Hiring Funnel Benchmarks 2026: Conversion Rates by Stage, Time to Fill and Applications per Job
What a normal hiring funnel looks like in 2026, from two large datasets: stage-by-stage pass-through rates, time to fill by seniority, applications per job and offer acceptance. Plus how to calculate your own funnel and where the numbers usually break.
Ece Badak • Oct 17, 2026 • 7 min read
Hiring funnel benchmarks answer a simple question: is our process normal, or is something broken? If 2% of your screened candidates reach an offer while similar teams see far more, you know where to look.
Below are hiring funnel numbers from two large 2026 reports, each linked to its source, followed by a way to calculate the same numbers for your own team. Benchmarks vary a lot by role, seniority and company, so treat them as a reference point rather than a target.
The two datasets behind this guide
| Report | Published | Data | What it measures well |
|---|---|---|---|
| Ashby, Recruiting Operations Benchmarks | May 2026 | 54M+ applications, 93K jobs, January 2021 to March 2026 | Stage pass-through rates, time to first fill, interview timelines, offer acceptance |
| Greenhouse, The Hire Standard | March 2026 | 640M+ applications, 6,000+ companies, 2022 to 2025 | Application volume, recruiter workload, time to fill over time |
Both come from software vendors and describe their own customers, who skew toward technology and knowledge-work companies. Retail, hospitality and other high-volume hiring often looks different.
Stage-by-stage pass-through rates
Ashby reports the share of candidates who move past each stage:
| Stage | All candidates | Referred candidates |
|---|---|---|
| Recruiter screen | about 35% | 52% |
| Onsite interviews | about 24% | 36% |
| After onsite | about 95% | 96% |
| Offer | about 81% | 85% |
Source: Ashby, May 2026.
Most candidates drop out at the screen and the onsite; once someone clears the onsite, they usually get an offer. Referred candidates pass every stage more often than the average.
What the rates imply for 100 screened candidates
Chaining the rates gives a rough picture. The arithmetic below is ours, and real funnels have more stages and more variation:
- 100 candidates reach the recruiter screen
- about 35 pass it and go on to the onsite interviews
- about 8 pass the onsite (24% of 35)
- about 8 remain after the onsite (95%)
- about 6 to 7 pass the offer stage (81%)
By this arithmetic, a team screens about 15 candidates for every hire, and reads far more applications before that.
Time to fill
| Measure | Value | Source |
|---|---|---|
| Time to fill, all jobs, 2022 | 43.6 | Greenhouse |
| Time to fill, all jobs, 2025 | 59.7 | Greenhouse |
| Time to first fill, junior roles (median days) | 52 | Ashby |
| Time to first fill, mid-level roles | 63 | Ashby |
| Time to first fill, senior roles | 71 | Ashby |
| Time to first fill, technical roles | 75 | Ashby |
| Time to first fill, business roles | 60 | Ashby |
| First to last interview, hired technical candidates (median days) | 17.9 | Ashby |
| First to last interview, hired business candidates | 14.4 | Ashby |
Greenhouse does not state the unit for time to fill on its public page; the values match days.
Greenhouse's numbers show a trend: time to fill rose by about 37% between 2022 and 2025. Ashby's show the spread: senior roles take about 37% longer than junior roles, and technical roles about 15 days longer than business roles.
Time to fill and time to hire are different measures. Time to fill usually starts when the job opens. Time to hire usually starts when the hired candidate applies or enters the process. Check the definition before you compare your number with any benchmark.
Application volume and recruiter workload
| Measure (Greenhouse) | 2022 | 2025 | Change |
|---|---|---|---|
| Applications per job | 116 | 244 | +111% |
| Annual applications per recruiter | 1,610 | 4,890 | +204% |
| Recruiters per organization | 10.4 | 4.6 | -56% |
| Monthly hires per recruiter | 2.2 | 4.9 | +122% |
Source: Greenhouse, The Hire Standard, March 2026.
Greenhouse sums up the period as teams that "had to manage twice as many applications with half as many recruiters." Under that load, teams that still read every CV by hand fall behind at the first stage, and good candidates wait or drop out.
Offer acceptance
Ashby reports median offer acceptance rates by source: 89% for referred candidates in business roles and 84% in technical roles, and 76% for sourced candidates in technical roles. For business roles, Ashby says referred, inbound and sourced candidates accept at similar rates.
A falling offer acceptance rate usually points to one of three things: compensation below the market, a process that took too long, or a mismatch between the role as described and the role as offered.
Where the time goes
Three stages often slow a funnel down:
- Application review. With about twice as many applications per job as in 2022, this is where candidates wait first.
- Interview scheduling. Ashby measures a median of 3.7 hours to confirm an interview with automated scheduling and 5 hours manually.
- Offer turnaround. Ashby compares the slowest and fastest quarters of companies by time to hire. The slow group takes on average about 3.3 more days (business roles) and 4.7 more days (technical roles) to create an offer once a candidate reaches the offer stage.
Ashby also measured how often interviewers agree. When two scorecards rate the same candidate, about 38% of those pairs differ by at least one point. Consistent criteria and written evidence reduce those gaps.
How to calculate your own funnel benchmarks
- Pick one role family and a closed period, for example all customer support hires in the last two quarters. Mixing roles hides problems.
- List your stages in order: applied, screened, interviewed, final interview, offer, hired.
- Count candidates entering and leaving each stage. Only include candidates whose outcome is final.
- Divide the number who moved on by the number who entered. That is the pass-through rate for the stage.
- Record the days each stage took, as a median, not an average; a few very slow cases distort averages.
- Compare stage by stage with your own previous period first, and with external benchmarks second.
A simple template:
| Stage | Entered | Moved on | Pass-through | Median days in stage |
|---|---|---|---|---|
| Application review | ||||
| Recruiter screen | ||||
| First interview | ||||
| Final interview | ||||
| Offer |
What to do when a stage is off
| Symptom | Likely cause | First fix |
|---|---|---|
| Very low pass rate at the screen | The job post attracts the wrong applicants, or the criteria are unclear | Rewrite the post around must-haves; agree criteria before screening |
| Long time in application review | Too many CVs for the team to read | Screen CVs against clear criteria, with tooling if needed |
| Many candidates drop out before interviews | Slow scheduling, or interviews only during office hours | Let candidates book or start interviews when they are ready |
| High pass rate at the screen, low at the final interview | The screen does not test what the final interview tests | Use the same criteria in every stage |
| Low offer acceptance | Pay, speed or role mismatch | Check pay bands; shorten the time from final interview to offer |
Firstview is built around the first three rows. CV screening reads every application against the role's criteria with a reason for each one, and the AI Interviewer runs first interviews whenever candidates are ready, against the same criteria the CV was read against. Whatever tools you use, measure the funnel stage by stage before and after you change something.
