Use retained search when the role is board-critical or confidential and the right people are not looking: you pay part of the fee upfront for one firm's full, mapped search. Use contingent search when the candidate pool is broad and active, because you pay only when someone is hired. Use an exclusive mandate when you want one committed firm for a fixed window without an upfront fee. Whatever you choose, judge the firm on how much of the market it actually maps.
Most founders pick a search model the way they pick a mobile plan. They look at the price, skim the terms and sign. Then the CTO search stalls at week ten and nobody can say why.
The model you choose decides who carries the risk, who gets paid when, and how hard the firm is motivated to work the search. It does not, on its own, decide how good the shortlist will be. That gap is where senior tech hires go wrong. Olofsson & Company, an executive search firm for AI and technology leadership, sees it from the other side of the table: the best Head of AI or VP Engineering for a given role is usually employed, well paid and not reading job ads. Any model that cannot reach that person is the wrong model, however cheap it looks.
Here is how the three models in the market actually work, and how to choose between them.
How do retained, contingent and exclusive search differ?
They differ on four things: exclusivity, when the fee is paid, how deep the search goes, and what happens if the role is hard to fill.
Contingent search works on a success basis. You pay nothing upfront, and the fee is owed only when a candidate the firm introduced accepts the offer and starts. Many employers run several contingent firms on the same role at once.[4] Fees are a percentage of salary, about 25% on average according to recruiters interviewed by BioHire, usually paid 30 days after the start date.[5]
Retained search is exclusive and funded in stages. The common pattern is a fee of 30% to 33.3% of first-year cash compensation, paid in thirds: one at engagement, one when the shortlist is presented, one when the candidate accepts.[5][3] Top-tier retained firms also set minimum fees of $75,000 to $100,000.[3] In return, the firm runs the whole process, and a retained shortlist typically holds three to five candidates, each interviewed in depth and referenced where possible.[2]
Exclusive search sits between the two. You commit to one firm for a defined period, typically 8 to 12 weeks, and pay no fee upfront. The firm still carries the financial risk, but it is the only firm on the brief, so it can afford to invest in it.[2] A close cousin is the hybrid or "container" model: a smaller flat fee upfront, and a contingent percentage if the role is filled.[5]
| Contingent | Exclusive | Retained | Hybrid ("container") | |
|---|---|---|---|---|
| Firms on the role | Often several | One, for a fixed window | One | One |
| Fee timing | On placement only | On placement only | Staged, from engagement | Small fee upfront, rest on placement |
| Typical fee | About 25% of salary[5] | Percentage on placement | 30% to 33.3% of first-year cash[3] | Flat fee plus percentage[5] |
| Firm's incentive | Speed | Quality within the window | Full process | Commitment without a full retainer |
| Best for | Broad, active candidate pools | Senior roles with an accessible pool | Confidential and board-level roles | Roles that need focus on a tighter budget |
Read the table as a map of who funds the effort. Contingent buys speed and costs nothing if it fails. Retained buys commitment and exclusivity, and you pay for it before anyone is hired. Exclusive and hybrid models try to get commitment without the upfront bill.
Which factors should decide the model for a senior tech hire?
Six factors do most of the work: how critical the role is, the compensation band, how passive the candidate pool is, confidentiality, urgency, and how comfortable you are with one firm.[1] Score your role against each before you speak to anyone.
- Criticality. A CTO or Chief AI Officer shapes the company for years. A wrong hire there costs far more than any fee, so pay for depth.
- Compensation band. Percentage fees rise with salary. One common rule of thumb says retained above $200,000 in total compensation and contingent below it.[1] Treat that as a starting point, not a rule.
- Passivity. Senior AI and engineering leaders are rarely on the market. If the person you need is in-seat and happy, a model that depends on active candidates will not find them.
- Confidentiality. Replacing a sitting CTO who has not been told, or building an AI team ahead of an announcement, needs a closed process. Where confidentiality is a hard requirement, retained is the only structural fit.[1]
- Urgency. Contingent firms are paid to move fast. But speed from a shallow pool gives you a fast hire from a narrow slate. Decide which risk you can live with.
- Exclusivity tolerance. If you want several firms competing, you have chosen contingent by definition. If you want one firm accountable, you need exclusive, hybrid or retained.
The pin.com analysis makes a fair point here: most advice collapses these six into a binary salary check.[1] Don't. A $180,000 Head of Machine Learning in a thin market can be a harder search than a $300,000 COO.
When does contingent search break down?
Contingent search works well for roles with a broad, findable pool, and plenty of tech hiring fits that description. It breaks down in three predictable places, and none of them is the recruiter's fault. They follow from how the model pays.
- Hard roles get dropped. A contingent recruiter on a difficult, thin-market role faces a real economic problem: the harder the search, the lower the expected return on the hours invested.[4] Firms that are not paid for search time may abandon difficult searches without notice in favour of easier work.[3]
- Confidential roles leak. Agencies juggle many mandates and show the same candidates to many clients. Nobody has to be careless for word to get out; the format itself spreads it.[1]
- Several firms crowd the market. Running three agencies feels like more coverage. In practice, the same candidates hear about your role from several recruiters, and senior candidates often withdraw from a crowded process.[2] For a CTO search, that can cost you the one person you wanted.
Our view is blunt. Do not put an unannounced succession search, or a niche AI leadership role, through a model that is designed to circulate candidates widely and walk away from hard work. Fix the incentive first.
What should you know about fees before you sign?
Ask how the fee is calculated, and what it covers, before you compare headline percentages.
Under the standard retained schedule, two-thirds of the fee is committed before anyone accepts an offer.[5] That is the price of a funded process. For a modest salary band, the $75,000 to $100,000 minimums at top-tier firms can make retained search uneconomic.[3] Paying for a full market study on a role with a broad, active pool is money you do not need to spend.
Watch the incentive built into percentage pricing. When the fee is a share of the candidate's salary, a higher package means a higher fee, which gives the search firm a reason to push compensation up.[3] Keep the offer negotiation in your own hands, and agree the fee basis (base only, or base plus bonus) in writing. For a fuller breakdown of what senior AI hires cost, see our guide to what it actually costs to hire AI leadership.
One firm or several? One role or a whole team?
For a single senior hire, use one firm. Running several in parallel rarely works.[2] It splits effort, multiplies outreach to the same people and leaves no one accountable when the search stalls.
Team builds are different, but the answer is the same. If you are hiring a Head of AI plus four senior engineers, or standing up a new engineering hub, one partner running parallel pipelines will know the market, the story and the comp benchmarks across every seat. Splitting the roles across agencies means five firms pitching your company five different ways. Negotiate the team as one engagement and agree how the roles are sequenced.
How do you judge a search firm under any model?
The model settles how the fee is paid. It does not settle how deep the search goes. What you are really buying in a committed search is discipline: a scoped brief, a documented market map, written candidate profiles and structured interviews.[2] Ask for those, whichever model you sign.
Four questions work under every model:
- How many people will you map, and how many are passive? A shortlist drawn from 40 active candidates is a different product from one drawn from 400 mapped leaders.
- Who does the work? Find out whether the consultant you meet runs the search or hands it to a junior team.
- What will you show me at week three? A market map and early calibration, not just CVs.
- What happens if the role changes? Briefs shift, especially in AI. Agree how that is handled.
This is where Olofsson & Company puts its weight. Its proprietary AI platform maps the full qualified market for each search, including the passive AI and technology leaders who never see a job ad, and its specialist consultants judge fit, motivation and timing. The platform makes the mapping faster. The consultants decide who is worth a conversation. See how that works on the Olofsson expertise page, or read seven signs you need an executive search partner.
Frequently Asked Questions
Is retained search always better for a CTO hire?
No. Retained fits when the role is confidential, board-critical or the pool is thin. For a CTO at an early-stage company with a well-known network, an exclusive mandate with a committed firm can work just as well. What matters is that someone maps the whole market and is accountable for the result.
Can I switch from contingent to exclusive partway through a search?
Yes, and it often helps when a contingent search has stalled. Agree a fixed window, typically 8 to 12 weeks, and pull the role from other agencies so the exclusive firm can approach candidates with a clean story.[2]
Sources
- pin.com: Six decision criteria, the $200K rule of thumb, and why contingent search structurally leaks confidential mandates
- oplu.com: Exclusive mandates of 8 to 12 weeks with no upfront fee, three to five candidate retained shortlists, and why parallel firms rarely work
- tgsus.com: 30% to 33.3% retained fees, $75,000 to $100,000 minimums, search abandonment, and the conflict in percentage-based fees
- stanleykent.com: How contingent search is paid and the economics of thin-market roles
- biohire.org: 25% average contingent fees, milestone retained schedules, and the hybrid container model