The agency vs in-house recruiting question is usually argued as a cost comparison, which is the least interesting version of it. Fees are visible and salaries are visible, so both sides produce a spreadsheet and the conversation stalls. The difference that actually matters is structural: the two models optimise for different outcomes, and that shapes everything from tooling to how long a sequence runs.
An agency is paid for a placement. An in-house team is paid for a hiring function. Those are not the same objective, and pretending otherwise is why so many partnerships sour in month four. This guide compares the workflows, the economics, the metrics and the tooling, then offers a decision framework you can apply per role rather than per company.
Agency vs in-house recruiting: where the incentives diverge
Agency economics reward submission speed and conversion. Revenue arrives when a candidate starts, so the rational behaviour is to work many requisitions in parallel, submit quickly, and move on from roles that stall. Nothing about that is cynical; it is the model working as designed.
In-house economics reward pipeline reuse and quality retention. Nobody pays a bonus for filling a role in nineteen days if the hire leaves in seven months. The rational behaviour is to invest in talent pools that serve future requisitions, protect the employer brand, and accept a slower fill on roles where a mis-hire is expensive.
Almost every observed difference in agency vs in-house recruiting traces back to those two sentences. Read the rest of this guide with them in mind.
Agency vs in-house recruiting workflows, side by side
| Stage | Agency | In-house |
|---|---|---|
| Intake | Client briefing, often 30 minutes by phone | Structured intake with scorecard and calibration profiles |
| Requisition load | 15–40 live roles across several clients | 8–15 roles for one employer |
| Sourcing horizon | Days; speed to first submission is the metric | Weeks; pipeline reused across requisitions |
| Sequence length | 2–3 touches, fast cadence | 4–5 touches, widening gaps |
| Employer brand | Represents many brands, sometimes anonymously | Single brand, long-term reputation exposure |
| System of record | Recruitment CRM with client and placement objects | ATS with requisition and offer workflows |
| Rejected candidates | Redeployed to another client immediately | Held as silver medallists for future roles |
| Success measure | Placements, fee revenue, submission-to-interview ratio | Time to fill, cost per hire, quality and retention |
Two rows deserve highlighting. The rejected-candidate row explains why agencies build enormous reusable networks while in-house teams build deep single-employer pools. And the sequence-length row explains why agency outreach reads as urgent while in-house outreach reads as patient — both are correct for their model.
Agency vs in-house recruiting economics, with actual numbers
Contingency agency fees typically run 15–25% of first-year salary, with retained and executive search reaching 25–33%. In-house costs are salary, on-costs, tooling and advertising, spread across whatever volume the team delivers. The comparison only becomes meaningful at a stated hiring volume.
| Scenario: 20 hires per year, £60,000 average salary | Agency | In-house |
|---|---|---|
| Placement fees at 20% | £240,000 | — |
| Recruiter salary plus on-costs | — | £69,000 |
| Sourcing data and outreach tooling | — | £6,000 |
| Job boards and advertising | — | £8,000 |
| ATS licence | — | £7,000 |
| Total annual cost | £240,000 | £90,000 |
| Effective cost per hire | £12,000 | £4,500 |
The catch sits in that last caption. At 20 hires, in-house wins decisively. At six hires, the same fixed cost produces £15,000 per hire and the agency looks cheap. In-house recruiters average roughly seven hires per quarter across all role types, dropping to around four for technical roles, so capacity planning determines the answer more than any philosophical preference.
Both models also carry costs that rarely appear on the invoice: vacancy cost while a role sits open, and replacement cost when a hire fails. SHRM’s recruiting benchmarking research is a reasonable external anchor when you build your own model. We break the calculation down in detail in our guide to cost per hire and time to fill.
Tooling: the stacks look similar and behave differently
Both models need search, contact data, sequencing and a system of record. What differs is how each component is configured.
- System of record. Agencies need a CRM that models clients, contacts, candidates and placements simultaneously. In-house teams need an ATS built around requisitions, approvals and offers. Forcing one to do the other’s job is the most common source of reporting pain, as covered in our ATS integration guide.
- Contact data. Agencies need breadth across industries because the next client is unknown. In-house teams need depth in a handful of functions and geographies. That difference should drive vendor selection, not database size.
- Sequencing. Agencies run many short sequences with high throughput. In-house teams run fewer, longer sequences with careful brand control and stricter suppression.
- Signals. Agencies benefit most from hiring-surge signals that identify clients with open needs. In-house teams benefit most from job-change signals that identify candidates in motion.
That last point is genuinely underrated. An agency using hiring signals is doing business development; an in-house team using the same platform is doing talent mapping. Same data, entirely different job.
Suppression and candidate experience
Suppression rules are stricter in-house, and the reason is exposure. If an in-house recruiter contacts the same engineer three times in a quarter, the damage lands on one employer brand permanently. An agency contacting the same person about three genuinely different clients is behaving normally, provided each approach is relevant.
Three rules protect both models. Maintain one global opt-out list across recruiting and sales, since candidates do not distinguish between your departments. Enforce a minimum re-contact interval, typically six months. And record every touch against the candidate record rather than leaving it in individual inboxes, which is the only way a team of five avoids stepping on itself.
Metrics each model should actually track
| Metric | Agency priority | In-house priority |
|---|---|---|
| Submission-to-interview ratio | Critical | Useful |
| Time to first submission | Critical | Low |
| Time to fill | High | Critical |
| Cost per hire | Low | Critical |
| Offer acceptance rate | High | Critical |
| Quality of hire at 12 months | Medium, tied to guarantees | Critical |
| Pipeline reuse rate | High across clients | High across requisitions |
| Source-of-hire mix | Medium | High |
When agency and client disagree, it is usually because they are watching different rows. Agreeing the shared metric at the start of an engagement prevents most of those arguments — typically submission-to-interview ratio, which measures brief quality and candidate quality at the same time.
Where agencies genuinely win
- Sudden volume. A new office or an unexpected reorganisation creates demand no fixed team can absorb.
- Genuinely rare skills. A specialist agency’s existing network beats a cold search almost every time.
- New geographies. Local market knowledge, salary norms and legal nuance take a year to build internally.
- Confidential searches. Replacing an incumbent requires distance the internal team cannot provide.
- Low, irregular hiring volume. Below roughly eight to ten hires a year, a fixed recruiter is hard to justify.
Where in-house genuinely wins
- Repeatable roles. The third support engineer costs a fraction of the first when the pipeline persists.
- Employer brand control. Every touchpoint is yours, which matters in small talent markets where reputation travels.
- Cultural assessment. Internal recruiters read team dynamics that no briefing document conveys.
- Long-term pipeline value. Silver medallists and nurture pools compound in value; agency networks do not compound for you.
- Data ownership. Your candidate history, outreach records and source attribution stay with you when a contract ends.
The hybrid model most teams settle on
In practice the agency vs in-house recruiting decision is rarely binary. Most organisations above fifty employees run a split, and the useful question is where to draw the line.
- In-house by default for roles you hire more than twice a year.
- Agency for spikes and specialisms — surge volume, confidential searches, and functions where you lack any network.
- Embedded or RPO for scaling phases, where you need agency capacity but in-house brand behaviour and data ownership.
- Always retain the data. Whichever model fills the role, candidate records and outreach history should land in your system.
That final point is the one most frequently lost. An agency engagement that leaves no pipeline behind has bought you a hire; one that leaves candidate records behind has bought you a hire and an asset.
Frequently asked questions about agency vs in-house recruiting
At what hiring volume does in-house become cheaper?
Around eight to twelve hires a year for mid-salary professional roles, using typical agency fees and fully loaded recruiter costs. Below that, fixed costs dominate. Above roughly twenty hires, in-house is usually cheaper by a wide margin.
Do agencies and in-house teams need different sourcing tools?
They need the same capabilities configured differently. Agencies weight breadth of coverage and multi-client suppression; in-house teams weight depth in specific functions and ATS write-back. A platform that handles both saves running two stacks.
How should we brief an agency to get better submissions?
Share three real profiles you would interview and three you would reject, with reasons. Calibration profiles outperform any written specification, and they cut the first submission cycle dramatically.
Can an in-house team run agency-style outreach?
Yes, and it is where most sourcing improvement comes from. Use the same search, data and sequencing discipline, but keep in-house suppression rules and longer cadences. Our outbound recruiting sequence guide covers the structure that transfers cleanly.
What happens to candidate data when an agency contract ends?
Whatever your contract says, so read it before signing. Agree at the outset that submitted candidate records, notes and outreach history transfer to your ATS. Without that clause, you lose the pipeline along with the partnership.
Decide per role, not per company
The strongest hiring functions stop treating agency vs in-house recruiting as an identity and start treating it as a routing decision. Repeatable roles go in-house because the pipeline compounds. Spikes, specialisms and confidential searches go external because speed and network beat fixed capacity.
Route each requisition on its own merits, keep the data whichever way it goes, and revisit the split every quarter as your volume changes.
Read next
- Cost per hire and time to fill — the numbers behind this decision.
- Candidate sourcing — the workflow both models share.
- ATS integration — keeping data yours whoever fills the role.
- The buying checklist — questions to ask before signing any data contract.
ZenBee supports both models from one platform: breadth of coverage and hiring-surge signals for agencies, deep function-level data and job-change alerts for in-house teams. Explore the ZenBee recruitment platform or compare plans on pricing.