Contingency vs Retained Engineering Search

Contingency: you pay on the hire

Under a contingency arrangement, the recruiter is paid only if you hire someone they introduced. No placement, no fee. The fee is typically calculated as a percentage of the hire's first-year compensation, and it falls due when the placement is confirmed.

The appeal is obvious: your financial risk is close to zero, and you can run more than one contingency firm on the same role. The consequence is equally structural. The recruiter carries all the risk, so they rationally prioritise roles most likely to close. If your role is harder than the other five on their desk, or if they suspect you are slow to decide, their attention goes elsewhere — and you will never be told that is what happened.

Retained: you pay in stages

A retained search is paid in instalments across the engagement, commonly an engagement fee up front with the balance tied to milestones or completion. The recruiter is being paid to run the search rather than to produce a hire, and the engagement is exclusive.

What that buys is committed effort and, usually, a genuine market map rather than whoever answered first. It is the sensible model for roles where the population of qualified people is small enough to enumerate — a principal-level structural engineer with a specific bridge speciality in a specific state, say — because the work is systematic identification rather than sourcing volume.

The trade is that you have paid before you know the outcome, and you have committed to one firm.

Which one your role needs

The useful question is not which model is better but how large the qualified population is and how hard it is to reach.

If there are hundreds of plausible candidates and the constraint is simply reaching them, contingency is efficient and you should not pay a retainer for it. If there are perhaps twenty people in the country who fit and half of them are not looking, contingency will underperform, because no recruiter working at risk will invest weeks in mapping a market they might not be paid for.

• Large qualified pool, standard requirements → contingency
• Small enumerable pool, senior or highly specialised → retained
• Confidential replacement of an incumbent → retained, for discretion
• Several similar roles at once → negotiate on volume in either model

The questions worth asking either way

Fee percentage is the least interesting term in a recruitment agreement. The terms that actually determine what you get are further down.

Ask what the guarantee period is and whether a departure inside it triggers a replacement or a refund — those are very different things. Ask what the fee is calculated on: base salary alone, or base plus bonus and other elements, which can move the number substantially. Ask how long candidate ownership lasts after an introduction. And ask how many recruiters will genuinely work the role, because "our team" and "one person between other searches" are both technically true descriptions of the same arrangement.

• What is the guarantee period, and is it replacement or refund?
• Is the fee on base salary or total first-year compensation?
• How long does candidate ownership last after an introduction?
• How many recruiters actually work the role?
• Who screens candidates before they reach me?

Where a network model sits

A crowdsourced or split-fee network is a third arrangement that is contingent in its economics but not single-desk in its reach: many specialist recruiters work the same brief, and the fee is still paid only on a confirmed hire.

That combination addresses contingency's core weakness — one recruiter's limited attention — while keeping your risk on the hire. The thing to verify is who screens. A network without a filter simply forwards you volume, which moves the sorting problem onto your desk. Ask specifically whether candidates are vetted before they reach you, and by whom.