The Interview Almost Never Changes the Decision
Somewhere between 30% and 50% of CEO transitions are poorly managed, and some research puts the figure as high as 70% (INSEAD, 21 Questions About CEO Succession, June 2026). The same work prices the cost of poor chief executive transitions across the S&P 1500 at close to US$1 trillion a year. A Stanford and Yale study of 3,000 listed US firms found that 29% of chief executive departures over a five year period were involuntary and only 23% were unambiguously voluntary, which leaves almost half in a middle category where nobody says out loud that the appointment did not work.
That is the top of the house. The economics are no kinder one level down, where a head of coverage carries a revenue line you can count to the pound and around half of all successions turn out to be effectively unplanned (INSEAD, 21 Questions About CEO Succession, June 2026). In the teams we recruit for in London and Paris, the senior hire who underdelivers for eighteen months before anyone intervenes is far more common than the dramatic exit, and a good deal more expensive once you add the lost year of coverage.
The awkward part is that the two inputs firms lean on hardest, the interview and the pedigree on the CV, aren't really any good at predictingperformance. Both feel rigorous because both are effortful. The evidence on the gap between effort and signal is now good enough to act on.
1. Senior Appointments Fail Slowly
Start with the measurement, because most firms have it wrong. A senior hire gets logged as a failure only if the person is dismissed, which keeps the recorded rate flatteringly low and means the process that produced the appointment never gets examined.
The succession research does not permit that comfort. Across 3,000 listed US firms over five years, 29% of chief executive exits were involuntary and 48% fell somewhere between voluntary and involuntary, leaving only 23% that were the executive's own decision (INSEAD, 21 Questions About CEO Succession, June 2026). Roughly half of successions were unplanned. Read that across to a division or a desk and the shape repeats: a large middle band of appointments that neither blew up nor delivered what the mandate promised, and which therefore never triggered a review.
Try this test. Could you name the last three senior hires your firm made and state what each was expected to deliver in year one? If the answer comes out hazy, nobody has measured the failure rate, which is a different thing from the failure rate being low.
2. The Interview Ratifies a Decision Already Taken
Senior interviewing is the most expensive stage of any process and, in most firms, the least structured. Six partners give ninety minutes each, form independent impressions, then converge in a room where the most senior voice settles it. That is a consensus mechanism dressed as a measurement one.
INSEAD faculty research with the assessment firm Athena tested this directly. Over 550 candidates completed both AI driven job simulations and static tests, and 83.4% were rejected by both methods, indicating the two structured approaches largely agreed with each other. Every candidate then sat a final human interview for compatibility. Those interviews, in the researchers' words, ended up practically never changing the original decision (INSEAD, How AI Can Improve Hiring, March 2025). The same paper notes that without careful structuring and design, conventional hiring methods suffer from low validity and fail to distinguish reliably between candidates.
An interview that never overturns the structured evidence is a social stage, and it should be budgeted and timetabled as one.
3. Pedigree Records a Fit With Somebody Else's Firm
The logo on a CV gets treated as portable evidence of quality. It tells you which platform selected this person, what balance sheet they were standing on, which clients were institutionally inherited and how much of the origination was theirs. Which building somebody sat in, in other words, rather than what they did once they were inside it.
Boards demonstrate a persistent bias towards external candidates despite mixed results, and only 20% to 30% of new CEO appointments in mature and emerging economies are external hires in the first place. Internal candidates consistently deliver longer tenure, better performance and smoother subsequent transitions. External appointments succeed mainly in turnaround situations, particularly where the incoming executive brings finance, accounting or operations depth from a different industry, while external star hires into stable contexts frequently disappoint (INSEAD, 21 Questions About CEO Succession, June 2026).
So the honest question at offer stage is whether the conditions that produced this person's track record exist in your firm, and if they don't, what replaces them.
4. What Travels Is Behaviour
If pedigree does not travel, something has to, otherwise senior hiring is a lottery. The portable part turns out to be behavioural, and narrower than most competency frameworks assume.
Aon's analysis of more than 750,000 leadership assessments, covering over 2,000 leadership skill profiles of which around 300 sit at senior and executive level, drawn from 2,500 client organisations assessing leaders across 31 countries, finds three competencies recurring most consistently across roles and markets: driving results, building relationships and coaching for performance (Aon, Leadership in Transition, 2025). Aon administers more than 30 million assessments a year in over 40 languages across 90 countries, which gives the base rates some weight (Aon, Pre-Hire Talent Assessment, 2026).
Coaching for performance is the competency that gets waved through in banking interviews, and it is the one that decides whether the firm has bought a producer or a business.
5. Rigour Has a Shape, and More Rounds Is Not It
Rigorous and lengthy are different things. A process that runs to nine meetings over four months is usually deferring a decision nobody wants to own, and losing good candidates while it does.
The direction of travel among large employers is towards structure earlier rather than volume later. Leadership assessments rose 85% between 2019 and 2022, and the share of organisations focused on leadership development climbed from 9% in 2020 to 33% by 2024 (Aon, A Better Approach to Succession Planning Using Assessment Data, October 2025). Where structure goes in at the front, interview load falls at the back: Deloitte recorded a 66% reduction in interviews needed after introducing structured pre-hire assessment (Aon, Pre-Hire Talent Assessment, 2026).
Ask less often how many partners have met the candidate. Ask more often what each meeting was designed to establish that the previous one could not.
A Note of Balance
None of this makes assessment a substitute for judgement, and the data has real limits. Only around 300 of Aon's 2,000 plus leadership skill profiles sit at senior and executive level, so the evidence base thins where the stakes are highest. Personality instruments were built for populations rather than for the one person who is going to run a Paris coverage team. Apply a mechanical process to a market of forty credible people and it will produce forty rejections and no hire.
The interview also does work that the validity literature undercounts. Senior hiring is a two way market. A managing director weighing a move is assessing your governance, your capital commitment and whether the person opposite is somebody they want to sit next to for a decade. That conversation is the sale.
Conclusion
The evidence points somewhere unfashionable. What predicts a successful senior hire is a clear specification of what the role must deliver, structured evidence gathered against that specification, and an honest account of which parts of a candidate's record depended on a platform they are about to leave behind. Not the brand on the CV, and not the number of partners who liked them.
None of that work can be done from an application pile. An inbound process produces a stack of CVs and no way to test them, so it over weights the one thing on the page that reads quickly, which is the logo. Establishing what somebody originated rather than inherited takes conversations with the people who were on the other side of the trade, with the client who signed the mandate and with the analyst who built the model. That is thirty or forty names across London and Paris, most of whom have no live interest in moving and will never appear in anyone's inbox. The rigour that predicts performance sits in the research, months ahead of the first interview.
Sources: INSEAD, 21 Questions About CEO Succession, June 2026; INSEAD, How AI Can Improve Hiring, March 2025; Aon, Leadership in Transition: Understanding the Leadership Pipeline Through Assessment Analytics, 2025; Aon, A Better Approach to Succession Planning Using Assessment Data, October 2025; Aon, Pre-Hire Talent Assessment, 2026.
