Recruiting agency metrics: definitions, formulas and worked examples

The recruiting agency metrics worth tracking follow a candidate from job order to placement and then to revenue: time-to-submit, submittal-to-interview, interview-to-offer, offer acceptance, fall-off rate, fill rate, time-to-fill and revenue per recruiter. Each is a simple ratio or a duration calculated from dated statuses in your ATS. The definitions matter more than the numbers: write each one down, count the same way every month, and compare against your own history before anyone else's.

This page defines each metric, gives the formula and works through the arithmetic with invented numbers. Putting them on one page every week is covered in the recruiter weekly report template.

The metrics at a glance

Funnel conversion: submittal to start

Conversion rates are only meaningful when the top and bottom of the fraction are the same group of candidates. There are two ways to count, and mixing them is the most common reason agency numbers jump around.

• Cohort counting: take every candidate submitted in June and follow what happened to those candidates, however long it took. Accurate, but June's numbers are not final until the slow processes finish.
• Period counting: count submittals in June and interviews in June. Fast, but June's interviews partly came from May's submittals, so small desks see rates swing wildly.

Use cohorts for monthly and quarterly reviews. For weekly reports, use period counts over a rolling four weeks, which smooths most of the mismatch.

What each rate points to

• Low submittal-to-interview usually means the brief and your screen disagree. Look at client rejection reasons for patterns before sending more candidates; see how to present candidates to clients.
• Low interview-to-offer points at fit the screen missed, or candidates arriving unprepared for what the client tests.
• Low offer acceptance almost always traces back to pay, or to a pre-close that did not happen. The offer call script starts with that call.
• High fall-off points at the weeks between yes and start: counteroffers, resignation handling, silence during notice.

Define "fall-off" before you count it. Some agencies count only accepted offers that never started; others also count placements that left within the guarantee period. Both are useful. Report them as two separate numbers.

Speed: time-to-submit and time-to-fill

Time-to-submit

"Time-to-submit" means two different things. Screen-to-submit runs from the end of the phone screen to the submittal and measures write-up speed. Time to first submittal runs from receiving the job order to the first candidate submitted and measures how fast sourcing produces. Pick one per report and label it. The screen-to-submittal workflow covers the first in detail.

Time-to-fill

Time-to-fill runs from the day you received the job order to the day a candidate accepted the offer. Some agencies end it at the start date instead; that adds the notice period, which neither you nor the client controls. Whichever you choose, count only filled roles and write the definition at the top of the report.

Report medians for durations, and split time-to-fill by role type or fee model. A contingent high-volume role and a retained search do not belong in one number.

Fill rate and job order quality

Fill rate is the share of closed job orders that your agency filled. The choice that changes the answer is what goes in the denominator.

• Count only closed orders. Open roles have no outcome yet; including them makes a growing desk look worse every month.
• Split by fee model. Exclusive, retained and contingent orders behave differently, and a blended fill rate hides which ones are worth taking.
• Split by client. A client whose orders you fill rarely, and who often fills them elsewhere, may be using you as a backup. That is a conversation, not a metric to hide.
• Track cancelled orders separately. A high cancellation rate for one client is a signal about the quality of their job orders, and a reason to ask more at the intake meeting.

Revenue per recruiter and contract margin

Revenue per recruiter

Divide fees billed in the period by the number of full-time-equivalent (FTE) recruiters in that period. Count a recruiter who joined halfway through as 0.5, and use your commission plan's splits for shared placements.

Decide whether you use fees billed, fees collected, or fees net of refunds under guarantees, and keep to it. For owners, revenue per recruiter only means something next to the cost of each recruiter, including salary, commission, tools and a share of overhead.

Contract desks: spread, margin and markup

For contractors, the hourly spread is the bill rate minus the pay rate minus the employer's costs of employing the contractor (payroll taxes, insurance and similar costs, which vary by state and arrangement). Gross margin and markup are often confused, so calculate both and label them.

Activity metrics, and why they come second

Calls, screens and submittals are inputs. They matter because they are what a recruiter controls on a given day, but they only make sense next to conversion. Two useful ways to link them:

• Screens per submittal: screens completed ÷ submittals. A rising number can mean better screening or a harder brief; a falling one can mean faster judgment or looser standards. Read it next to submittal-to-interview.
• Submittals per open role: submittals ÷ open job orders. Tells you whether effort is spread across the desk or piled on one search.

Setting activity targets without outcome metrics produces activity. A recruiter who doubles submittals and keeps the same number of interviews has halved their submittal-to-interview rate, and the client has read twice as many profiles that did not fit.

Calculating them from your ATS or a spreadsheet

Every metric above comes from a small set of dated fields. If your ATS records when statuses change, export them; if not, a spreadsheet with one row per submittal and one row per job order is enough.

SUBMITTALS (one row per candidate submitted to a job)
SubmittalID | JobID | Client | Recruiter | ScreenEnd | SubmittedAt |
FirstInterviewDate | OfferDate | AcceptedDate | StartDate |
FellOff (Y/N) | Fee | RecruiterSplit

JOB ORDERS (one row per job)
JobID | Client | FeeModel (contingent/exclusive/retained) |
ReceivedDate | ClosedDate | Outcome (filled by us / filled elsewhere /
cancelled) | RoleType

Formulas that work in Excel 365 and Google Sheets, using named ranges for the columns:

Submittal-to-interview for June cohort
=COUNTIFS(SubmitMonth,"2026-06",FirstInterviewDate,"<>")
 / COUNTIFS(SubmitMonth,"2026-06")

Offer acceptance, all offers
=COUNTIFS(AcceptedDate,"<>") / COUNTIFS(OfferDate,"<>")

Median time-to-fill, filled roles
=MEDIAN(FILTER(AcceptedDateByJob - ReceivedDate, Outcome="filled by us"))

Fill rate excluding cancelled
=COUNTIFS(Outcome,"filled by us")
 / (COUNTA(Outcome) - COUNTIFS(Outcome,"cancelled"))

SubmitMonth is a helper column holding the submittal month as text, such as 2026-06. AcceptedDateByJob is the acceptance date of the filling candidate, looked up onto the job order row. Leave Outcome blank while a role is open, so the fill rate formula counts closed orders only.

Benchmarks, targets and common mistakes

Published agency benchmarks exist, but they rarely state the definitions, desk mix, fee model or sample behind them. A submittal-to-interview figure from an agency that submits two candidates per exclusive search is not comparable to one from a high-volume contingent desk. Set targets from your own trailing three to six months, by client and role type, and move them when the trend moves.