Why most side projects never ship

I have a graveyard of unfinished projects. Half-built apps sitting in private repos, domain names I bought in a burst of enthusiasm that now auto-renew quietly every year. If you're a developer, you probably have something similar.

Over time I started noticing a pattern. The projects that ship and the ones that don't have almost nothing to do with the quality of the idea or how excited I was at the start. The difference comes down to a few specific behaviors that I can now predict within the first week.

The excitement trap

Every new project starts the same way. You get an idea, usually in the shower or while half-asleep, and suddenly you can see the whole thing. The landing page, the features, the onboarding flow. You open your laptop and start building immediately.

This is the most dangerous phase. You're high on possibility and you make decisions that feel productive but aren't. You set up a monorepo when a single folder would do. You configure CI/CD before you have anything to deploy. You spend an afternoon choosing between three icon libraries.

None of this moves you toward a working product. It moves you toward a well-configured project that doesn't do anything yet.

Scope is the killer

The number one reason side projects die is scope. Not technical complexity, not lack of time, not losing interest. Scope.

When you're planning in your head, everything seems small. "It's just a dashboard with some charts." But a dashboard needs auth. Auth needs email verification. Email verification needs a transactional email setup. The charts need data, which needs an API, which needs a database schema, which needs migrations. Suddenly your "simple dashboard" is a two-month project.

The projects that ship are the ones where I ruthlessly cut scope before writing a single line of code. Not after I've already built half of it. Before.

I ask myself: what is the absolute minimum this needs to do to be useful? Not impressive. Not polished. Useful. Then I cut that in half, because my first answer is always too ambitious.

Time boxing works

The best trick I've found is giving myself a hard deadline. Not "I'll try to finish this in two weeks" but "this ships on Friday or it doesn't ship at all."

When you have a real constraint, you make different decisions. You skip the custom design system and use defaults. You hardcode things that could be configurable. You deploy to Vercel instead of setting up your own infrastructure.

These shortcuts feel wrong in the moment. But a shipped product with hardcoded values beats an unshipped product with a beautiful config system every time.

Build the core loop first

Every product has a core loop. The one thing the user does repeatedly. For a todo app, it's adding and completing tasks. For an analytics tool, it's viewing a dashboard. For a social app, it's posting and reading.

Build that loop first. Not the onboarding. Not the settings page. Not the billing. The core loop.

If the core loop isn't interesting, no amount of polish will save the product. And if it is interesting, everything else can be rough and people will still use it.

I shipped the first version of 21metrics with no landing page, no onboarding, and a login flow that was literally just "enter your email." The core loop worked, and that was enough to get the first 50 users.

The finish line problem

The last 10% of a project takes 50% of the time. This is where most side projects die, because you've solved the interesting problems and all that's left is the tedious stuff. Error handling. Edge cases. Mobile responsiveness. Terms of service pages.

You have two options here. Either push through the tedium because you're committed to shipping, or accept that this project isn't going to ship and move on to the next thing. Both are valid.

What doesn't work is sitting in the middle. Spending a few hours every weekend on something you've lost enthusiasm for, making incremental progress that never quite gets you to done.

Ship something this week

If you have an idea you've been thinking about, here's my challenge: build something and ship it in one week. Not a proof of concept on your laptop. Something with a URL that another person can visit.

It doesn't need to be good. It doesn't need to be complete. It just needs to exist in the world where someone else can see it.

The gap between "I could build this" and "I built this" is enormous, and the only way to close it is to ship.