CEO Selection Criteria: What Actually Predicts Performance

Most lists of chief executive characteristics are written backwards. They start from people who succeeded, describe them, and present the description as a specification. The result is a set of qualities that sound right, are impossible to assess in a process, and would have been true of plenty of executives who failed. Vision, presence, drive. Nobody has ever selected against them.

This page takes the other direction. It starts from what a board can actually observe before an appointment, sets out which of those things tend to predict the first three years, and is candid about which widely used criteria are mostly noise. It is written for boards, nomination committees and the search partners who serve them, particularly in regulated financial services, where the consequences of the appointment reach the regulator as well as the shareholders. For critical senior appointments, the search itself also needs a deeper process than a standard vacancy, particularly where confidentiality and board-level assessment matter.

Which criteria genuinely predict performance?

Decisiveness under incomplete information, a record of building the team beneath them, and evidence they have changed their mind when the facts moved.

Decisiveness is the one most boards name and few test properly. It is not speed. A chief executive who decides quickly and wrongly is worse than one who is slower, and every board has met the first kind. What predicts performance is the ability to decide with roughly seventy per cent of the information, act, and correct course when the remaining thirty arrives. That is observable in a candidate history if the process asks for it.

The second criterion is who they have developed. A candidate whose former teams are now running things elsewhere has demonstrated something a candidate with a brilliant personal record has not: that the organisation was stronger for their having been in it, rather than dependent on them. Reference conversations should ask this directly.

The third is the least comfortable to assess and the most telling. Ask for a decision the candidate reversed, and why. Executives who cannot produce one either have never been wrong, which is unlikely, or do not register their own reversals, which is a problem a board will inherit.

Which criteria do boards over weight?

Sector tenure, presentation, and the size of the previous organisation.

None of these should be ignored. Each is worth something. The error is treating them as primary when they are corroborating, and the cost of that error is an appointment that looked obviously right and was quietly wrong for two years before anyone said so.

What is different about selecting a chief executive in financial services?

The regulator is a party to the decision, and the candidate has to be able to hold a position under supervision as well as under commercial pressure.

In most sectors a board appoints and the market judges. In regulated financial services a supervisor will form a view of the individual, sometimes formally, and will continue to form one throughout the tenure. That changes the specification in two ways.

The first is that a chief executive who treats the compliance function as a constraint on the commercial job, rather than as part of it, is a liability whatever their other qualities, and boards should test for the attitude rather than assume it. The second is that the candidate will at some point have to tell a supervisor something unwelcome, and tell the board something the supervisor has said that the board does not want to hear. Ask for a time they did either. The answer, or its absence, is informative.

How should the criteria be tested in the process?

With evidence from the record and from references, not with self description.

Candidates at this level are practised at describing themselves, and a process built on interviews alone measures that practice. The corrective is to draw evidence from three places and check them against one another.

• The record. What actually happened at each organisation in the years after they arrived and, revealingly, in the two years after they left.

• References chosen by the board, not only those supplied. Former direct reports and former board members see different things from former peers.

• A structured session on real decisions. Not hypotheticals. Specific choices they made, what they knew at the time, what they got wrong, and what they did about it.

Where the three disagree, the disagreement is the finding. A candidate whose self account, record and references all align is rare and worth noticing. One whose references are warmer than the record supports, or whose account of a decision does not match what a former colleague remembers, has told the board something useful before the appointment rather than after it.

How long should a board give a new chief executive before judging?

Long enough to see a full cycle of the decisions the role exists to make, and no longer than it would give anyone else.

Boards err in both directions here and the second error is more expensive. Judging too early punishes an appointee for the state they inherited, and executives brought in to fix something often make the numbers worse before they make them better, because the fix involves writing down what the predecessor left. A board that panics at the first bad quarter has not selected badly. It has failed to decide in advance what the first year was supposed to look like.

Judging too late is the quieter failure. The signs that an appointment is not working are usually visible within eighteen months: senior departures the chief executive did not anticipate, decisions deferred rather than made, and a board hearing about problems from outside before it hears from the top. Boards tend to wait for the numbers to confirm what those signs already said, and the cost of the wait lands on the organisation rather than on the individual.

The discipline is to write down, before the appointment, what the board expects to see at six, twelve and eighteen months, and to hold the review against that rather than against whatever the most recent quarter happened to produce.

What should the board settle before the search starts?

What the organisation needs in the next three years, which is not always what it needed from the last chief executive.

Boards frequently specify the previous incumbent with the weaknesses removed. That produces a search for a person who does not exist, and it assumes the next three years look like the last three. A better starting point is to name the two or three things the organisation must get through, whether that is a regulatory change, a succession problem beneath the top, or a market that has stopped growing, and to write the criteria against those. The person who fits that specification may look quite unlike the last one, and the board should be prepared for that before the shortlist arrives rather than surprised by it. How the people function earns its place in that conversation is covered in our note on building credibility with the executive team.