I run a one-person studio. There is no project manager, no co-founder, no engineering team to absorb the boring parts. The whole stack — code, design, music, words, marketing — is me, in a spare bedroom, after the kids are asleep. The single most useful skill I've developed over the last few years isn't a programming language or a DAW shortcut. It's a workflow for actually finishing things.
Most side projects die in the middle. Not at the start (the start is fun) and not at the end (the end is rare). They die in week three, when the novelty has worn off and the problem turns out to be three problems wearing a trenchcoat. Here is the workflow I use to drag them across the finish line anyway.
1. Scope to a single sentence
Before I write a line of code or record a single note, I force myself to write the project as one sentence with a verb and an object. "A web app that finds the actual listing agent for any property." "An ambient album that helps you fall asleep." "A book of bad poetry written by a fictional duck." If I can't finish the sentence without using the word "and", the project is too big and I split it.
This sounds obvious. It isn't. The default mode of a curious person is to keep adding — "and it should also have a forum, and a leaderboard, and an iOS app, and dark mode, and…". The single-sentence rule kills the additions before they sprout.
2. Pick the ugliest version that still works
The first shippable version of anything I build is embarrassing. It has no logo. The colors are wrong. The copy is placeholder. The audio is one take with no mastering. That's the goal. The ugly version exists to prove the thing can exist at all. Polish is cheap once the core works; rebuilding the core because you polished the wrong thing is expensive.
For software, "ugly but works" looks like: one route, one form, one button, one happy path, no auth, no edge cases, no analytics. For music, it's a rough mix exported straight from the DAW. For books, it's the manuscript pasted into a plain template.
3. Ship in public, even if no one is watching
I deploy every web project to a real URL on day one. Not a localhost demo, not a screenshot in a slack channel — a live URL I can text to my brother. The act of putting something at a real address triggers a different part of my brain. Suddenly the placeholder logo is intolerable. Suddenly the typo in the title is urgent. Public is a forcing function.
For music, "shipping in public" means uploading the rough mix to a private SoundCloud link. For books, it's exporting a PDF and reading it on my Kindle. Get it out of the environment where you made it.
4. Cap the project at one weekend of polish
After the ugly version works, I give myself one weekend — about ten focused hours — to polish. Logo, colors, copy, the empty states, the README, the meta tags, the social card. When the weekend is over, the project is done in the sense that matters: it can be shown to a stranger without an apology.
Anything beyond that ten hours is feature work, not polish, and it goes on a separate list to be triaged later. Most of that list never gets done, and that's fine. The project is alive in the world.
5. Write the boring landing page
Every project gets a landing page that explains, in plain language, who it is for and what it does. No clever copy, no animated hero, no "reimagining the way you…". A title, a sentence, a screenshot, a button. The boring landing page is what makes the project searchable, shareable, and explainable. It's the difference between a project and a thing.
6. Move on, but keep the lights on
Once a project ships, I move to the next one. I do not iterate forever. I keep the lights on — security updates, broken-link fixes, the occasional bug — but I don't try to grow every project into a company. Most of mine are happy as small, weird, useful things. The point of a one-duck studio isn't scale. The point is to keep shipping.
The meta-point
None of this is novel. Every productivity book ever written has a version of "ship small, ship often, kill scope creep". What's hard isn't knowing the rules. It's actually following them when the project is yours and there's no one to tell you no. The workflow above is mostly a set of pre-commitments I make to myself so that future-me, who is tired and over-caffeinated at 11pm, can't talk past-me out of finishing.
If you're stuck in the middle of something right now, try the single-sentence test on it. See if it survives.