The retained-versus-contingent question gets debated as if it were a question about fees. It isn't. The fee structure is the visible part — the part that ends up in the contract, the part finance teams compare on spreadsheets. The actual difference is about something else: who is taking the risk, what behaviours that creates, and whether the search you're paying for is the search you think you're getting.
Most companies, even sophisticated ones, choose the wrong model for the wrong reasons. They choose retained because it sounds serious. They choose contingent because it sounds efficient. Both choices, made on those grounds, tend to produce worse outcomes than the choice would have produced if it had been made on the basis of what the search actually requires.
What each model is
A contingent search is, structurally, a sales transaction. The recruiter is paid only if a hire is made. They have no obligation to do the search exclusively, no obligation to do it at all if other opportunities look more promising, and no recourse if the client decides not to hire after candidates have been presented. Their incentive is to find someone, anyone, who is hireable, quickly, before another recruiter does the same. Multiple firms may be working the same search simultaneously, none of them aware of the others.
A retained search is structurally a service engagement. The recruiter is paid in instalments — typically a third on signing, a third at shortlist, a third on placement — regardless of whether a hire is ultimately made. They work exclusively, taking the search off the market to other firms. Their incentive is to do the work properly, because the work itself is what's being paid for. The placement is the outcome, not the trigger.
These are not two flavours of the same thing. They are different products, structured to solve different problems.
What contingent search is actually good for
The contingent model works well in one specific situation: when the role is reasonably defined, the candidate pool is large and active, and speed matters more than calibration. A mid-level engineering hire in a major tech market. A regional sales rep. A finance manager in a city with twenty thousand of them. In these situations, the search is essentially a sourcing problem. The candidates exist, they are findable, and the question is which firm can put the right CV in front of the right hiring manager first.
In that environment, paying only for outcomes is rational. The recruiter's incentive — to move quickly — aligns with the client's interest. The competition between firms drives speed. The work is not particularly complex, and the absence of deep calibration matters less because the candidate pool itself filters for adequacy.
Contingent search struggles, badly, outside this zone. The further a search moves from "well-defined role, large active pool, speed matters most" — toward "ambiguous role, narrow pool, calibration matters most" — the worse the contingent model performs.
Why contingent fails at senior levels
Senior hires are, almost by definition, not standardised. The pool of people who could be the next CFO of a 200-person company is not large, not actively looking, and not adequately described by any job specification. The work required to find them, persuade them, and assess them is not sourcing work. It is research, relationship-building, calibration, and judgement.
The contingent model cannot pay for this work, because it doesn't pay for work at all. It pays for placements. A contingent recruiter doing senior search is being asked to do weeks of research with no guarantee of any return. Rationally, they don't. They do the version of the search that fits their economics — they call people they already know, they recycle candidates from other searches, they present whoever can be persuaded to take a call quickly. Whether this matches the actual brief is secondary.
Worse, when multiple contingent firms are working the same senior search, candidates start receiving multiple approaches from different recruiters for the same role. This is, to a senior candidate, an immediate signal that the search is not being run seriously. The strongest people disengage. The search ends up with the candidates who didn't notice or didn't mind — which is, again, not the population the client thinks they're hiring from.
Why retained search isn't automatically the answer
The instinct to respond to all this by saying "use retained search for senior roles" is correct as a default but wrong as an absolute. Retained search has its own failure modes, and a lot of retained engagements deliver bad outcomes that the structure of the engagement makes hard to see.
A retained firm has been paid the first third before any work begins. The second third arrives at shortlist, which is a milestone the firm controls — they decide when the shortlist is "ready," and the definition of "ready" is, in practice, "the moment we have three people who can be presented without embarrassment." If the search is hard, the temptation to lower the standard for what counts as a presentable shortlist is real, and the financial structure rewards it.
The placement fee, the final third, is contingent on a hire — but by that point, the firm has been paid two-thirds regardless. The financial pressure to actually place is much weaker than it appears. A retained firm that does sloppy work still gets paid for two-thirds of the engagement, and many do.
The other retained-search failure mode is the opposite: the firm has been paid up front and now needs to look busy. They produce activity — long progress reports, presented candidates who don't fit, market intelligence the client didn't ask for — that creates the appearance of work without necessarily moving the search forward. Sophisticated retained clients can tell the difference. Less sophisticated ones often can't, and assume the activity is the value.
The hybrid models, and why they exist
Between pure contingent and pure retained sit a series of hybrid arrangements that have emerged because both of the pure models have problems. Engaged search — sometimes called retainer-light — pays a smaller fee up front to begin work, with the bulk of the fee on placement. This aligns incentives better than contingent for harder searches, while keeping more of the risk on the recruiter than full retained.
Fixed-fee arrangements decouple the fee from the salary entirely, charging a flat rate regardless of the hire's compensation. These tend to be favoured by clients who feel the percentage-of-salary structure overpays the recruiter for higher-comp hires that aren't proportionally harder to find.
Container fees split a search into a retained research phase and a contingent placement phase, paying for the front-end work explicitly while leaving the placement at risk.
Each of these exists because the market is trying to solve the same underlying problem: how do you pay for work that needs to be done thoroughly, without paying for work that isn't done at all? There is no clean answer. Every fee model rewards some behaviour and punishes others, and the question is which trade-off matches the search at hand.
How to choose
The honest framework for choosing between models isn't about seniority or compensation level. It's about three questions:
How well-defined is the role? If the brief is clear, agreed by all stakeholders, and matches a real market category, contingent or engaged is workable. If the brief is genuinely unclear — if part of the work is figuring out what you're actually hiring for — only retained makes sense. Contingent recruiters cannot afford to do the discovery work.
How active is the pool? If the people you want are likely to be on the market, contingent works. If you need to reach people who are not looking and persuade them to consider the role, retained is the only model that pays for the persuasion work. Passive candidate sourcing is expensive in time, and time is the resource the contingent model cannot fund.
How exposed is the search to confidentiality risk? If the search is genuinely confidential — replacing a current incumbent, exploring a sensitive market entry, hiring against a competitor — multiple contingent firms talking to multiple candidates is a leak waiting to happen. Retained is structurally safer because the search is exclusive.
If two or three of these point toward retained, the contingent model will probably waste your time. If two or three point toward contingent, paying retained fees is probably overpaying for a service you don't need.
The questions worth asking either way
Whichever model you choose, there are questions worth asking that most clients don't ask, and that change the quality of the work either way.
Of a contingent recruiter: how many other searches are you working on right now? How many of them are similar to ours? If two of your clients want the same candidate, who gets her?
Of a retained recruiter: what happens if you don't find the right candidate in four months? Do you keep working? Do we get a partial refund? What is the off-limits agreement — which of your other clients' companies can you not recruit from? How many other retained mandates does the partner running our search have right now?
These questions matter more than the headline fee. A retained firm with a partner running fourteen simultaneous mandates is not giving you the attention the fee structure implies. A contingent firm working two searches is, in practice, giving you most of the focus of a retained engagement.
The market is not as binary as it pretends
The other thing worth knowing is that the line between retained and contingent has been blurring for years, and most of the largest firms now do both — sometimes within the same client relationship. The terms in the contract matter less than the actual practice. A "retained" engagement with a firm that has lost interest is worse than a contingent engagement with a recruiter who has reason to focus.
What you are actually buying is the attention, the judgement, and the network of the specific people doing your search. The fee model determines the financial structure of the engagement, but it does not, by itself, determine the quality of the work. That is determined by who is doing the work, how much of their time you are getting, and what their incentives actually look like in practice.
What this means before signing
The decision is not retained-versus-contingent in the abstract. It is: for this specific role, with this specific level of definition, in this specific market — which structure best aligns the recruiter's behaviour with what we need? And then, separately: of the firms that work in that structure, which of them is actually going to do the work, with the right people on it, at a level of focus we can verify?
Both questions are answerable. Most clients only ask the first.