The candidates we assess for top-level executive leadership positions are invariably intelligent, socially sophisticated professionals. They sport impressive resumes, impeccable references, have attended the world’s best business schools, and command a depth and variety of experience that have Boards and CEOs salivating. They are more than capable of navigating the obstacle course of interviews and presentations that are set up in a vain attempt to separate the “excellent” from the “good”. It makes you wonder why 8 out of 10 of those selected struggle to achieve even average performance and why within 2 years, 50% will fail.

It is often not the fault of the candidates – they are simply playing the game by the rules. It’s the fact that selection committees possess no special powers of insight or evaluation that set them apart from the rest of us. Put simply, humans are terrible at picking other humans in a manner that correlates to success. We have known this for a long time but pretend that all the failings in human selection don’t apply to us. But being a CEO or top executive doesn’t immunize you from basic errors of judgment, especially when it comes to recruiting high quality candidates.

When Tom Rath and I wrote Strengths Based Leadership we referenced a remarkable Gallup statistic: 96% of Americans rate themselves at- or above-average in their ability to lead. While publicly and reluctantly admitting that we might be “pretty good” leaders, we privately believe we are better than just about everyone else.

This test holds true for top executives, too. We conducted a research study of 100 executives – a mixture of CEOs, CFOs, CIOs CMOs etc. Two questions were on the table:

On a 1-5 scale with 5 high, how strongly do you rate yourself in hiring great leaders?

For every 10 leaders you hire, how many turn out to be great performers?

Even leaving “great leaders” and “great performers” undefined does nothing to moderate the strength of respondents’ opinions. On the first question, 73 gave themselves a straight “5” on their ability to hire great leaders. The lowest score was “3” and only 7 executives rated themselves this low. The overall average was 4.7. With such a high level of capability, doesn’t it make you wonder why so few companies break ahead of the pack?


Responses to the second question were slightly more reserved. On average, 68 out of every 100 executive-level hires were claimed to be “great performers”. 11 executives claimed a 100% record in picking great performers, which would be remarkable if true. What becomes obvious, looking at the different responses to these two questions, is that our 100 executives saw a difference between “great leadership’ and “great performance”. Apparently a leader can be “great”, regardless of the performance they produce.

But there is no point in asking these questions unless we already know the answers, and in this case with these 100 executives, we do. The “executive hit rate” (the percentage of hires who became top quartile performers) had already been calculated. Far from the average of 68% these executives claimed for themselves, the very best executives barely approached a hit rate of 1 in 5. The top 3 executives selected 18% top quartile performers, while the bottom 17 selected zero. The average was 12%, barely 1 in 10.


Let’s talk methodology. Our comparative calculation of executive performance was based on a statistical calibration of three independent, objective performance measures. One of these measures was common to all – group, divisional, sectional or functional employee engagement data as measured and re-measured over time. This was an indexed “change” metric. The two remaining measures were role- and company-specific, such as productivity, efficiency, revenue growth, operating margin, and/or product/process quality metrics. We scaled these measures historically against the leaders these executives had replaced and then indexed these metrics to enable effective comparison between internal and external peers.

Psychology literature is full of research demonstrating that, left to their own intuition, leaders pick people who resemble themselves in personality and behavior. Furthermore, they are prone to a whole set of cognitive biases that make it difficult for them to see or accept their shortcomings. We all remember our one or two fantastic successes and conveniently forget or rationalize our more numerous failures. Few of us take the time to quantify our effectiveness.

For the two questions we asked above, three cognitive biases skew the thinking of many executives: confirmation bias, self-serving bias, and the availability heuristic. In each case these biases significantly influence recalled memory. An executive can nearly always remember a “superstar” they picked and confirmation bias convinces them that this is a consistent pattern – in effect the executive is thinking “I picked a great person therefore I am great at selecting people.” It is relatively easy to discount failures and then self-serving bias kicks in and attributes these failures to factors beyond the executive’s control.

It is not only uncomfortable recalling selection failures, it also requires greater cognitive effort. This effort is too much for many executives and the availability heuristic favors happier memories that more easily come to mind. The consequence is that executive leaders have a far higher opinion of their selection abilities than the evidence supports.

The solution to this problem – let’s call it “selection blindness,” – can only come with an acceptance of fallibility. In our next article we will describe the methodology for subverting these selection failings and enabling more accurate statistical predictions of future leadership performance. In short, we will outline how to reduce the executive selection error rate and discover more outstanding executive performers.