name the durable layer more and more explicitly — a line in the description, a formal “Core Competencies” list, sometimes the entire substance of a role. Writing it down didn’t make it measurable. The screen still checks the degree and the years, and takes the durable layer on faith every time.

Try it on your own posting.

Paste any job description below. We’ll split it the way this brief splits every role — what a screen can verify against what it can’t. Nothing leaves your browser.

or try a sample
0
verifiable requirements
0
durable requirements, unmeasured
screen can verifyrequired, nothing measures it

What the durable gap costs you.

A rough, transparent estimate — every assumption is visible and yours to change. Pick a role and adjust the sliders.

3.2
fellows a year who don’t convert — after you’ve carried the full mentoring cost
$5,760
mentoring hours spent on fellows who didn’t work out (at $45/hr loaded)

Assumes ~$45/hr loaded mentor cost. This counts only mentoring time on non-converts — it excludes the 75% offer-commitment risk and the opportunity cost of a slot that could have gone to someone you’d keep. The point isn’t the exact figure; it’s that the gap has a price, and seeing the durable layer first shrinks it.

06 · who already has the durable layer

Military Veterans arrive with layer two built. The screen sees only the half they’re still completing.

This is why the gap matters more for military talent than almost any other group — and why it’s an opportunity, not a problem, once you can see it.

Durable layerbuilt in service
Credential layerstill completing
What the screen seescredential only

A service member spent years leading under pressure. The résumé shows the one thing they’re still finishing.

Composure when it matters, coordinating a team through ambiguity, resourceful problem-solving under real constraints — the durable layer is what military service builds, at a depth a civilian résumé rarely reaches. It’s the layer employers say they want most.

But the credential — the specific degree, the civilian job title, the industry certification — is often the part still in progress at transition. So the screen sees the half they’re missing and misses the half they’ve already mastered. The strongest thing they bring is invisible at exactly the moment you’re deciding.

If you host veterans through SkillBridge, you’re already betting on the durable layer — just without a way to see it first.

See it before you commit →

07 · the gap even the good tools leave

Simulations work. Veteran-native ones that measure durable skills don’t exist at scale.

Three things each solve part of the problem. Nothing on the market combines all three — and that intersection is exactly where a transitioning service member’s value lives.

interactive simulation durable-skill measurement veteran-native design PowerTechs

Simulation alone → not veteran-native

Generic VJTs measure the work, but not in military context — the veteran still has to translate.

Veteran programs alone → don’t scale

SkillBridge and fellowships see the person in real life, but at 12 weeks and one employer at a time.

Skills inference alone → not demonstrated

Inference from records guesses at skills; it doesn’t let a candidate show them under pressure.

The intersection → unoccupied

A 15–60 minute simulator that measures durable skills in military context delivers fellowship-grade signal at screen-scale.

08 · your roles, in your own words

The six scores aren’t our vocabulary. They’re yours.

The objection to any new assessment is fair: “it outputs its own scores — what do I do with them?” Here’s the answer. Each PowerTechs skill maps directly onto requirement language your ATS already screens for, drawn verbatim from live veteran-track postings.

PowerTechs score
The requirement language it maps to
Analytical Reasoningcritical thinking
Analyzing shift and site performance against business objectives; investigating incidents to determine the extent of a compromise; fraud protection and analytics.Amazon Area Manager · Lockheed Cyber Defense Analyst · JPMC Pathways
Clear Communicationstakeholder comms
Excellent communication and interpersonal skills; consistent, effective communication across the team; briefing complex technical material to varied stakeholder levels.Amazon Area Manager · Lockheed SOC postings
Team Coordinationcross-functional
Managing 20+ (up to 50–75) associates; developing subordinate managers; collaborating with Safety, Engineering, HR and QA teams.Amazon AM/OM · JPMC Pathways rotations
Resourceful Problem-Solvingproblem solving
Dealing with operational issues on shift; driving process improvement (Kaizen / Six Sigma); developing solutions under mission constraints.Amazon Operations Manager · Lockheed engineer reqs
Adaptive Learninggrowth mindset
Staying current on evolving tools, techniques, and threat trends; rotating across two distinct business roles in 24 months.Lockheed SOC · JPMC Pathways structure · Cisco pathway
Composure Under Pressureresilience
Real-time decision-making on the FC floor; 24/7 incident detection, containment, and response; strategic planning and risk mitigation.Amazon careers framing · Lockheed SOC · JPMC qualifications

The same map, on live vacancies you can pull today

Three priority verticals, verified postings, job IDs cited so your team can open the full descriptions. Switch verticals below.

Vertical: Logistics & OperationsHighest MOS transferDirect simulation-thesis validation
Military & Veterans Area ManagerJob ID 3122451
Team CoordinationComposure Under Pressure

Leading a team keeping operations safe, efficient, and high-quality; analyzing shift performance against business objectives.

Area Manager — Veterans EncouragedJob ID 3105098
Analytical ReasoningClear Communication

Leading 20+ hourly employees; requires performance-metrics / process-improvement experience and excellent communication and interpersonal skills.

Military & Veterans Operations ManagerJob ID 3122441
Team CoordinationAdaptive Learning

Managing and developing a team of Area Managers; leading operational teams on shift and dealing with issues in real time.

the one non-skill gate to flag

Amazon accepts 6+ years equivalent military experience in place of a bachelor’s degree. It already runs a 15-minute Redeployable simulation for these roles — direct validation of the thesis, and the depth benchmark to beat.

Vertical: Cybersecurity / DefenseHighest salary upliftClearance is the gate
Cyber Defense Analyst, Level 3 (SOC)Job ID 707709BR · Orlando
Composure Under PressureAnalytical Reasoning

SOC support, investigating security incidents, maintaining consistent and effective communication across the team. Levels 1–2 also open (720403BR, 707708BR).

Information Systems Security OfficerJob ID 725768BR · Littleton
Analytical ReasoningClear Communication

Classified cybersecurity, TS/SCI with poly; salary band $76,500–$134,895.

Cyber Security EngineerJob ID 715313BR
Adaptive LearningResourceful Problem-Solving

Coordinating with multiple teams in an Agile environment; TS/SCI; salary band $101,000–$178,135. Entry-level openings also live (729152BR).

the one non-skill gate to flag

Clearance is the gating asset. A veteran with active TS/SCI skips the biggest hiring bottleneck — which makes this the highest salary-uplift pathway in the product.

Vertical: Financial ServicesSkills-first, rotationalOfficer/NCO tenure gate
US Military Pathways Rotational AssociateColumbus · NY · Wilmington · Plano
All six skills

Two one-year rotations across Consumer & Community Banking, Corporate Functions, Asset & Wealth Management, and Commercial & Investment Banking.

Risk / Fraud / Analytics tracks (within Pathways)rotational
Analytical ReasoningComposure Under Pressure

Rotations develop project/business management, relationship management, analytics, and fraud protection. Requirements name “organizational and operational readiness, strategic planning and risk mitigation” — near-verbatim to the Analytical Reasoning + Composure pair.

the one non-skill gate to flag

Tenure threshold: 5+ years as a Commissioned Officer or ~8+ as an NCO, plus demonstrated strategic-planning and risk-mitigation experience.

where PowerTechs fits

Let veterans demonstrate the durable layer — before you commit a slot.

A transitioning service member usually arrives with composure, coordination, and problem-solving built well past what a résumé shows, while the credential is the part they’re still completing. Standard screens see only the half they’re missing. PowerTechs measures the half that decides who succeeds.

Interactive, not self-report

A game-like simulation GenAI can’t complete for the candidate — on the right side of the 2024–2026 shift.

Measures the durable six

Communication, coordination, leadership, adaptability, problem-solving, composure — mapped to recognized frameworks.

Veteran-native by design

Built around military context, so veterans demonstrate rather than translate — and you see it before you host.

Book 20 minutes →

On the call: we read your own postings the way this brief reads them — where the durable layer sits in each role, and how you’d see it in a candidate before you commit. No cost, no obligation.

PowerTechs · Austin, TX · a field brief for employers and hiring partners · calendly.com/powertechs