A recruitment agency steers on a handful of numbers: time-to-fill, conversion per process step, return per sourcing channel and margin per placement. More KPIs make the dashboard fuller but the steering weaker. In a market with 93 vacancies per 100 unemployed people (CBS, 2026), the agency that makes speed and return measurable wins — not the one with the most charts. This article defines the core KPIs, shows how to measure them without manual work and where the pitfalls are.

Which KPIs really matter for an agency?

For a recruitment or staffing agency, steering comes down to a handful of numbers: time-to-fill, conversion per process step, return per sourcing channel and margin per placement — supplemented by the health of your candidate pool. More than seven KPIs per team works against you in practice: a KPI is only a KPI if someone can act on it. The rest is reporting.

The temptation is to measure everything the ATS can export. Do the opposite: start from the three decisions you make each month — where do I deploy recruiters, which channel gets budget, which client deserves priority — and choose, per decision, the numbers that support it. A good check per KPI: does it have a clear definition, a target value and an owner? If not, it's a chart, not a steering instrument (see also the term KPI).

Time-to-fill and conversion per stage

Time-to-fill — the number of days from vacancy open to candidate signed — is the speed KPI for your entire process, and it weighs more heavily the tighter the market is. That context remains substantial: at the end of 2025 there were 380,000 vacancies open in the Netherlands, with 93 vacancies per 100 unemployed people (CBS, 2026). Candidates have a choice; the fastest agency wins more often. Bullhorn also calculates that 80 per cent of candidates expect to be placed within twenty days (GRID Industry Trends Report, 2025).

The total figure only becomes useful when broken down by stage: request to first shortlist, shortlist to interview, interview to offer, offer to start. Per transition you measure both conversion and lead time — then you see whether the days leak away in the search or in the waiting (for client feedback, for scheduling). Define start and end points unambiguously in your ATS, otherwise you compare apples with pears; the term time-to-fill explains the definition choices.

Source ROI: what does each channel deliver?

Source ROI measures, per sourcing channel — job boards, own site, referrals, sourcing — not how many candidates it delivers, but how many interviews and placements. That distinction is the whole point: a channel that delivers many CVs and few placements costs money and recruiter time. With source ROI, the annual job board negotiation becomes a calculation instead of a gut feeling.

The precondition is automatic source registration: every candidate is tagged with their channel on arrival, without a recruiter having to fill it in. You arrange that in the integration between your channels and your ATS — multiposting with source attribution is the practical route for this. Manual source registration rarely reaches the data quality you need for this KPI.

Margin per placement and revenue per recruiter

Ultimately you steer on money: margin per placement (fee or hourly-rate margin minus direct costs) and revenue or gross margin per recruiter. These numbers make visible what the operational KPIs hide — a recruiter with many placements at low margins can contribute less than a colleague with few, high-value placements.

The data challenge here lies in the connection between ATS and back office: margin only emerges once placement, hours and invoicing are linked to one another. Agencies that keep this in sheets by hand have, by definition, outdated numbers. Connecting ATS to Power BI turns it into a live dashboard; how the flow from placement to invoicing is itself automated is covered in the recruitment automation playbook.

How do you measure this without manual work?

The only sustainable route is automatic measurement: KPIs that are compiled by hand each week die after three months. Concretely, that means three agreements. One: the ATS is the single source for process data — what isn't in it doesn't exist (the single source of truth agreement). Two: source registration and status transitions are enforced by the system, not by discipline. Three: the dashboard refreshes itself from the ATS and the back office, instead of from exports.

Comparison material for your own numbers can be found in recruitment statistics. And feel free to first work out what the manual work in your current reporting costs — for that there's the free ROI calculator.

In short

  • Steer on three to seven KPIs per team: time-to-fill (per stage), conversion per step, source ROI and margin per placement.
  • Context 2025/2026: 380,000 open vacancies and 93 vacancies per 100 unemployed people (CBS, 2026); 80% of candidates expect a placement within 20 days (Bullhorn GRID, 2025).
  • Source ROI requires automatic source registration — manually filled source fields don't reach the necessary data quality.
  • Margin per placement only emerges once ATS and back office are connected; sheets mean, by definition, outdated numbers.
  • Every KPI needs a clear definition, target value and owner — otherwise it's a chart, not a steering instrument.

Further reading

Frequently asked questions

What is a good time-to-fill?

There is no universal norm: time-to-fill varies strongly by role, sector and region. More important than an external benchmark is your own trend per stage: is the lead time getting shorter or longer, and in which step is the change happening? For context: 80 per cent of candidates expect to be placed within twenty days, according to Bullhorn (GRID, 2025).

How many KPIs should an agency track?

Three to seven per team. Every KPI should have a clear definition, a target value and an owner — someone who can intervene if the number drifts. More indicators make the dashboard fuller but the steering weaker.

What data do I need for source ROI?

Per candidate, the channel they came in through, and per channel the flow-through to interview and placement. That requires automatic source registration on arrival — via multiposting with source attribution or an integration between your channels and the ATS. Manually filled source fields are almost always too incomplete.

Can't my ATS just report this itself?

Process data (time-to-fill, conversions) often, yes. Margin per placement usually not, because that data lives in your back office: hours, rates, invoicing. For the complete picture you connect ATS and back office to a dashboard — see our page on connecting recruitment data to Power BI.

How often should I look at these numbers?

Operational KPIs (pipeline, time-to-fill per stage) weekly in the team meeting; financial KPIs (margin, revenue per recruiter) monthly. A live dashboard makes the frequency less important by itself: the number is always there, and the meeting gets a fixed rhythm.

Written by Hugo Eleveld · Updated . This article is informational; for tailored advice book an intro call.

Prefer these kinds of insights in your inbox?

Leave your email address and we'll add you to the list and email you as soon as the next edition on IT, automation and dashboards comes out. You can unsubscribe at any time with a single email.

We'll only use your email address for this — see the privacy statement.