Home/Insights/RPO and embedded hiring 10 min read · Updated September 2026
RPO & EMBEDDED HIRING · HIRING MODELS

RPO vs In House Recruiting: Which One Fits Your Team?

The choice between RPO and an in house recruiter is arithmetic, not philosophy. It comes down to how many roles you will close in a year and whether that volume holds after this one. Below roughly twelve to fifteen hires the permanent seat rarely pays for itself; above forty it is usually the cheapest option you have. Here is what each model costs once you load it properly, and the hybrid most scaling teams end up running.

PM
Pratik Mokashi
COO, Talhive · 40+ India mandates for US and EU clients
Key takeawaysThe whole piece in five lines
01The decision is driven by two variables only: how many roles you will close in a year, and whether that volume continues after this year.
02An in house recruiter costs far more than salary once you load tooling, ramp and the quiet months.
03Embedded wins on a spike or a new market. In house wins on steady volume over years.
04Most scaling teams run a hybrid, and that is usually the right answer rather than a compromise.
05Both models fail the same way: nobody owns the outcome. Decide who is accountable for the close before you decide the model.
Discuss a mandate →

Should you use RPO or hire an in house recruiter?

It comes down to volume and duration. Below roughly 12 to 15 hires a year an in house recruiter rarely pays for itself, because the seat costs the same in the quiet months as the busy ones. Above about 40 hires a year with that volume continuing, in house is usually cheaper and builds knowledge you keep. Between those two numbers, and for any hiring spike with an end date, embedded RPO is the better economic answer.

"We are about to hire our first in house recruiter. Our VP People says we need one, our CFO says agencies are cheaper. We are hiring maybe fifteen people next year. Who is right?"
COO, Series B SaaS · 90 people, scaling engineering and product

If you are deciding this week, read only this.

Under 12 hires a year
Agencies or dedicated search
Per-hire fees cost less than a permanent seat at this volume.
12 to 40 hires, sustained
Embedded RPO
Capacity pricing beats per-hire fees, without the fixed headcount.
40+ hires, continuing for years
In house team
Cheapest per hire at this scale, and the knowledge stays with you.
A spike with an end date
Embedded, not in house
Do not create a permanent seat to solve a temporary problem.

The argument between RPO vs in house recruiting usually gets framed as a philosophy question, as though one model is more committed to quality and the other is a shortcut. It is not. It is an arithmetic question with two inputs: how many roles you will close in the next twelve months, and whether that number holds after this year.

What an in house recruiter actually costs.

The number most teams carry in their head is the salary. That is roughly two thirds of the real figure, and the missing third is where the comparison usually goes wrong.

~65%
Share of the true annual cost that salary represents, before tooling and management time
3-4mo
Typical ramp before a new in house recruiter is closing at full rate
12-15hires
Annual volume at which the permanent seat starts to pay for itself

Add the sourcing licences, the job board spend, the applicant tracking system, and the hiring manager hours the process consumes. Then add the part nobody models: the quiet months. A permanent seat costs the same in a quarter with two open roles as it does in a quarter with twelve.

What embedded hiring costs, and what you actually get.

Embedded and RPO models price on capacity rather than per placement, which changes the shape of the cost. You are buying a number of open roles worked in parallel for a number of months, not a fee per person who signs.

The practical difference is that the cost per hire falls as volume rises, and stops entirely when the programme does. That is the whole argument for the model, and it is also the argument against using it for steady long-run volume, where the fixed seat eventually wins.

Agencies price per hire, embedded prices per month, in house prices per year. Match the pricing shape to the shape of your hiring plan and the decision makes itself.

The comparison, at three hiring volumes.

ANNUAL HIRESAGENCYEMBEDDED RPOIN HOUSE
5 rolesCheapestUnderused capacitySeat idle most of the year
15 rolesFees stack up fastBest fitBreaking even at best
40 rolesProhibitiveWorks, costs more per hireCheapest per hire

Tell us your hiring plan for the year and we will tell you which model costs less on your actual numbers.

Discuss a mandate →

Where in house genuinely wins.

Steady volume over several yearsECONOMICS

Once the seat is busy every month, the fixed cost is the cheapest way to buy recruiting capacity. Nothing external competes at that point.

Watch for: the plan that assumes this year's volume repeats. Check it against your actual headcount plan, not the optimistic one.
Deep employer brand workCOMPOUNDING

Careers content, referral programmes, the relationships with candidates who say no this year and yes in two. This work compounds and it belongs to someone who stays.

Watch for: whether anyone is actually doing this, or whether the seat is fully consumed by req firefighting.

Where embedded wins.

A hiring spike with an end dateTIMING

You raised, you have twenty roles to close in nine months, and then you return to normal. Creating a permanent seat for a temporary problem leaves you with a redundancy conversation.

Watch for: spikes that quietly become permanent. Review at month six.
A market you have never hired inKNOWLEDGE

Your in house recruiter may be excellent and still have no network in Bengaluru, no read on which companies produce the calibre you want, and no sense of what an offer should look like.

Watch for: assuming recruiting skill transfers across markets. It transfers far less than people expect.

The hybrid most teams actually run.

After the first year, most scaling companies land in the same place, and it is worth going there deliberately rather than by accident. The in house recruiter owns the repeatable roles and the employer brand. Embedded capacity is added for the spike, the new market, or a seniority band the in house team has not hired before.

That split works because it matches each model to what it is good at. It fails when nobody is clear who owns the close, which brings us to the real failure mode.

How both models fail.

01Nobody owns the outcome

The single most common failure, and it is independent of model. Two parties are involved, both are busy, and the close belongs to neither.

WHAT IT COSTS
Roles stay open for months while everyone reports activity and nobody reports a signature.
THE FIX
Name one person accountable for the close on each role, before the model conversation.
02Comparing salary against fees

The CFO comparison that starts the argument is almost always salary against agency fee, which is not the comparison.

WHAT IT COSTS
A permanent seat created on a number that was understated by roughly a third.
THE FIX
Load both sides properly: tools, ramp, management time and the idle months against real cost per hire.

How we structure embedded hiring.

We scope embedded work against the number of roles open in parallel and the months the programme runs, not per placement, and we say plainly when the volume does not justify it. On executive and founding roles we run dedicated search on milestone-based fees instead, because those searches are a different exercise.

If your plan is fifteen roles and half of them are senior, the honest answer is usually a split, and we will tell you which half belongs where.

Send the hiring plan. We will come back with the cost per hire under each model on your own numbers.

See how embedded works →

Frequently asked questions.

As a working rule, somewhere around 12 to 15 closes a year at the point the recruiter is fully productive, and the volume has to be sustained. A single year of high volume followed by a quiet year usually does not justify the permanent seat, because you are then carrying salary through months with nothing to fill.
Usually, once volume is involved. Agency fees are charged per hire, so ten hires cost ten fees. Embedded and RPO models are priced on capacity rather than per placement, so the cost per hire falls as volume rises. At one or two hires a year the agency model is cheaper.
Mostly scale and formality. An embedded recruiter is one person working inside your team on your tooling and your employer brand. RPO usually means a larger scoped programme with process, reporting and often several people. The commercial logic is the same: you are buying capacity rather than individual placements.
Yes, and most scaling teams end up here. The common split is an in house recruiter owning the steady, repeatable roles and the employer brand, with embedded capacity added for a spike, a new market, or a level of seniority the in house team has not hired before.
Salary is roughly two thirds of it. Add sourcing tools, job board spend, an applicant tracking system, the hiring manager time the role consumes, and the months where the seat is funded but the pipeline is thin. Compare that total against the model you are considering, not the salary on its own.