The company has outgrown its founder, and everyone knows it except the org chart. Decisions queue outside one office. Executives manage upward instead of outward. Strategy has narrowed to the set of moves the founder can recognize as extensions of past victories. Nobody says any of this, because the founder is also the founder: the reason the place exists, the largest shareholder of its mythology, and often the person who signs off on whether the problem exists.
The staying-too-long founder is a stock character, easy to mock from outside and almost impossible to dislodge from inside. It is worth being precise about why, because the mechanism is not vanity. It is physics.
The attribution machine
In 1985, Meindl, Ehrlich and Dukerich named the force: the romance of leadership, our systematic tendency to over-attribute organizational outcomes to the individual at the top. In their studies, the same executive was rated far more charismatic when the company happened to perform well; the attribution intensified when other explanations were scarce. Founders live inside this machine longer and more intensely than anyone. Every success for thirty years has been narrated, by employees, journalists, and eventually the founder, as evidence of personal magic. Asking such a person to conclude they are now the constraint is asking them to override three decades of unanimous feedback.
The tragedy is that the machine also runs in reverse after the overstay: the same attribution error that inflated the founder's indispensability will assign them the blame for the stagnation, personally and totally.
What does the record say about the succession itself? Schepker and colleagues' meta-analysis of 13,578 CEO successions found a short-term performance dip and no significant direct long-term effect; long-run outcomes ride on the successor's origin and the changes that follow, with insiders on average improving long-term performance. The founder's fear ("the place will fall apart without me") and the board's hope ("a savior will transform everything") are both, on the evidence, romances. Companies survive successions routinely. What they survive less well is the years of drift purchased by postponing one.
The structural remedy
If overstay is systemic, the correction has to be installed in the system, preferably by the founder, preferably early, when the founder's power is at its peak and the clause costs nothing.
Semco offers the sharpest existing example of what installation looks like. As Ricardo Semler has told it, everyone there evaluated their boss anonymously twice a year, and a low score twice in a row meant losing the function, a rule that applied to Semler himself. Whoever controls other people's days, the logic ran, has to keep deserving it. The detail that matters for founders is the inclusion: the owner submitted his own managerial role to the same recurring test as every supervisor. The role, not the ownership; capital and function were cleanly separated, which is exactly the joint at which founders typically weld themselves in place.
A founder governed by such machinery does not need to win an argument about whether they have stayed too long. The machinery holds the argument on schedule, twice a year, with data.
Short of the full Semco apparatus, the installable parts are: a real evaluation of the founder-as-executive by the people who experience the management, feeding to a board empowered to act on it; a designated successor given genuine load early, on the relay pattern the succession research favors; and a pre-committed exit design, written while leaving is hypothetical, because the founder who negotiates their departure at 72, under pressure, negotiates against themselves.
The founder's actual asset
The final reframe is about what the founder is protecting. The overstay treats the role as the asset. But the role is the one thing the founder holds that depreciates with every additional year of holding it. What appreciates is everything else: the ownership, the institutional memory, the moral authority to bless the successor's changes, the standing to say "we don't do that here" once a year and be heard. A founder who leaves the role on time keeps all of it. A founder who overstays converts it, slowly, into the polite silence of people waiting.
The question is not whether the founder is still capable. Often they are. The question is what the role could do in other hands, and that question has a fair judge only if the founder built the courtroom before they were the defendant.
Sources
- Meindl, Ehrlich & Dukerich (1985). The Romance of Leadership (Administrative Science Quarterly 30, DOI 10.2307/2392813). Observers systematically over-attribute organizational outcomes to leaders; perceived charisma tracks firm performance; attribution intensifies when alternative explanations are scarce. https://doi.org/10.2307/2392813
- Schepker, Kim, Patel, Thatcher & Campion (2017). CEO succession, strategic change, and post-succession performance: a meta-analysis (The Leadership Quarterly, DOI 10.1016/j.leaqua.2017.03.001). 13,578 successions: short-term dip, no direct long-term effect; insider successors on average improve long-term performance. https://doi.org/10.1016/j.leaqua.2017.03.001