Can you screen narcissism out of executive hiring? Not with the tools most organisations use: assessments entertain, references defend, and interviews actively select for the pattern, because interviews reward confident self-presentation, which is the pattern’s native sport. What works is behaviour observed in context: trial assignments with the real team, structured biography over time, and stakeholder evidence from previous roles. The goal is not to filter narcissism out but to see it accurately and place it deliberately.
Every few months a CHRO asks me for the assessment that detects narcissists before they are hired, and I have to deliver the same disappointing news: the instrument exists, you are holding it, and it is called the hiring process, currently running in reverse. Executive selection as commonly practised does not fail to filter narcissistic patterns. It selects for them, enthusiastically, and then acts surprised. Understanding why is the first half of fixing it; the second half is accepting what the fix actually requires, which is more contact with reality than most selection processes are designed to tolerate.
Why the funnel selects for the pattern
Start with the mechanism I described in the field guide: under stress, humans choose competence signals over warmth, and boards hire under stress. The dentist experiment from Rescue the Alpha Wolf applies without modification to executive search: when the pain is acute, the committee picks the marinier with the Harvard wall over the warm practitioner with the trembling hands, every time, and the candidate who radiates certainty owns the room precisely because the room is anxious.
Now look at what each stage of the funnel measures. The interview measures confident self-presentation under observation, which is not a proxy for leadership but is a near-perfect proxy for the narcissistic skill set; as I put it in Let’s Talk Leadership, assessments entertain, references defend, and interviews seduce, but none of them predict behaviour like behaviour. The self-report personality test measures the candidate’s theory of the desirable answer, and any bright candidate games it; the pattern games it sincerely, because the projector writes flattering copy about itself without ever experiencing the writing as dishonest. References measure the candidate’s curation ability, since nobody supplies the three people they burned. And the executive-search narrative interview measures storytelling, at which the pattern is, again, professionally excellent: the heroic turnaround, the resistance overcome, the visionary alone against the sceptics. Notice the composition of that story. It is the narcissistic circle rendered as a competency.
Then promotion mechanics compound the intake bias. Internal advancement flows toward visibility, credit capture and upward management, three activities the pattern performs as breathing, while the reciprocal leaders who make their teams larger than themselves are, by exactly that virtue, harder to see from above. Organisations then wonder aloud why the top of the pyramid behaves the way its own incentives shaped it.
The reframe: filter is the wrong word
Here is where I part company with the standard advice, which promises detection and removal. Two objections. First, at intake volumes the traits are not rare contaminants; drive, confidence and self-belief cluster in exactly the population that applies for these jobs, and a filter tight enough to exclude every narcissistic pattern excludes your next great crisis leader with it. Second, the premise is wrong about value: as I argued under talented narcissism, the pattern arrives bundled with genuine, sometimes irreplaceable capability, and its cost is phase-dependent, not absolute. In fire you want an alpha wolf; in consolidation you want a Director; and there is no universal CEO, only scripts matched or mismatched to phases. The wrong archetype in the wrong season creates pathology, as I put it in the book: a warrior in a stable phase creates drama, a consolidator in a crisis breeds paralysis.
So the serious objective is not exclusion but accurate sight plus deliberate placement. See the profile clearly, price its costs honestly, match it to the phase that needs it, and contract the governance before the honeymoon, not after the first invoice.
What accurate sight requires
Behaviour in context, before the contract. The strongest single upgrade available, straight from Let’s Talk Leadership: give finalists a real, time-boxed team task with the people they would actually lead, titles stripped from the room, and score what happens. Nothing dissolves a rehearsed narrative faster than two hours of unscripted collaboration; the credit ledger, the interruption pattern, the response to being corrected by a junior, all of it surfaces because none of it can be performed for long. Then ask the involved team plain questions about strengths, pitfalls and advice. It is astonishing, I wrote, how quickly the noise disappears when you listen to facts instead of narratives.
Biography over time, read structurally. Not the achievement tour, but the timeline of rupture and resolve: how did each role end, what happened to the teams afterward, where are the former lieutenants now and what do they say unsupervised? One burned bridge is a career; a lattice of them is a pattern. Attend also to the phase question hiding in the references: a candidate praised entirely for crisis heroics is being accurately described as an alpha wolf, which is only good news if you are actually on fire.
Stakeholder evidence, not curated referees. Structured, cross-checked conversations beyond the supplied list, focused on behaviour and consequence rather than adjectives. The methodological principle is the same one that governs my coaching practice: self-perception is the least reliable witness in the building, and the aggregated experience of the people in the room is the only instrument the charm cannot reach.
The correction test, live. Somewhere in the process, ensure the candidate is contradicted, correctly, on something that matters, and watch the machinery rather than the words. Updating under correction is the single most load-bearing behaviour you can observe in two hours; its absence is the single most expensive thing you can hire.
Succession: the filter’s final exam
Everything above intensifies at succession, because the sitting leader is now part of the instrument. An unpolished pattern at the top curates successors as it curates information: heirs are selected for reflection rather than capability, genuine rivals are exported years before the vacancy, and the board inherits a shortlist that is itself a symptom. The countermeasure is the same principle applied upward: independent stakeholder reality, board contact below the CEO as routine, and succession criteria written against the phase the company is entering rather than the phase the incumbent conquered. And when the succession involves a founder, the psychology thickens further, because the apparatus was built to need him; that resistance has its own anatomy, and boards that ignore it get succession plans that are excellent in every respect except occurring.
The quotable version: interviews are auditions for a play that never opens; if you want to know who you are hiring, stop watching the audition and watch two hours of the actual job.
Frequently asked questions
Is there a test that detects narcissism in candidates? Validated instruments exist in research settings, but in high-stakes selection self-report is gameable and the pattern games it sincerely. No questionnaire outperforms structured observation of real behaviour with the real team plus rigorous, non-curated stakeholder evidence from past roles.
Should we simply not hire people with narcissistic traits? You could not staff an executive floor that way, and you would exclude your best crisis leadership with the bathwater. The workable policy is sight, placement and governance: know the profile, match it to the phase, contract the mirrors and boundaries in advance.
What is the biggest red flag in an executive interview? Less any single answer than the ledger across the whole conversation: every success personally authored, every failure externally caused, predecessors incompetent, former colleagues ungrateful. And the live correction test: contradict the candidate correctly and watch whether you get an update or a counterattack, however charming.
How do we protect promotion decisions from the same bias? Promote on stakeholder-experienced behaviour rather than visibility: skip-level input, team-health data over time, regretted attrition under the candidate. The leaders who make others larger are systematically underweighted by upward-facing processes; correct for it deliberately.
Who should own this in the organisation? The board for CEO selection and succession, with independent facilitation; the CHRO for the executive layer, with explicit licence to run behavioural trials and non-curated referencing. Delegating it entirely to search firms returns you to the audition.
Arvid Buit is a master executive coach (ICF, EMCC, APECS, Marshall Goldsmith SCC), founder of TRUE Leadership, and author of Let’s Talk Leadership, Red de Alfawolf (with Martin Appelo) and Wahlberg. Organisational psychology for selection and governance contexts, not clinical advice; nothing here licenses diagnosing candidates. The full field guide · For boards facing the sitting-CEO version