Short answer: Recruiter productivity at a staffing agency is best measured as recruiter hours per placement, split by step. Placements per recruiter show output; hours per step show where the time goes, and how much is judgement rather than chasing, asking again and checking twice. Measure two normal weeks, then change the process before you judge any recruiter.
Key takeaways
- Placements or margin per recruiter tell you what came out. They do not tell you where the hours went or what to change.
- Measure recruiter hours per placement by step, with a short time diary or random checks over two normal weeks.
- Sort every hour into three kinds: judgement, necessary work done once, and repeated work such as chasing documents or asking the same question again.
- Repeated work is the part a process change can remove. Judgement is the part worth protecting.
- Report per branch, per client and per step rather than per named recruiter. The aim is to fix the process, not to watch individuals.
On this page: What is recruiter productivity at a staffing agency? · Where do the hours go in a frontline placement? · How do you measure recruiter hours per placement? · Which recruiter productivity metrics should a staffing agency track? · Why do activity counts mislead? · Should you measure individual recruiters? · Why does repeated work pile up in frontline staffing? · What can you change once you know where the hours go? · What can a time study not tell you? · How Radius helps · Frequently asked questions
What is recruiter productivity at a staffing agency?
It is useful output per recruiter hour, counted in placements that start.
Most definitions of recruiter productivity divide output by capacity: placements, fee income or gross margin per recruiter in a period. That answers "how much came out?" It cannot answer "where did the hours go, and which could we get back?" Two branches can make the same number of placements while one spends its time talking with candidates and clients and the other spends it chasing certificates.
Two definitions keep the measure honest for frontline staffing.
Count a placement when the candidate starts the first shift. A candidate who does not turn up on day one used the same recruiter hours and produced nothing.
Separate recruiter effort from candidate waiting time. Time to fill counts calendar days from order to start. Hours per placement counts the working hours your recruiters put in.
The pressure is real in the Netherlands. At the end of the second quarter of 2026 there were 375,000 open vacancies, or 95 for every 100 unemployed people, according to preliminary figures from CBS (30 July 2026) (NL). In a market that tight, an hour spent chasing a document is an hour not spent on the next candidate, and recruiter hours do not grow when applications do, as we argue in Frontline Hiring Does Not Have a Knowledge Problem. It Has a Capacity Problem.
Where do the hours go in a frontline placement?
They go into nine recurring steps between the client's order and the first shift.
| Step | What the recruiter does | Where repeated work creeps in |
|---|---|---|
| 1. Client order | Agrees the role, shifts, requirements and certificates with the client | Requirements left vague, so questions return later |
| 2. Posting and sourcing | Posts the role, searches past applicants, replies to new ones | The same person found twice, in two places |
| 3. First contact | First call or message, availability, basic questions | Missed calls and messages, callbacks, voicemail |
| 4. Interview | Asks about shifts, travel, experience and the language set for the role | Answers kept in personal notes, so the next person asks again |
| 5. Documents | Asks for identity, right to work and certificate documents | Documents requested after the client says yes, one at a time |
| 6. Checks | Confirms identity, right to work, certificates and expiry dates | The same document checked again by someone else or for another client |
| 7. Submission | Sends the candidate and the evidence to the client | The client asks what the interview already answered |
| 8. First shift | Confirms time, place, what to bring and who to report to | Shift details sent late or never confirmed |
| 9. Replacement | Starts again when a candidate drops out or does not turn up | The whole chain runs a second time |
Within every step, an hour falls into one of three kinds.
- Judgement. A person weighs evidence and decides: whether a candidate meets the requirements of the role, how to explain a night shift pattern honestly, what to advise a client whose requirements do not match the market.
- Necessary work done once. Posting a role, booking an interview, preparing a contract, a check carried out and recorded.
- Repeated work. Chasing a document or an answer, asking a question someone already asked, typing the same details into a second system, checking a document a second time, restarting after someone did not turn up.
The aim is to protect the first kind and shrink the third. Repeated work adds hours without adding information.
How do you measure recruiter hours per placement?
Log two normal weeks by step, then divide by placements started.
- Fix the codes before anyone logs. Use the nine steps above, or your own list, plus the three kinds of time. Write one example of each kind per step so that everyone codes alike.
- Choose a method. In a time diary, each recruiter notes in 15 minute blocks which step and which kind of time they worked on. Random checks, also called work sampling, ask less: at random moments, each recruiter records only what they are doing at that moment.
- Pick two normal weeks. Avoid peak season or a client launch, unless that is what you want to study.
- Match hours to starts. Hours logged this week produce starts next week. Follow every order opened during the two weeks until it is filled or closed.
- Calculate per branch and per client type. Hours per placement = recruiting hours ÷ placements started from that work, overall and for each step. Share of repeated work = repeated work hours ÷ all recruiting hours.
If you use random checks, you need enough of them. To estimate a share of about 30 per cent to within 5 percentage points either way, with 95 per cent confidence, you need about 323 observations: 1.96 × 1.96 × 0.3 × 0.7 ÷ (0.05 × 0.05) ≈ 323. This follows from the normal approximation confidence interval for a proportion in the NIST/SEMATECH statistics handbook. For example, four recruiters over ten working days would need about eight random checks each per day.
Log as you go rather than from memory.
Running intake for several clients and branches at once? See how Radius Hire is set up for staffing agencies.
Illustrative example: one branch, two weeks
Illustrative example. The figures are invented to show the calculation and do not describe any real agency.
A branch has four recruiters working for three warehouse clients, mostly order picker and forklift roles. Each recruiter logs about 32 hours of recruiting work a week, with the rest going to client meetings and other duties. Over two weeks that is 4 × 32 × 2 = 256 recruiting hours. From the orders opened in those weeks, 24 candidates started their first shift.
Hours per placement: 256 ÷ 24 ≈ 10.7.
Hours per placement at one branch, over two weeks
| Step | Hours | Share of hours | Hours per placement |
|---|---|---|---|
| Client order | 16 | 6% | 0.7 |
| Posting and sourcing | 30 | 12% | 1.3 |
| First contact | 52 | 20% | 2.2 |
| Interview | 34 | 13% | 1.4 |
| Documents | 46 | 18% | 1.9 |
| Checks | 22 | 9% | 0.9 |
| Submission | 20 | 8% | 0.8 |
| First shift | 18 | 7% | 0.8 |
| Replacement | 18 | 7% | 0.8 |
| Total | 256 | 100% | 10.7 |
Figures per placement are rounded, so the column adds up to slightly more than the total.
| Kind of time | Hours | Share of hours | Hours per placement |
|---|---|---|---|
| Judgement | 62 | 24% | 2.6 |
| Necessary work done once | 96 | 38% | 4.0 |
| Repeated work | 98 | 38% | 4.1 |
| Total | 256 | 100% | 10.7 |
The branch spends more hours per placement on repeated work than on judgement. Most of the repeated work sits in three steps: chasing documents, calling back candidates at first contact, and replacing people who did not start.
Where the 256 hours went, by kind of time
The branch spends more hours on repeated work than on judgement
If a process change halved the repeated work, the branch would get back 98 ÷ 2 = 49 hours over two weeks, about one and a half recruiter weeks (49 ÷ 32 ≈ 1.5). With the same 24 starts, hours per placement would fall from about 10.7 to about 8.6, because (256 − 49) ÷ 24 ≈ 8.6. Whether a given change achieves anything like that is what your next measurement will show.
Which recruiter productivity metrics should a staffing agency track?
Track a small set that links hours, output and quality.
| Metric | How to count it | What it tells you | Watch out for |
|---|---|---|---|
| Recruiter hours per placement, by step | Hours on the step ÷ placements started | Where the time goes | Work and starts happen in different weeks; use matched orders |
| Share of repeated work | Repeated work hours ÷ all recruiting hours | How much a process change could free | People code differently; agree examples first |
| Placements per recruiter week | Placements started ÷ recruiter weeks worked | Output | Client mix and role difficulty differ by branch |
| Gross margin per recruiter hour | Gross margin on placements started ÷ recruiting hours | What an hour of recruiting earns | Pay rates and contracts differ per client; compare like with like |
| First shift attendance | Candidates who started ÷ candidates confirmed for a first shift | Whether speed still brings the right person | Pay, travel and shift times also drive it |
Read hours per placement next to first shift attendance and, later, early leavers. Our guide to measuring quality of hire in frontline recruitment covers the weeks after the start, and The Cost of a Frontline Hire in the Netherlands shows how to put a euro figure on the hours.
Why do activity counts mislead?
Calls and messages count effort, and effort includes the repeated work.
A dashboard of calls, messages and profiles submitted misleads in two ways. The recruiter who sends five reminders for one certificate looks busier than the colleague who asked for it once at the start. And people work towards what is counted: count calls, and calls go up.
Chasing also interrupts other work. In a 2008 laboratory experiment in which participants answered emails, those who were interrupted finished in less time with no difference in quality, but reported more stress, frustration, time pressure and effort (Mark, Gudith and Klocke, 2008). A recruiter who keeps pace through constant interruptions may be paying for it in ways no activity report shows.
Activity counts still work as a warning light: a sudden drop in first contacts in one branch is a reason to go and look.
Should you measure individual recruiters?
Measure branches, clients and steps; individual figures mostly reflect desk mix.
A recruiter whose client needs night shifts and certificates will log more hours per placement than a colleague whose client does not. That comparison says more about the clients than about the people. The study is there to find repeated work in the process, not to watch named staff. Keep diaries anonymous within the branch, report only per branch, client and step, and agree this with the team before anyone logs.
Involve the people affected early: recruiters often know where the repeated work sits before the figures show it.
Why does repeated work pile up in frontline staffing?
Many clients, many requirements and late checks multiply the same questions.
- Every client has its own list. Shifts, travel, certificates such as a forklift certificate or VCA. When the list is unclear at intake, the questions come back later. Telling candidates the hard facts early helps, as we argue in Tell Candidates About the 05:30 Start.
- Information lives in several places. Interview notes, message threads and the client's own system, so the planner and the client site ask again.
- Documents arrive after the yes. A missing certificate found after the client accepts means chasing under time pressure, and sometimes a replacement. Our article on which checks can happen before the offer explains the options.
- Every drop out restarts the chain for the same order.
What can you change once you know where the hours go?
Remove repeated work at its cause, then measure again.
| Repeated work | Common cause | Change to test |
|---|---|---|
| Chasing documents | Requested after the client says yes, one at a time | Ask for every document the role needs at the start, and check before the offer |
| Asking the same question again | Answers kept in personal notes | Ask every applicant the same structured questions and store each answer against the requirement it relates to |
| Entering details twice | Intake, interview and checks in separate tools | Keep sourcing, interviews, checks and contracts on one candidate record |
| Checking a document twice | No record of who checked what, and when | Record the source and date of every check where the next person can see it |
| Replacing people who did not start | Shift details and start information not confirmed | Share shift facts early and confirm the first shift the day before |
Change one thing in one branch at a time, and repeat the two week measurement six to eight weeks later. If hours per placement fall while first shift attendance holds, the change worked. If hours fall and attendance falls with them, the work has moved rather than gone.
Do not cut judgement to improve the number. In a 2022 review of selection research, structured interviews came out top of the selection procedures studied for predicting job performance (Sackett, Zhang, Berry and Lievens, 2022). Those are averages across many studies, but the time spent asking good questions and weighing the answers is the time worth keeping. Our guide on how to run structured interviews at high volume shows how to keep that structure when the week gets busy.
What can a time study not tell you?
It shows where the hours went, not the reasons behind them.
- Logging changes behaviour. People work differently when they know they are recording their time.
- Two weeks may not be typical. A new client or a sick colleague changes the mix of steps. Note what was unusual.
- Coding differs between people. Agreed examples reduce this but do not remove it.
- Placements lag the work. Without matched orders, the result swings from one fortnight to the next.
- Quality sits outside the number. Pair it with first shift attendance and early leavers.
- A before and after comparison is not an experiment. Seasons, clients and pay rates change at the same time as your process.
How Radius helps
Radius Hire is the AI hiring platform for frontline roles. Its AI Interviewer, Match and Radius Passport sit on one candidate record, and a recruiter makes every decision.
- One pipeline. The AI Sourcer is built to bring applicants from the channels you connect, such as job boards, and from your own candidate records into one pipeline, with the source recorded.
- The same questions for every applicant. The AI Interviewer invites applicants to a short, structured voice interview they can take on their phone, whenever it suits them, in the language set for the role. Each interview can be kept as a recording, a transcript and a summary that the next person can read.
- Evidence per requirement. For each requirement, Match shows whether it is backed by a document or a check, rests only on what the candidate said, or is still missing.
- Checks before the offer. Where the employer uses Radius Passport, identity and right to work are checked with a trusted verification partner and certificates are recorded with their expiry dates, so a recruiter can review the evidence before an offer. It does not replace the employer's own identity check.
- A person decides. Radius never rejects a candidate on its own. A recruiter reviews the evidence and makes every hiring decision. Candidate data is hosted in the EU; details are on the trust page.
Radius is built for staffing agencies that recruit for several clients and sites. How clients, branches and permissions are set up is agreed in the demo. Radius is new, so this article explains how the product works rather than claiming results.
Frequently asked questions
How do you calculate recruiter productivity?
Divide useful output by recruiter time over the same period. For a staffing agency, a practical pair is placements started per recruiter week, which shows output, and recruiter hours per placement split by step, which shows where the time goes. Count a placement when the candidate starts the first shift.
How many recruiter hours should a frontline placement take?
We know of no reliable public benchmark, so start from your own baseline. The hours depend on the client's requirements, shifts, certificates and drop outs. Compare each branch with itself over time, and clients of a similar kind with each other.
How long should a recruiter time study run?
We suggest two normal working weeks as a first baseline. Avoid peak weeks for the baseline, and repeat the same study six to eight weeks after you change something, so you compare like with like.
What is the difference between recruiter productivity and time to fill?
Time to fill counts calendar days; recruiter productivity counts working hours. A role can fill quickly and still take many recruiter hours, or fill slowly with few. Track both, because candidates feel the days and your margin feels the hours.
Does new recruitment software automatically raise recruiter productivity?
Only if it removes repeated work rather than moving it elsewhere. Measure hours per step and first shift attendance before and after the change. If chasing simply moves from the recruiter to the planner or the client site, the agency is no better off.
Conclusion
Placements per recruiter tell you how much came out, not where the hours went. Measure recruiter hours per placement by step for two normal weeks, sort the hours into judgement, necessary work and repeated work, and fix the largest block of repeated work at its cause. Measure again and check that first shift attendance holds. Protect the hours your recruiters spend asking good questions and making decisions: those are the hours that place the right person.
Sources
- CBS (30 July 2026). "Werkloosheid gedaald in tweede kwartaal 2026". Statistics Netherlands, preliminary seasonally adjusted figures for the second quarter of 2026 (NL). https://www.cbs.nl/nl-nl/nieuws/2026/31/werkloosheid-gedaald-in-tweede-kwartaal-2026 (checked 22 October 2026).
- Mark, G., Gudith, D. and Klocke, U. (2008). "The cost of interrupted work: more speed and stress". Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI 2008), 107 to 110. https://doi.org/10.1145/1357054.1357072 (checked 22 October 2026).
- Sackett, P. R., Zhang, C., Berry, C. M. and Lievens, F. (2022). "Revisiting meta-analytic estimates of validity in personnel selection: Addressing systematic overcorrection for restriction of range". Journal of Applied Psychology, 107(11), 2040 to 2068. https://doi.org/10.1037/apl0000994 (checked 22 October 2026).
- NIST/SEMATECH. "e-Handbook of Statistical Methods", section 7.2.4.1, "Confidence intervals" (for a proportion). National Institute of Standards and Technology. https://www.itl.nist.gov/div898/handbook/prc/section2/prc241.htm (checked 23 October 2026).