The Hiring Crisis: By The Numbers

Hiring in 2025–2026 isn't just frustrating. It's structurally broken — and the data shows both sides of the table are losing.

Candidates spend months applying to roles that may not exist. Recruiters spend weeks screening applicants they'll never hire. Companies bleed money on unfilled positions while qualified people sit in limbo. And the tools built to fix this? Most of them made it worse.

This report compiles findings from SHRM, Harvard Business School, Greenhouse, Revelio Labs, CareerBuilder, the Bureau of Labor Statistics, and other sources to map where the system is failing — and where AI-driven assessment can actually move the needle.

60% of applicants abandon applications before finishing

According to SHRM and CareerBuilder, 60% of job seekers quit mid-application because the process is too lengthy or complicated. The data gets worse the deeper you look:

• 92% of people who click "Apply" never complete the full application
• 73% abandon if the process takes longer than 15 minutes
• 40% abandon applications that aren't mobile-friendly
• Applications with 50+ questions see only a 5.7% completion rate — meaning 94.3% walk away

The top complaint: redundant data entry. Candidates upload a resume, then manually re-enter the same information into form fields. Many do this across dozens of applications per week.

Greenhouse's analysis found that nearly 1 in 5 job postings are ghost jobs — listings with no active hiring behind them. In some industries, the rate climbs to 1 in 3.

• 40% of companies posted fake listings in 2024
• 62% of companies admit to posting ghost jobs to make current employees feel replaceable
• Hires per job posting have halved since 2019: from 8 hires per 10 postings to just 4
• The Wall Street Journal reported ghost jobs now account for 18–22% of active listings, up from 12–15% in 2022

The result: candidates apply in volume because they can't trust that any individual listing is real. This creates a flood of applications that overwhelms recruiters — which makes them rely more on blunt automated filters — which screens out qualified people — which makes candidates apply to even more jobs. A self-reinforcing cycle.

• 61% of job seekers report being ghosted after interviews
• 80% of candidates who experience a negative recruitment process share that experience with others
• 63% will reject a job offer outright due to a bad candidate experience
• Only 26% of North American job seekers say they had a great candidate experience

80+ hours per hire on manual screening

The average corporate job posting receives 250 applicants. In high-volume industries, that number can be far higher. Recruiters spend the majority of their time on initial screening — reading resumes, cross-referencing requirements, scheduling calls — before ever assessing actual capability.

• The average time to fill a role is 42+ days and rising, reaching 68.5 days in some 2025 analyses
• Every unfilled role costs an estimated $500/day in lost productivity
• 31% of HR leaders admit their own hiring technology doesn't work

Meanwhile, Talent Acquisition teams face their own churn. By the time a new TA specialist is trained to screen effectively for a technical role, they've often moved on — resetting the cycle.

98.4% of Fortune 500 companies use an Applicant Tracking System. The popular claim that "75% of resumes are auto-rejected by ATS" is largely debunked — only 8% of employers enable content auto-rejection.

But the real problem isn't automated rejection. It's that ATS systems organize and sort applications based on criteria that humans set poorly:

• 88% of employers believe they're losing qualified candidates because their ATS filters are too rigid
• One hiring manager submitted their own CV under a fake name and was auto-rejected — because HR had configured the search for "angular js developer" instead of "Angular"
• 94% of recruiters say ATS has a positive impact — yet the candidate experience data tells a completely different story

The disconnect is clear: ATS works well as an organizational tool for recruiters, but fails as a screening tool for quality. It optimizes for keyword matching and form processing, not for understanding whether a candidate can actually do the job.

The failures above share a common root: hiring tools were built to process volume, not to evaluate people. ATS systems manage documents. They don't assess capability, potential, or fit.

This is where automated assessment — done correctly — changes the equation for both sides.

For candidates

• A structured, transparent process where the criteria are clear
• Assessment that evaluates actual skills, not just resume formatting
• Faster feedback loops instead of weeks of silence
• Reduced need to apply to hundreds of listings to get one response

For recruiters and hiring managers

• Screening that understands context — recognizing that "Angular" and "AngularJS" describe the same skill, that a career pivot signals adaptability, not instability
• Objective, evidence-based scoring that reduces unconscious bias and gut-feel decisions
• Assessment across multiple dimensions — not just keywords, but soft skills, practical ability, communication, and cultural fit — weighted by what actually matters for the role
• Hours of manual screening compressed to minutes, without sacrificing evaluation depth

For organizations

• Shorter time-to-fill reduces the $500/day cost of vacant roles
• Better initial matching drives lower turnover — companies using structured assessment see 40% less new-hire turnover
• Consistent, auditable evaluation criteria across every candidate
• Integration with existing HR systems so assessment data flows into hiring decisions, not into a silo

Upstack AI was built from the premise that hiring should work for everyone in the process — not just the side with the budget for enterprise software.

Contextual CV Screening, Not Keyword Matching

Rather than filtering resumes by exact keyword match, Upstack's AI reads CVs the way a skilled hiring manager would — extracting evidence of actual capability across five weighted dimensions:

Every score comes with an evidence trail — specific quotes and observations from the CV — so recruiters can see why a candidate scored the way they did. Red flags (vague descriptions, buzzword-heavy content with no specifics) and green flags (quantified impact, self-initiated projects, mentoring) are surfaced automatically.

Multi-Stage Assessment Beyond The Resume

Resumes tell you what someone claims. Assessment tells you what they can do. Upstack provides a configurable pipeline that goes well beyond document screening:

• Qualifying screening — fast eligibility verification (availability, location, certifications) so no one wastes time on basic mismatches
• AI conversational interviews — behavioral and communication assessment through natural dialogue, not rigid form-filling
• Technical assessment — quizzes, coding challenges, system design, and case studies tailored to the role
• Practical VM environments — candidates demonstrate skills in real tools (databases, cloud platforms, design software, CRM systems) in a controlled environment
• ID verification — biometric liveness detection and document matching for trust and compliance

Each stage gates the next. Candidates aren't asked to complete hours of assessment before basic fit is confirmed.

Real-Time Analytics & Candidate Ranking

Recruiters see live candidate progress, scored and ranked across every assessment dimension. Comparison matrices surface who leads in soft skills, technical depth, experience, and potential — giving hiring managers structured data instead of stacked resumes.

HRM Integration

Assessment results sync directly with BambooHR (and expanding) — candidates, scores, and recommendations flow into existing hiring workflows rather than creating another disconnected tool.