How to finish side projects (instead of starting new ones)

You finish a side project by cutting it down to the smallest version you'd still call finished, writing down exactly what "done" means, and giving it a fixed weekly slot that doesn't depend on inspiration. Then you ship it while it still embarrasses you a little, or you end it on purpose. Most side projects simply fade, because the scope keeps growing and nothing ever forces the finish.

Why side projects die in the middle

The start of a project is the best part. The idea is fresh, every session shows visible progress, and you get to imagine the finished thing. Then comes the middle: the bug that won't die, the chapter that doesn't work, the bridge of the song you can't fix. Progress slows down, and a brand-new idea shows up looking exactly the way this one did in week one.

You'll also misjudge how long it takes, like everyone. In a 1994 study by Buehler, Griffin and Ross, students predicted they'd finish their senior thesis in about 34 days. On average it took them about 55, a week longer than their own worst-case estimate, and only around 30 percent finished by the date they'd predicted. Programmers have a joke about this, credited to Tom Cargill of Bell Labs: the first 90 percent of the code takes the first 90 percent of the time, and the remaining 10 percent takes the other 90 percent.

None of this means you're bad at projects. It means the default plan, "I'll work on it when I have time," is built to stall right in the middle.

That shiny new idea is your real competition, so give it somewhere to go. Keep an ideas file, and the moment a new project shows up, write it there with one line about why it's exciting. Then go back to the current one. One rule makes this work: only one side project is active at a time, and a new one starts only when the current one is finished or deliberately ended. If the new idea still excites you after that, it was a good one. Plenty won't, and the file will show you which.

Cut the scope until it stings

The most important finishing skill is cutting. List everything the project "needs", then cross out everything it doesn't need to count as finished:

• An app becomes one screen that does one thing well.
• A novel becomes a short story with the same characters.
• An album becomes three songs, properly recorded.
• A YouTube channel becomes five videos published on a schedule.
• Redoing the bedroom becomes painting one wall and fixing the shelf.

Everything you cut goes on a "version two" list. Parking ideas there instead of deleting them makes letting go much easier, and if the first version works, that list becomes your roadmap.

Write down what done looks like

"Finished" needs a finish line someone else could see. "When it's good" is a mood. Done sounds more like this:

• the app is live at a link you can send;
• the story is submitted to one magazine or contest;
• the three songs are uploaded;
• the online course is completed and the certificate downloaded.

Write it in one sentence, put a date next to it, and keep it where you'll see it. When the definition starts drifting ("well, it should also have..."), that's scope creep, and the new idea goes on the version two list.

Give it a slot, not your leftovers

Side projects usually get leftover time, whatever remains after work, errands and everyone else. Leftover time is unpredictable and usually tired. Give the project a fixed slot instead: the same day and time every week, small enough to survive a bad week. Two 45-minute sessions every week will carry a project further than a heroic Saturday that happens whenever the stars align.

End every session by writing the very next step in one line: "fix the login error", "record the bridge", "rewrite the opening scene". The next session then starts in seconds instead of twenty minutes of "where was I?" Protecting that slot through the boring weeks is most of the game, and how to be consistent covers that part in detail. If you want a nudge with teeth, set the project up as a goal in Toxic Coach with a deadline, and the reminders will go from mild to savage as the date gets closer.

Announce the date, not the dream

Telling people helps, if you tell them the right thing. In a 2009 paper, Peter Gollwitzer and colleagues found that when other people noticed someone's intention tied to who they wanted to be, that person then acted on it less, as if the recognition gave them a premature sense of already being there. The effect showed up in people who were strongly committed to that identity.

So skip the big announcement about who you're becoming. Tell one or two people a date and a deliverable instead: "I'm sending you the first three chapters on March 1." A date someone is waiting for works as a deadline. Applause for a dream is a reward you've already collected.

Ship it before it feels ready

The last stretch is where perfectionism hides. There's always one more feature, one more pass, one more thing to fix before anyone can see it. Pick a ship date and ship whatever exists on that date, as long as it meets your one-sentence definition of done. If going public feels like too much, start small: five friends, one online community, the people who said they wanted to see it.

Finished and a bit rough teaches you more than polished and imaginary. You'll learn what people actually notice, which is often not what you obsessed over, and version two gets built on real feedback. If the endless polishing loop sounds familiar, perfectionism and procrastination is for you.

Finish it or end it on purpose

Some projects shouldn't be finished. The interest is gone, your life changed, or you already got what you came for: you learned the language, the framework, the chords. That's fine. What drains you is the third option, the project you never work on and never end, sitting in the back of your mind like a message you haven't answered.

Watch out for the sunk cost trap. Psychologists Hal Arkes and Catherine Blumer described it in 1985: the more money, effort or time you've put into something, the more likely you are to keep going, partly because stopping would feel like admitting the investment was wasted. So ask one question: if this project didn't exist yet, would you start it today? If the honest answer is no, close it deliberately. Write down what you learned, archive the files, and tell anyone who was waiting. Ending it on purpose frees the slot for something you actually want to finish. If you notice you drop most things halfway, why do I give up so easily digs into that pattern.

This week's move

Open the project you feel guiltiest about. Give it a one-sentence definition of done and a date. Move everything else to the version two list. Book two slots in your calendar for this week. Or, if you already know the answer, end it today, properly, and enjoy the space it leaves behind.