Several days before I was removed as CEO of a company I had built, someone close to the science told me the move was being prepared. I had time. Not much, but enough. I spent it convincing myself there was nothing to be done.
I promised that the final article in this series would be the red flags I missed. I have been trying to write that article for weeks, and I kept producing lists. Categories of warning signs. Frameworks for response. All of it true, none of it honest. Because the red flags I missed were not signals I failed to detect. They were signals I detected and chose not to act on, because acting would have meant having conversations I did not want to have, with people I had decided were too powerful to challenge.
So this final article is not a list. It is the story of one phone call, the reasons I never made it, and what I have since understood about when the price of that call was actually set.
The Warning
Here is what happened, in the only version I can tell.
A company I led was approaching a governance crisis. I knew, because someone with no reason to lie to me said so, plainly, in the days before it arrived. The move was being prepared. The outcome was not yet fixed.
I want to be precise about what I had available to me in those days, because the precision is the point. There was a board member whose investment and whose trust I had earned over years, who would have stood with me if I had asked. There was a director whose financial position made him quiet, but quiet is not the same as unreachable; he saw what was happening and had never been given a reason to believe anyone would back him if he said so. There was a coalition possible. A counter-position possible. A real contest possible, with a real chance of a different outcome.
I did not pick up the phone. Not to any of them.
What I did instead is what I have since learned most CEOs in that position do. I treated the warning as a description of something already settled. I rehearsed the reasons the other side would win: more experience, better networks, more standing with the institutions that mattered. All of that was true. None of it was a reason not to fight. I converted a warning into a verdict, and then I waited for the verdict to be read out.
The company did not survive the sequence that followed. Neither did the science. I will come back to what that cost, because it is the only part of this story that ultimately matters.
Why Smart People Freeze
I have spent a long time on the question of why I did nothing, partly because the obvious answers are wrong. I was not naive. I was not inexperienced. I had sat in boardrooms for many years and could read power as well as anyone in the building. The freeze was not a failure of perception. It was a failure with a structure, and I have since seen the same structure in other CEOs often enough to describe its parts.
The first part is status asymmetry. When the threat comes from someone senior to you in experience, reputation, or network, your own assessment of the situation quietly downgrades itself. You assume their position is stronger than it is because their CV is stronger than yours. I have watched capable CEOs defer to reputations in moments when the reputation was the only thing on the other side of the table. The voice that sounds unanswerable in the days before a crisis is almost never as unanswerable as it appears. You only discover that by answering it.
The second part is what I now think of as the settled-outcome delusion. A warning arrives, and instead of hearing it as an invitation to act, you hear it as advance notice of a result. This is a category error with a deadline. Red flags whisper before they shout, and the whisper is the entire value of the warning: it is the interval in which the outcome is still being assembled, consensus is still being gathered, and a single well-placed conversation can change what gets assembled. Treat the whisper as a verdict and you have voluntarily skipped the only phase in which you had moves.
The third part is hope. Maybe it will resolve itself. Maybe the relationship will recover. Maybe I am misreading this. Hope is a fine thing in drug development, where the data will eventually tell you whether it was justified. In governance, hope is not a strategy. It is a sedative. Every CEO I have known who was destroyed by a governance failure saw the warning signs. The pattern is consistent: recognition arrives early, and action arrives late or never, with hope filling the gap between them.
I had all three running at once. The phone stayed on the desk.
The Price Was Set Years Earlier
For a long time I told this story to myself as a story about those few days. The warning, the freeze, the outcome. A failure of nerve at the decisive moment.
I no longer believe that is the right account. Here is what I’ve learned: the phone call felt impossible not because of the days I had, but because of the years I had not used. By the time the whisper came, every structural decision that would have made the call unnecessary, or made it winnable, had already been taken. Most of them had been taken casually, in periods when governance felt like paperwork, by people focused on the science and the money.
This series has described five failure modes. Walk each of them backwards and you arrive at the same place: a moment, early in the company’s life, when the failure became probable, and when preventing it would have cost almost nothing.
The board that cannot resist a power play was composed at the term sheet. I wrote in the first article of this series that the cap table creates the board, and that boards designed for the convenience of the financing protect incumbents rather than missions. Nobody ambushed by their board was ambushed by a board built deliberately. They were ambushed by a board assembled as a by-product of a fundraise, seats allocated like loose change at the end of a valuation negotiation. The composition conversation costs an awkward hour before signing. Its absence costs you the room when the room is all that matters.
The Chairman relationship that turns hostile was never defined when it was friendly. I’ve watched CEO-Chairman relationships fail because both parties operated from different unspoken models of the role, each convinced the other was overreaching, each behaving correctly by their own definition. The fix at the start is a conversation and a page of writing: let us write down what each of us thinks the other does. A Chairman who declines that conversation has answered a question you had not yet thought to ask. By the time the models collide, there is no page to point to, and the contest is settled by power rather than agreement.
The scientific voice that gets silenced was never given decision rights when the founder had leverage. I’ve seen a scientific founder go from controlling shareholder to excluded from his own research, and the exclusion was not accomplished in the moment it became visible. It was accomplished at every earlier moment when his authority over scientific direction could have been written into the structure and was not, because writing it down felt unnecessary among people who trusted each other. Trust is not a structure. The third article in this series exists because of what that costs.
The platform that gets amputated in the name of focus was never structurally defended. When the word arrives in a boardroom, the question of what is being given up should already have an owner, a forum, and a paper trail. Where it does not, focus stops being a strategic choice and becomes a story imposed by whoever wants the simpler narrative. I have watched optionality worth more than the company’s entire raise disappear in a single meeting because nobody had ever been made accountable for articulating it.
The founder destroyed in transition signed away the protections years before the transition began. Board seat, voting rights, equity protections, decision rights: every one of them is cheap to negotiate when the founder is the company, and unobtainable once the founder is a problem to be managed. The previous article in this series told the story of what gets lost. All of it was losable because none of it was secured.
Five failures. Five early moments. In every case, the conversation that would have prevented the catastrophe was available years before the catastrophe, at a price measured in discomfort rather than money.
The Cost Curve
This is the part I most want discovery-stage and Series A founders to take away, because you are reading this at the only point in your company’s life when the following is true.
Governance has a cost curve, and it compounds in both directions. At incorporation, the cost of getting it right is essentially zero: some thought, some writing, some conversations that feel premature. At the term sheet, the cost is awkwardness: questions that feel rude to ask of investors and directors you are trying to close. At Series B, the cost is a crisis: renegotiating structures that now protect people with the power to refuse. After that, the cost is the company.
Most founders price governance at the moment they encounter it, which is to say they price it at zero, because in the early months it produces nothing and prevents nothing visible. The compounding is invisible until it is unpayable. This is the inversion the industry does not say out loud: the period when governance feels least necessary is the period when it is being permanently determined.
The conversations themselves are not complicated. Three of them, if early and repeated honestly, would have changed the ending of nearly every failure described in this series. With a director, before appointment: walk me through the last time you disagreed with a CEO you backed, and what you did about it. With a Chairman, at the start: let us write down what each of us thinks the other does, and revisit it every six months. With your scientific team, always: tell me what I am wrong about. And when information arrives in your boardroom by a route you did not provide, the line is not aggressive, merely curious: help me understand the path that information took to you. None of these costs more than a few minutes of discomfort. I have now watched what the alternative costs.
What the Whisper Requires
I said I would come back to what my silence cost, and I owe you that before the end.
The science did not survive what followed the call I never made. Work that should have become medicine became an entry in an insolvency filing. Patients who needed what we were building did not get it, and will not, because the years and the data and the team cannot be reassembled. The people I could have called have, I hope, forgiven me. I am not sure the patients would, if they knew what had been possible.
What I carried out of that experience is not a framework. It is a sentence. When you hear the whisper, you have less time than you think, and more standing than you believe. I run a company now with that sentence as background music, and the structures this series has described, the ones I once treated as paperwork, are the first things I built.
If you have read this series from the beginning, you have walked through five ways good science dies in the boardroom. Underneath all five is the same shape: someone saw something early, and the conversation that could have changed the outcome never happened. The cost of the difficult conversation is almost always lower than the cost of the conversation you never had. That is the only governance principle that matters. Everything else in this series follows from it.
Every biotech that fails through governance is a medicine that never reaches the patients who needed it. The science is hard enough. Build the structures while they are cheap, and make the call while it is still a whisper.
This is the final article in a six-part series on biotech governance. To read the series from the beginning, start with Board Composition That Actually Works.
I write Beards on Biotech because I have seen what happens when governance fails, and because I am still in the arena trying to do better. I have told you about the call I didn’t make. If you have one of your own, I would like to hear about it.