Why Executive Hires Fail

The most instructive failure I know of looked, on paper, like the safest hire of the year.

The company was a family-owned industrial group, second generation, preparing for its first outside capital. The role was CFO. The candidate was close to perfect: a sector veteran with two successful capital raises behind him, an auditor's early training, flawless references from two listed companies, and a technical interview performance that left the owner's advisors with nothing to ask. Three people interviewed him. All three conversations converged on the same subject, his experience, because his experience was magnificent and everyone in the room enjoyed talking about it. He was hired quickly, at a premium, with relief.

Fourteen months later he was gone. Not because a covenant was breached or a model was wrong; his models were the best the company had ever had. He left because he could not live inside a company where the owner walked into his office without an appointment, reversed a decision over Sunday lunch with his sister, and expected the CFO to treat the family's real-estate holdings as an extension of the group's balance sheet. He had run finance functions where the border between governance and management was drawn in ink. Here it was drawn in blood, and nobody had assessed whether he could work on that terrain. Nobody had even discussed it. Every conversation had tested what he could do. None had tested who he was, what he needed, or what this particular company would ask him to tolerate.

The failure was not his. It belonged to the process, and the process failed in the most common way executive assessment fails: it measured one kind of thing three times and another kind of thing not at all.

This book is about those two kinds of thing, the two registers of evidence about a senior candidate, and about how to assess each of them with methods that fit it. Before building anything, though, we should look honestly at the record, because the record is what justifies the book.

The record, audited

Everyone in this industry has heard the famous statistic: "46 percent of new hires fail within eighteen months." It comes from a 2005 study by a training company, it described new hires in general rather than executives, it defined failure loosely, and it has been repeated for twenty years mostly because it is memorable. This book will not lean on it. If a claim about failure is going to carry the argument for redesigning how you assess people, the claim itself has to survive assessment.

Here is what survives. Across the more careful bodies of evidence, the longitudinal studies of executive turnover, the corporate-board surveys, the transition literature, and the annual analyses of forced CEO departures, a consistent picture emerges, and it is bad enough without exaggeration. Organizations should expect something in the range of 20 to 30 percent of external senior hires to be terminated or pushed out within roughly two years, with a further 20 percent or so remaining in the seat while seriously underperforming. Taken together: roughly 40 to 50 percent of external executive appointments fail or struggle badly inside their first two years. Corporate Executive Board research put the struggling-or-failing share of transitioning executives at about half, and the figure has been replicated in spirit, if not in decimal places, everywhere the question has been studied with any rigor.

Two features of this record matter more than the headline range.

First, the failures are expensive out of proportion to their number. A failed senior hire costs multiples of salary in severance, search, and interim cover, and those are the small items. The large items are the eighteen months of strategic drift, the departures the failure triggers one level down, and the board's diminished appetite for the bold external candidate the next time, when the bold external candidate might be exactly what the situation needs.

Second, and this is the finding on which everything in this book stands: the failures are not, in the main, failures of competence.

What actually kills them

When researchers and practitioners perform honest post-mortems on failed executive transitions, technical incapacity almost never leads the list. The consistent finding, stable across decades and methodologies, is that upwards of two-thirds of executive transition failures trace to relational, political, and cultural causes: the new leader misread the culture, mishandled the politics, failed to build the alliances the role actually ran on, alienated the team they inherited, or misdiagnosed the situation itself, arriving with the playbook for a turnaround when the company needed steady scaling, or the playbook for scaling when the company was quietly on fire. The Center for Creative Leadership's derailment research established this pattern in the 1980s; the transition literature has confirmed it ever since. Executives are almost never fired because they lack the skills that got them hired. They are fired for how they land, relate, read, and adapt, or fail to.

The trend line makes the point sharper. The leading annual analyses of forced CEO departures now find that ethical lapses and behavioral misconduct account for a larger share of forced dismissals — approaching 40 percent — than financial underperformance does. Boards have become more forgiving of a bad quarter and less forgiving of conduct. Read as evidence about assessment, this is remarkable. The causes that now dominate executive failure, conduct and character under pressure and relation to oversight, are precisely the ones that standard hiring processes assess least, and least well.

So the failure data contain a structural message, not just a cautionary one. If executive failure were mostly about capability, the remedy would be more rigorous capability assessment: harder technical interviews, deeper track-record analysis. Necessary work, and Part II of this book is devoted to doing it properly. But it would not move the failure rate much, because capability is not where most of the failures come from. Most executive failure is a fit-and-conduct phenomenon occurring in people whose capability was genuinely high. That is why the second register exists.

The portability illusion

There is a tempting objection: surely proven performance is the answer. Hire the person who has demonstrably done it, and the rest takes care of itself.

The evidence says otherwise, and the studies here are among the strongest in the field. Boris Groysberg and colleagues followed over a thousand star investment analysts, professionals with individually measurable, publicly ranked performance, as they moved between firms. The stars' performance dropped after the move, the decline persisted for years, and the receiving firms' market value suffered. The stars who kept their standing were disproportionately those who moved with their teams, which is the finding that explains the finding. What looked like individual excellence was substantially embedded excellence: performance woven into a specific platform, specific colleagues, specific internal networks, and firm-specific capital that did not travel with the CV.

Matthew Bidwell's research on internal versus external staffing completes the picture from the buyer's side. External hires into comparable roles were paid substantially more, performed worse for their first two years, and exited at higher rates, both voluntarily and involuntarily, than internal people promoted into the same kind of job. The external candidate's polished dossier commands a premium precisely because it is legible; the internal candidate's firm-specific effectiveness is real but hard to see in writing. The market pays for the visible register and is repeatedly disappointed by the invisible one.

Neither finding says "never hire externally". External hires bring options internal ones cannot, and Bidwell's externals had stronger observable credentials for a reason. The findings say something more precise: a track record is evidence about a person-in-a-context, and the context does not come with the person. An assessment process that treats past performance as a portable quantity is systematically overpaying for the register it can see.

When strengths turn

One more body of evidence belongs in the diagnosis, because it explains the failures that surprise everyone: the hires who fail because of what made them impressive.

The derailment tradition, from the CCL studies through Hogan's work on the "dark side" of personality, documents a pattern every experienced assessor eventually sees in the flesh: the qualities that propel an executive upward mutate, under stress, into the mechanisms of their undoing. Boldness and confidence become arrogance and the inability to hear bad news. High standards become micromanagement. Diligence becomes rigidity. Charm becomes manipulation. These reversals are not rare edge cases; derailment researchers consistently estimate that a large fraction of managers exhibit at least one significant derailment risk, and the risks activate exactly when transitions activate them: new scope, new scrutiny, ambiguity, isolation, concentrated power.

The assessment implication is uncomfortable and important: a strengths-focused interview is structurally blind to derailment, because derailers present, in calm conditions and polished candidates, as strengths. You do not find them by asking about achievements. You find them, and Part IV will show how, in patterns across decisions, in how the candidate talks about opposition and oversight, in what former subordinates say carefully, and in what happens to the narrative when it is gently pressed.

The market inefficiency

If the causal structure of failure is this well documented, why do hiring processes keep assessing the wrong register? Part of the answer is method: the tools most firms use cannot see the second register, and later chapters take that up. But part of the answer is preference, and it has been measured.

Steven Kaplan, Mark Klebanov, and Morten Sorensen analyzed detailed assessments of over three hundred CEO candidates in private-equity situations and then followed what happened. Candidate profiles varied along two broad dimensions. One contrasted execution-oriented capability (efficiency, persistence, holding people accountable, moving fast) with interpersonal polish (openness, treating people well, communicative charm), and subsequent firm performance related clearly to the execution side, and to general ability, rather than to the interpersonal factor. Yet hiring processes, boards and investors and interviews alike, demonstrably tilt toward the polished, communicative candidate. The charisma is visible in the room. The execution lives in the record. Rakesh Khurana's study of CEO selection at large corporations described the same tilt sociologically: the "irrational quest" for the charismatic outsider, the corporate savior whose presence outruns the role-specific evidence.

Hold this finding carefully, because it is subtler than "personality doesn't matter." Identity, meaning motivation and values and conduct under pressure and relation to authority, matters enormously; the failure data just showed it. What the Kaplan results expose is that interview-room charisma is evidence of neither register. It is not execution capability, which lives in the verified track record. And it is not identity, which does not disclose itself in ninety minutes of practiced warmth. Charisma is a third thing, the performance of the encounter itself, and unstructured executive interviews are optimized to collect it. The typical process, in other words, doesn't merely under-assess the register that kills most hires. It actively substitutes a pleasant proxy that predicts neither.

The two registers

Everything the failure record shows can now be said in one distinction, and the distinction is the book.

Evidence about a senior candidate comes in two registers. The first is capability: what the person can do, meaning technical mastery, management craft, cognitive horsepower, the behavioral skills of the executive trade. Capability leaves traces: results, artifacts, verifiable episodes, observable performance on realistic tasks. It can be assessed with methods that examine traces, whether track-record forensics, structured behavioral interviews, work samples, or technical references, and Parts II and III of this book build that machinery.

The second is identity: who the person is such that the doing happens. Motivation, values, relation to authority and oversight, what a possible future means to them, how they behave when the situation stops flattering them. Identity leaves different traces, and reluctantly. It does not answer direct questions, because every senior candidate knows the right answers to direct questions. It appears in relation: across multiple encounters, in the pattern of real decisions, in what a person defends and avoids and treats as obvious, in the space between their account and what their references saw. It requires a different process architecture, with separated conversations, decisional evidence, disciplined reference triangulation, and psychometrics used as prompts rather than verdicts, and Part IV builds that.

The two registers answer different questions. Capability answers: can this person do this work at this level? Identity answers: can this person own this role, in this organization, with this owner, board, and moment — and remain themselves doing it? At senior level, capability is usually the threshold and identity the differentiator. The shortlist all clears the capability bar, and the failure data show where the remaining variance lives.

The central error of executive assessment, the error in the CFO story, the error the statistics price at 40 to 50 percent, is running one process for a two-register problem. Three interviews about experience are three measurements of the same register, each one redundant with the last, while the register that will decide the outcome goes unmeasured. Worse, the single blended conversation invites the proxy. When a process has no method for identity, interviewers assess it anyway, by feel, which means by charisma and similarity, which the evidence says predicts nothing and discriminates plenty.

This book's claim is not that identity assessment should replace capability assessment, nor that judgment should replace structure, nor that anything here makes hiring safe. The claim is narrower and more useful: the two registers require different evidence, collected by different methods, in a deliberately designed sequence — and a process that does this honestly can reclaim a large share of the failures the record describes. The failures are not random. They cluster exactly where the standard process is blind, and a blindness that specific can be engineered against.

Instrument: the register self-diagnostic

Before reading further, audit your own current process — the one you actually ran on your last three senior hires, not the one in the policy document. For each question, answer with what happened, not what was intended.

1. Across all interviews for the last senior hire, what share of total conversation time addressed the candidate's past experience and achievements? (If above roughly 70 percent, you ran a one-register process.)
2. Did any two interviewers assess deliberately different things — with different questions, by design — or did every conversation converge on the same terrain?
3. Was any claim on the CV independently verified beyond references the candidate supplied? (Scope, P&L, team size, actual decision rights?)
4. Did any part of the process require the candidate to do anything — analyze, present, decide — rather than describe?
5. Could anyone on the panel state, in writing, what the candidate would need from an employer to thrive — their motivational profile — and what evidence supports it?
6. Was the candidate's relation to authority and oversight — how they handle being challenged, overruled, or monitored — assessed by any deliberate method?
7. Did references answer specific questions about specific episodes raised in interviews, or did they deliver general endorsements?
8. Did anyone probe why this move, why now, tested against the facts of the candidate's situation rather than their stated answer?
9. Where did "culture fit" get assessed — against criteria defined in advance from the organization's side, or as a feeling after the meetings?
10. If the hire failed eighteen months in, does your file contain evidence that would let you diagnose why — or only evidence that everyone was impressed?

Score it simply. Questions 1–4 and 7 audit your capability process; questions 5–6 and 8–9 audit your identity process; question 10 audits whether you are running an evidence process at all. Most organizations discover they have a partial capability process, a charisma process where the identity process should be, and no learning loop. The rest of this book is the repair, in order: what evidence counts (Chapter 2), the capability register in its technical and behavioral halves (Parts II and III), the identity register (Part IV), and the assembly of both into a judgment someone owns (Part V).

Where the rules run out

An honesty note before the toolkit, in keeping with a discipline this book will maintain throughout: each chapter tells you where its evidence is thin.

The failure literature is thinner than its fame suggests. Definitions of "failure" vary across studies; many of the most-quoted numbers come from consultancy surveys rather than peer-reviewed longitudinal work; executive-specific samples are small; and causal attributions in post-mortems are retrospective, which invites tidy stories. The ranges given in this chapter are the ones that survive an audit of provenance, stated as ranges precisely because false precision is its own kind of dishonesty. Nothing in the book's argument requires the failure rate to be 47 rather than 41 percent. The argument requires only what the evidence robustly supports: that executive failure is common, expensive, concentrated in the identity register, and poorly predicted by the assessments most organizations actually run. That much is solid ground.

Notes and sources

Evidence grades: [M] meta-analytic or systematic; [L] longitudinal/peer-reviewed primary; [S] survey or practitioner analysis, credible but not peer-reviewed; [A] audited claim — widely cited figure assessed for provenance in the research underlying this book.

• "46% fail within 18 months": Leadership IQ (2005) — general new hires, loose failure definition; cited here only to set aside. [A]
• Executive transition failure/struggle rates (40–50% within ~2 years, composed of 20–30% exit plus ~20% underperformance): Corporate Executive Board / Gartner transition research; corroborating estimates across succession and transition studies. [S/A]
• Two-thirds-plus of failures from relational, political, cultural, and situational-diagnosis causes: Center for Creative Leadership derailment research (McCall & Lombardo and successors); executive-transition literature (Watkins and successors). [L/S]
• Forced CEO dismissals for ethical lapses/misconduct approaching 40% and exceeding financial underperformance: annual CEO-success analyses of forced departures (Strategy&/PwC tradition). [S]
• Star portability: Groysberg, Chasing Stars (2010) and related studies of >1,000 ranked analysts. [L]
• External-hire penalty: Bidwell, "Paying More to Get Less" (Administrative Science Quarterly, 2011). [L]
• Derailment and dark-side dynamics: Hogan & Hogan; Hogan Development Survey research tradition; CCL studies. [L/M]
• CEO characteristics and the execution/interpersonal contrast: Kaplan, Klebanov & Sorensen, "Which CEO Characteristics and Abilities Matter?" (Journal of Finance, 2012); Kaplan & Sorensen, "Are CEOs Different?" (Journal of Finance, 2021). [L]
• Charismatic-savior dynamics in CEO selection: Khurana, Searching for a Corporate Savior (2002). [L]