What Most Recruitment Agencies Get Wrong About Their Own Profit

Sit down with a recruitment agency founder and ask about the month's placement count, and the answer comes immediately. Ask about the fall off rate during the guarantee period, the share of placements that trigger a rebate or a replacement before that window closes, and the answer usually does not come at all. Not because it is being hidden. Because nobody has ever built the number in the first place.

The Gap

A placement fee looks completely earned the moment it clears. The invoice is real, the payment is real, and there is nothing about that specific transaction that looks wrong on a normal pass through the books. Sitting underneath nearly every placement contract, though, is a rebate or replacement guarantee, a window during which a candidate leaving triggers a refund, a prorated repayment, or a free replacement search carried out at the agency's own cost. That clause exists in the contract from day one, regardless of whether it ever actually gets used.

Most growing agencies book the fee in full anyway. It is rarely a deliberate choice. It is simply what happens by default when nothing forces a different question to get asked.

The same blind spot usually extends one step further, into how recruiter commission gets structured. Commission tied to the placement itself, paid out the moment the fee is invoiced, creates a second exposure sitting on top of the first. If that placement later falls through inside the guarantee window, the agency is refunding a fee it has already paid commission against, out of cash that has already left the business twice over.

Sizing the Actual Exposure

It is worth working through what this gap is genuinely worth, using cautious, clearly stated assumptions rather than a worst case scenario.

Take an agency placing around 8 candidates a month, roughly 96 a year, at a typical fee near €8,000. That's about €64,000 a month in placement fees moving through the business, or €768,000 a year. Jobvite's 2018 Job Seeker Nation survey found that 33% of new hires leave a role within their first 90 days across industries broadly, but that figure spans everything from high volume entry-level hiring to senior specialist roles, and a business placing vetted candidates into guaranteed professional positions should reasonably expect to sit well below it.

Even using a deliberately conservative 8% fall off rate, that works out to 8 placements a year running into some form of rebate. The average clawback per claim, once full refunds on early leavers are blended with prorated repayments on later ones, typically lands around 47% of the fee, not 100%, and not 10%. At 47%, that's just over €30,000 annually that has likely been counted as profit without ever being tested against what the business actually experiences. That is not a worst case figure. It is closer to a floor.

Why Growth Makes This Worse, Not Better

The natural assumption is that a stronger placement month is simply a better month. For an agency carrying this kind of exposure, a stronger month is also a month where more fees are being booked in full, which means more unprovisioned risk enters the business at exactly the point it looks like it is performing best.

That risk does not disappear if it goes unaddressed. It sits on the balance sheet unrecognised, invisible until a candidate actually leaves during a guarantee window and the agency has to fund a refund or a free replacement search out of cash that has, on paper, already been counted and spent as profit.

Building the Number, Step by Step

The provision only becomes credible once it comes from the business's own numbers, not a percentage borrowed from an article or a competitor's stated policy. The actual process is fairly direct.

Start by pulling the full placement history, not a recent sample, ideally the last 2 to 3 years, long enough to see a genuine pattern rather than a handful of coincidences. Every placement then needs one factual tag next to it: did it run the full guarantee period without a claim, or did it trigger a rebate or replacement, and if so, at what point inside that window.

A single blended percentage across the whole book tends to hide more than it reveals. The next step is breaking that fall off rate down by whatever actually drives the difference in this specific business, commonly seniority level, since junior and senior placements rarely behave the same way, sometimes client type or industry vertical, sometimes which consultant handled the placement. A blended average of 8% might really be 3% on senior roles and closer to 15% on junior ones, and that difference changes where the real exposure sits.

From there, each segment gets its own two figures: how often a clause actually gets triggered, and on average, how much of the fee is actually clawed back when it does, since a full refund, a 50% prorated repayment, and a free replacement search all carry a different real cost, and a free replacement needs to be converted into its approximate cost equivalent to be counted properly.

Applying that forward is the last step. Take current placement volume and fees by segment, apply each segment's own rate and average clawback, and hold that amount back from what gets called distributable profit each month or quarter, as a genuine provision rather than a rough guess.

This is not a one time exercise. As the business's client mix or seniority mix shifts, or as enough new data accumulates, the rate needs revisiting, since a number built on placements from 3 years ago will not necessarily reflect what the business looks like today.

Choosing How to Treat It

There are two defensible ways to handle this once the rate is known. Some businesses defer the fee entirely, only recognising it once the guarantee period has fully expired. Others recognise the fee immediately but hold the calculated provision against it. Neither is automatically correct, what matters is choosing one deliberately, on purpose, rather than defaulting into full recognition simply because nothing forced the question.

It is usually worth reviewing commission structure at the same time, since tying payout to the placement surviving some portion of the guarantee period, rather than to the invoice alone, stops the same risk from being taken twice.

Getting the Real Number

This is one of the more specific pieces of work I do with recruitment and placement businesses, building a fall off rate the business can actually trust from its own history, then designing a provision, and where it makes sense, a commission structure, around what the business genuinely experiences rather than an assumption borrowed from somewhere else.

If placement volume is growing and the monthly numbers never seem to move, that consistency is worth checking rather than simply enjoying. A rebate risk that never shows up in the numbers has not necessarily disappeared. It may simply never have been provisioned for in the first place.

Book a Discovery Call to talk through what your own placement history is actually saying.