Week 51 Day 2: The Project You Should Have Killed Sooner

Think about the project. What were the early warning signs you ignored? What kept you from pulling the plug sooner -- sunk cost, political pressure, personal attachment? And what finally made you stop? The gap between when you should have stopped and when you actually stopped is the cost of the lesson. The question is whether you have internalized the lesson well enough to stop the next one sooner.

Here is the anatomy of a project that should have been killed sooner, and the specific forces that keep dead projects walking. Force one -- sunk cost fallacy. You have already invested six months, three engineers, and a significant amount of organizational credibility in the project. Killing it means all of that investment is 'wasted.' Except it is already wasted -- the investment is gone whether you continue or stop. The only question is whether you invest more into a failing project or redirect that investment toward something with a better return. The sunk cost fallacy is one of the most well-documented cognitive biases, and it affects leaders disproportionately because leaders have more personal identity attached to the projects they sponsor. Force two -- escalation of commitment. Related to sunk cost but psychologically distinct. Escalation of commitment occurs when you increase your investment in a failing course of action specifically because you have already invested. The logic is: 'We have come this far, we cannot stop now.' 'One more sprint and we will get it working.' 'If we just add one more engineer...' Each additional investment raises the stakes, which makes subsequent abandonment feel even more costly, which produces even more investment. The cycle continues until the project fails catastrophically or someone outside the project forces the decision. Force three -- optimism bias. You believe the next milestone will be the one where things turn around. The team will figure it out. The technology will work. The market will shift. This optimism is often genuine -- you are not lying to yourself, you are miscalibrating your probability estimates in a self-serving direction. Force four -- identity threat. You championed this project. Your judgment is on the line. Killing the project feels like a public admission that your judgment was wrong, which threatens your identity as a competent leader. Force five -- team loyalty. The team has been working hard. Killing the project feels like telling them their effort was meaningless. You keep the project alive partly to protect the team's morale, not realizing that the team already knows the project is failing and is demoralized by the refusal to acknowledge reality. The antidote to all five forces is a pre-commitment decision framework -- a set of conditions you define at the start of the project that will trigger a kill decision. 'If we have not achieved milestone X by date Y, we will stop.' 'If the market research shows less than Z demand, we will pivot or stop.' 'If the technical prototype does not meet the performance threshold by week 8, we will stop.' These pre-commitments are made when you are rational (project inception) and enforced when you are emotional (mid-project), which is exactly when cognitive biases are strongest. Add this lesson to your Leadership Operating Manual: every significant project should have explicit kill criteria defined before work begins.

The sunk cost fallacy in organizational contexts is documented by Arkes and Blumer (1985) in their seminal paper 'the psychology of sunk cost,' which demonstrated through multiple experiments that individuals consistently factored irrecoverable past investments into future decision-making, even when they understood intellectually that sunk costs should be irrelevant. Escalation of commitment was first documented by Staw (1976) in 'knee-deep in the big muddy: a study of escalating commitment to a chosen course of action,' and has been extensively studied in organizational contexts by Staw and Ross (1987), who identified four determinants of escalation: project determinants (the difficulty of recognizing failure before it is complete), psychological determinants (sunk cost and self-justification), social determinants (saving face and maintaining consistency), and organizational determinants (political dynamics and institutionalized inertia). Their research found that escalation was most severe when the decision-maker had personal responsibility for the initial investment decision -- exactly the situation of a leader who championed the project. The pre-commitment solution implements what behavioral economists call 'Ulysses contracts' (Elster, 1979) -- named after Odysseus binding himself to the mast before encountering the Sirens. The principle is that individuals can improve their decision quality by making binding decisions in a rational state that constrain their behavior in a future emotional state. Research by Rogers, Milkman, and Volpp (2014) demonstrated that pre-commitment mechanisms improved decision quality by 30-50% in contexts where cognitive biases predictably distort judgment, because the pre-commitment shifts the default from 'continue unless I actively decide to stop' (which favors escalation) to 'stop unless the predefined criteria are met' (which favors rationality).