I Ranked Every Side Project I've Ever Built — Half of Them Taught Me Nothing
The Side Project Trap
Every developer I know has a graveyard. A GitHub full of half-finished repos, a domain they bought in a moment of inspiration, a Notion page titled "app ideas" that hasn't been opened since 2021. I'm no different — except I finally sat down and audited mine.
I listed every personal project I've built over the last eight years, estimated the hours I sank into each, and scored them on one question: did this actually make me better at my job?
The results were humbling. Roughly half were worth it. The other half were expensive lessons in what not to build. Here's the breakdown, winners first.
The Seven That Paid Off
1. The URL Shortener (Skill score: 8/10, Hours: ~40)
Yes, it's a cliché. It's also the perfect first backend project. Building one taught me routing, database design, redirects, caching, and — when I put it on a real domain — DNS, SSL, and deployment. If you've never shipped anything, ship this. The cliché exists for a reason.
2. A Discord Bot for My Friend Group (9/10, ~60 hours)
This one surprised me. A bot that pulled sports scores and managed game nights for my friends taught me async programming, rate limiting, and third-party API auth. The secret ingredient: real users who notice when it breaks. Nothing teaches error handling like twelve people complaining in a chat channel.
3. A Personal Finance Dashboard (9/10, ~80 hours)
I imported my own bank exports and built charts. Learned data pipelines, CSV parsing, and enough about time zones to develop a mild phobia of them. Also the only project that's saved me actual money.
4. Rebuilding My Portfolio From Scratch (7/10, ~50 hours)
Every rebuild forced me to learn the current hotness — and more importantly, to finish something and put it online. Shipping a public site that represents you is a different skill than shipping a private repo.
5. A CLI Tool for My Own Workflow (8/10, ~25 hours)
Small, boring, and the best ROI on this list. I automated a task I did weekly at work. It made me faster every single week, and "I built internal tooling" reads great in performance reviews.
6. A Recipe App With My Mom as the Only User (6/10, ~70 hours)
The app is mediocre. But building for a real, non-technical user taught me more about UX than any tutorial — mainly that my assumptions about what's "intuitive" are consistently wrong.
7. Contributing to Open Source (10/10, ongoing)
Not technically my project, but the single highest-leverage thing on this list. Reading other people's code, navigating a maintainer's review process, and writing a good PR description are all job skills you can't learn alone.
The Three That Weren't Worth It
8. The SaaS Idea (3/10, ~200 hours)
The big one. I spent five months building a niche productivity tool nobody asked for. I learned React patterns I'd have learned in three weeks on a smaller project, and I learned that building the product is maybe 20% of a SaaS. Marketing, onboarding, support — none of it was interesting to me. If your goal is skills, this was a wildly inefficient way to get them. If your goal is a business, do customer interviews before writing code. I did them after. Bad trade.
9. My Own Framework (2/10, ~90 hours)
"How hard can it be?" Very hard, turns out, and the world did not need a tenth JavaScript framework. The one redeeming skill: I finally understood what React and Express are actually doing under the hood. But I could've gotten 80% of that from reading source code for a weekend.
10. A Crypto Price Tracker (4/10, ~30 hours)
Built it because crypto was hot in 2021, not because I cared. Motivation died the moment the market did. The lesson: never pick a project based on a hype cycle. Pick something you'd still want to maintain when it's unfashionable.
My Rubric, So You Can Steal It
Before starting anything, I now ask four questions:
- Does it have real users, even if they're just me or my friends? Feedback loops are everything.
- Will it take under 50 hours to something usable? Big-bang projects die. Small ones ship.
- Does it use at least one skill I want for my next job? Alignment with career goals turns hobby time into resume material.
- Would I still care about this in six months? If the answer is hype-driven, pass.
The Bottom Line
Side projects are the best professional development money can't buy — if you pick them deliberately. The graveyard is fine; every developer has one. Just make sure your next repo is a future war story and not another headstone.
