Why I Shut Down Zapply
Zapply was my first real startup — a wild experiment to see if I could build something useful, make a little money, and learn what it takes to launch.
The idea was simple:
Software engineers pay $19/month, and a bot applies to remote jobs for them automatically.
Over a few months, I got deep into it:
- Scraped jobs from Working Nomads
- Hacked Greenhouse and Lever ATS
- Built a backend deployed to Fly.io
- Started a JavaFX desktop app
- Integrated SSL.com, Apple dev certs, install4j, and Stytch for auth
- Ran marketing experiments (Google Ads, Twitter, LinkedIn)
- Collected 500+ email signups
- 133 of them said they’d pay
We had traction. We had code. We had excitement.
But we also had a big problem.
❌ Why I Shut It Down
In short: the solution just wasn’t good enough.
- Job matching sucked. Keywords aren’t enough — a human can instantly tell if a job is relevant. Bots can’t.
- Limited ATS support. We only automated Greenhouse and Lever. The rest? Still a chaotic mess.
- Too many arbitrary questions. Every job asks different things. Either I had to set fake answers (bad), or ask users to fill forms (also bad).
Altogether, it added up to a product that was worse than just applying manually.
And if I wouldn’t use it myself, why ask others to?
So I shut it down.
✅ What I Learned (and Why It Was Worth It)
This wasn’t a waste — it was founder bootcamp.
- I learned to validate faster (next time: 2 weeks, not 2 months)
- I learned to run lean marketing tests
- I learned how hard real automation is
- I picked up new skills: auth systems, certificates, installers, app deployment
- I proved to myself I could launch something and get attention
So yes — it ended. But I’m walking away better, sharper, and ready for the next one.
Zapply is done. But I’m just getting started. Stay tuned!