What I’m Left With After Winding Down a Three-Month Side Project
Yesterday I officially shut down a side project I started late last year. It got up to 40 users, but it just wouldn’t grow beyond that.
At first I went with “ship first” and put it up right away, but after I added auth, I stopped touching it. Working alone, I’d fix the frontend and the backend would break; fix that and the frontend would break again. Once that kept repeating, it got a lot less fun.
Still, I got something out of it. I can now set up a CI/CD pipeline on my own, and I’ve developed the habit of checking logs. Next time, I’m planning to write some tests from the start.
If anyone has wound down a project in a similar way, how did you start your next one?
9 answers
Got a source? "It didn’t grow past 40 users" just sounds like they didn’t do any marketing lol
Shutting it down is a decision too, and deciding after only three months might have been a bit hasty. But hey, if it stopped being fun for you, that's the most important signal. Best of luck with your next project.
True, deployment is half the battle, seriously.
40 users isn't a small number, is it? With that many, I feel it would've been a bit of a shame to give up.
Speaking of tests—trying to hit 80% coverage from the start is its own kind of hell. Start by just adding tests for a few core pieces of logic.
If you can now set up CI/CD on your own, haven't you already gotten your money's worth out of that side project? lol
This is just how it goes when you work alone... I hate having the backend break, so I switched to not building a frontend at all and just shipping API docs. Maintenance dropped dramatically after that.
Well, I’m not so sure about that. 40 users in 3 months is actually pretty good; the problem seems to be that you gave up after adding authentication, not the project itself. “Ship first” has spread like a myth, but in reality, if you add things like authentication and payments before meeting your first users, you lose all your motivation. Next time, ship an MVP with no authentication at all within two weeks, see how people respond, and then add it.
I also dropped a similar project last year, and what I realized then was: "A solo project doing both frontend and backend is doomed once it goes past three features." So next time I started with something backend-only, like a CLI tool or a crawler. With no UI, there's nothing to break lol. You don't even need to add auth. I recommend it.