Build log
Building in public. Every idea I research ends up here, including the ones I kill, with the reason. My bar: I'd use it myself, people already pay for something like it, customers can find it without ads, it needs less than three hours a week of upkeep, and it carries low legal risk.
November 2026: seven ideas I won't build
I researched these over a few weeks. None passed all five rules. Here's why.
-
Payment provider to accounting sync
- The idea
- Sync payouts and fees from payment providers into small-business accounting software.
- Why I stopped
- Existing integrations already cover the broad version well, and some are built by the payment providers themselves. Only a very narrow niche would be left.
- What now
- I'd only revisit a narrow variant after ten people say they would pay for it.
-
Document portal for accounting firms
- The idea
- A secure portal where clients upload documents for their bookkeeper.
- Why I stopped
- It's already built into the software these firms use, and the data includes national ID numbers. High risk, little room.
- What now
- Not unless I already know ten firms that want it.
-
E-invoicing for small businesses in Belgium
- The idea
- Send and receive Peppol e-invoices for small companies.
- Why I stopped
- The market filled up fast once e-invoicing became mandatory, with free options everywhere. A mistake could also cost a customer a fine.
- What now
- Done. A solved problem.
-
White-label uptime and SSL monitoring
- The idea
- Monitoring that agencies can resell under their own brand.
- Why I stopped
- Several good tools already offer white-label, and monitoring means being on call 24/7. That breaks my 3-hours-a-week maintenance rule.
- What now
- Done.
-
PDF generation API
- The idea
- An API that turns HTML into PDFs.
- Why I stopped
- A commodity: strong free open-source options exist, and big cloud providers bundle it.
- What now
- Done.
-
Stock and crypto signal app
- The idea
- Alerts when an asset hits a signal I care about.
- Why I stopped
- I use it myself, but selling signals is regulated financial advice in the EU, and market-data licences are expensive.
- What now
- Stays a personal hobby project.
-
Media player for IPTV
- The idea
- A polished player app for IPTV playlists.
- Why I stopped
- The category is closely tied to piracy, and app stores and payment providers treat it as high risk.
- What now
- Only as a white-label app for licensed providers, and probably not even then.
What I learned
- If an idea is obviously good, someone has usually already built it well.
- "I'd use it" isn't enough: I also need a way for customers to find it without paying for ads.
- Anything that needs me on call at 3 a.m. doesn't fit a one-person, part-time setup.