The program was three years old and, by the honest reckoning of nearly everyone in the room, finished. It had missed its second set of targets, its original sponsor had moved to another division, and the market it was built to capture had reshaped itself twice since the business case was signed. The quarterly portfolio review reached its line item late in the afternoon, when attention was thin and everyone wanted to be somewhere else. There was a slide. There were numbers arranged to be survivable. There was a brief, tired discussion in which two people said careful things about giving it another couple of quarters to prove itself, and nobody said the sentence everyone was thinking, which was that it should have been shut eighteen months ago. Then it was renewed, and the meeting moved on to the next line, and the budget for the coming year quietly absorbed a project that no one in the building actually believed in.
Notice what did not happen. Nobody argued for it. Nobody, in any real sense, decided to continue it. Continuation was simply what occurred when no one was willing to be the person who ended it. The program did not survive because it won an argument. It survived because killing it required a volunteer, and the room was fresh out.
An organization knows how to start. It does not know how to stop.
Every initiative in a company arrives with a full set of accountability scaffolding built around its birth. Someone championed it. Someone signed the case. Someone's name is attached to the win it was supposed to deliver, and that person has every incentive to see it through, because its success is their success. Starting is a well-owned act. It has a proud parent and a paper trail, and both of them show up to defend it.
Stopping has none of this. To end a live program is to stand up in a room and, in effect, author a short document titled the original bet was wrong. Nobody wants their name on that document, least of all the person who made the original bet, who is frequently still in the building and frequently now more senior than they were when they made it. So the machinery that assembles so eagerly around a launch simply fails to assemble around a shutdown. The decision to stop has no natural owner, no sponsor, no scaffolding. It is an orphan. And orphaned decisions do not get made. They get deferred into next year's budget, where they are perfectly safe, because a renewal implicates no one while a cancellation implicates someone specific.
The reversal has a price, and nobody wants to pay it.
This is the asymmetry that quietly bankrupts portfolios. Continuing a failing project is the default, and the default requires no signature, no courage, and no confession. Ending it requires all three. The person who proposes the stop does not get credited with the money they save, because saved money is invisible and abstract and never lands as a line anyone can point to. What they get, visibly and immediately, is ownership of the admission that a thing the company committed to, staffed, and defended in public was a mistake. They absorb the reversal so the organization can keep the savings. Almost no incentive structure on earth rewards that trade, so almost nobody takes it.
And so the failing thing acquires a strange kind of immortality. Not because it is strong, but because it is protected by the discomfort of the very people who would have to be the ones to kill it. Each quarter it is easier to renew than to bury. Each quarter it becomes a little more embedded, a little more someone's whole job, a little more a thing that other things now quietly depend on. Sunk cost is usually explained as a glitch in the individual brain, a bias you can train people out of with a good workshop. That is only half of it, and the smaller half. The deeper reason failing projects run forever is structural. Continuation is nobody's decision, and termination is somebody's fault, and an organization will always drift toward the option that leaves no fingerprints.
The company was not funding the project anymore. It was funding its own reluctance to write the obituary.
You are not bad at killing projects. You made killing them unsurvivable.
Here is the inversion that should ruin a portfolio review or two. Leaders tend to believe their organization has a discipline problem, that it needs to be more ruthless, to prune harder, to develop some killer instinct about its own commitments. This gets the diagnosis exactly backward. The problem is not a shortage of ruthlessness. It is that the organization has, without ever meaning to, turned the act of stopping something into a personal career risk while leaving the act of continuing it a completely free pass. You did not build a company that cannot decide. You built one in which one direction of the decision is fully insured and the other is standing outside in the rain with no coverage at all. Your people are not cowards. They are reading the incentives correctly, and the incentives are unambiguous: the person who ends the doomed program will be remembered as the one who ended it, while the person who lets it limp along for another year will not be remembered at all. Anonymity is the safer bet, and capable people take the safer bet every single time.
What it looks like to design a stop.
Rebuilding this does not begin with more conviction, which is only willpower asked to do the work that architecture should have done. It begins by giving the stop the same scaffolding the start always had. Every serious commitment should be born with its own kill criteria written down at the outset, back when nobody is defending anything yet, naming the specific conditions under which the organization will end it and the specific person whose job it is to make that call when those conditions arrive. A predefined stop has an owner before the project even exists, which means executing it later is not an act of individual bravery but the routine completion of a decision the organization already made together. Continuation should be forced to cost something too. If every renewal demanded a live signature and a real argument each year, the way the launch once did, the failing programs would stop surviving on silence, because silence would no longer be enough to keep them breathing.
This is the pattern that surfaces when SSUNDAR takes apart how an organization actually governs its own commitments, as opposed to how its stage-gate diagram insists it does. The revealing question is never how do you approve things. Every company can answer that one fluently, with a flowchart. The revealing question is who in this building is allowed to stop something, and what does it cost them personally when they do, and the answer is almost always a long silence followed by a name that nobody in the room wants to be. Fix that, and the judgment that was in the room the entire time stops being held hostage by the fear of owning a reversal. Leave it unfixed, and the portfolio will go on filling quietly with the undead, each one renewed by no one, ended by no one, and paid for by everyone.
Any organization can start something. The ones worth trusting are the ones that can still bury it.