One Finished Thing Beats Ten Rough Ones

I went through an old drive in January and counted nineteen project folders.

Four of them were finished. The other fifteen were somewhere between an idea and an apology.

The thing that got me was not the number. It was that I could remember being excited about all fifteen, and I could not remember abandoning a single one. There was no decision. They just got quiet, one at a time, while something newer got loud.

The math is worse than it looks

Ten projects at eighty percent are not worth eight projects.

They are worth zero.

Value gets delivered at the end, all at once, and not one minute before. Nobody drives across eighty percent of a bridge. Nobody uses the app that almost handles logins. A book at ninety percent is not a book, it is a document, and the difference between those two things is the entire last chapter.

This is genuinely counterintuitive because effort feels like it accumulates. You put in the hours, the hours were real, the work got better. It just does not convert into anything until the thing crosses the line.

Which means the fifteen folders on that drive represented a real chunk of my life converted into precisely nothing.

Why the last ten percent is so hard

Starting is the best part and everyone knows it. Blank page, no constraints, no evidence yet that your idea has problems. Pure upside.

The last ten percent is the opposite of all of that. It is naming things. It is the edge case where somebody puts an apostrophe in the field. It is writing the two paragraphs of instructions, fixing the layout on a phone, deciding what happens when it fails.

None of it is fun and none of it makes a good story. And critically, none of it changes what the thing is. It only changes whether it works, which is a much less exciting axis to move along.

So it gets deferred. And a project you have deferred twice is a project you have started to resent, and resentment is where things go silent.

What finished actually means

I needed an actual definition, because “done” is negotiable and I will negotiate.

Here is the one I use: someone other than me can use it without me explaining it.

That is the whole standard. Not perfect, not polished, not everything I imagined in week one. Just usable by a person who was not in my head while I built it.

It is a brutal line because it moves the goalposts from my experience to somebody else’s. I can convince myself something is basically done. I cannot convince myself that a stranger will figure out the unlabeled button.

It also kills scope creep neatly. Once “usable by someone else” is the bar, every extra feature has to justify itself against a finish line I can actually see.

Unfinished things are not free to hold

This is the part I underrated for years. I assumed an abandoned project was neutral. It sits there, it costs nothing, maybe I come back to it.

It is not neutral. Each one takes up a small amount of standing rent in your head. You think about it in the shower. It comes up when someone asks what you are working on and you give the answer that includes it, which is a small lie you then have to keep telling.

Fifteen of those running at once is a meaningful background load, and you do not notice it until you clear a few and feel how much quieter it gets.

Killing counts as finishing, by the way. That is the piece I want to be loud about. Deleting the folder and admitting it is over closes the loop just as cleanly as shipping does. What does not close the loop is leaving it there in case you feel like it later.

The lessons live in the last ten percent

Here is the argument that actually changed my behavior, more than any productivity reasoning.

You do not learn the hard parts by starting. You learn them by finishing.

Ten rough starts teach you the same lesson ten times, which is how to begin. You get very good at the exciting part and you never once find out what happens next.

One finished thing teaches you what nothing else can. What breaks under real use. Which of your clever decisions was actually a problem. How long the boring part takes, which is the single most useful number a builder can have, and you can only get it by living through it.

Every ounce of judgment I have about building things came from the four folders, not the fifteen. It is not close.

What I do now

One at a time, to the definition, before the next one starts. If a new idea shows up and it is genuinely better, then the current thing gets formally killed rather than quietly parked. Out loud, on purpose, so I know the count.

I also stopped trusting my own sense of how close something is. Eighty percent done means about half done, every time, without exception. I plan for that now instead of being annoyed by it.

And I keep the definition posted where I can see it, because on the day the work stops being fun I will absolutely try to redefine “finished” to mean whatever I have in front of me.

A finished thing goes off and has a life of its own. An unfinished one just quietly spends yours.