Two different products
A contingent recruiter is paid only if their candidate is hired, and typically competes with other agencies and the client's own channels. Their economics reward speed and volume: the first credible CV across the line often wins. This model serves the middle of the hiring market well, where roles are numerous and candidate pools are broad.
A retained firm is engaged exclusively and paid in stages regardless of outcome. Its economics reward thoroughness: it is contracted to map the entire relevant market, approach the strongest individuals whether or not they are looking, and provide the board with evidence rather than options.
What changes for candidates
Senior candidates experience the two models very differently. A contingent approach is often one of several identical-sounding calls about the same role, which cheapens the opportunity and signals that the employer is shopping. A retained approach carries weight precisely because it is singular: the candidate knows the organisation has committed to a serious, confidential process, and that their own candidacy will be handled with discretion.
For passive candidates — the ones boards most want — this distinction decides whether they take the call at all.
Assessment depth
Because a retained firm is paid for the process rather than the placement, it can afford to slow down where it matters: competency-based interviewing, structured referencing, psychometric and leadership assessment, and honest counsel when the evidence points away from a popular candidate. Contingent processes rarely carry that infrastructure, because the commercial model cannot fund it.
When each model is right
Contingent search remains a sensible choice for mid-seniority roles with large, active talent pools, and where several hires are needed quickly. Retained search is the appropriate instrument for board appointments, C-suite roles, succession-critical positions, confidential replacements, and any mandate where the cost of a wrong appointment dwarfs the fee. The question to ask is not 'what does it cost?' but 'what does failure cost?'