The sunk cost trap of side projects, and when to stop building
A reflective essay on the silent killer of indie hackers. How to spot the trap, do the opportunity cost math, and give yourself permission to let go.
I once spent fourteen months on a project that I knew, somewhere underneath the daily commits, was already dead. I just could not say it out loud. Every Saturday morning I would open the laptop, scroll past the unread emails about it, and tell myself the same thing: just two more weekends.
The graveyard of indie projects is mostly built out of those weekends.
The silent killer
Sunk cost is the quiet voice that tells you the four hundred hours you have already poured in mean something. Logically you know they do not. Those hours are gone. The only question that matters is what the next hour is worth, and whether this project is the best place to spend it.
But sunk cost does not show up as logic. It shows up as identity. After enough weekends, the project is not a thing you are building anymore. It is a thing you are. To stop is to admit something about yourself that you are not ready to admit.
That is why the trap is so effective. It does not feel like a trap. It feels like commitment.
The signs you are in it
I have a small collection of patterns from founders who eventually listed on Failedups, and from my own past projects. The signs cluster:
You avoid telling anyone what you are working on. Not because it is a secret. Because explaining it has started to feel embarrassing. The pitch you wrote eighteen months ago has not survived contact with reality, and you have not written a new one.
You have more “almost there” moments than ship dates. The launch is always six weeks out. It was six weeks out in October, and it was six weeks out in February.
Your weekend coding has become a chore. The thing you used to look forward to after the day job is now the thing you procrastinate on. You finish the week tired, sit down on Saturday, and feel something close to dread.
You are rebuilding the auth flow for the third time. Or the database schema. Or the design system. The work you are choosing is the work that feels like progress without requiring you to face the actual question, which is whether anyone wants this.
If two of those are true, you are probably not building a product anymore. You are paying interest on a debt you took out a long time ago.
The math nobody wants to do
Here is the calculation I avoided for fourteen months.
Every hour I spent on that project was an hour I did not spend on something else. That is the part founders intellectually understand. What they avoid is naming what the something else actually is.
For me it was sleep. It was a daughter who wanted to play Lego on Saturday afternoons. It was the next idea, the one that kept tugging at me from a corner of a notes app, that I would not let myself open because I had already “committed.”
If you genuinely do the math, the question is not “is this project worth my time?” The question is “is this project worth more than my best alternative use of that time?” And the best alternative is almost never another version of the same project. It is your kids. Your sleep. The idea you have not let yourself start. A long walk where your brain finally gets to defragment.
That math is brutal, which is why most of us refuse to do it.
The freedom on the other side
Failedups exists, in part, because of how often I have heard the same sentence from founders after they list: I feel lighter than I have in two years.
Not “I am rich now.” The deals close in the $500 to $10k range. Most of the lightness is not financial. It is the closing of a tab in your head that has been open since 2023. The fact that the project, finally, belongs to someone else’s brain. Someone who is excited the way you used to be.
A few sellers tell me they cried after the listing went live. Not from grief, from relief. There is something genuinely tender about handing a thing you built to someone who wants it more than you do. It is a better ending than the slow rot of an unmaintained Vercel deployment.
You do not even need to sell. Closing it counts. Archiving the repo, redirecting the domain, writing a short post about what you learned, and walking away clean is a perfectly honorable ending. Listing it just gives the project one more chance, and gives you a small amount of money to mark the moment.
How to actually decide
I wrote a piece a few weeks ago about why most side projects die, and the three that survive. The short version: surviving projects have a tight distribution loop, a meaningful first user inside the founder’s own life, and brutal scope discipline. If you cannot honestly tick those three, and you cannot add them in a single weekend of focused work, the project is telling you something.
Listen to it.
The decision is not “is this idea good?” The idea was probably fine. The decision is “is the version of me, on the calendar I actually have, going to ship this?” That is a different and more honest question.
If the answer is no, you are allowed to stop. You do not need permission, but in case you were waiting for some, here it is: you can let go. The hours you spent are not wasted. They taught you things, and they will compound into whatever you build next, on a project you are not already exhausted by.
The graveyard is full. You do not have to add another headstone with your own hands.
If you are ready, list your project. One conversation, one honest founder note, and the tab in your head finally closes.