“At one time, he was the leading, perhaps the only, candidate for CEO.”
A senior executive at a Fortune 500 company was describing a colleague to a researcher. The colleague had then been handed something he had never faced before, an unprofitable operation. After that, the executive said, he “seemed to go on a downward spiral,” more remote each day, unable to work with key subordinates.
Why had he derailed? Partly luck, the executive said. Partly the specialists he had surrounded himself with. “And some of it was that he never learned to delegate. He had no idea of how to lead by listening.”
Morgan McCall and Michael Lombardo of the Center for Creative Leadership collected about 20 stories like this one, and about 20 of executives who made it to the top. They published the results in January 1983. The quick explanations, they noted, were the Peter Principle or a fatal flaw. What they found was that those who derailed “differed only slightly from those who reached the top.”
Six turns
Both groups were, in the report’s words, “incredibly bright,” spotted early, with outstanding track records. The derailed were often remembered as technical geniuses. What tripped them was the job changing underneath them. “After a certain point,” the authors wrote, “managers cease to do the work themselves, and become executives who see that it is done. Some never made this transition.”
Eighteen years later, Ram Charan, Stephen Drotter and James Noel turned that transition into a map. Their book, The Leadership Pipeline, describes six passages: from managing yourself to managing others, then to managing managers, then to running a function, a business, a group of businesses and finally a whole enterprise. The model, they write, was first developed by Walter Mahler, a human resources consultant, and called Critical Career Crossroads.
Each step up changes which work counts as yours, and what you have to let go.
The authors break each passage into three shifts: skills, how you spend your time, and what they call work values, your sense of what counts as real work. A first-line manager who still counts only code as real work may be the first to cancel one-on-ones when a deadline looms. Two levels up, the same habit can show up as a boss who reaches past his managers to direct their people. The CCL report has one: an external affairs executive who knew little about government regulation tried to direct a subordinate with 30 years in the field. The expert balked, and the boss lost a battle, the report said, “that should never have begun.”
The question that skipped a level
Imagine you now run three team leads, after years of leading one team yourself. An engineer from one of those teams stops you in the hallway with a design question. You know the answer. It takes you five minutes.
His lead hears about the decision a week later, in a standup. The next month two more engineers from that team bring their questions to you, and a fourth stops asking anyone. You have become the fastest route to an answer, and your lead has become a step people skip.
The bottom rung
The pipeline assumes a first passage: someone who did the work well enough to be asked to manage it. That intake may be narrowing.
Erik Brynjolfsson, Bharat Chandar and Ruyu Chen have been tracking payroll records from ADP, which cover millions of American workers. In their August 2026 revision, employment of 22-to-25-year-olds in AI-exposed occupations stood 19% below where it would be had it kept pace with less-exposed peers. Experienced workers in the same occupations showed no comparable gap. The shortfall came mainly from fewer young people being hired, and it was concentrated where AI is used to do tasks rather than help people do them.
The authors are explicit about what this isn’t. These are “early, descriptive indicators,” not causal estimates. The pattern weakens when they control for education, some divergent trends began before generative AI arrived, and it is stronger in their sample than in national surveys.
For the pipeline, the open question is where the next first-line managers learn the work they will one day manage. You can check that in your own team. List what your two most recent junior hires did in their first six months. Mark the tasks an AI tool now does. For each mark, write down where a newcomer now sees that work done from start to finish. A blank line is a place to look.
Where the map runs out
The Leadership Pipeline is a practitioners’ framework built from the authors’ consulting and corporate work, not a model tested against outcomes. The 1983 report is closer to evidence, but thin: about 19 insiders recalling about 40 careers at a handful of companies, with every derailed executive a man. Its authors refused to turn it into a formula. “There is not one best way to succeed (or even to fail),” they wrote.
They did notice one pattern. The executives who derailed had usually succeeded at the same kind of job again and again. The ones who arrived had moved from line jobs to staff jobs and back, or started a business from scratch.
One derailed executive, in a colleague’s words, was “a talented guy and not a bad manager either.” His overmanaging, abrasive style “never allowed his colleagues to develop and never allowed him to learn from them.” He was passed over.