Cross-industry benchmarks converge on the same picture: hiring is getting slower, applicant volumes are rising, and the gap between process length and candidate availability is widening every year.
Financial services runs longer. The sector averages 40 to 48 days, with UK business and finance roles taking 5.5 to 6 weeks. London financial services saw a 12% quarter-on-quarter hiring increase from Q4 2024 to Q1 2025. CFO searches typically run 60 to 90 days, with many retained searches taking 8 to 14 weeks. The role scope has expanded into ESG, digital transformation, and data strategy, narrowing the candidate pool and complicating assessment.
The 10-day window. LinkedIn research found that top candidates are typically off the market within 10 days. The average hiring process takes 41 to 45+ days. This gap is where organisations lose their strongest candidates to faster competitors. 52% of companies report losing top talent to faster-moving rivals.
A slow process loses the best candidates. A rushed process produces bad hires. The evidence reveals a striking paradox at the centre of both failures.
The cost of a bad hire. SHRM puts replacement cost at 50 to 200% of annual salary depending on seniority. For senior and strategic roles, the full organisational impact can reach five times annual salary. For a £150,000 role, that ranges from £450,000 to £750,000. Executive average cost per hire has reached $35,879, up 113% from 2017.
The paradox is structural. More qualified candidates are less willing to tolerate prolonged processes. Among candidates who declined in favour of a competitor, 71% reported the competitor moved at least two weeks faster at the offer stage. Offers extended within one week of the final interview have acceptance rates approximately 20% higher than those extended after two or more weeks.
The most data-backed finding in hiring research: four interviews capture virtually all the predictive signal. A fifth adds less than 1% improvement.
Google's transformation. Todd Carlisle, Google's organisational psychologist, analysed five years of internal hiring data. The result: Google cut from 15 to 25 interviews per candidate down to four, reducing median time-to-hire from 90 to 180 days to 47 days, saving approximately 50,000 engineering hours annually.
Practitioner consensus. ERE.net recommends three rounds: pre-employment screening plus phone interview, face-to-face with hiring manager and key stakeholders, and a final interview if needed. Their position: additional rounds after the third yield negligible new information. Most companies follow 3 to 5 stages depending on seniority. If a company cannot decide after five rounds, the system is broken.
Schmidt and Hunter's 1998 landmark meta-analysis, updated by Sackett et al. in 2022, provides the definitive validity hierarchy for selection methods. The findings have transformed which tools deserve investment.
| Method | Validity (r) | Practical Implication |
|---|---|---|
| Structured interviews | .51 → top-ranked | The best tool available. Use at every stage. Sackett et al. 2022 confirmed as potentially the single strongest predictor. |
| Cognitive ability (GMA) | .51 → reduced by .10–.20 | Still valuable but no longer the gold standard it was assumed to be. |
| Work sample tests | .54 → .33 | Strong for task performance but does not predict contextual performance. |
| Integrity tests | .41 | Predict counterproductive behaviours: theft, absenteeism, disciplinary problems. |
| Unstructured interviews | .38 | Roughly half the validity of structured. The default at most firms. |
| Conscientiousness | .31 | The only Big Five trait that predicts across all occupational groups. |
| Situational judgement tests | .26–.34 | High face validity, moderate predictive validity, low adverse impact. |
| Reference checks | .26 | Better than education or experience but substantially weaker than interviews. |
| Job experience (years) | .18 | Poor predictor. Length of tenure tells you almost nothing about capability. |
| Education (years) | .10 | Nearly worthless as a standalone predictor. |
Timing determines whether an assessment filters effectively or alienates strong candidates. The MBTI should never be used in hiring. Here is what the evidence supports and when to deploy it.
The MBTI question: never use it. The Myers-Briggs Company itself states the MBTI should not be used in hiring and selection. Up to 50% of people are reclassified into different types on retest. It has no demonstrated predictive validity for job performance. The APA has raised concerns. MBTI is designed for self-reflection and team development only.
| Fairness Tier | Assessment Methods | Candidate Perception |
|---|---|---|
| Most preferred | Interviews and work sample tests | High face validity. Candidates see a clear connection between the assessment and the job. |
| Moderately preferred | CVs, cognitive tests, references, biodata, personality inventories | Accepted when the purpose is explained. Positive testing experiences boost subsequent performance by 30%. |
| Least preferred | Honesty/integrity tests, personal contacts, graphology | Low face validity. Candidates who perceive unfairness are 25% less likely to refer others. |
The diversity dimension. Cognitive ability tests produce approximately 1.0 standard deviation difference between demographic groups, the largest adverse impact of any common selection method. Berry et al. (2024) argued this trade-off should be reframed given the revised, lower GMA validity estimates. Structured interviews show nearly a third less bias than unstructured interviews: standardised questions and rubrics reduce bias effects from d=.59 to d=.23. The BPS published new guidance on Neurodiversity and Psychometric Testing in November 2024, covering proactive testing preparation and reasonable adjustments for neurominorities. Under the Equality Act 2010, employers must ensure reasonable adjustments for disabled candidates taking psychometric tests.
Finance and insurance have the highest candidate resentment rates globally. The process itself is a selection variable, whether you design it that way or not.
Passive candidates operate on a different calculus. 70% of the global workforce are passive candidates. At the executive level, that figure rises to 70 to 80%. These are people who are not actively job-seeking but would consider the right opportunity. Creating genuine interest takes 4 to 5 weeks, but once interest exists, the window to convert it into action is narrow. 30% of accepted offers for executive-level hires fail due to counter-offers from the current employer.
The UK notice period factor. UK senior candidates typically serve 1 to 3 months' contractual notice, compared to the US convention of 2 weeks. The total elapsed time from first interview to day one can easily exceed 4 months. This makes process speed before the offer even more critical: every week wasted in the interview process is a week added to an already long conversion cycle.
The most robust and replicated finding in all of personnel selection research. Structured interviews are the strongest single predictor of job performance. It is also the most ignored finding in practice.
Panel interviews. McGill University research found structured panel interviews have the highest predictive validity, with accuracy improving 20 to 30 percentage points compared to one-on-one. An effective alternative is the "team method": two interviewers, one asking questions for two-thirds of the time while the second takes notes, then the second conducts focused follow-ups. This preserves the conversational feel while gaining multi-assessor benefits.
Culture fit: the most dangerous stage. When interviewers assess "culture fit" through unstructured conversation, research shows they are measuring whether the candidate reminds them of themselves, not whether the candidate shares organisational values. The consensus: replace "culture fit" with "culture add" and assess values alignment only through structured questions with scoring rubrics.
Decision-making gates. The recommended hybrid: use hard knockout criteria at screening only for genuine non-negotiables, then accumulate information through subsequent stages with a final holistic evaluation. Oxford's Review of Economic Studies found that the sequential nature of interviews creates assessment distortion, where evaluators judge the current candidate relative to the previous one rather than against an absolute standard. Independent scoring before group discussion is critical.
Based on the convergence of evidence across all research streams, the optimal hiring process for senior finance roles follows four stages completed within 3–4 weeks from first interview to offer.
UK notice periods add to the elapsed time. UK senior candidates typically serve 1–3 months of contractual notice, compared to the US convention of two weeks. This means that even after offer acceptance, employers face a 4–12 week gap before the candidate starts. The total elapsed time from first interview to day one can easily exceed four months. This makes process speed before the offer even more critical: every week wasted in the interview process is a week added to an already long conversion cycle.
The recruiter's structural advantage. A recruiter who can tell a passive CFO exactly what the process looks like, how many stages, who they will meet, and a committed timeline, has a structural advantage over every firm that says "we'll be in touch." Transparency about process is a differentiator precisely because so few firms offer it.
Implications for hiring process design. The Employment Rights Act 2025 shifts the cost-benefit analysis firmly toward investing in rigorous selection upfront rather than relying on the ability to dismiss within a long qualifying period. For buy-side firms hiring regulated individuals (compliance officers, MLROs, executive directors), the FCA's regulatory layer is unavoidable and makes the interview process timeline even more critical. Right to work checks must be completed before an employee's start date, with fines of £45,000 per illegal worker for a first breach and £60,000 for repeat offences.
The FCA has set comply-or-explain targets for listed companies: at least 40% of board members should be women, at least one senior board position held by a woman, and at least one board member from a minority ethnic background. Only 12% of fund managers are female. These targets create additional pressure on firms to ensure their hiring processes do not inadvertently screen out diverse candidates through biased assessment methods.
Gartner identifies four trends reshaping talent acquisition through 2027. AI-first hiring for high-volume roles, early-career strategy rethink, candidate fraud prevention (GenAI-generated applications), and recruiter role transformation. Their prediction: by 2027, 75% of hiring processes will include AI proficiency certifications or tests. Only 19% of employers worldwide believe their hiring process is excellent.
The reality gap in skills-based hiring. 60% of companies adopted some form of skills-based hiring by 2024, up from 40% in 2020. 79–90% of organisations implementing it reported reduced mis-hires. Yet 43% of job postings requiring bachelor's degrees could be performed with alternative credentials. The gap between aspiration and practice is vast. Skills-based assessment methods (work samples, structured interviews, technical exercises) are demonstrably more predictive than proxies like degrees and years of experience. But saying you do skills-based hiring and actually doing it are different things.
The practical position for specialist recruitment. For senior finance hiring in the London buy-side market, where candidate pools are small, relationships are long-term, and every interaction shapes reputation, AI screening tools carry more risk than benefit. The strongest evidence still points to structured interviews conducted by skilled humans as the single best predictor of job performance.
AI may have a legitimate role in high-volume screening at the application stage. But for the calibre of hire this market demands, the human judgment layer is irreplaceable. Efficiency and accuracy are different things, and for senior hires the cost of a false positive dwarfs any screening time saved.
The regulatory environment is tightening. The EU AI Act (entered into force August 2024) classifies all AI used in recruitment as high-risk. Since February 2025, biometric categorisation, emotion recognition in the workplace, and social scoring are prohibited. Full obligations for high-risk systems apply from August 2026, requiring mandatory risk assessments, bias testing, human oversight, and transparency disclosures. The Act applies extraterritorially to any company whose AI outputs are used in EU hiring contexts. The UK Government published "Responsible AI in Recruitment" guidance in March 2024, taking a less prescriptive but directionally aligned approach.