2026 State of Hiring — Benchmarket

01 · Where the data agrees

Multiple studies. Same findings.

Applications per hire

3×

The volume isn't a temporary spike. It's the new baseline.

Ashby Gem CareerPlug

Offer acceptance

82%

Closing isn't where hires are lost. The funnel breaks earlier.

Gem Employ SmartRecruiters

Candidate deception

91%

Recruiters who spotted deception in the past year. Fraud is the new pattern, not the edge case.

Greenhouse Gartner Willo §03 →

Time-to-hire trend

60%

got slower

Only 11% got faster. The default direction is now slower.

GoodTime 2026 SHRM Mitratech

Two-sided ghosting

53% vs 41%

Candidates · Recruiters

Both sides report it. Silence runs both ways.

SHRM 2025 iHire

Quality of hire

89% vs 20%

Say it matters · Measure it

The most-cited priority is the least-measured outcome.

LinkedIn SHRM Ashby Starred

02 · Trust

Trust got worse on both sides.

2021

Pre-AI baseline.

Candidates

Resumes, cover letters, networking

Recruiters

ATS keyword search, human review

Both sides on shared ground. The signal each side sent was the signal each side read.

2022

ChatGPT lands.

Candidates

Resume drafting with ChatGPT, cover-letter generation

Recruiters

JD writing, sourcing message drafts

Public AI generation arrives. Both sides start using it. Neither tells the other.

2023

AI sourcing & resume tools mainstream.

Candidates

Teal, Kickresume, Rezi — resume polish at scale

Recruiters

HireEZ, Fetcher, Gem — autosourcing at scale

Both sides now route communication through AI. Pattern-matching begins.

25% of US candidates use AI

2024

AI screening meets interview prep.

Candidates

Final Round AI, Interview Coder — live interview assist

Recruiters

Eightfold, Paradox, AI resume screeners

Each side is now screening with AI and being screened against AI.

2025

Agents arrive on both sides.

Candidates

LazyApply, Sonara — autonomous applying

Recruiters

Metaview, BrightHire — AI interview scoring

Each side's automation is now adversarial to the other's.

74% now use AI

2026

Trust collapsed.

Candidates

Deepfake interview tools, prompt injection

Recruiters

Identity verification, liveness checks, deepfake detection

Each side's AI is the other's adversary. Verification has become a vendor category.

Only 8% believe AI hiring is fair

2021

Pre-AI baseline.

CANDIDATES

Resumes, cover letters, networking

RECRUITERS

ATS keyword search, human review

Both sides on shared ground. The signal each side sent was the signal each side read.

2022

ChatGPT lands.

CANDIDATES

Resume drafting with ChatGPT, cover-letter generation

RECRUITERS

JD writing, sourcing message drafts

Public AI generation arrives. Both sides start using it. Neither tells the other.

2023

AI sourcing & resume tools mainstream.

CANDIDATES

Teal, Kickresume, Rezi — resume polish at scale

RECRUITERS

HireEZ, Fetcher, Gem — autosourcing at scale

Both sides now route most communication through AI. Pattern-matching begins.

25% of US candidates use AI to apply

2024

AI screening meets AI interview prep.

CANDIDATES

Final Round AI, Interview Coder — live interview assist

RECRUITERS

Eightfold, Paradox, AI resume screeners

Each side is now simultaneously screening with AI and being screened against AI.

2025

Agents arrive on both sides.

CANDIDATES

LazyApply, Sonara — autonomous applying

RECRUITERS

Metaview, BrightHire — AI interview scoring

Each side's automation is now adversarial to the other's.

74% of US candidates now use AI

2026

Trust collapsed.

CANDIDATES

Deepfake interview tools, prompt injection

RECRUITERS

Identity verification, liveness checks, deepfake detection

Each side's AI is the other's adversary. Verification has become a vendor category.

Only 8% believe AI hiring is fair

The candidate's view

"I can't trust this process."

46%

of US job seekers say trust in hiring has dropped in the past year. 42% blame AI directly.

of job seekers believe AI in hiring is fair.

87%

say employer AI-use transparency matters, but only 27% have ever seen an employer policy on it.

41%

admit using prompt injections to bypass AI filters, because they don't trust the process to see them otherwise.

The employer's view

"I don't actually know if it's working."

70%

of hiring managers say AI helps them make faster, better decisions, but most can't prove it.

56%

of HR professionals don't formally measure AI investment success at all.

83%

are still experimenting with AI, not operationalizing it.

25%

of TA leaders feel confident they can measure quality of hire. 89% say it matters most.

Both sides are just trying to do their best, but the collective result is worse for everyone.

— Daniel Chait, CEO, Greenhouse · 2025 AI in Hiring Report · n=4,136

03 · Fraud

Candidate fraud.

Each side's AI is the other's adversary. Evasion improves faster than detection.

What candidates confess

"I tweaked things a bit."

admit to interview fraud: posing as someone else, or having someone pose for them.

28%

used AI to generate fake work samples.

32%

claimed AI skills they don't have.

36%

altered appearance or voice in video interviews.

What recruiters detect

"It's everywhere now."

91%

spotted candidate deception in the past year: AI-generated resumes, fake skills, deepfakes.

37%

of employers still see resumes as reliable indicators. Perfect matches now signal AI, not fit.

~1%

of 300M resumes Greenhouse processes annually contain prompt injections.

1 in 4

candidate profiles could be fake by 2028.

Gartner forecast (projection, not observation)

04 · The Hidden Cost Nobody Tracks

The Interviewer Tax.

When trust erodes and fraud amplifies, teams add "safety rounds" to compensate.

2021: 14 rounds (avg)

→

2026: 20 rounds (avg · 35–36 in tech)

+6 "safety rounds"

Gem 2026 · 35–36 interviews per technical hire ^[1]

Rounds

+42%

Interview rounds per hire
since 2021 ^[1]

Time

38%

Recruiter time spent
scheduling these rounds ^[2]

Cost

~$43K

Yearly per recruiter that schedules their own interviews ^[3]

Estimated cost per technical hire

Interview time is only half the cost.

Per hire · per activity · loaded

Cost

×1

Interview time itself

35 interviews × 45 min × $150/hr

$3,938

×2

Scheduling overhead

2.8-day lag × 15 min coordination per round

$1,313

×3

Feedback & scorecards

35 interviews × 10 min written feedback

$875

×4

Debrief meetings

3 debriefs × 30 min × ~5 attendees

$1,125

Estimated interviewer tax · per technical hire

$7,251

The meter keeps running

Every extra round adds days to time-to-fill, now 38 days globally, 60–61 days at large orgs. At roughly $500/day in lost productivity per open role, each round carries a second price tag no one is reading. ^[4]

Each extra round costs

$2,000

$4,000

$6,000

$8,000

$10,000

~4 days × $500/day per added round

^[1] Gem reports interviews per hire up 33% overall since 2021, with technical roles averaging 35–36 interviews; the 42% figure (14 → 20 rounds) is from Gem's 2025 benchmarks. Ashby separately reports a +52% increase for technical roles over the same period.

^[2] GoodTime 2026 Hiring Insights Report (surveyed 500+ TA leaders at 1,000+ employee companies, Nov 2025): “Recruiters spend 38% of their time scheduling interviews.”

^[3] 38% × ~$108K loaded recruiter cost. Loaded cost derived from $80–110K median total comp (Glassdoor, Built In, LinkedIn 2025–26) at a 1.3x benefits-and-overhead multiplier. Applies to teams where recruiters handle scheduling directly; orgs with dedicated coordinators or scheduling automation (GoodTime, Modern Loop, Cronofy, Paradox) shift this cost rather than eliminate it.

^[4] Time-to-fill: 38 days global average (GoodTime 2026, surveyed 500+ TA leaders); 60–61 days at 10,000+ employee orgs (Mitratech 2025). $500/day cost of vacancy: SHRM industry estimate, widely cited; reflects lost productivity, overtime coverage, and team strain. Conservative figure for non-technical roles; can run 2–4× higher for specialist or engineering roles.

Verified inputs: 35–36 interviews per technical hire (Gem 2026), 2.8-day scheduling lag (candidate.fyi 2025). Modeled assumptions: 45-min interview, 15-min scheduling, 10-min feedback, $150/hr loaded cost, $500/day cost of vacancy. Substitute your own rates.

05 · Benchmarks

The Benchmarks. The Variance.

Without precision benchmarks, comparing can mislead.

Metric	Highest Reported Avg.	Lowest Reported Avg.
Applications / Hire	369Software & Tech	57Education
Time to First Interview	12 days50–200 emp · Technology	4.9 daysAll industries · SMB
Offer Acceptance Rate	91%Manufacturing	80.8%Software & Tech
Interview-to-Hire Ratio	39Smallest firms, 1–10 emp	10Best-performing cohort
Apps / Recruiter Increase	+410%Greenhouse cohort · 2022–2025	+93%Gem 2026 · global avg
Active Recruiter Decline	~30% below peakIndeed Software Dev Postings · Feb 2026	−14%Gem 2026 · all industries
Time to Fill	60–61 days5,000+ employee orgs	38 daysGlobal median

01 · Range spread

Ranges are 2–6× wide. "Industry average" is a fiction without segmentation.

02 · The tech outlier

Tech is the outlier on every dimension. Don't benchmark against other industries.

06 · AI Legal & Security Exposure

AI moved fast. Legal is catching up.

Vendors and teams built faster than oversight could follow.

Regulatory landscape

Now and next.

2023–2025

Law

NYC · CA · CO

Bias audits, disclosure, human review.
Mar 2025

Case

ACLU v. HireVue

Disability + race · video interview AI.
May 2025

Case · Class certified

Mobley v. Workday

Vendor liable as "agent."
Aug 2025

Case

Harper v. Sirius XM

Class action · third-party AI screening.
Jan 2026

Law

Ontario WFW Act

AI disclosure, 45-day notification.
Aug 2, 2026

Law · Scheduled

EU AI Act

High-risk · transparency · audits. Some obligations subject to proposed delays.

⚠

The internal risk

When AI writes your tools, nobody reviews the code.

AI code generation pushed internal tooling into the hands of non-engineers: HR, recruiting ops, people analytics. The tools work; the review process built for hand-written code doesn't catch what gets shipped. For systems touching candidate data or ATS integrations, that's a category of exposure that didn't exist eighteen months ago.

What to do now

Inventory the AI tools touching your candidate data.

Inventory every AI tool touching candidate data. Document what each one does, what it decides, and how it gets reviewed. If you can't list them, you can't defend them.

07 · The Agentic Pivot

The AI evolution. The adoption lag.

AI can technically handle most steps in recruiting. Few teams trust it to own one.

Step 1

Sourcing Agent

Identifies candidates from ATS, external databases, and network signals autonomously

Step 2

Personalization Agent

Writes personalized outreach using the candidate's public portfolio and signals

Step 3

Scheduling Agent

Manages calendar logistics across time zones without human coordination

Step 4

Interview Intelligence

Captures, summarizes, and scores interviews in real time

Step 5

Human Orchestrator

Tunes agents, handles exceptions, owns judgment calls AI cannot make

Source: Ai4 2026 · Transform 2026 · practitioner communities

08 · Acquisitions

Recruiting tech is consolidating.

Pattern 01 · Strategics

Legacy buying innovation.

SAP → SmartRecruiters

ATS

Workday → Paradox

Conversational AI

Workday → Sana

AI Learning

iCIMS → Apli

Frontline AI

Radancy → myInterview

Video Interviews

HireVue → Hireguide

Structured Interviews

Pattern 02 · Adjacency

Expanding scope.

Paychex → Paycor

HCM / Payroll

Payoneer → Boundless

Global EOR

Docebo → 365Talents

Internal Mobility

PayScale → Datapeople

JD Optimization

Phenom → Be Applied

Equitable Hiring

Perceptyx → Lyceum AI

EX Analytics

Pattern 03 · Private equity

PE buying control.

Thoma Bravo → Dayforce

HCM Platform

EQT/CPP → NEOGOV

Public-Sector HR

EQT → Remember (KR)

Korea HR Tech

TA Associates → OneSource Virtual

Payroll Services

General Atlantic → SmartHR

Japan HR Tech

Tritium → Loxo

Recruiting CRM

09 · Tools changing the funnel

New players are disrupting.

While incumbents consolidate, focused AI-native tools are reshaping every stage of the funnel: from sourcing to scheduling to verification.

Companies disrupting the space · Q2 2026

Publisher

Benchmarket

The intelligence layer for recruiting performance. Sits on top of any ATS. Continuous analysis, AI-driven recommendations, automated stakeholder updates.

Juicebox

PeopleGPT: natural-language sourcing across 800M+ profiles. $80M Series B at $850M valuation. Used by Ramp, Perplexity, and leading AI labs.

End-to-end recruiting OS by the founder of Interseller. 600+ customers in 40 days. In-house foundation models, 850M profiles, 70% pipeline acceptance.

Metaview

Captures, summarizes, and analyzes interviews. Explicitly does not decide who to hire. $50M raised, GV-led. 2,500+ customers including Sony, Brex, Deel, ElevenLabs.

BrightHire (Zoom)

Pioneered the interview-intelligence category. Used by Canva, Duolingo, Instacart, Ramp. Acquired by Zoom in December 2025 to power AI hiring in Workplace.

GoodTime

Automates up to 90% of interview management, even for the most complex panels. Used by Databricks, HubSpot, Aon, HelloFresh, Canva. Orchestra AI agent handles auto-replace.

Company descriptions based on public materials and announcements as of Q2 2026.

10 · Meet Benchmarket

Meet Benchmarket.

We understand hiring data. Benchmarket runs continuously on your ATS, connecting every signal, diagnosing what's driving performance, and giving your team a clear read on what to do next. Continuously. Automatically. On demand.

What Benchmarket does

Saves your team time

Continuous performance analysis

What used to be a quarterly analyst project (pulling data, reconciling sources, rebuilding the deck) now runs continuously in the background. Your team gets back the days they used to spend assembling the report.

Improves recruiting performance

Dual benchmarking

Compare yourself to your own history and to the industry at the same time. You see exactly where you stand, exactly where you're improving, and exactly where the gap is widening.

Catches problems early

Risk detection

Stalled roles, slipping recruiters, breaking patterns. Surfaced before they show up in a quarterly review. You see the problem when you can still solve it.

Makes your team strategic

AI recommendations

Specific, prioritized actions. Not generic dashboards. Benchmarket tells you which job is at risk, which recruiter is stalling, and what to fix first. The strategic conversation is already started by the time the meeting begins.

Speaks to the whole business

Stakeholder updates

Hiring managers, leadership, and finance get the read they need, in the format they understand. No slides to build. No Friday-afternoon spreadsheet pulls. The update is already sent.

Any question, any time

On-demand analysis

Custom investigation on demand. For board reports, strategic reviews, or any question your team needs answered. Analysis that used to require a consulting firm or three weeks of pulling data.

11 · Why companies adopt Benchmarket

Always-on diagnosis.
Running on your hiring.

Benchmarket sits on top of your ATS and never stops working, so your team always knows what's happening, what it means, and what to do next.

— Built by a recruiting operator who lived this problem.

Hours reclaimed.

Every reporting cycle is three jobs in one: collecting, interpreting, and figuring out how to present data. Benchmarket handles all three automatically, without leaving a single data point behind.

The why behind every metric.

Your ATS shows you what happened. Benchmarket shows you what's driving it: diagnosing performance across every role, recruiter, and funnel stage so your team understands the cause, not just the number.

Prioritize with precision.

What needs attention in hiring changes constantly. Benchmarket's continuous analysis surfaces the scenarios where focused effort will drive the most improvement.

Proactive, not reactive.

A stalled pipeline at day 45 shouldn't be a surprise. Benchmarket surfaces risks as they emerge, giving your team time to course-correct before small issues become costly ones.

Full-picture intelligence.

Recommendations you can trust because they're built on all of your data, not a slice of it. Benchmarket analyzes every data point, every angle, continuously.

Strategic confidence at every level.

Recruiters know where to focus. Managers know where to lean in. Executives get strategic clarity. Benchmarket gives every level the intelligence to show up with confidence.

About the founder

Kevin Serwatka

Founder & CEO, Benchmarket

LinkedIn →

Why I built this.

After recruiting and leading talent teams at Google, Facebook, Robinhood, and later advising startups as a VC Talent Partner, I kept seeing the same problem: recruiting teams had plenty of data, but no consistent way to turn it into better decisions across the full hiring process.

I didn't know SQL, how to build a data pipeline, or how to integrate data sources. The difference was I simply cared about the data and found ways to listen to it.

I also geeked out on ATS's (cool, right?!), how they were set up, how to search them, and even helped shape the development of two enterprise ATS's.

That's where the idea of Benchmarket started.

Every part of Benchmarket has been tested in real environments that I was personally recruiting in.

Now any team can connect their ATS (we connect to 55+!) and get objective, automated, and on-demand intelligence for everyone involved in hiring. We have been building with one goal in mind: turn recruiting data into improved recruiting performance for all of our users.

If this sounds interesting, I'd love to personally show you how Benchmarket can help elevate your recruiting team.

What you can do now

Five things you can start this quarter.

Each action is supported by a finding in this report. None of them require new tools or new headcount — just a decision to start.

Audit your interview rounds against your offer-accept rate.

If you're at 20+ rounds for a tech hire and your offer accept rate isn't 90%+, you're paying the safety-rounds tax for no return. Cut from the bottom of the loop.

§04 →

Publish your AI-use policy to candidates.

87% of candidates say transparency matters. Only 27% have ever seen a policy. Low-cost trust win — and increasingly a legal one.

§02 / §06 →

Build your baseline before benchmarking the industry.

Time-to-fill, drop-off-by-stage, and offer-accept by recruiter — measured against your own last 4 quarters, not someone's global median.

§05 →

Inventory every AI tool touching candidate data.

Document what each tool does, what it decides, and how it gets reviewed. EU AI Act high-risk obligations land August 2026. If you can't list them, you can't defend them.

§06 →

Define quality-of-hire before you measure it.

89% say it matters; 20% measure it. Pick two leading indicators (90-day review + 12-month retention) and start tracking. Anything beats nothing.

§01 →

Tool · Benchmark calculator

How does your hiring compare?

Benchmarks are one signal. Performance has many.

Company size

Function you're hiring for

Your average time-to-hire (days)

Days from first application to offer accepted.

Interviews per hire (avg)

Total interviews across every candidate, to make one hire.

The read · diagnosis

—

Metric 01 · Time-to-hire

—

Days from first application to offer accepted.

Metric 02 · Interviews per hire

—

Total interviews across every candidate, to make one hire.

A note on these numbers

Across the major recruiting benchmark sources, time-to-hire averages disagree by up to 26%, and interview load varies even more by industry segment. The number you compare against shapes the goal you set. This calculator picks the comparison that best matches your segment, but the spread itself is the story §05 of this report told.

The full diagnosis

Two metrics. You have ten more.

Time-to-hire and interview load are symptoms. The full read identifies which lever is dragging your numbers (sourcing, interviewing, or closing), broken down by recruiter, by job, and by stage. ↑ Back to Benchmarket

● Aggregated from publicly available recruiting industry research

All cited market numbers are sourced. Modeled estimates and illustrative examples are labeled.

Research Sources

Drawn from publicly available recruiting benchmark reports, surveys, legal filings, and market announcements. Modeled estimates are labeled separately.

Gem 2025 & 2026 Recruiting Benchmarks (140M applications) CareerPlug 2025 (10M applications) Employ 2025 (6,640 companies) SmartRecruiters 2025 (90M applications, 95 countries) SHRM 2025 (n=2,040 HR professionals) NACE 2025 Ashby Talent Trends LinkedIn Future of Recruiting 2025 LinkedIn Talent Insights 2025 Starred 2026 GoodTime 2025 Indeed UK 2024 (n=3,000 employers + workers) TestGorilla 2025 Harvard / Burning Glass Institute Stanford / USC arXiv 2025 SSRN Voice AI 2025 ERE Talent Board CandE 2024 MIT Sloan Feb 2026 University of Washington Nov 2025 Ghosting Index 2025 (50+ studies) RecruitBPM 2026 Greenhouse 2025 AI in Hiring Report Greenhouse AI Principles Framework 2026 Gartner Future of Hiring 2026 Korn Ferry Talent Acquisition 2025 Bullhorn GRID Industry Trends 2025 Workday Recruiting Insights 2025 iCIMS Class of 2026 Report CIPD Resourcing & Talent Planning 2025 Aptitude Research Hiring Trends 2025 McKinsey Talent Trends 2025 Deloitte Global Human Capital Trends 2025 EU AI Act Final Text (Reg 2024/1689) NYC Local Law 144 Bias Audits California AB 2930 / FEHA Updates Mobley v. Workday court filings Pitchbook M&A Database Benchmarket Benchmark Database

Multiple studies. Same findings.

Trust got worse on both sides.

Candidate fraud.

The Interviewer Tax.

The Benchmarks. The Variance.

AI moved fast. Legal is catching up.

The internal risk

What to do now

The AI evolution. The adoption lag.

Recruiting tech is consolidating.

New players are disrupting.

Meet Benchmarket.

Always-on diagnosis.Running on your hiring.

Why I built this.

Five things you can start this quarter.

Audit your interview rounds against your offer-accept rate.

Publish your AI-use policy to candidates.

Build your baseline before benchmarking the industry.

Inventory every AI tool touching candidate data.

Define quality-of-hire before you measure it.

How does your hiring compare?

—

—

—

Two metrics. You have ten more.

Always-on diagnosis.
Running on your hiring.