Contingency versus retained is not a pricing question, it is a question of who is working on your role when it stops being easy. Contingency fits a single critical hire and carries no cost until someone starts. A retained arrangement fits a hard, quiet or confidential search, where exclusivity and a committed fee beat paying only on success.
Ask most contractors why they buy construction recruiting on contingency and the answer is some version of no placement, no fee. It is a good answer. It is also a description of your risk position, not of the recruiter’s behavior, and the two get conflated constantly.
Here is what it looks like from the other side of the desk. A contingency recruiter is carrying somewhere between eight and fifteen live roles, none of which are guaranteed to pay. Their rational move on any given Thursday is to work the role most likely to close soonest. If yours is a project manager for a metro they know, with a client who moves fast, you get real attention. If yours is a quiet replacement for a seated chief estimator in a market they have never worked, you are the role they get to when the easy ones are done.
Nobody is behaving badly in that story. The incentive is doing exactly what it was designed to do. The question worth asking is whether your particular role is one that model serves, and for a lot of construction hires, it is.
Is contingency or retained construction recruiting better for a general contractor?
Neither, in general. Contingency fits a single critical hire where you want zero risk until someone starts. Retained fits a hard, quiet or confidential search, where a dedicated consultant on an exclusive mandate is worth more than paying only on success. Match the model to the search in front of you, not to a preference.
The industry argues this as a moral question, which is why it never resolves. Retained firms say contingency produces volume without commitment. Contingency firms say retainers charge you for work that might not produce anything. Both critiques are accurate about the failure mode of the other model, and neither is an argument about your situation.
Your situation has three variables: how hard the seat is to fill, how exposed the search is if it becomes public, and whether you want exclusivity badly enough to commit money before placement.
What changes when a retainer is on the table?
Not the recruiter’s effort in some abstract sense. What changes is their capacity allocation. A retainer removes your role from the queue where it competes with every other unpaid mandate on the desk, which matters most on searches that are slow, quiet, or genuinely difficult.
That is the honest version, and it is narrower than the usual retained pitch. A retainer does not make anyone smarter about your market or better connected in your metro. It buys the ability to spend three weeks mapping a market where nobody is applying, without that time competing against a placement that could close on Friday.
Which is why the model question tracks so closely to the search type. Roles where the candidate pool is visible and active, and where several qualified people are open to a move, close fine on contingency. Roles where the right person is employed, not looking, and unreachable through anything you could post are the ones where the queue position decides the outcome.
How often you land in the second category is a market condition, not a preference. Of the 1,041 firms answering the salaried-hiring question in the 2025 AGC workforce survey, 91.7 percent reported difficulty filling salaried positions, monthly hiring and separations data is published in BLS JOLTS, and the construction volume those seats are staffed against is tracked by the Census Bureau. A contractor filling senior seats in that market is running hard searches, which is the condition the retained model was built for.
| Contingency Search | Retained Search | |
|---|---|---|
| When you pay | On placement only | In three stages: commencement, qualified shortlist and placement |
| Relative percentage | The higher of the two | The lower of the two |
| Best fit | One critical or one-off mandate | A hard, quiet or confidential mandate |
| Exclusivity | Exclusive as standard; non-exclusive at a higher rate | Exclusive |
| Guarantee | Replacement only | Replacement only |
Both models above are published, because a comparison that hides one side’s terms is a sales page. Fees are calculated on first-year cash compensation: base salary, plus any contractual or target bonus stated in the offer letter, plus any signing bonus. Discretionary bonus, commission, equity, allowances and benefits are excluded. Most US search firms quote percentages within a few points of each other, so the guarantee terms deserve as close a read as the rate. What the percentage is applied to matters as much as the number, which is the subject of our note on the recruiting fee base.
What does retained construction recruiting cost against contingency?
Retained Search carries a lower percentage than Contingency Search, so it costs less in total if the search completes. The price of that saving is commitment: most of the retained fee is billed before placement, at commencement and qualified shortlist, and it is owed even if the search stalls.
That comparison is worth running on your own numbers. Multiply the likely first-year pay by the percentage each model carries, and set the difference against the money you would commit before placement. The gap widens with salary: on higher-compensation seats the difference between the two percentages does more work. For a reference point, federal data puts the national median wage for construction managers at $114,990 and for cost estimators at $78,740, both as of May 2025 (BLS OEWS and cost estimators).
The arithmetic is not the whole case, though, and pretending otherwise would be the same mistake the fee-percentage argument makes. What a committed search buys that a queue position does not is time and exclusivity: a consultant who can spend three weeks mapping a market where nobody is applying, whether that is a Texas metro or a Southeast state, without another agency racing to the same candidate. That is a claim about process, and the process is what to test when you compare firms.
Against that, a retained search is a genuine commitment and it should read as one. Most of the fee is billed before a placement, a minimum fee applies whatever the salary, and a search that stalls still costs money, softened only by a credit if the search restarts. An easy role that a posting would fill should not be retained.
Which model handles a search that cannot be posted?
Either can, but the exposure is different. A contingency desk working your confidential role alongside a dozen open ones has more surfaces where the mandate can leak. A single named search partner working a defined list has fewer. The controlling factor is not the fee model, it is how many people know.
This is where the two questions get mixed up most often, so it is worth separating them. Confidentiality is a function of who is briefed, how candidates are approached, and whether the firm operates under presumed NDA. A retained arrangement tends to produce fewer touchpoints as a byproduct, not as a feature. You can run confidential executive search for mandates that cannot be publicly posted under either model, and the thing to interrogate is the process, not the invoice structure.
The interrogation is short. How many people at your firm will know the client’s name? Is the approach made by the person who will run the search, or by a researcher working from a script? What happens to the candidate list if we do not proceed?
A firm that answers those cleanly is a safe pair of hands on a quiet search whichever way it bills. A firm that cannot is not made safe by a retainer.
Who this is not for
A contractor filling a senior seat that a posting would fill should use contingency and stop reading here. The retained case rests on the search being hard, quiet or high-stakes enough that exclusivity pays for itself, and an easy, visible role is none of those. Committing most of a fee before a placement, on a role that would have closed anyway, is buying an option you have no use for.
It also does not apply to trades, craft labor, or temporary field staffing. Those are different markets with different economics, and Selah Talent Partners does not work in them.
And if your last three searches all closed quickly on contingency with candidates who are still in seat, the model is working. The reason to look at this at all is a pattern of searches that stall, go quiet, or produce shortlists that do not survive a technical conversation.
The useful question is not which model is better. It is which one your next hire actually needs, and whether the firm you are asking will tell you when the answer is the cheaper one. Our published terms are on pricing, and how to choose between construction search firms covers what to ask before you sign either.
Sources and further reading
- BLS OEWS construction manager wage data, SOC 11-9021
- BLS OEWS cost estimator wage data, SOC 13-1051
- AGC results national survey response data
- BLS JOLTS hiring and separations data
- Census Bureau construction spending data