Leadership Assessment: What Predicts Executive Success
Structured interviews predict senior performance twice as well as unstructured ones. Which leadership assessment methods work, and which boards should drop.
Leadership assessment works when it tests candidates on the job itself, not on how they come across in a meeting. The strongest evidence, Sackett and colleagues' 2022 re-analysis of decades of selection research, ranks structured interviews first at a validity of .42, followed by job knowledge tests, biodata and work samples. Unstructured interviews score .19 and years of experience .07. Most board-level processes still lean on exactly those two.
What is leadership assessment in executive hiring?
Leadership assessment, in executive hiring, is the structured evaluation of candidates for a senior role against criteria the board sets before it meets anyone. It combines interviews, work simulations, psychometric instruments and referencing into one body of evidence. The aim is prediction: which candidate will perform in this role, in this business, over the next three to five years.
Leadership assessment is not the same as two things that share its vocabulary. The GMAC Executive Assessment is an admissions test for executive MBA programmes, which is why so much search traffic for "executive assessment" is about practice papers. The 360-degree leadership assessment is primarily a development tool for people already inside an organisation. Both have their place. Neither answers a board's question about an external CEO or CFO candidate.
The distinction matters because most senior hiring in India still runs on a sequence of conversations: a founder lunch, a dinner with the chair, two or three panel meetings. Each feels rigorous. Very little of it is structured, scored or compared across candidates. The research on what that costs is unusually clear.
Which leadership assessment methods actually predict performance?
Structured interviews predict job performance best, according to the most rigorous recent evidence. Paul Sackett, Charlene Zhang, Christopher Berry and Filip Lievens re-analysed decades of selection meta-analyses in the Journal of Applied Psychology in 2022, correcting a statistical overcorrection that had inflated earlier figures. Their revised operational validities, where 1.0 would be perfect prediction, are:
- Structured interviews: .42 (previously estimated at .51)
- Job knowledge tests: .40
- Empirically keyed biodata: .38
- Work sample tests: .33
- Assessment centres: .33 (revised in 2023 from .29 for managerial roles)
- General cognitive ability tests: .31 (previously .51)
- Conscientiousness, measured in general terms: .21
- Unstructured interviews: .19 (previously .38)
- Years of job experience: .07 (previously .18)
Two points follow for boards. First, the conversational interview that dominates senior hiring predicts less than half as well as a structured version of the same meeting. Second, the thing most executive CVs are built around, years of experience, barely predicts performance at all.
Sackett's team is candid that these are averages across many jobs, not executive roles alone. The 80% credibility interval for structured interviews runs from .18 to .66, so a badly designed structured interview can do little better than an unstructured one. Their conclusion is the right frame for a nomination committee: "one might propose structured interviews as the focal predictor against which others are evaluated."
Why do boards still rely on unstructured interviews?
Boards rely on unstructured interviews because they feel like judgement, and judgement is what directors are appointed to exercise. A good conversation with a confident candidate produces a strong impression, and a strong impression is easy to mistake for evidence. The difficulty is that the traits which dominate those conversations are not the ones that predict results.
The CEO Genome research by the leadership advisory firm ghSMARTpublished in Harvard Business Review in 2017, analysed in-depth assessments of 17,000 executives. It found that charisma, confidence and pedigree had little bearing on CEO success. What distinguished high performers were four behaviours: deciding with speed and conviction, engaging stakeholders for impact, adapting proactively, and delivering reliably. None of those is reliably visible over dinner.
The cost of getting it wrong is not abstract. The same HBR article cites Conference Board data showing that about a quarter of CEO departures in the Fortune 500 between 2000 and 2013 were involuntary. Our analysis of why executive hires fail found the same pattern at CXO level: most failures founder on politics, culture and people rather than on technical competence.
What should a CEO or CXO assessment actually measure?
A CEO or CXO assessment should measure the few capabilities the role's next three years will actually demand, written down before the long list is drawn. That means translating strategy into a scorecard: the outcomes the executive must deliver, the competencies those outcomes require, and the organisational conditions they will have to work within.
The best evidence on which capabilities matter comes from Steven Kaplan, Mark Klebanov and Morten Sorensen's study of CEO candidates in buyout and venture-backed companies, published in the Journal of Finance in 2012. Candidates were assessed on more than thirty individual abilities. Investors were more likely to hire candidates with strong general ability, but subsequent company success was more strongly related to execution skills than to team-related and interpersonal skills.
That finding does not make interpersonal skill irrelevant. It means boards tend to overweight it in the room, because warmth and fluency are what a meeting surfaces most easily. A scorecard corrects for that by making each assessor score execution evidence separately from impression: what the candidate actually delivered, with what resources, against what obstacles.
For a CEO appointment, the scorecard belongs with the nomination and remuneration committee before any search opens. We set out the full board process in how to hire a CEO, and the internal side of the same question in CEO succession planning.
Do psychometric tests work for senior leadership hiring?
Psychometric tests work for senior leadership hiring as one input, not as a verdict. Cognitive ability tests carry a revised validity of .31 in Sackett's 2022 analysis, and a meta-analysis of 113 twenty-first-century studies by Griebe and colleagues, reported by Sackett's team in 2023, puts the figure lower still, at .23.
There is a practical reason the signal weakens further at the top. By the time a candidate reaches a CXO shortlist, they have been filtered on ability many times over: by education, by successive promotions, by the search itself. The differences between shortlisted candidates on a reasoning test are small, and a test cannot discriminate much within so narrow a band. Statisticians call this range restriction.
Personality instruments behave in a similar way. Conscientiousness measured in general terms shows a validity of .21; measured in a work context, with items framed as "at work, I..." rather than "I...", it rises to .25. For senior hiring, the useful output of a personality instrument is a set of hypotheses for the interview panel to test, not a score to rank candidates by.
Are assessment centres worth it for executive roles?
Assessment centres are worth it for executive roles when the exercises mirror the actual job. Sackett's team revised their validity estimate for assessment centres from .29 to .33 in 2023, after accounting for the lower reliability of performance ratings in managerial jobs. That places them level with work samples among the five strongest predictors.
The traditional assessment centre, with in-tray exercises, group discussions and role-plays over two days, rarely suits a CXO candidate who is currently employed and discreet about the search. The executive equivalent is a compressed, role-specific work sample. A CFO candidate reviews two years of redacted accounts and presents the three questions they would put to the audit committee. A COO candidate walks the board through how they would sequence the first 90 days across the plant network.
A work sample of this kind shows how the candidate thinks about this business rather than business in general, which is exactly what it is for. Confidentiality governs the design: material shared should be proportionate, covered by a non-disclosure agreement, and limited to what the exercise needs.
Does a 360-degree leadership assessment help when hiring?
A 360-degree leadership assessment helps when the candidate is internal and does not work for external candidates. The method gathers structured ratings from a person's manager, peers and direct reports, which is only possible for someone already inside the organisation. For an internal successor, a recent 360 is valuable evidence and should sit alongside the external field.
For external candidates, the substitute is structured multi-source referencing: the same competency questions put to a former chair, a peer and a direct report, with answers scored against the scorecard rather than summarised as "strongly recommended". Referencing should be backed by discreet verification of the claims that matter, including credentials, directorships, litigation and regulatory history. That is the work of persona due diligence, and it catches what polite references do not.
How should a board combine leadership assessment methods?
A board should combine leadership assessment methods into one pre-agreed process, because combinations predict better than most single tools. Averaging across every combination of up to five widely used predictors, Sackett's team found a mean validity of .47 on their revised figures, close to the .51 that older research had promised. The sequence we recommend to nomination committees is:
- A scorecard agreed before the search opens: outcomes, competencies and deal-breakers, signed off by the committee.
- Structured interviews by at least two assessors: the same core questions for every candidate, scored independently before any discussion.
- A role-specific work sample: real, redacted business material and a presentation to the panel.
- Psychometric instruments as hypothesis generators: interpreted by a qualified psychologist, never used as pass-or-fail gates.
- Structured referencing and background verification: on the final two candidates, before the offer.
- A decision rule set in advance: how the evidence is weighted, and who breaks a tie.
The last step matters most. When the committee agrees the weighting before it meets candidates, a strong dinner finds it much harder to override a weak work sample. This is the discipline behind our retained executive search mandates, and the assessment work in our leadership consulting practice for boards that want an independent view of an internal or external shortlist.
What does India's DPDP law mean for assessing candidates?
India's Digital Personal Data Protection framework means that candidate assessment data held in digital form, including psychometric reports, interview scores and reference notes, is personal data that must be handled for a stated purpose. The DPDP Rules were notified on 14 November 2025 with an eighteen-month phased compliance period, according to the Press Information Bureau.
The DPDP Rules require a separate consent notice explaining the specific purpose for which personal data is collected, and they give individuals the right to access, correct and seek erasure of their data, with requests answered within ninety days. For a board, the practical steps are simple: tell candidates what will be assessed and by whom, keep reports within the committee and its advisers, and agree a retention period with the search firm. Senior candidates notice how their information is handled, and it forms part of their view of the employer.