Cost per hire and time to fill are the two numbers every hiring function reports and almost nobody defines the same way twice. One team counts agency fees and job board spend. Another adds recruiter salaries, referral bonuses and the ATS licence. A third quietly excludes the roles that took nine months. All three produce a figure, and none of the figures can be compared.
That matters because these metrics drive budget. If your cost per hire is understated, sourcing investment looks expensive. If your time to fill excludes the approval delay before the requisition opened, the recruiting team absorbs blame for a finance bottleneck. This guide sets out the standard formulas, the 2026 benchmarks, the vacancy cost most models omit, and the levers that genuinely move each number.
Cost per hire: the formula and what it hides
The standard calculation is straightforward. Add all internal recruiting costs to all external recruiting costs, then divide by the number of hires in the period.
The difficulty is never the arithmetic. It is deciding what counts, because the boundary determines whether your figure lands at £2,000 or £9,000 for identical activity.
| Internal costs | External costs |
|---|---|
| Recruiter and sourcer salaries plus on-costs | Agency and search fees |
| Hiring manager and interviewer time | Job board and advertising spend |
| Referral bonuses paid | Sourcing and contact data licences |
| ATS and recruiting tooling licences | Background checks and assessments |
| Careers site and employer brand costs | Relocation and signing bonuses |
| Recruiting team travel and events | Job fairs and campus programmes |
Two inclusions separate a serious model from a decorative one. Interviewer time is a real cost — twenty interviews per hire at an hour each, across salaried staff, frequently exceeds the job board spend for the same role. And referral bonuses belong in the calculation even though referrals reduce cost overall, because excluding them makes your cheapest channel look free rather than merely cheap.
2026 benchmarks, and why the numbers disagree
Published benchmarks vary wildly, and the variation is not random. Medians and averages behave very differently in hiring data, because a small number of executive and hard-to-fill searches drags the mean upward.
| Measure | Non-executive roles | Executive roles |
|---|---|---|
| Median cost per hire | ~$1,300 | ~$15,000 |
| Average cost per hire | ~$5,475 | ~$35,879 |
| US median time to fill | ~44 days | 70–120 days |
| Technical roles time to fill | 48–76 days | — |
| Interviews per hire | 11–18 | 15–25 |
| Hires per recruiter per quarter | 5.0 business, 3.8 technical | 1–2 |
The median-to-average gap for non-executive roles is roughly four-to-one, which is why two teams can both cite credible research and still argue. Report your own median and average side by side, and state which one you are comparing against. SHRM publishes the underlying methodology in its recruiting benchmarking research, and it is worth reading before adopting anyone’s headline number.
Context matters too. US time to fill has risen roughly a third since 2021, while applications per hire have nearly tripled. More applicants and slower hiring at the same time tells you the bottleneck moved from attraction into screening and decision-making.
Time to fill, time to hire and the difference that matters
These are routinely used interchangeably, and they measure different failures.
- Time to fill. Requisition approved to offer accepted. This is the business-facing number, and it captures your process end to end.
- Time to hire. Candidate’s first engagement to offer accepted. This measures candidate experience and decision speed, independent of how long sourcing took.
- Time to productivity. Start date to full performance. Rarely tracked, and the number that actually tells you whether the hire was good.
Track the first two together, because the gap between them is diagnostic. A long time to fill with a short time to hire means sourcing is slow. A short time to fill with a long time to hire means you found people quickly and then lost weeks in scheduling and deliberation. Those problems have nothing in common and no shared fix.
Also record where the days go. Splitting the timeline into sourcing, screening, interviewing, decision and offer negotiation converts a vague complaint about slowness into a specific stage you can attack.
The vacancy cost that belongs in every model
The metric measures what you spent to fill a role. It says nothing about what the empty seat cost while you were spending it, and for most commercial roles the second number is larger.
A workable estimate divides annual revenue per employee by 260 working days, then multiplies by days vacant. At £250,000 revenue per employee, a role open for 60 days costs roughly £57,000 in foregone contribution. Against that, a £12,000 agency fee that closes the role three weeks earlier is straightforwardly profitable.
Non-revenue roles need a different proxy — overtime paid to cover the gap, delayed project delivery, or attrition risk on the overloaded team. The precise method matters less than including the number at all, because a model that only counts recruiting spend will always conclude that recruiting should spend less.
Where the two metrics pull against each other
Optimising one of these numbers in isolation reliably damages the other, and it is worth naming the trade-offs explicitly before someone sets a target.
- Agencies cut time to fill and raise cost per hire. The right call when vacancy cost is high, the wrong call when it is not.
- In-house sourcing cuts cost per hire and can raise time to fill, at least until the pipeline matures.
- More interview stages improve quality and lengthen everything. Each additional stage adds roughly three to five days, mostly to scheduling.
- Cheap contact data raises both figures. Poor coverage means more profiles for the same pipeline and more recruiter hours per hire.
Set a target on one metric and a guardrail on the other. “Reduce time to fill to 40 days without cost per hire exceeding £4,500” is a usable objective. “Reduce both” is a wish.
Levers that genuinely move cost per hire
- Shift channel mix toward referrals and internal mobility. These yield hires at many times the rate of job boards, and they are the largest single lever available.
- Rebuild from your own database first. Silver medallists and past applicants convert faster and cost almost nothing to reach.
- Reduce agency dependency on repeatable roles. Any role you hire more than twice a year should have an internal pipeline.
- Cut interview stages that do not change decisions. Audit your last twenty hires and find the stage that never once produced a rejection.
- Improve contact data coverage. Raising contactability from 65% to 85% removes roughly a quarter of the profiles you need to work per hire.
The last one is the least visible and among the most effective. We cover coverage testing and verification in detail in our guide to candidate contact data.
Levers that genuinely move time to fill
- Start sourcing before approval where you can. Much of the delay sits before the requisition officially opens, and a warm pipeline collapses that stage.
- Pre-book interview slots. Scheduling, not deciding, is where most of the calendar disappears. Reserve recurring panel slots in advance.
- Set a reply service level. Interested candidates answered within four working hours convert substantially better than those answered in two days.
- Run a proper intake with calibration profiles. Three yes-profiles and three no-profiles prevent the two-week detour that a vague brief guarantees.
- Use AI for list building and drafting. Teams applying AI assistance to sourcing report roughly a quarter off their timelines, largely by compressing search and first-draft outreach.
Most of these are process changes rather than purchases, which is why time to fill often improves faster than cost per hire once a team starts measuring stage by stage.
Building a dashboard that survives scrutiny
Report these metrics in a way that anticipates the obvious challenges, because they will be raised the first time a number moves in the wrong direction.
- Segment by role family. A blended figure across sales, engineering and warehouse roles is arithmetic without meaning.
- Publish median and average together, so one difficult search cannot be mistaken for a trend.
- Show stage-level timing, which locates delay outside recruiting when that is where it sits.
- Include source of hire, which requires clean attribution — see our guide to ATS integration for how that breaks.
- Add quality of hire at twelve months. Without it, both headline metrics reward speed and cheapness regardless of outcome.
That final line is the one that protects the function. A cost per hire target with no quality measure eventually produces exactly the hiring you would expect.
Frequently asked questions about cost per hire and time to fill
What is a good cost per hire?
It depends entirely on role and market, which is why external benchmarks are weaker than your own trend. Medians near $1,300 and averages near $5,475 for non-executive roles are reasonable anchors, but a 20% improvement against your own baseline is a more useful goal than matching anyone else’s figure.
Should recruiter salaries be included?
Yes. Excluding them makes in-house recruiting appear nearly free and distorts every agency comparison you will ever run. Include fully loaded costs, including on-costs and tooling.
When does the time to fill clock start?
At requisition approval, by convention. If approval routinely takes three weeks, track that separately as time to approve so the delay stays visible rather than being absorbed into the recruiting number.
How do agencies affect these metrics?
Agencies typically shorten time to fill and increase cost per hire, and whether that trade is worthwhile depends on vacancy cost. Our agency vs in-house recruiting comparison works through the volume thresholds where each model wins.
Can sourcing tools reduce cost per hire?
They can, through two mechanisms: reducing agency spend on repeatable roles, and improving contactability so fewer profiles are needed per hire. Both effects only appear in reporting if source of hire is captured accurately, which brings you back to attribution.
Define it once, then measure the trend
Cost per hire and time to fill are only useful when the definitions are fixed, the segments are honest, and the vacancy cost sits alongside them. Chasing an external benchmark computed on someone else’s role mix produces confident decisions built on nothing.
Write down your formula, segment by role family, publish median and average together, and then improve against your own baseline quarter after quarter.
Read next
- Candidate sourcing — the funnel arithmetic behind both metrics.
- Agency vs in-house recruiting — the cost model compared side by side.
- ATS integration — attribution that makes this reporting possible.
- Outbound recruiting sequence — where sourcing time is actually spent.
ZenBee reduces the profiles needed per hire with verified emails and direct mobiles across 700M+ contacts, plus hiring and job-change signals that tell you where to spend the week. Explore the ZenBee recruitment platform or compare plans on pricing.