Backdoor Hires and Missed Fees: What We Found in 56 Recruitment Agency Databases

A backdoor hire happens when a client employs a candidate an agency introduced, but the agency never receives the fee it was owed. The introduction was valid, the candidate is in the role, and no invoice was ever raised. It also goes by missed fee, fee leakage, or circumvention.

Most agency directors assume it happens to them occasionally. Almost none have measured it, because measuring it means checking every CV send against where that candidate actually ended up, sometimes two years later. No consultant has the hours for that.

Illumini ran the exercise across 56 recruitment agency databases. 89% of them contained at least one missed fee.

Key findings

• 89% of the 56 recruitment agency databases analysed contained at least one missed fee

• The median agency had three missed fees recovered in a single year

• Missed fees occurred at a rate of one in every 2,139 CV sends

• The average fee recovered on a successful claim was £13,200

• 84% of cases originated on the client side, 16% were agency process failures

What is a backdoor hire?

A backdoor hire is a placement where an agency made a valid introduction, the client hired the candidate, and no placement fee was paid. It is a different thing from a candidate ownership dispute, where two agencies claim the same introduction, and from a rebate, where a fee was paid and later clawed back.

We counted a case where four things were true. A CV was sent to the client. The candidate subsequently worked at that client. No invoice was received. No other agency was paid first.

That includes cases where the agency never issued terms of business, or issued them and never chased. Those are still fees nobody collected, and they are the ones an agency can fix without a difficult conversation.

How often do backdoor hires happen?

In Illumini's analysis of 56 recruitment agency databases over a one-year period, 89% contained at least one missed fee, with a median of three per agency.

The rate is the more useful number, because it scales to any agency regardless of size. A missed fee occurred once in every 2,139 CV sends.

CV sends are speculative introductions rather than deal attempts, so the volumes are large. The largest agencies in the sample send at six-figure volumes annually. An agency sending 20,000 would expect about nine missed fees annually. At 100,000 sends, closer to 47.

The distribution matters as much as the rate. Cases were spread thinly across nearly every database rather than concentrated in a handful of agencies with bad clients. This is not a small number of firms being systematically cheated. It is drift, accumulating quietly across a large client base, showing up almost everywhere in ones and twos.

What is a missed fee worth?

Where an agency pursued a missed fee and claimed it successfully, the average amount recovered was £13,200.

At the median of three cases a year, an agency that recovered all three would collect around £40,000. An agency sending 20,000 CVs a year, expecting nine cases, would be looking at roughly £120,000 in recoverable fees.

Whether a case turns into cash depends on three things: whether terms of business were in place when the introduction was made, whether the hire falls inside the validity period those terms specify, and whether the agency wants to open a fee dispute with a client it still trades with. That last one decides more cases than the first two.

Why do backdoor hires happen?

Around 84% of cases were client-side. The remaining 16% were agency process failures.

Agency-side means terms were never sent, never accepted, or an introduction went quiet and nobody followed up. Entirely within the agency's control, and the cheapest 16% of the problem to eliminate.

Client-side covers the rest. The candidate was hired without the agency being told. The hiring manager who received the CV left before the hire completed. The CV was forwarded internally and the origin dropped off it somewhere along the way.

A hiring manager who genuinely forgets where a CV came from eight months later leaves the same trace as one who acted deliberately. That is why we stopped using the phrase backdoor hire internally. It names a decision, and what the record actually shows is an outcome. The commercial consequence is identical either way, which is the part that matters to whoever is running the P&L.

Does it count if the client hires into a sister company?

Around 23% of client-side cases involved the candidate being hired by a different legal entity within the same group as the client the CV went to.

This is its own problem with its own fix. The CV goes to one operating company and the candidate turns up at a sister company or the parent. Sometimes that is genuine, because the entities run separate hiring processes and separate PSL arrangements. Often it is just where the introduction gets lost.

Whether the introduction fee is payable comes down to how the terms of business define the client. Terms naming a single trading entity leave the agency exposed the moment a hire happens anywhere else in the group. Terms extending the definition to subsidiaries, parent companies and entities under common control close it.

Any agency working with large multi-entity clients should know which of those two positions its contract takes, because a fifth of client-side leakage sits in that gap.

How do agencies find missed fees?

Manual detection does not survive contact with a real database. Establishing that a candidate ended up at a client means knowing where that candidate works now, then matching it back to a CV send that may be two years old.

Automated detection compares an agency's own send history against current employment data and flags where the two line up. It produces a list of cases to look at. Each one still needs a person to check the terms, the validity period, and whether another agency got there first.

The sensible starting point is a historical sweep rather than ongoing monitoring. A one-year lookback on data the agency already holds will surface whatever is sitting there, and gives a real number to decide against.

How this research was carried out

Sample: 56 recruitment agency databases, comprising Illumini clients globally. All had more than 50 employees and typically sent over 10,000 CVs a year.

Period: each database was analysed on a one-year lookback from the date it was run. The cases in this research were collected across two years of running that analysis.

Method: CV send records were matched against current employment data to identify candidates working at clients they had been introduced to. Billing status and prior agency involvement were confirmed with each agency directly.