Recruitment Agency KPIs: 12 Metrics That Predict Revenue

Most recruiting KPI frameworks are built around activity metrics: calls made, CVs sent, interviews arranged. These measure effort. They say almost nothing about whether your agency is actually healthy or heading for trouble.

The metrics that actually predict revenue growth — or warn you about a coming cliff — are different. They're harder to collect, less satisfying to track week-to-week, but they're the ones that explain why some agencies compound quietly for years while others have a great Q2 followed by a crisis.

This is a framework for agency owners and operations leads who want to run the business like a business. Twelve metrics, benchmarks where data exists, and a clear explanation of what each one actually tells you.

A Note on Data Collection

Before diving into the metrics: most of these require either a purpose-built ATS with analytics, or considerable manual aggregation. If you're tracking placements in a spreadsheet and client BD in another spreadsheet, some of these will be impractical to measure accurately right now. That's useful information in itself — the inability to measure a metric is often a signal that the process it represents isn't systematised enough to scale.

A recruiting platform like Yena's Hiring OS or a properly configured alternative should surface most of these automatically. The ones that require manual calculation are noted below.

The 12 KPIs

1. Fill Rate

What it measures: The percentage of job orders your agency successfully fills.
Formula: Placements ÷ Total Job Orders × 100
Benchmark: 55-70% is typical for contingency agencies; 85-95% for retained search firms.

Fill rate is the most fundamental efficiency metric in recruiting. A low fill rate doesn't just mean lost revenue — it means consultants are investing time in searches that generate nothing. Every unfilled role is a cost: the candidate pipeline built, the client calls made, the hours spent.

The diagnostic question is why you're not filling. Is it unrealistic role briefs from clients? Too-narrow sourcing channels? Weak offer management at the close? Each root cause requires a different fix.

2. Time-to-Shortlist

What it measures: Days from job order receipt to first qualified candidates presented to client.
Benchmark: 3-7 days for exec search; 1-3 days for contingency volume recruiting.

Time-to-shortlist is increasingly a competitive differentiator. Clients — particularly those who've had a bad experience with slow searches — use it as a proxy for agency capability. Being first with a strong shortlist wins business.

It also predicts revenue timing. Agencies with consistently short time-to-shortlist have more predictable cash flow because placement cycles are faster and more foreseeable. According to Bullhorn's GRID Report, the gap between top-quartile and median agencies on time-to-shortlist has widened significantly since 2023, partly driven by AI-assisted matching.

3. Client Reorder Rate

What it measures: Percentage of clients who place a second job order within 12 months of the first placement.
Benchmark: Above 50% is strong; below 30% signals a satisfaction problem.

"Winning a first placement is a sale. Earning a second one is a business."

— Common framing among agency growth consultants, via ERE Media

Client acquisition costs in recruiting are significant — business development calls, pitches, proposals, reference checks. If a client doesn't come back after a successful first placement, you've essentially subsidised their hiring without building an asset.

Low reorder rate is often a quality-of-hire problem that takes 6-12 months to surface. The placement looked fine, the candidate passed probation, but the client isn't thrilled with performance 9 months in. They look elsewhere next time. Tracking this metric with a 12-month lag is essential for diagnosing quality issues early.

4. Candidate NPS

What it measures: Net Promoter Score from candidates post-process, whether placed or not.
Benchmark: Above 40 is good; above 60 is exceptional for agencies.

This one is underused in agency recruiting, probably because it doesn't feel directly connected to revenue. It is. Candidate NPS predicts two things: referral volume and employer reputation in candidate communities.

In tight talent markets — particularly in DACH and the Nordics, where specialised professional communities are small — word travels fast. A candidate who had a poor experience with your agency's process doesn't just walk away. They tell colleagues. NPS gives you a leading indicator of reputation trends before they show up in inbound candidate quality.

The mechanics are simple: a three-question survey sent 48 hours after a process concludes (placement or rejection). SHRM's candidate experience research consistently shows that agencies in the top quartile on candidate experience have 23% higher referral rates than those in the bottom quartile.

5. Fee per Placement

What it measures: Average revenue per successful placement, and distribution across the range.
Benchmark: Highly variable by market; track trend over time more than absolute number.

Average fee per placement tells you whether your pricing strategy is working and whether your client mix is moving in the right direction. A growing average fee means you're either moving upmarket (higher-value roles) or negotiating better terms.

The distribution matters as much as the average. If your mean is €15K but your range is €3K-€45K, you have a wildly inconsistent pricing practice that makes revenue forecasting nearly impossible. Tightening the distribution — pricing more consistently based on role level and sector — is often a significant revenue improvement on its own.

6. Pipeline-to-Placement Ratio

What it measures: Total candidates submitted to clients vs. candidates placed.
Formula: Total Submissions ÷ Placements
Benchmark: 4:1 to 6:1 is typical; below 3:1 is excellent; above 10:1 suggests sourcing quality problems.

High submission-to-placement ratios are expensive. Every candidate submitted requires preparation, client communication, and follow-up. If you're submitting 15 people to place 1, you're doing a lot of work that clients aren't hiring.

The causes are usually one of two things: either the initial brief quality is poor (you don't fully understand what the client actually wants) or sourcing quality is weak (you're submitting candidates that don't genuinely fit). Both are fixable, but they require different interventions. Tracking ratio by consultant and by client type often reveals where the specific problem lives.

7. Sourcing Channel ROI

What it measures: Which sourcing channels (LinkedIn, job boards, database, referrals) produce placed candidates, and at what cost.

Most agencies know they're paying for LinkedIn Recruiter seats and job board credits. Very few track which channels actually produce placements versus which ones generate application volume that never converts.

The math is straightforward: divide your annual spend on each channel by the placements that originated from it. In most agencies, internal database sourcing (candidates you've already spoken to) and referrals produce placements at a fraction of the cost of paid channels. LinkedIn is typically mid-range. Generic job boards often produce the worst ROI for agency recruiting, yet attract significant budget.

This analysis typically produces immediate budget reallocation decisions. It also highlights the value of database quality — using your existing candidate relationships more effectively is often the highest-ROI sourcing investment available.

8. Consultant Utilisation

What it measures: Percentage of each consultant's billable time spent on revenue-generating activity vs. admin, meetings, and overhead.
Benchmark: 60-75% billable time utilisation is realistic; above 80% risks burnout.

Admin overhead in recruiting is consistently underestimated. CV formatting, email chasing, data entry, scheduling — a 2024 LinkedIn Talent Solutions study found that recruiters spend an average of 14 hours per week on tasks that could be automated. That's roughly 35% of a working week.

Low utilisation is often an ATS problem, not a people problem. Consultants working in platforms that don't integrate smoothly with LinkedIn, email, and calendar end up doing manual data entry that erodes their productive time. The right tooling doesn't just improve speed — it increases the billable percentage of every hour worked.

9. Client Concentration Risk

What it measures: Percentage of total revenue attributable to your top 1, 3, and 5 clients.
Benchmark: Top client below 20% of revenue; top 3 below 40%.

This is the metric that agency owners are most likely to not want to measure, because the numbers are often uncomfortable. A client that represents 30% of your revenue isn't an asset — it's a liability. When their hiring freezes (and it will, eventually), you have a crisis.

Client concentration risk is a slow-building problem. It feels fine quarter to quarter until it doesn't. Tracking it explicitly, and setting targets to diversify below the concentration thresholds above, turns it from an existential risk into a manageable growth objective.

10. Placement Falloff Rate

What it measures: Percentage of placements where the candidate leaves (or is let go) within the guarantee period.
Benchmark: Below 5% is strong; above 10% is a serious quality signal.

Every falloff is financially painful — you're typically obligated to fill the role again, for free. But the real damage is to the client relationship. Clients remember falloffs. They're the stories told in the board meeting when someone asks whether to use the same agency again.

Agencies with low falloff rates typically have two things in common: strong candidate qualification (they're honest about fit mismatches rather than pushing candidates through) and good offer management (they surface concerns before acceptance rather than after). Tracking falloff rate by consultant identifies both good practice and concerning patterns.

11. Speed-to-First-CV

What it measures: Time from job order receipt to first CV sent to the client.
Benchmark: Under 24 hours for contingency; under 48 hours for retained.

Speed-to-first-CV is distinct from time-to-shortlist. This is about responsiveness signalling — clients form impressions of agency capability within the first 48 hours of a brief. A well-qualified CV sent within 24 hours communicates that you understand the brief and have relevant talent on hand. A response that takes five days communicates the opposite, regardless of eventual quality.

The agencies hitting sub-24-hour speed-to-first-CV consistently are invariably the ones with clean, well-structured databases and AI-assisted matching. When you can query your existing database intelligently and surface strong matches in minutes, the first CV is a database hit rather than a new sourcing exercise.

12. Gross Margin per Consultant

What it measures: Total gross profit (placement fees minus direct costs) divided by number of consultants.
Benchmark: £80-150K gross margin per consultant annually is typical for perm agencies; higher for exec search.

This is the metric that finance-minded agency owners watch most closely, and it's the most direct predictor of whether you can profitably hire your next consultant. If your current consultants aren't generating sufficient margin, adding headcount worsens the problem rather than solving it.

The levers are fill rate, average fee per placement, and utilisation. Improving any one of them improves margin per consultant. All three together compound.

KPI Summary Table

Which Metrics to Start With

Tracking all twelve at once is a recipe for analysis paralysis. If you're building your first real KPI dashboard, start with the four that provide the widest diagnostic view:

Fill rate tells you whether delivery is working. Client reorder rate tells you whether quality is meeting client expectations. Gross margin per consultant tells you whether the business model is financially sound. Client concentration risk tells you whether you have a resilience problem hiding in plain sight.

Those four, tracked quarterly, will surface the most important issues in most agencies. Add the others as operational maturity increases.

"The agencies that outperform over a decade aren't necessarily better at finding candidates. They're better at knowing what's working and doing more of it."

— RecruitingDaily, Agency Performance Analysis 2025

GDPR Considerations for European Agencies

Several of these metrics require storing and analysing candidate data over extended time periods — candidate NPS, client reorder rate, and falloff rate all depend on longitudinal tracking. Under GDPR, you need a lawful basis for retaining candidate data after a search concludes.

Legitimate interest is commonly used by agencies for database retention, but it requires a documented balancing test. For agencies operating in Germany, Austria, or Switzerland, Works Council agreements may also govern how candidate performance data is tracked and used internally. If you're building a KPI framework for the first time, a brief legal review of your data retention policy is worth the investment.

Frequently Asked Questions

How often should we review these KPIs?

Short-cycle operational metrics (time-to-shortlist, speed-to-first-CV, consultant utilisation) benefit from weekly review. Medium-cycle metrics (fill rate, pipeline-to-placement ratio, sourcing channel ROI) are better reviewed monthly — enough time for patterns to emerge. Longer-cycle metrics (client reorder rate, falloff rate, client concentration) need quarterly review at minimum, because they require 6-12 months of data to be meaningful.

Can these metrics be tracked without dedicated software?

Some can, with disciplined spreadsheet management. Fill rate, fee per placement, and gross margin per consultant can be calculated manually. Time-based metrics (time-to-shortlist, speed-to-first-CV) require either automated logging or manual timestamps that are easy to forget or fudge. Candidate NPS requires a survey tool. For most agencies beyond 5 consultants, the manual approach creates enough overhead that the metrics don't actually get tracked consistently. That's the hidden cost of under-investing in systems.

What's the single most predictive KPI for revenue growth?

Client reorder rate. It captures quality, relationship health, and pricing discipline in a single number. An agency with a 60%+ client reorder rate is compounding — each year's client base becomes next year's baseline with less BD spend. An agency with a 25% reorder rate is essentially starting from scratch each year. The gap in long-term growth between those two profiles is enormous.

How do we improve metrics that are currently weak?

Diagnose before intervening. A weak fill rate has completely different causes than a weak falloff rate, and the same metric can be weak for different reasons in different agencies. Start with root-cause analysis: break the metric down by consultant, by client type, by role level, and by time period. The pattern usually points clearly to where the problem lives.

Are these metrics the same for all agency types?

The metrics are universal but the benchmarks differ. A high-volume contingency agency and a boutique retained search firm will have very different fill rate and time-to-shortlist norms. The table above gives ranges — calibrate your targets to your model and to peer benchmarks in your market segment. Bullhorn's annual GRID Report publishes benchmarks segmented by agency type and is worth using as a calibration point.

Most of these KPIs can be tracked automatically inside a good ATS. Yena's Hiring OS surfaces fill rate, pipeline velocity, and placement history natively, with consultant-level reporting that makes these reviews a 20-minute weekly exercise rather than a half-day spreadsheet exercise. Start a free trial and see what your current numbers actually look like.