Every week, I talk to leaders who are stuck in the same spot. They have a project that is not working. They know it. But they cannot bring themselves to kill it, and they cannot bring themselves to fully commit to it either. So they do the worst possible thing: they keep it on life support, too weak to win but too visible to die.
The project stays on the budget but gets half the resources it needs and none of the attention. Enough to keep it breathing, never enough to find out if it could actually work. In portfolio management circles, these are called zombie projects, and most organizations have at least one.
I have been on both sides of this. I have killed things too early and regretted it. I have held on to things too long and paid for it. And I have sat in that middle ground where you are not really building and not really quitting, just slowly bleeding time and money while telling yourself you are being patient.
That middle ground is where most of the damage happens.
The Sunk Cost of Not Deciding
Leaders tend to frame this as a binary: keep going or quit. But there are actually three options on the table at any given time, and each one has a price tag. Most conversations about sunk cost fallacy focus on the cost of staying too long. That is only one third of the picture.
The cost of staying too long
Resources locked into something that is not producing. Team energy draining. And the quiet cost that is harder to measure: what those same resources could be doing if they were pointed somewhere else. There is also a credibility problem. Your team can usually see what you cannot, or will not, and every month you keep pushing a dead initiative, you spend a little more of their trust.
The cost of pivoting too often
Every pivot flushes the context your team has built up: the customer conversations, the product instincts, the hard won understanding of what works and what does not. It burns trust because people start to wonder if anything they build will survive longer than a quarter. Organizations that pivot constantly never develop deep competence in anything.
The cost of indecision
You are not fully in and not fully out. The project gets enough budget to exist but not enough to succeed. You are paying the costs of staying and the costs of leaving while adding a third tax on top: the confusion and paralysis that comes from nobody knowing the plan. And your best people hate working on zombie projects. High performers want their work to matter.
Your job as a leader is not to avoid cost. You cannot. Your job is to choose which cost you are willing to pay and pay it on purpose rather than by default.
The hardest part of avoiding the indecision tax is that we tend to treat projects as a single, indivisible bet. "Is this working or not?" But to make a clean call, you have to take the engine apart.
A Decision Making Framework: Validate Your Original Thesis
When something is underperforming, the instinct is to evaluate it as a whole. That question is too blunt to be useful. You need to break your original bet into its components and look at each one separately, because the whole thing is rarely all valid or all invalid.
When you started this project, you were making a set of assumptions, whether you wrote them down or not:
1. The Problem
Does your audience still experience this pain the way you originally understood it? Has the problem shifted, or shrunk, or been solved by someone else while you were building?
2. The Audience
Are you still serving the people you set out to serve? Or have you drifted toward a different group without noticing?
3. The Mechanism
Is your approach to solving this problem still sound, or has enough changed around you that your delivery model needs rethinking?
4. The Timing
Were you early, late, or right on time? A valid thesis with bad timing looks identical to a bad thesis, and the two require very different responses.
5. The Economics
Can this work at a cost structure that makes sense? Sometimes the thesis is right but the unit economics will never close, and no amount of persistence changes that.
The reason this matters is that it gives you the ability to make partial decisions instead of all-or-nothing ones. Maybe the problem and the audience are still right, but the mechanism needs to change. That is not quitting. That is evolving with clear information. Or maybe the problem itself has disappeared, and continuing is not courage, it is avoidance.
Try it now
Walk through each of those five components. Mark each one as "still true," "partially true," or "no longer true." If most of the thesis holds, you are probably in persistence territory. If the foundation has shifted underneath you, staying the course is just a habit wearing the costume of a strategy.
How to Know When to Pivot and When to Persevere
The pivot or persevere question is the part that makes the decision hard. When you are in the middle of it, both persistence and stubbornness feel like commitment. Both feel like doing the right thing. The difference only becomes obvious in hindsight, which is not helpful when you need to make the call now.
There are two diagnostic questions I come back to.
Is the underlying thesis still valid?
Not "do I still believe in this," because belief is sticky and unreliable. When you look at the evidence from the last six months, does the data support the idea that you are solving a real problem for real people?
Are you learning or just enduring?
Persistence generates new information with each cycle. Stubbornness looks like doing roughly the same thing with minor variations and waiting for different results. If your last three iterations have not produced meaningfully new insight, that is a signal.
Set a Kill Criteria Before You Need One
The single most useful thing you can do when you are in this gray zone is set a concrete milestone before the emotional pressure of the decision arrives. I am talking about a specific test with a specific timeline and a specific consequence.
Most leaders skip this step, and the reason is not laziness. It is fear. Setting a kill criteria means accepting, in advance, that you might be wrong. But that is precisely why it is valuable. A pre-set criteria turns a potential failure into a completed experiment, and those are very different things.
Here is how to make it work.
Define it now, not later
If you wait until you have seen the results to decide what counts as success, you will rationalize. Write it down today. The act of committing something to paper forces a clarity that keeping it in your head does not.
Make it falsifiable
"We need to see more traction" is not a milestone. It is a wish. "We need 15 paying customers at our target price point within 90 days" is something you can actually measure.
Decide what happens if you miss it
"If we do not hit X by this date, we sunset the project and reallocate the team." Without that pre-committed consequence, the milestone is just another number you renegotiate when things get uncomfortable.
Watch the leading indicators
The milestone is your destination, but you should know what the road looks like on the way there. Are conversations getting warmer? Is the pipeline building? Or is every signal flat?
The milestone is not a deadline for the project to succeed. It is a deadline for you to have enough information to make a clear decision.
That reframe matters, because it turns a pass/fail moment into an information gathering exercise, and that is psychologically easier to commit to honestly.
Stuck on a decision like this?
A Clarity Day gives you the structured conversation and outside perspective to make the call with confidence — in a single day.
Learn About Clarity Day